NASA/TP—2004–213089

July 2004
Survey of Technologies Relevant to Defense
From Near-Earth Objects
R.B. Adams, R. Alexander, J. Bonometti, J. Chapman, S. Fincher, R. Hopkins,
M. Kalkstein, and T. Polsgrove
Marshall Space Flight Center, Marshall Space Flight Center, Alabama
G. Statham and S. White
ERC, Inc., Huntsville, Alabama
Since its founding, NASA has been dedicated to
the advancement of aeronautics and space
science. The NASA Scientific and Technical
Information (STI) Program Office plays a key
part in helping NASA maintain this important
role.
The NASA STI Program Office is operated by
Langley Research Center, the lead center for
NASA’s scientific and technical information. The
NASA STI Program Office provides access to the
NASA STI Database, the largest collection of
aeronautical and space science STI in the world. The
Program Office is also NASA’s institutional
mechanism for disseminating the results of its
research and development activities. These results
are published by NASA in the NASA STI Report
Series, which includes the following report types:
• TECHNICAL PUBLICATION. Reports of
completed research or a major significant phase
of research that present the results of NASA
programs and include extensive data or
theoretical analysis. Includes compilations of
significant scientific and technical data and
information deemed to be of continuing reference
value. NASA’s counterpart of peer-reviewed
formal professional papers but has less stringent
limitations on manuscript length and extent of
graphic presentations.
• TECHNICAL MEMORANDUM. Scientific and
technical findings that are preliminary or of
specialized interest, e.g., quick release reports,
working papers, and bibliographies that contain
minimal annotation. Does not contain extensive
analysis.
• CONTRACTOR REPORT. Scientific and
technical findings by NASA-sponsored
contractors and grantees.
• CONFERENCE PUBLICATION. Collected
papers from scientific and technical conferences,
symposia, seminars, or other meetings sponsored
or cosponsored by NASA.
• SPECIAL PUBLICATION. Scientific, technical,
or historical information from NASA programs,
projects, and mission, often concerned with
subjects having substantial public interest.
• TECHNICAL TRANSLATION.
English-language translations of foreign scientific
and technical material pertinent to NASA’s
mission.
Specialized services that complement the STI
Program Office’s diverse offerings include creating
custom thesauri, building customized databases,
organizing and publishing research results…even
providing videos.
For more information about the NASA STI Program
Office, see the following:
• Access the NASA STI Program Home Page at
http://www.sti.nasa.gov
• E-mail your question via the Internet to
help@sti.nasa.gov
• Fax your question to the NASA Access Help
Desk at 301–621–0134
• Telephone the NASA Access Help Desk
at 301–621–0390
• Write to:
NASA Access Help Desk
NASA Center for AeroSpace Information
7121 Standard Drive
Hanover, MD 21076–1320
301–621–0390
The NASA STI Program Office…in Profile
i
NASA/TP—2004–213089
R.B. Adams, R. Alexander, J. Bonometti, J. Chapman, S. Fincher, R. Hopkins,
M. Kalkstein, and T. Polsgrove
Marshall Space Flight Center, Marshall Space Flight Center, Alabama
G. Statham and S. White
ERC, Inc., Huntsville, Alabama
July 2004
National Aeronautics and
Space Administration
Marshall Space Flight Center • MSFC, Alabama 35812
Survey of Technologies Relevant to Defense
From Near-Earth Objects
ii
Available from:
NASA Center for AeroSpace Information National Technical Information Service
7121 Standard Drive 5285 Port Royal Road
Hanover, MD 21076–1320 Springfield, VA 22161
301–621–0390 703–487–4650
Acknowledgments
The authors would like to thank Langley Research Center for supporting this worthwhile endeavor through Revolutionary
Aerospace System Concepts (RASC) funding, and the Marshall Space Flight Center astronomy group
for the technical assistance of Roy Young.
.
TRADEMARKS
Trade names and trademarks are used in this report for identification only. This usage does not constitute an official
endorsement, either expressed or implied, by the National Aeronautics and Space Administration.
iii
EXECUTIVE SUMMARY
Nearly 300 million yr ago a comet with a diameter of ≈10 km struck the Earth near Chicxulub in
present-day Mexico. The impact created a crater nearly 300 km in diameter and drove very large quantities
of dust and debris into the atmosphere. The dust prevented much of the Sun’s radiation from reaching the
Earth’s surface and the resulting Ice Age caused the extinction of ≈50 percent of the existing animal species—
including, most famously, the dinosaurs. Approximately 50,000 yr ago a stony-iron asteroid with a diameter
of ≈150 m struck the Earth in what is now north-central Arizona. This impact created a crater 1 mi in
diameter and the resulting shockwave killed every large mammal within a radius of 24 km. In 1908, a small
comet or asteroid with a diameter of ≈50 m entered the Earth’s atmosphere over eastern Russia. The extreme
heat and aerodynamic pressures generated during entry caused the object to disintegrate explosively at an
altitude of ≈8 km. This explosive burst—since termed an airburst—occurred above Siberia, near the town
of Tunguska. The airburst took place very close to the altitude at which the maximum amount of ground
damage would result. It left a zone of destruction nearly 40 km in diameter around the point of disintegration.
The scientific community now accepts that these events are just major examples of the continuous
ongoing bombardment of the Earth by a wide variety of objects, most of them fragments of either asteroids
or comets in orbit around the Sun. Many feel that, because there appears to have been no loss of life due to
cosmic bombardment, this threat can be ignored. Against this it can be argued that, as our knowledge of the
solar system in general—and its minor bodies in particular—is very recent, many unexplained catastrophes
in the past may actually be attributable to the impact of asteroids or comets. Additionally, the exponential
growth of world population, combined with our increasing technological dependence, makes humanity
much more vulnerable to the consequences of a near-Earth object (NEO) impact. Despite these worrying
trends, effective means of defense against the NEO threat are available to us. That is the topic of the
Marshall Space Flight Center-led study reported in this Technical Publication (TP).
This TP is divided into nine sections. Each section is intended to gradually introduce the reader to
the various facets of the problem—quantifying the NEO threat, developing mitigation options, and evaluating
their effectiveness. After the introduction in section 1, section 2 presents the threat of asteroids and comets,
and outlines our growing understanding of NEO impacts. Section 3 discusses various mission configurations
that might be selected as part of a defense strategy. Section 4 describes in detail the propulsion technologies
that were considered as candidates to transport defense hardware out to an Earth-bound NEO. Section 5
reviews the actual defense or mitigation technologies that would be used to either fragment or deflect an
Earth-bound NEO. Section 6 introduces the trajectory modeling tools used in the study. Section 7 describes
the means by which all of the above tools were incorporated into a single “master” design environment and
used to search for optimal solutions to the planetary defense problem. Section 8 contains results from
parametric analyses that were conducted using the tools and data described in sections 2 through 7. Finally,
conclusions and recommendations from this project are presented in section 9.
Section 2 gives an overview of our current understanding of asteroids and comets, as well as the
threat that they pose to the Earth. Methods of categorization for asteroids and comets, by both orbital
iv
parameters and composition, are introduced. Next, the frequency, location, and sequence of known and
suspected impacts on the Earth are presented. A timeline is then given, describing the development of
human knowledge of asteroids, comets, and the impacts from these bodies. This timeline helps to explain
why our historical record does not contain more information on possible impacts. Current knowledge of
the NEO population within the solar system is then discussed. The immediate physical effects of various
types of impact are then described, as are the longer-term consequences. Finally, some thoughts are offered
on the current lack of significant public credibility regarding this threat.
There are many different options that are potentially available to defend against an incoming NEO.
Grouping these options into categories allows us to develop analysis processes applicable across an entire
category, rather than having to establish a different process for each option. This categorization approach is
new to this study; evidence of its previous use was nowhere in the literature.
At the highest level, mitigation options can be divided into two categories: deflection and
fragmentation. The deflection option leaves the object largely intact, but changes its velocity by a small
amount, sufficient to ensure that it will miss the Earth by a distance greater than or equal to a certain
minimum value—in the literature, this is usually set at 3 Earth radii. The fragmentation option breaks the
NEO into pieces, each small enough to burn up in the Earth’s atmosphere. There is some uncertainty over
the maximum fragment size that would not threaten the Earth; a diameter of 10 m was selected here as a
first approximation. The means by which energy is delivered—to either fragment or deflect an NEO—can
also be categorized. A remote station would project a beam or fire a projectile at the incoming NEO. An
interceptor would travel across the inner solar system and actually impact on the NEO. Finally, under the
rendezvous option, the mitigation hardware would be transported out to the NEO and would match orbits
with it. This allows the vehicle to operate on or near to the NEO for an extended duration, delivering its
deflection or fragmentation energy over an extended time instead of in one short impulse.
Several propulsion systems were considered as candidates to place the mitigation vehicle onto
either an intercept or a rendezvous trajectory with the NEO. A multistage liquid oxygen/liquid hydrogen
rocket was selected as the baseline option due to its technical maturity. The nuclear thermal rocket was
retained as an additional option because of its superior specific impulse. A scaled-down derivative of the
ORION concept of the 1960’s was also retained—referred to as the nuclear pulse technique. For trajectories
in the inner solar system requiring large inclination changes, solar sails are highly competitive and so were
likewise considered. Finally, a derivative of the solar sail—the solar collector—was considered.
In addition, several very interesting concepts for threat mitigation were considered. Fragmentation
of the incoming NEO using nuclear devices was assessed. The use of nuclear devices to deflect the NEO
was also considered. The solar sail was considered, but was rejected, as it was found to be impractical for
all but the smallest asteroids. The solar collector, however, showed remarkable promise for all but the
largest asteroids and comets. The novel option of deflection by use of a rapidly growing, pulsed magnetic
field was considered. Furthermore, the use of a mass driver, transported out to and installed on the incoming
asteroid, was considered and was modeled in some detail. Finally, deflection of the NEO by purely kinetic
means—utilizing an ultra-high velocity inert projectile—was considered.
Both the outbound and inbound trajectories were modeled using tools developed under this study.
The outbound model solved the Gauss problem for high thrust trajectories. The inbound model calculated
v
backwards in time from the point of impact with the Earth in order to determine the instantaneous ∆V
necessary to achieve the minimum required deflection. The inbound trajectory calculations accounted for
the influence of both the Sun and the Earth. Inclusion of the Sun’s gravitational field in these calculations
yields significantly different results than the two body approximations found in the literature. Some of our
propulsive techniques do not provide high thrust levels and, as they neglect gravity losses, these trajectory
calculation techniques are only approximate. However, for this first attempt at a solution, we were willing
to accept the consequent level of uncertainty in our answers. We intend to revisit this issue and model the
effects of gravitational losses at the earliest opportunity in the future.
Our threat parametric builds on the results of previous studies in modeling the overall consequences
of an impact. Our model uses the current knowledge of the asteroid and comet population to execute Monte
Carlo simulations to establish the probability of impact. These results are combined with estimates of the
average number of fatalities for an asteroid or comet of a given size and composition and are then used to
determine the average number of deaths resulting from this threat over a given time period. Our various
concepts for threat mitigation can then be evaluated by the percentage of this threat that each can defeat,
and ultimately, by the mean number of lives that would be saved.
The next objective was to combine the tools and concepts described above to determine the optimal
configurations to defeat the threat. Time and funding constraints necessitated a less ambitious approach.
Assuming a baseline asteroid orbit, we integrated our trajectory, propulsion, and threat mitigation tools to
quantify the relationship between required system mass and size of object deflected. These results are
summarized for the concepts evaluated in the study in the table below. Based on these results, as well as
qualitative comparisons documented here, we considered the nuclear pulse option the most viable for the
overall threat. However, recognizing our limited modeling capability for this study, we strongly recommend
a broad spectrum of deflection technologies, including all options evaluated here, be considered for any
future work.
System Maneuver
Time Before Impact
(days)/Outbound
Travel Time (days)
Total System Mass
at SOI (t)
for Different
Asteroid Diameters
(m)
Maximum Diameter
of Asteroid*(m)/
Total System Mass
at Earth SOI (t)
Staged chemical/
mass driver
Staged chemical/
nuclear deflection
Staged chemical/
kinetic deflection
Nuclear pulse
Solar collector
Rendezvous
Rendezvous
Intercept
Intercept
Rendezvous
Rendezvous
(≈3 yr)
Rendezvous
(≈10 yr)
2,900/2,400 NA NA NA
100 1,000 10,000
50/6,849
80/6,918
73.8 NA NA
29.7 41.8 1,240
0.637 1.07 167
0.550 0.636 34.6
0.847 8.27 1,300 9,000/1,000
9,000/1,000
1,000/1,000
260/1,000
5.62 568 87,800
1,509/910
1,025/800
2,170/1,200
1,076/65**
3,635/115**
1,075/132
§
§
* Maximum was constrained to a total system mass at Earth SOI of 1,000 t.
** Times are for 100-m-diameter chondrite. Rendezvous times are greater for larger asteroids, although total missions times change little.

§
The solar collector system is limited more by solar collector size than by total system mass. See figure 112.
vii
TABLE OF CONTENTS
1. INTRODUCTION....................................................................................................................... 1
2. THREAT ................................................................................................................................... 3
2.1 Categorization of Asteroids and Comets .............................................................................. 3
2.2 Earth’s Impact Record .......................................................................................................... 7
2.3 History Related to Near-Earth Objects ................................................................................. 13
2.4 Measuring the Near-Earth Object Population ...................................................................... 16
2.5 Damage Mechanisms............................................................................................................ 18
2.6 The Credibility Problem ....................................................................................................... 25
2.7 The Torino Scale ................................................................................................................... 29
3. MISSION CONFIGURATIONS ................................................................................................. 31
3.1 Deflection Versus Fragmentation ......................................................................................... 31
3.2 Remote Station Versus Interception Versus Rendezvous ..................................................... 33
4. OUTBOUND PROPULSION ..................................................................................................... 37
4.1 Staged Chemical ................................................................................................................... 38
4.2 Nuclear Thermal Rocket ....................................................................................................... 43
4.3 Nuclear Pulse ........................................................................................................................ 46
4.4 Solar Sail .............................................................................................................................. 53
4.5 Solar Collector ...................................................................................................................... 60
5. THREAT MITIGATION ............................................................................................................. 62
5.1 Nuclear Fragmentation ......................................................................................................... 62
5.2 Nuclear Deflection................................................................................................................ 66
5.3 Solar Sails ............................................................................................................................. 71
5.4 Solar Collector ...................................................................................................................... 74
5.5 Magnetic Flux Compression................................................................................................. 76
5.6 Mass Driver .......................................................................................................................... 81
5.7 Kinetic Deflection................................................................................................................. 93
6. TRAJECTORY MODELING...................................................................................................... 102
6.1 Outbound .............................................................................................................................. 102
6.2 Inbound ................................................................................................................................. 103
viii
TABLE OF CONTENTS (Continued)
7. THREAT PARAMETRIC ........................................................................................................... 110
8. PARAMETRIC RESULTS.......................................................................................................... 115
8.1 Integrated Analysis ............................................................................................................... 115
8.2 Architecture Options ............................................................................................................ 117
8.3 Parametric Performance ....................................................................................................... 118
9. CONCLUSIONS AND RECOMMENDATIONS ..................................................................... 144
9.1 Public Awareness ................................................................................................................. 144
9.2 Statistical Problem ............................................................................................................... 144
9.3 Funding of Future Work ....................................................................................................... 145
9.4 Development and Deployment of Mitigation Systems ........................................................ 145
9.5 Accomplishments ................................................................................................................. 146
9.6 Assessment of Mitigation Options ....................................................................................... 147
9.7 Future Work ......................................................................................................................... 148
9.8 Summary Conclusion ........................................................................................................... 149
APPENDIX A—CURRENT NEAR-EARTH OBJECT SEARCH PROGRAMS ........................... 150
A.1 SpaceWatch .......................................................................................................................... 150
A.2 Spaceguard ........................................................................................................................... 150
A.3 Lincoln Near-Earth Asteroid Research ................................................................................ 150
A.4 Near-Earth Asteroid Tracking .............................................................................................. 151
A.5 Lowell Observatory Near-Earth Object Search ................................................................... 151
APPENDIX B—SOLAR ARRAY CALCULATIONS ..................................................................... 152
B.1 Method I ............................................................................................................................... 152
B.2 Method II.............................................................................................................................. 152
APPENDIX C—MASS DRIVER..................................................................................................... 154
C.1 Model of the Forces on a Bucket Coil Due to the Nearby Drive Coils ............................... 154
C.2 Drive and Bucket Coil Currents ........................................................................................... 158
C.3 Analysis of Bucket Kinetic Energy and Acceleration .......................................................... 161
C.4 Drive Coils ........................................................................................................................... 163
C.5 Braking Coils ....................................................................................................................... 166
C.6 Bucket Design ...................................................................................................................... 166
C.7 Interference Between Adjacent Drive Coils......................................................................... 170
C.8 Effect of Bucket Coil Motion on Stationary Coil Circuit Operation ................................... 176
ix
TABLE OF CONTENTS (Continued)
C.9 Drive Coil ........................................................................................................................... 180
C.10 Braking Coil ....................................................................................................................... 181
C.11 Self-Induced Coil Stresses ................................................................................................. 182
APPENDIX D—COIL FORCE MODEL ......................................................................................... 185
REFERENCES .................................................................................................................................. 190
x
LIST OF FIGURES
1. Organization of RASC FY 2002 activities ............................................................................ 1
2. Orbits of 2062 Aten, 1862 Apollo, and 1221 Amor relative to Earth and Mars ................... 4
3. Orbits of representative short-period comets relative to the outer planets.
Halley’s Comet represents the Halley class and Comet Crommelin
represents the Jupiter class .................................................................................................... 5
4. Orbits of representative short- and long-period comets relative to the solar system.
Comet Hale-Bopp represents long-period comets ................................................................ 6
5. Location of known impact craters noting diameters as of 1998.
Age distribution of these craters is also included.................................................................. 8
6. Projected area affected from the Tunguska blast of 1908. Arrows depict
the location and direction trees were knocked down from the blast ..................................... 9
7. Tunguska impact area superimposed over Madison County, Alabama, USA.
Several hundred thousand casualties can be expected from such an impact ......................... 10
8. Aerial view of the Barringer Impact Crater in Arizona, USA............................................... 11
9. Calculated impact areas from the Barringer meteor .............................................................. 12
10. Extinction events versus time according to the fossil record ................................................ 13
11. Location of known minor planets on March 2, 2002, plotted relative
to the inner planets. NEOs are red, other asteroids are green, and comets are blue.............. 16
12. Number of known NEAs versus time. Note the rapid increase in discoveries
in recent years due to the use of CCDs and increased interest in the asteroid
and comet threat .................................................................................................................... 17
13. Deep-water wave height at 1,000 km distance versus initial meteor radius
for soft stone meteor .............................................................................................................. 20
14. Deep-water wave height at 1,000 km distance versus initial meteor radius
for Fe meteor ......................................................................................................................... 20
xi
LIST OF FIGURES (Continued)
15. Blast wave damage versus impact energy ............................................................................. 21
16. Density and optical depth of atmospheric dust versus impact energy .................................. 23
17. Blast wave damage versus impact energy ............................................................................. 24
18. Mass of water lifted into the atmosphere versus impact energy ........................................... 25
19. Predicted world population in the last century. Population is extrapolated
through the middle of this century ........................................................................................ 27
20. Human population density graph for all continents (except Antarctica).
Greenland and Iceland are not represented as well as some Pacific islands ......................... 28
21. Illustration of the various category of threat under the Torino Scale .................................... 29
22. Categories within the Torino Scale ....................................................................................... 30
23. Illustration of deflection method of threat mitigation ........................................................... 32
24. Illustration of fragmentation method of threat mitigation..................................................... 32
25. Delivering deflection or fragmentation energy by the remote station mode......................... 34
26. Delivering deflection or fragmentation energy by the interception mode ............................ 34
27. Delivering deflection or fragmentation energy by the rendezvous mode ............................. 35
28. Two-stage lox/LH
2
vehicle. Image produced by INTROS ................................................... 38
29. Regression curve fit of lox/LH
2
launch vehicle stages ......................................................... 39
30. Schematic of a nuclear thermal rocket .................................................................................. 43
31. Regression curve fit of NERVA program-developed engines ............................................... 44
32. U.S. Air Force ORION spacecraft (1964) ............................................................................. 47
33. NASA Gabriel spacecraft (1999) .......................................................................................... 47
34. EPPP concept vehicle ............................................................................................................ 50
xii
LIST OF FIGURES (Continued)
35. Shock absorber operations .................................................................................................... 51
36. Artists concept of a billowing, square solar sail .................................................................... 54
37. Schematic and dimensions for a square solar sail ................................................................. 54
38. Sail trajectories relative to lightness number ........................................................................ 58
39. Solar collector configuration ................................................................................................. 60
40. Blast yield—explosive placed at center of body—required for fragmentation
as a function of asteroid radius .............................................................................................. 63
41. Device mass versus explosive yield ...................................................................................... 64
42. Geometric position of the pulse unit to the planetary body and cone
half-angle definition .............................................................................................................. 69
43. Nuclear pulse rocket delivery system sketch ........................................................................ 70
44. Deflection ∆V imposed on 10-m-diameter asteroid .............................................................. 72
45. Deflection ∆V imposed by 6 gm/m
2
solar sail with varying areas
on varying diameter asteroids ............................................................................................... 73
46. Schematic of solar collector .................................................................................................. 74
47. ∆V imposed on varying asteroid sizes by 100-m-diameter solar collector ........................... 75
48. Critical disk diameter ............................................................................................................ 78
49. Magnetic flux compression generato .................................................................................... 80
50. E-bomb magnetic flux compression generator ...................................................................... 80
51. Coil gun conceptual design ................................................................................................... 83
52. Relative current flow directions of bucket and immediately adjacent drive coils ................ 83
53. Relative current flow directions of bucket and nearest drive coils ....................................... 84
54. Relative current flow directions of bucket and surrounding drive coils ............................... 85
xiii
LIST OF FIGURES (Continued)
55. Relative current flow directions of bucket and nearby drive coils ...................................... 85
56. Mass driver operation .......................................................................................................... 86
57. Relative current flow directions of bucket and braking coils.............................................. 86
58. Mass driver system schematic ............................................................................................. 87
59. Mass driver system view ..................................................................................................... 89
60. Main components of the mass driver .................................................................................. 89
61. System following discharge—bucket returns for reloading................................................ 90
62. Bucket about to be reloaded and cooled prior to next discharge......................................... 90
63. Loaded bucket about to enter mass driver ........................................................................... 91
64. Bucket containing expellant under acceleration within mass driver ................................... 91
65. Discharge in progress—bucket is decelerating while expellant mass
exits mass driver at high speed............................................................................................ 92
66. Following discharge—decelerated bucket exits mass driver
and joins return system........................................................................................................ 92
67. Interception geometry ......................................................................................................... 94
68. Impact and ejection geometry ............................................................................................. 95
69. Fragmentation data and curve fit results ............................................................................. 100
70. Outbound trajectory ∆Vs for 3,600-day total mission duration .......................................... 103
71. Illustration showing a typical NEO orbit. The velocity of the planetary
body at impact for this case is (–40,0,0)
T
km/s, parallel to the x axis
of the Heliocentric-Ecliptic system..................................................................................... 104
72. M1999JT6 orbit plot ............................................................................................................ 105
73. Required impulsive ∆V for 42 km/s velocity for various maneuvers
to avoid collision with Earth, showing the benefit of the UP maneuver
when impact is only a few days away ................................................................................. 107
xiv
LIST OF FIGURES (Continued)
74. Required impulsive ∆V for 42 km/s velocity for various maneuvers to avoid
collision with Earth, showing the benefit of the DECEL and OUTSIDE
maneuvers when impact is several weeks away................................................................... 107
75. Required impulsive ∆V for 35 km/s velocity for various maneuvers
to avoid collision with Earth, showing the benefit of the DECEL maneuver
when impact is only a few days away .................................................................................. 108
76. Required impulsive ∆V for various maneuvers to avoid collision with Earth
for planetary body with velocity of 35 km/s ........................................................................ 108
77. Required impulsive ∆V for various maneuvers to avoid collision with Earth
for planetary body with velocity of 35 km/s (long lead time) .............................................. 109
78. Average deaths from single asteroid impact versus size, and the average
total number of deaths prevented if all impacts of equal or less energy
can be avoided ...................................................................................................................... 114
79. Proposed analysis process for assessing total amount of threat mitigated ........................... 116
80. Staged chemical/mass driver model ..................................................................................... 119
81. (a) Optimal deflection direction, (b) optimal deflection direction—detailed view,
and (c) optimal deflection direction—detailed view—minimal deflection ∆V.................... 120
82. Variation of mass driver total system mass with required asteroid
deflection ∆V for a 50-m-diameter chondrite ....................................................................... 121
83. Staged chemical/mass driver vehicle mass at Earth departure ............................................. 122
84. Staged chemical/mass driver vehicle mass at Earth departure (expanded view) ................. 123
85. Optimal staged chemical/mass driver mission ..................................................................... 123
86. Staged chemical/mass driver vehicle mass versus chondrite asteroid diameter .................. 124
87. Mass driver deployed mass and total operating time versus
chondrite asteroid diameter (50–100 m) .............................................................................. 124
88. Mass driver deployed mass and total operating time versus chondrite asteroid
diameter (100–1,000 m) ....................................................................................................... 125
xv
LIST OF FIGURES (Continued)
89. Diagram of the ModelCenter setup for the staged chemical/nuclear deflection option ........ 125
90. Total system mass for the staged chemical/nuclear blast option versus
total mission time for various rendezvous times. Here, the staged chemical
system does not match the asteroid’s orbit at encounter. ....................................................... 126
91. Total system mass for the staged chemical/nuclear blast option versus
total mission time for various rendezvous times (zoomed). Here, the staged
chemical system matches the asteroid’s orbit at encounter. .................................................. 127
92. Minimum total system mass for the staged chemical/nuclear blast option,
showing the optimum rendezvous and total mission times for intercept .............................. 127
93. Minimum total system mass for the staged chemical/nuclear blast option,
showing the optimum rendezvous and total mission times for rendezvous .......................... 128
94. Optimum intercept trajectory for the staged chemical/nuclear deflection option ................. 128
95. Optimum intercept trajectory for the staged chemical/nuclear deflection option ................. 129
96. Minimum total system mass for the staged chemical/nuclear blast option versus
chondrite diameter for both intercept and rendezvous .......................................................... 130
97. Minimum total system mass for the staged chemical/nuclear blast option versus
chondrite diameter for the smaller chondrites, showing both intercept and rendezvous ...... 130
98. Staged chemical/kinetic deflection model ............................................................................. 131
99. (a) Staged chemical/kinetic deflection vehicle mass at Earth departure,
(b) detailed view, and (c) detailed view—minimum mass solution ...................................... 133
100. Optimal staged chemical/kinetic deflection mission ............................................................. 134
101. Staged chemical/kinetic deflection vehicle mass versus chondrite asteroid diameter .......... 135
102. Projectile mass versus chondrite asteroid diameter ............................................................... 135
103. Diagram of the ModelCenter setup for the nuclear pulse option .......................................... 136
104. Minimum total system mass for the nuclear pulse option, showing
the optimum rendezvous and total mission times.................................................................. 137
105. Optimum rendezvous trajectory for the nuclear pulse option ............................................... 137
xvi
LIST OF FIGURES (Continued)
106. Minimum total system mass for the nuclear pulse option versus chondrite diameter ......... 138
107. Minimum total system mass for the nuclear pulse option versus chondrite diameter
for the smaller chondrites ..................................................................................................... 138
108. Plot of the difference between required outbound and inbound solar collector sizes.
Negative values mean that the inbound requirement dominates. ......................................... 139
109. Combinations of total mission time and rendezvous time where inbound and outbound
required solar sail sizes are equal, and the associated total system mass for
500-m-diameter chondrite .................................................................................................... 140
110. Diagram of the ModelCenter setup for the solar collector option........................................ 141
111. Optimum rendezvous trajectory for the solar collector option for a
100-m-diameter chondrite .................................................................................................... 142
112. Minimum total system mass and size for the solar collector option versus
chondrite diameter ................................................................................................................ 142
113. Minimum total system mass and size for the solar collector option versus chondrite
diameter for the smaller chondrites ...................................................................................... 143
114. Drive coils included in and omitted from the analysis ......................................................... 154
115. Drive coils included in the analysis...................................................................................... 155
116. Elliptical function curve fits—K(m) and E(m) versus m...................................................... 157
117. Bucket coil current directions .............................................................................................. 159
118. Bucket and drive coil current directions during acceleration............................................... 160
119. Ideal current versus time profile for a single drive coil ....................................................... 163
120. Sinusoidal current versus time profile for a single drive coil .............................................. 164
121. Oscillating drive circuit—charging capacitor ...................................................................... 164
122. Oscillating drive circuit—discharging capacitor .................................................................. 164
123. Real current versus time profile for a single drive coil ........................................................ 165
xvii
LIST OF FIGURES (Continued)
124. Bucket conceptual design .................................................................................................... 166
125. Bucket handling through a complete cycle ......................................................................... 168
126. Discharge of isolated drive coil ........................................................................................... 170
127. Current flow directions of two adjacent drive coils ............................................................ 172
128. Simple model of a drive coil circuit .................................................................................... 176
129. The effect of bucket acceleration ........................................................................................ 181
130. Self-induced magnetic field and resulting force ................................................................. 182
xviii
LIST OF TABLES
1. Description of asteroid and comet compositions and representative predicted densities ..... 6
2. Comparable terrestrial events for NEOs of various diameters .............................................. 7
3. Timeline of scientific discoveries relevant to humanity’s knowledge
of asteroids and comets ......................................................................................................... 15
4. Recent near misses by comets and asteroids. By comparison, the distance
between the Earth and the Moon is ≈240,000 mi. ................................................................. 18
5. Consequences of impact by NEOs of various sizes .............................................................. 26
6. Data on in-service and historical launch vehicles that have lox-/LH
2
-powered stages ........ 40
7. Qualitative considerations for outbound propulsion using chemically-powered rockets ..... 42
8. Data on nuclear rocket engines developed under the NERVA program................................ 44
9. Qualitative considerations for outbound propulsion using nuclear thermal rockets ............. 45
10. Qualitative considerations for outbound propulsion using nuclear pulse ............................. 48
11. Optical, billowing force and other parameters used in solar sail analysis ............................ 56
12. Qualitative considerations for outbound propulsion using solar sails................................... 59
13. Qualitative considerations for outbound propulsion using solar collectors .......................... 61
14. Nuclear device masses .......................................................................................................... 64
15. BLU–113 penetrator characteristics ...................................................................................... 65
16. Qualitative considerations for threat mitigation using nuclear fragmentation ...................... 65
17. Qualitative considerations for threat mitigation using nuclear deflection ............................ 67
18. Chemical rocket assumptions ................................................................................................ 71
xix
LIST OF TABLES (Continued)
19. Qualitative considerations for threat mitigation using a solar sail ...................................... 73
20. Qualitative considerations for threat mitigation using a solar collector .............................. 76
21. Qualitative considerations for threat mitigation using magnetic flux compression ............ 81
22. Qualitative considerations for threat mitigation using the mass driver ............................... 93
23. Relative size of largest fragment at various collisional energies ........................................ 99
24. Qualitative considerations for threat mitigation using kinetic deflection ........................... 101
25. Original and modified orbital elements of 1999JT6 ........................................................... 104
26. Explanation of the different maneuvers available for use in the program PBI ................... 106
27. Types of planetary bodies examined in the Monte Carlo simulation and their average
contribution to the total number of deaths over the next century........................................ 110
28. Architecture options considered in this study ..................................................................... 118
29. Causes of death and associated probabilities for a U.S. resident ........................................ 145
30. Summation of parametric results for mitigation concepts .................................................. 147
xx
LIST OF ACRONYMS AND SYMBOLS
AANEAS Anglo-Australian Near-Earth Asteroid Survey
au astronomical unit
carb carbonaceous chondrite
CCD charged-coupled device
CF catastrophic fragmentation
DOD Department of Defense
EMP electromagnetic pulse
EPPP external pulsed plasma propulsion
Fe iron
GEODSS ground-based electro-optical deep space surveillance
GMD global missile defense
H hardness parameter for asteroidal material
H
2
O water
HOPE human outer planet exploration
ICBM intercontinental ballistic missile
ICR inductance capacitance resistor
JPL Jet Propulsion Laboratory
K-T Cretaceous-Tertiary (extinction event)
LaRC Langley Research Center
LCR inductance-capacitance-resistance
xxi
LIST OF ACRONYMS AND SYMBOLS (Continued)
LEO low-Earth orbit
LH
2
liquid hydrogen
LINEAR Lincoln near-Earth asteroid research
LN
2
liquid nitrogen
LONEOS Lowell Observatory near-Earth object search
lox liquid oxygen
LP long period (comet)
Ma mega annum (million years)
MSFC Marshall Space Flight Center
MSSS Maui Space Surveillance Site
MTM momentum transfer mechanism
NEA near-Earth asteroid
NEAT near-Earth asteroid tracking
NEO near-Earth object
NERVA nuclear engine for rocket vehicle applications
Ni nickel
NO nitrogen oxide
N
2
O
4
nitrogen tetroxide
NTR nuclear thermal rocket
PBI planetary body intercept
PBM planetary body maneuvering
xxii
LIST OF ACRONYMS AND SYMBOLS (Continued)
PBO planetary body—outbound
PHA potentially hazardous asteroid
RASC revolutionary aerospace systems concept
SOI sphere of influence
SP short period (comet)
TD30 Advanced Concepts Department
TNT tri-nitro-toluene
TP Technical Publication
UDMH unsymmetrical dimethyl hydrazine
USA United States of America
xxiii
NOMENCLATURE
A area—solar array, solar sail, or cross sectional
a semimajor axis of orbit (astronomical units or kilometers)
a
0
solar sail characteristic acceleration
a
B
radius of bucket coil
a
D
radius of drive coil
B magnetic flux density; magnetic field
B
b
non-Lambertian coefficient for backside of sail
B
f
non-Lambertian coefficient for front side of sail
C electrical capacitance
C
1
,C
2
,C
3
solar sail force coefficients
C
LN
2
specific heat of liquid nitrogen
c speed of light
D spherical diameter
D
standoff
standoff distance
E elliptical integrals of the second kind; electric field; energy
E
P
energy of collision
e orbital eccentricity
F number of expected average fatalities per year for all impacts of some given energy
or less; total radial force acting on a current-carrying coil; solar flux; force; thrust
F
an
absorptive force normal to solar sail centerline
F
at
absorptive force tangential to solar sail centerline
F
E
effectivity factor
F
e
emissive force tangential to solar sail centerline
F
inert
inert mass fraction
F
n
solar sail force normal to solar sail centerline
xxiv
NOMENCLATURE (Continued)
F
norm
total force adjusted for billowing
F
rn
reflective force normal to solar sail centerline
F
rt
reflective force tangential to solar sail centerline
F
t
solar sail force tangential to solar sail centerline
f number of expected average fatalities per year for an impact of a given energy; radial
force per unit length of coil
f
i
substitute variable (used in bucket coil motion analysis); i = 1,2,3,4
G constant of gravitation
g
0
gravitational constant at Earth’s surface
H asteroid “hardness”
H
c
total thermal capacity of bucket coils and their LN
2
coolant
H
vap
heat of vaporization
h bucket internal height
I orbital inclination (degrees); electrical current
I
0
initial current
I
B
current flowing in bucket coil; solar sail front side emission coefficient
I
D
current flowing in drive coil
I
D0
maximum current flowing in drive coil
I
j
current flowing in coil j
I(r) luminous intensity
I
sp
specific impulse
I
sp
n
specific impulse of the nth stage
J current density
K elliptical integrals of the first kind; thermal conduction constant between bucket LN
2
reservoir and external cold plate
KE kinetic energy
k variable used to evaluate elliptical integrals
L electrical self-inductance; conductor length
xxv
NOMENCLATURE (Continued)
L
s
solar luminosity
l diameter of solar collector or curved sail; length; orbital inclination (degrees)
M mutual inductance between two current-carrying, single-turn coils; mass of solar array;
magnification factor for solar collector
M
B
mass of loaded bucket
M
body
mass of body
M
ej
mass of material ejected due to the interceptor impact
M
f
final mass
M
f
n
final mass of nth stage
M
i
mass of interceptor; initial mass
M
ij
mutual inductance between bucket coils i and j
M
i
n
initial mass of nth stage
M
i
(x) mutual inductance between one turn of a bucket coil and one turn of drive coil I when the
bucket coil is a distance x from the drive coil
M
L
mass of largest fragment
M
LN
2
mass of liquid nitrogen used to cool the bucket superconducting coils
M
NEO
mass of near-Earth object
M
n
mutual inductance between one turn of a bucket coil and one turn of stationary coil n
MR
n
mass ratio of nth stage
M
S
mass of Sol (the Sun)
M
T
mass of target NEO
m mass
mass flow rate of ejecta
m
pay
payload mass
m
pay
n
payload mass of nth stage
m
p
n
propellant mass of nth stage
m
reac
NTR reactor mass
m/s mass relative to Earth
e
m&
xxvi
NOMENCLATURE (Continued)
m
s
n
inert mass of nth stage
N number of circuits in an electrical system; number of turns in coil
N
B
number of turns per bucket coil
N
D
number of turns per drive coil
N
DC
total number of drive coils
N
i
number of turns in coil i
P electrical power
P
ej
momentum of ejected material
P
jet
jet power
p
o
total pressure
Q,q electrical charge
R resistance of coil; radius of coil
R
a
asteroid orbital radius
R
c
crater radius
R
E
Earth’s orbital radius
R
S
radius of Sun
r bucket internal radius; radius of vehicle from Sun; distance from coil to the target
′ r
reflection coefficient; idealized radius of target body
˙˙ r acceleration of orbiting body
r
o
wire radius
r position vector
r
E/S
position vector of the Earth relative to the Sun (other objects use similar notation; time
derivatives use the overdot notation)
S distance between adjacent stationary coils; standard intercoil distance
′ s
spectral reflection coefficient
T temperature; tension
T
cp
temperature of cold place
T
H
maximum temperature of liquid nitrogen used to cool bucket superconducting coils
xxvii
NOMENCLATURE (Continued)
T
L
minimum temperature of liquid nitrogen used to cool bucket superconducting coils
t time
U total magnetic energy
U
e
total electrostatic energy
U
M
magnetic energy
V relative speed of collision; bucket internal volume; electrical potential
V
E/S
velocity vector of the Earth relative to the Sun (other objects use similar notation)
v velocity
v
e
velocity of ejecta
v
i
interceptor speed; mass driver bucket speed as it passes drive coil i
v
min
minimum speed of ejected material
W required blast yield
x distance between bucket and drive coils
Y NEO material strength
α Lagrange multiplier; angle of attack
β sail lightness number
′ β
gas expansion factor
∆KE change in kinetic energy experienced by mass driver bucket as it moves between two
adjacent drive coils
∆Q generated in mass driver bucket during one operational cycle; total amount of heat; total
amount of heat generated in mass driver bucket during one operational cycle
δ skin depth
p δ
coefficient, vertical component of momentum
ε overall efficiency of solar electric power system; induced electromagnetic field
ε
b
emission coefficient for back
ε
f
emission coefficient for front
ε
n
inert mass fraction
xxviii
NOMENCLATURE (Continued)
η ratio of bucket internal height to internal radius; efficiency of solar sail
λ variable used in equation solutions
λ
n
payload mass fraction for nth stage
µ
0
permeability of free space
ν ejecta speed; final mass driver bucket speed mass driver stationary coil current oscillation
frequency
θ solar sail cone angle
ϕ variable of integration used to evaluate elliptical intgrals; angle between impact and ejecta
ρ NEO material mass density; resistivity
σ solar array mass per unit area or solar sail loading parameter; conductivity
′ σ
Stefan-Boltzmann constant
τ time during which bucket must remain in contact with external cold plate in order for its coils
to be cooled after one operational cycle
Φ magnetic flux
Φ
i
total magnetic flux through bucket coil i
φ solar sail centerline angle
Ω longitude of the ascending node (degrees)
ω argument of Perihelion (degrees); oscillation frequency of coil current
l
r
coil path vector
Subscripts
E
denoting the Earth
PB
denoting the planetary body
S
denoting the Sun
1
TECHNICAL PUBLICATION
SURVEY OF TECHNOLOGIES RELEVANT TO DEFENSE FROM NEAR-EARTH OBJECTS
1. INTRODUCTION
In FY 2002, the revolutionary aerospace systems concepts (RASCs) activity, managed from
Langley Research Center (LaRC) selected a broad range of projects for the year’s activities. These projects
were organized into five groups as shown in figure 1. Marshall Space Flight Center (MSFC) participated in
two of these groups, led from TD30/Advanced Concepts. The work completed for group 2—human outer
planet exploration (HOPE)—is documented in NASA/TP—2003–212691.
1
MSFC’s participation in
group 4 activities is documented in this Technical Publication (TP). Several projects were funded under
group 4; however, MSFC’s activities were confined to planetary body maneuvering (PBM).
Figure 1. Organization of RASC FY 2002 activities.
RASC
Group 1:
Human/Robotic
Exploration
Group 2:
Human Exploration
of the Solar System
Beyond Mars
Human outer planet
exploration (HOPE)—
develop revolutionary
aerospace systems concepts
for human exploration
of the outer planets
Group 3:
Aerospace
Mobility Studies
Group 4:
In-Space
Remote Sensing
Group 5:
Tool Development
Planet body maneuvering
(PBM)—
develop, investigate, and
evaluate techniques for
maneuvering potentially
Earth-threatening
planetary bodies
(asteroids, comet nuclei)
2
Work under the PBM project was confined to defense of the Earth from collisions from asteroids
and comets. Many of the technologies developed for protective maneuvering of planetary bodies are also
applicable to maneuvering these bodies for resource utilization. Asteroids and comets can be maneuvered—
in a careful and controlled way—close to the Earth to be mined for structural materials and water. The mass
of these bodies could also anchor rotating tethers and skyhooks. Finally, such bodies could be intentionally
targeted to impact Mars or Venus in order to alter rotation speed and/or atmospheric composition.
Despite these other potential applications for maneuvering technologies, this project concentrated
solely on planetary defense as the most critical mission—a uniquely suitable one for NASA. This can be
seen immediately from the NASA mission statement:
To understand and protect our home planet
To explore the Universe and search for life
To inspire the next generation of explorers
… as only NASA can.
It has been suggested that the mission of planetary defense is best suited to the Department of
Defense (DOD). This argument is based on the DOD’s extensive experience in the interception of high-
speed objects. However, the very high energies necessary for deflection of massive planetary bodies com-
bined with the unique problems of operation in interplanetary space suggest that NASA will have a major,
if not a leading, role to play. The above mission statement suggests very strongly that the Agency should
address this threat. NASA’s unique capabilities may well make it the most uniquely qualified organization
in the world to take on the daunting task of protecting the planet from this threat.
Research conducted by the TD30 PBM team to understand and categorize the threat of impact by
an asteroid or comet is summarized in this TP. Using the limited knowledge currently available on the solar
system’s asteroid and comet population, an analytical tool was developed to estimate the number of human
lives that could potentially be lost because of this threat over a specified period of time. Propulsion tech-
nologies suitable for reaching the approaching object were then researched and deflection methods inves-
tigated. Analytical tools were developed to model the actual deflection techniques. These various tools
were then linked with an additional set of tools capable of modeling both inbound and outbound trajecto-
ries. Parametric results could then be generated using the linked propulsion, deflection, and trajectory tools
to calculate optimal deflection techniques for use against specific threat scenarios. Finally, these paramet-
ric results are presented and a set of conclusions established as to the effectiveness of each deflection
method.
3
2. THREAT
2.1 Categorization of Asteroids and Comets
Asteroids and comets are both categorized as minor planets. This designation is appropriate on the
grounds of both size and orbital parameters. An additional level of categorization can also be established
on the basis of composition. These categorizations are discussed further below.
Distinguishing between asteroids and comets is not entirely straightforward. Asteroids are small
objects (<1,000-km diameter), usually in eccentric, low-inclination solar orbits. The majority of asteroids
have both aphelia and perihelia between the Martian and Jovian orbits, although there are exceptions. They
are thought to be largely of rocky and/or metallic composition.
By comparison, comets can be distinguished from asteroids by both orbit and composition. Cometary
orbital eccentricities are usually higher than those of asteroids (e > 0.35 for most comets, with some even
having e > 0.9). Cometary aphelia are usually located beyond the orbits of the gas giants. Comets are
ice-rich bodies that become visually prominent when heat from the Sun causes their trapped volatiles to
sublimate. The most visible and distinctive features of comets are the coma and tail, produced by the
release of these volatile compounds, and also by dust. Most of the mass of a comet is contained within a
comparatively tiny central nucleus, now thought likely to have been formed directly from the primordial
solar nebula.
Asteroids can also be categorized according to their orbital parameters. Each category is usually
named after its first representative. Hence, the asteroids 2062 Aten, 1862 Apollo, and 1221 Amor are all
significant because they name their respective categories. Asteroids in the Aten class have orbits with
apehelia <0.983 au and semimajor axes less than that of the Earth. Apollo asteroids have semimajor axes
greater than Earth’s (<1 au) and perihelia <1.017 au. Finally, Amors have orbits between Earth and Mars
(1 au < r < 1.3 au). Even though these orbits do not necessarily cross that of the Earth, asteroids in this class
could easily be perturbed into a collision trajectory. The orbits of Aten, Apollo, and Amor class asteroids
are illustrated in figure 2.
4
Their orbits also categorize comets. Comets are generally listed as short or long period. Short-
period comets are further broken down into Halley and Jupiter classes. Jupiter class comets have perihelia
within Jupiter’s orbit. Although they can have longer periods, clearly shown by the 76-yr period of the
example from which they take their name, Halley’s comets remain within the solar system. The orbits of
Halley’s Comet and a representative Jupiter class comet are shown in figure 3. Short-period comets are
believed to originate from the Kuiper belt and generally have inclinations <30∞. Additionally, their orbital
periods are <200 yr.
Figure 2. Orbits of 2062 Aten, 1862 Apollo, and 1221 Amor relative to Earth and Mars.
5
Long-period comets are believed to come from the Oort cloud; they have periods well above
200 yr and appear to have no preferred orbital inclination. The orbit of a representative long-period comet,
Hale-Bopp, is found in figure 4.
Figure 3. Orbits of representative short-period comets relative to the outer planets.
Halley’s Comet represents the Halley class and Comet Crommelin represents
the Jupiter class.
–50
–50
–40
–30
–20
–10
0
10
20
30
40
50
–40 –30 –20 –10 0 10 20 30 40 50 Sun
Earth
Mars
Jupiter
Saturn
Uranus
Neptune
Pluto
Crommelin
Halley
6
Finally, minor planets can be organized according to their composition. A list of composition types
with approximate densities is given in table 1. These densities are highly conjectural, having been deduced
from samples found after atmospheric entry. Note that chondrites comprise 88 percent of the asteroid
population. Previous studies indicate that ª80 percent of impact-related deaths are likely to be caused by
chondrites and short-period comets.
Density
(g/cm
3
)
Density
(g/cm
3
) Asteroids
Chondrite
Achondrite
Iron
Stony irons
Mesosiderite
Pallasite
Comets
Short period
Long period
3.6
3.2
7.9
5.0
4.3
1.4
1.1
Figure 4. Orbits of representative short- and long-period comets relative to the solar
system. Comet Hale-Bopp represents long-period comets.
Table 1. Description of asteroid and comet compositions and representative predicted densities.
–350
–350
–250
–150
–50
50
150
250
350
–250 –150 –50 50 150 250 350
Sun
Earth
Mars
Jupiter
Saturn
Uranus
Neptune
Pluto
Crommelin
Halley
Hale-Bopp
7
2.2 Earth’s Impact Record
Table 2 lists the impact frequency for progressively larger near-Earth objects (NEOs). Crater diam-
eters and terrestrial events likely to inflict comparable damage are also listed. Note that even the smallest
diameter objects are capable of causing very major damage. A 23-m-diameter object can cause destruction
equivalent to the nuclear weapon used at Hiroshima at the end of World War II. At the other end of the size
spectrum is the Chicxulub impact, which is widely believed to have initiated an ice age
at the boundary between the Cretaceous and Tertiary periods and the consequent extinction of over
50 percent of the then existing species of flora and fauna, including the dinosaurs.
NEO
Diameter
Yield
(TNT
Equivalent)
Impact
Frequency
(Per Myr)
Crater
Diameter Comparable Terrestrial Event
2 m
4 m
6 m
23 m
55 m
250 m
500 m
1 km
1.5 km
10 km
500 ton
4,500 ton
20,000 ton
1 Mton
11 Mton
1,4000 Mton
10,000 Mton
87,000 Mton
310,000 Mton
8.7E7 Mton
250,000
69,000
28,000
2,700
540
35
10
2.9
1.4
0.007
35 m
75 m
120 m
450 m
1.1 km
5 km
10 km
20 km
31 km
200 km
Minimum damaging earthquake (M=5)
Largest chemical explosion (Heligoland Fortifications, 1947)
Atomic bomb explosion (Hiroshima, Japan, 1945)
Typical hydrogen bomb explosion (1 Mton)
Barringer Meteor Crater, Arizona; Tunguska explosion, Siberia, Russia
Gardnos, Norway; Goat Paddock, Australia
Lake Mein, Sweden; Bosumtwi, Ghana; Oasis, Libya
Haughton Dome, Canada; Rochechouart, France; Ries Crater, Germany
Total annual energy released from Earth (seismic, volcanic, etc.)
Sudbury, Canada; Vredefort, South Africa; Chicxulub, Mexico
Figure 5 illustrates the location of 145 known impact craters distributed around the world. The
actual number of Earth impacts is thought to be much higher, but most of the evidence has been
destroyed or covered by geological processes and vegetation. Note that the majority of the craters are
<50 Ma (Mega annum, or million years) old. Additionally, most crater diameters are in the 50- to
100-km range. This evidence supports the theory that wind and water erosion, seismic events, vegetation,
and the like are constantly erasing crater sites. Somewhat perversely, craters >100 km are not always
simple to find because their effects are so widespread as to not be easily recognized as impact craters. For
instance, the Chicxulub crater was eventually only identified from radar density mapping performed by a
petroleum company owned by the Mexican government. Sometimes there is circumstantial evidence that
indicates that a major impact crater is present. In the case of Chicxulub, a large number of sinkholes were
found around the periphery of the impact crater. As an interesting side note, these sinkholes contained
potable water, without which it may have been impossible for the Spaniards to explore that portion of the
continent in the 1700’s.
Table 2. Comparable terrestrial events for NEOs of various diameters.
2
8
Figure 5 also illustrates which parts of the Earth have been subjected to the most thorough search
for evidence of extraterrestrial impacts. Many more impacts have been found in Europe, North America,
and Australia than in other regions. There is no obvious reason why these continents would have
received a higher impact flux than the others. It seems likely that impact structures exist in equivalent
numbers, but as yet, undiscovered, on the other continents. Finally, note that few impact structures have
been found underwater. It is expected that cratering is mitigated by the cushioning effect of the oceans and
that the erosion rate is higher for submerged craters. In addition, it is clearly more difficult to find craters
in deep water.
50
40
30
20
10
0
50 150 250 350 450 550 >600
Age (Ma)
N
u
m
b
e
r

o
f

C
r
a
t
e
r
s
Crater Diameter
0.02–5 km
5–20 km
20–50 km
50–100 km
100–300 km
Figure 5. Location of known impact craters noting diameters as of 1998.
Age distribution of these craters is also included.
2
9
The last significant impact on Earth was the Tunguska event of 1908. This impact is believed to
have been caused by a 30- to 60-m object that detonated at a height of ª8 km above the Earth’s surface. The
estimated blast point is illustrated in figure 6. Investigators who explored the area during a series of expe-
ditions between 1958 and 1965 carefully recorded the direction in which trees had fallen as a result of the
blast. These directions are mapped in figure 6; they clearly indicate the location of the center of the event.
Most strikes by large NEOs do not reach the Earth’s surface. Instead, the combination of heat and
stress, which the object experiences as it travels at very high speeds through the atmosphere, usually causes
it to disintegrate explosively. Unfortunately, modeling and empirical evidence suggest that the heights at
which such explosive blasts are most likely to occur are similar to those determined—by nuclear weapons
experts—to cause maximum surface damage.
Figure 6. Projected area affected from the Tunguska blast of 1908. Arrows depict
the location and direction trees were knocked down from the blast.
3
Scale (km)
0 5 10
2
N
S
10
Figure 6 indicates the total ground area affected by even this relatively small object. Living crea-
tures inside this area are not thought to have survived the event. Fortunately, Tunguska is an unpopulated
area in Russian Siberia. A similar strike in a populated area would have caused widespread devastation.
Consider figure 7—the Tunguska event superimposed over Madison County, Alabama, in the United States
of America (USA), the authors’ residence and location of MSFC. An impact of this magnitude would
devastate the county, killing the majority of its 250,000 inhabitants. Superimposed over a more densely
populated area, such as a large city like New York City or London, the devastation would cause the deaths
of several million people.
Figure 7. Tunguska impact area superimposed over Madison County, Alabama, USA.
Several hundred thousand casualties can be expected from such an impact.
Scale
10 km
5 miles
11
Figure 8 depicts an aerial view of the Barringer meteorite crater in Arizona, USA. It is ª1 mi in
diameter and ª570 ft deep and is believed to have been caused by an object ª150 m in diameter that
impacted between 25,000 and 50,000 yr ago. This object did survive the transit through the atmosphere and
physically impacted the Earth’s surface. This scenario is characteristic of large nickel (Ni)-iron (Fe) aster-
oids that can survive the thermal and stress experienced during atmospheric entry.
Figure 8. Aerial view of the Barringer Impact Crater in Arizona, USA.
4
The projected devastation from this impact is shown in figure 9. As can be seen, human injuries and
fatalities are expected up to 24 km from the impact point. Additionally, the impact would cause hurricane-
force winds with resulting damage up to 40 km away. Although this damage is less than that expected from
the airburst of a non-Ni-Fe object of similar diameter, the destruction is still far from trivial.
12
Since impacts are concealed or erased due to natural processes, consideration must be given to
other types of evidence in order to determine the Earth’s impact history. Figure 10 shows the estimated
percentage of all species that were driven to extinction as a function of time. Many of the spikes in this
graph coincide closely in time with major known impact craters. Other peaks in this graph coincide with
possible stratigraphic evidence of impact—material in the geologic strata that could be due to ejecta
distributed around the world following an impact.
Figure 9. Calculated impact areas from the Barringer meteor.
4
Meteor Crater
Fireball Scorches Plants
and Animals Out to 10 km
West Sunset Mt.
Winslow
Meteor City
Large Animals Killed
or Wounded by Pressure Pulse
and Air Blast up to 24 km
From Impact
Hurricane-Force
Winds up to 40 km
East Sunset Mt.
13
Possible additional evidence of impacts is provided by the record of occasional geomagnetic field
reversals contained in frozen lava flows in the Earth’s crust. The Earth’s magnetic field periodically
reverses naturally; however, there is evidence of several reversals that cannot be explained by the normal
sequence of periodic changes. The Earth’s field could be reversed by a sharp impact of sufficient magni-
tude; the mechanism is similar to that which acts when a ferromagnet is struck with a hammer to realign its
magnetic field.
2.3 History Related to Near-Earth Objects
The concept that the Earth is threatened by impacts from space is not universally accepted. It is
entertaining to look at some remarks made through the ages concerning impacts from extraterrestrial
objects. The first quote is frequently attributed to President Thomas Jefferson:
50
40
30
20
10
0
0 100 200 300 400 500
Million Years
E
x
t
i
n
c
t
i
o
n

(
g
e
n
e
r
a
)

(
%
)
59.5
Large (Diameter >70 km) Dated Crater
Stratigraphic Evidence of Impact
Possible Stratigraphic Evidence of Impact
*
Figure 10. Extinction events versus time according to the fossil record.
5
14
“I could more easily believe two Yankee Professors would lie than that stones would fall
from heaven.”
—President Thomas Jefferson, 1807
To be fair, it is unclear whether President Jefferson actually said these words. Additional comments
by him in the same time period suggest that he was more open to the audacious theory that meteoroids were
of extraterrestrial origin.
Carl Sagan considered the idea of moving asteroids and comets—both for resource utilization and
planetary defense—as a potential unifying endeavor for humanity:
“Since hazards from asteroids and comets must apply to inhabited planets all over the
Galaxy, if there are such, intelligent beings everywhere will have to unify their home worlds
politically, leave their planets, and move small nearby worlds around. Their eventual choice,
as ours, is spaceflight or extinction.”
—Carl Sagan, Pale Blue Dot, 1994
Here Sagan suggests that planetary defense from NEOs is a strong justification for the continued
exploration of space. Additionally, he believed that the development of systems needed to defeat this threat
should not be shouldered by one nation, but would require participation of all of humanity. This premise
will be addressed in section 9.
Edward Teller, renowned physicist and developer of the hydrogen bomb, also commented on the
threat posed by asteroids and comets:
“Here is the situation that, to my mind, is a scandal, and I think people can understand that
it is a scandal: There is a probability of a few percent in the next century of the arrival of a
stony asteroid … approximately 100 m in diameter. It is a practical certainty that … it will
come completely unannounced. We won’t have any indication of it. Yet such an object is apt
… to do a lot of damage. … Just in dollars it could be billions, and in lives it might reach
millions.”
—Edward Teller, LLNL, 1995
Teller illustrates the threat represented by NEOs. Some astronomers have strongly suggested that
the threat necessitates a larger investment than is currently being made in the business of searching for and
categorizing NEOs. Teller indicates that under current circumstances there is likely to be little or no warn-
ing before a catastrophe occurs.
An effective search program clearly requires the use of telescopes in both the Northern and South-
ern Hemispheres. Efforts in the Southern Hemisphere were adversely affected by the Australian government’s
decision to withdraw funding. At the time, the Australian Minister for Science made the following
comment:
15
“I’m not going to be spooked or panicked into spending scarce research dollars on a fruit-
less attempt to predict the next asteroid. I’m just not convinced that the hype and alarm and
even fear-mongering is enough to justify an instant investment.”
—Peter McGauran, Australian Minister for Science,
60 Minutes Interview, 2002
This action threatened to end all survey efforts in the only participating Southern Hemisphere
country. Today, the Australian survey efforts continue but are funded by sources in the United States and
Europe. One prominent member of the Australian survey team responded to the Minister’s actions in a
direct manner.
“The dinosaurs did not have a space program. That’s why they died.”
—Duncan Steel, 2002 former member, Anglo-Australian
Near-Earth Asteroid Survey (AANEAS),
now Professor, University of Salford, England
The continued efforts to survey the population of NEOs are summarized in appendix A. In consid-
ering popular skepticism about the threat of Earth impacts, it is instructive to consider the development of
human knowledge about comets and asteroids. Table 3 contains a list of notable and relevant scientific
discoveries. As can be seen, our understanding of asteroids only dates back about two centuries. Although
knowledge of the existence of comets predates written records, most of human history comets were
revered (or feared) as omens and not regarded as objects of scientific curiosity. The Alvarez theory, identi-
fying the Chicxulub impact as being responsible for the Cretaceous-Tertiary (K-T) extinction, has only
been widely publicized since 1980.
Table 3. Timeline of scientific discoveries relevant to humanity’s knowledge
of asteroids and comets.
300 B.C.
1543
1608
1609
1609
1687
1705
18th C
1794
1801
1932
19th/20th C
1980
1994
Aristarchus theorizes circular orbits around Sun
Copernicus proposes Sun-centered system
Hans Lippershey invents the telescope
Galileo makes first astronomical telescope observations
Kepler develops first two laws of planetary motion
Newton publishes The Principia
Halley reports findings on cometary trajectories
Existence of asteroid belt theorized by Bode, Kant, et al.
Chladni suggests extraterrestrial origin of meteorites
Giuseppe Piazzi discovers Ceres
Reinmuth discovers 1862 Apollo
Discovery of Atens, Apollos, etc.
Alvarez theorized asteroid impact for K-T extinction
Recorded impact of Comet Shoemaker–Levy 9 into Jupiter
16
While an impact between a significantly sized NEO and Earth has not been directly recorded, a
cometary impact with Jupiter was recently observed. On July 16, 1994, Comet Shoemaker–Levy 9 was
observed impacting Jupiter by the approaching Galileo spacecraft. Results of the impact were recorded
by telescopes all over the world and in Earth orbit. David Levy, co-discoverer of the comet, documented
this historic event in reference 6.
2.4 Measuring the Near-Earth Object Population
To determine the overall threat posed by the solar system’s asteroids and comets, one must develop
a proper understanding of the populations of these two types of bodies. Unfortunately, neither population is
well understood. It is believed that orbital parameters are currently known for only
ª10 percent of the total NEO population. Also, since larger objects are easier to detect, our knowledge
of the known NEO population is biased toward these larger objects.
Using the known NEO population, the relative location of these objects can be plotted together
with the orbits of the inner planets. This plot is shown in figure 11 and gives the location of all known
objects on March 2, 2002. The green circles are minor objects that are not considered candidates for Earth
impact. The red circles are minor objects that have perihelia <1.3 au. Blue squares represent periodic
comets. The planets are shown as crosshair circles on their orbits. Figure 11 illustrates the large population
of NEOs around the Earth at any time, however, does create a misleading impression. Due to the finite
pixel size, the inner solar system—particularly the asteroid belt—appears to be full of NEOs; in fact, of
course, it is overwhelmingly empty.
Figure 11. Location of known minor planets on March 2, 2002, plotted
relative to the inner planets. NEOs are red, other asteroids
are green, and comets are blue.
17
(1) The recent acceptance of the Alvarez hypothesis, and the theory that other Earth impacts have
also affected the planet in the past, have led to a significant increase in the time and resources available to
locate these objects.
(2) The development of charged-coupled devices (CCDs) has computerized the previous manual
process of searching for new celestial objects. Before the advent of CCD technology, such searches were
conducted by the human study of photographic plates.
(3) Use of the Internet has facilitated international coordination and data sharing from sky surveys.
Our knowledge of the NEO population has increased significantly in the past few years. The num-
ber of known near-Earth asteroids (NEAs) is plotted shown as a function time in figure 12. The rapid
increase in the number of objects identified over recent years can be explained in a number of ways:
Figure 12. Number of known NEAs versus time. Note the rapid increase
in discoveries in recent years due to the use of CCDs and
increased interest in the asteroid and comet threat.
7
1995 LLNL Planetary Defense Workshop
Alan B. Chamberlin (JPL)
14 January 2002
1,800
1,600
1,400
1,200
1,000
800
600
400
200
0
1980 1983 1986 1989 1992 1995 1998 2001
1980/01–2001/12
Year
N
u
m
b
e
r
All NEAs
Large NEAs
18
Note that, over recent years, the total number of asteroids located has increased more rapidly than
the number of large asteroids. This is due to the fact that CCDs enable the detection of smaller
objects more than was previously thought possible.
Several movies and television programs have been released in recent years, giving the general
public an appreciation, however scientifically distorted, of the NEO threat. The increase in NEO detection
rates, coupled with the increase in public awareness, has raised the level of coverage given in the general
media to this threat. Recently, major news outlets have published several articles detailing the Earth’s close
encounters with NEOs; table 4 gives some details. To our knowledge, several asteroids have passed the
Earth at a distance of less than twice the Moon’s orbital radius from the Earth in the past year. In at least one
instance, the detection was made after the asteroid had already passed the point of closest approach.
Table 4. Recent near misses by comets and asteroids. By comparison,
the distance between the Earth and the Moon is ª240,000 mi.
2.5 Damage Mechanisms
The potential damage mechanisms resulting from an impact are several and varied; some are straight-
forward while others affect the Earth’s ecosphere in a more indirect manner. Some of the mechanisms
presented here are a little speculative, but this is to be expected. There is, fortunately, not a lot of applicable
empirical data upon which to draw.
The mechanisms can be broadly categorized as either directly or indirectly linked to the impact
itself. Within the directly linked category are the tsunami and the blast wave. Within the indirectly linked
category are the effects of releasing large quantities of dust, nitric oxide, and water vapor into the atmo-
sphere, as well as the possible geopolitical outcome of an impact.
300- to 400-m diameter
January 7, 2002, flew 375,000 mi from Earth
70-m diameter
March 8, 2002, flew 288,000 mi from Earth
800-m diameter
August 18, 2002, will fly 330,000 mi from Earth
1.2-km diameter, 28 km/s
Will fly by Earth on February 1, 2019
2001 YB5
2002 EM7
2002 NY40
2002 NT7
19
Before considering the damage mechanisms, it is important to discuss the two types of impact
events that can occur—air burst and surface impact. During passage through the atmosphere, an incoming
body will experience a considerable ram pressure force on its leading surface. This force, in addition to
decelerating the body, will place it under considerable internal stress, with some resulting strain. A suffi-
ciently friable body, such as one comprised largely of stony material, is likely to fragment. The potential
energy imparted by the stress forces being suddenly released causes the fragments to fly apart explosively.
This large-scale disintegration, while the body is still at some altitude, is referred to as an airburst. A body
with a more solid internal structure, particularly one with a largely metallic composition, will be able to
withstand the ram pressure-induced stresses and will strike the surface—either land or water—largely
intact. This is referred to as a surface impact.
Although the kinetic energy released locally is clearly greater for a surface impact, damage can
extend over a wider area from an airburst, particularly one at the optimum altitude. The Tunguska meteor-
ite produced an airburst at an altitude of ª8 km while traveling at ª20 km/s. Although the level
of destruction immediately below the airburst was less than would have been produced by a surface
impact; i.e., no crater was produced, the cumulative damage over an extended area was greater.
3
Now, regarding the resulting damage mechanisms, both airbursts and surface impacts can give rise
to all of those detailed in sections 2.5.1 through 2.5.6. There will undoubtedly be some difference in the
magnitude and extent of a particular mechanism if produced by an airburst rather than by a surface impact,
but at the present high level of detail, such differences are not important.
2.5.1 Tsunami
Both an oceanic impact and an airburst over the ocean will generate high-energy water waves.
These waves are two-dimensional disturbances whose height diminishes inversely with distance from their
point of origin. The key to development of a tsunami is the relationships between wave speed and water
depth, and among wave energy, height, and speed. Shallower water results in a lower speed, but slower
waves become steeper due to simple energy conservation. As a wave approaches the continental shelf, the
shallower water slows it and increases its height. By these mechanisms, a wave of moderate amplitude in
the deep ocean can increase in height by an order of magnitude as it comes ashore. With increased wave
height, the potential for causing damage ashore increases proportionately.
Tsunamis have caused great damage and loss of life in coastal areas, particularly low-lying regions.
In July of 1998, a 30-ft-high tsunami came ashore in Papua, New Guinea, killing more than 2,100 people.
The cause of the tsunami was determined to be an underwater landslide that took place more than 2,000 mi
distant.
Recent modeling studies indicate that the surface impact of a 400-m-diameter body—at any point
in the Atlantic Ocean—would devastate both the American and European/African coastlines with final
wave heights in excess of 60 m.
3
The height of deep-water waves 1,000 km distant from the impact point
of a soft stone meteor of varying size
8
is shown in figure 13. Figure 14 shows the height of a deep-water
wave 1,000 km distant from the impact point of an Fe meteor of varying size.
8
20
Figure 13. Deep-water wave height at 1,000 km distance versus initial
meteor radius for soft stone meteor.
Figure 14. Deep-water wave height at 1,000 km distance versus initial
meteor radius for Fe meteor.
10,000
1,000
100
10
1
0.1
100 1,000 10,000
v
o
= 11.2 km/s
v
o
= 15.0 km/s
v
o
= 20.0 km/s
v
o
= 22.0 km/s
v
o
= 25.0 km/s
v
o
= 30.0 km/s
Radius (m)
H
e
i
g
h
t

(
m
)
10,000
1,000
100
10
1
0.1
10 100 1,000 10,000
v
o
= 11.2 km/s
v
o
= 15.0 km/s
v
o
= 20.0 km/s
v
o
= 22.0 km/s
v
o
= 25.0 km/s
v
o
= 30.0 km/s
Radius (m)
H
e
i
g
h
t

(
m
)
21
2.5.2 Blast Wave
Blast waves are produced by both airbursts and surface impacts. The destructive potential of the
wave is determined by the total explosive energy of either the airburst or the surface impact as well as the
altitude at which the explosion occurs.
The blast wave consists of a shock wave followed by a substantial wind. The shock is characterized
by the peak overpressure. Even a 2-psi (13.972-kPa) overpressure will create a severe hazard due to flying
debris. A 4-psi (27.944-kPa) overpressure corresponds to hurricane-force winds of 70 mph.
Figure 15 (taken from ref. 9) consists of a graph showing the total damage area, defined as the area
that experiences an overpressure of 4 psi or higher, as a function of the impact energy. Within the graph, the
following abbreviations are used:
SP = short-period comet
LP = long-period comet
Carb = carbonaceous chondrite.
The surface explosion curve is derived on the basis that only 3 percent of the impact energy goes into shock
waves; i.e., e = 0.03.
Figure 15. Blast wave damage versus impact energy.
10
–1
10
0
10
1
10
2
10
2
10
3
10
4
10
5
10
6
10
7
10
8
10
3
10
4
10
5
10
6
10
7
10
8
Impact Energy (Mton)
D
a
m
a
g
e

A
r
e
a

(
k
m
2
)
Area of Earth
Area of U.S.
Airburst at
Optimum Height
K-T
Impact-Generated
Vapor Cloud
Surface Exploration
at = 0.03 ε
Carb
Fe
Tunguska
Stone
SP
LP
22
2.5.3 Atmospheric Dust Loading
Submicron dust reaching the stratosphere as a result of an impact will adversely affect the regional
and global climate as well as disrupting important biochemical mechanisms, such as photosynthesis.
The sequence of events is complex and four distinct steps are required to model the effects of dust
loading:
(1) The amount of dust lofted by the impact must be derived, based upon the impacting body’s
composition, its kinetic energy, and the nature of the impact.
(2) The rate at which dust disperses around the globe must be derived, taking into account the
settling process—some dust will leave the atmosphere.
(3) The effect of the atmospheric dust upon Earth’s radiation balance must be calculated.
(4) The effect of low light levels on surface temperature, precipitation, and photosynthesis must be
established.
It is anticipated that a large impact could cause a drop of several degrees in global temperature and
the loss of one or more year’s crop with resulting starvation, mass migration, social disorder, and possibly
warfare. Quantification is difficult without extremely complex numerical modeling. Most work to date has
concentrated upon the atmospheric dust loading, resulting from the use of nuclear weapons. Modeling
results indicate that a 5,000-Mton nuclear exchange could reduce continental interior temperatures to
–25 ˚C within as little as 2 wk. In practice, heat transfer from ocean water would probably mitigate some-
what, resulting in minimum temperatures of only about –5 ˚C.
5
Figure 16 shows the density of atmospheric dust and the resulting optical depth as functions of
impact energy.
9
23
2.5.4 Nitrogen Oxide Production
Four significant mechanisms for the production of nitrogen oxides (NOs) during an impact have
been identified:
(1) The formation of NO in the shock waves that accompany both atmospheric entry and explosive
disintegration.
(2) If a surface crater is created, NO forms if the resulting ejecta plume moves through the atmo-
sphere at more than 2 km/s.
(3) If the ejecta plume leaves the atmosphere, further NO will be formed within the subsequent
reentry shocks.
(4) In the case of an impact from an extremely large body, the hot ejecta dispersed through the
atmosphere could briefly bring the local temperature up to ª1,500 K, at which point additional NO would
be created directly from atmospheric gases.
The most direct consequence of elevated atmospheric NO levels are the production of acid rain. In
addition to elevated atmospheric NO levels, sulfate- or carbonaceous-rich impacting bodies may also
produce sulfur dioxide and carbon dioxide, respectively. Elevated NO levels in the atmosphere would also
Figure 16. Density and optical depth of atmospheric dust versus impact energy.
10
2
10
0
10
–2
10
–4
10
–6
10
–8
10
–10
10
–5
10
–3
10
–1
10
1
10
3
10
5
10
0
10
2
10
4
10
6
10
8
Total Mass
Submicron Dust
K-T Clay Layer
Pinatubo Optical Depth
Impact Energy (Mton)
O
p
t
i
c
a
l

D
e
p
t
h

o
f

S
u
b
m
i
c
r
o
n

D
u
s
t
M
a
s
s

(
g
m
/
c
m

2
)
24
threaten the integrity of the Earth’s ozone layer. An NO volume mixing ratio; i.e., volume of NO as a
fraction of total air volume, as low as 2¥10
–7
, if mixed uniformly throughout the atmosphere, would render
the current ozone ultraviolet screen ineffective. Figure 17, taken from Toon et al.,
9
contains a brief sum-
mary of the effects of NO on both the ozone layer and oceanic acidity levels.
Figure 17. Blast wave damage versus impact energy.
2.5.5 Water Vapor Injection
Under normal circumstances, the upper atmosphere has very low humidity levels. A large
oceanic impact would inject large amounts of water into the upper atmosphere. The resulting increase in
water levels above the tropopause would give rise to a “greenhouse” effect and would substantially
increase Earth’s temperature. There are two mechanisms by which water can reach the upper atmosphere:
direct splash to high altitude and water vaporization.
Analysis indicates ª150 million kg of water would be vaporized per megaton of impact energy.
4
The current water vapor level above the tropopause is approximately 2¥10
–4
to 6¥10
–4
gm/cm
2
; this means
that impacts as low as 104 Mton are capable of doubling that level.
The impact of additional water vapor at high altitude is uncertain, mainly because the mechanisms
by which water leaves the upper atmosphere are not properly understood. However, the resulting green-
house effect could significantly increase global temperatures. Figure 18 shows the mass of water injected
into the atmosphere as a function of impact energy.
9
10
16
10
15
10
14
10
13
10
12
10
10
10
9
10
8
10
7
10
0
10
1
10
2
10
3
10
4
10
5
10
6
10
7
10
8
10
9
10
–9
10
–8
10
–7
10
–6
10
–5
10
–4
10
–3
10
–2
10
–1
10
11
NO From Global Ejecta:
2-mm Particles
NO From Plume
at Impact Site
NO From Global Ejecta:
200- m Particles
NO Mostly in Stratosphere
or Mesosphere
NO Well-Mixed in Atmosphere
Removes Ozone UV Screen
N
O

(
m
o
l
e
s
)
Impact Energy (Mton)
F
r
a
c
t
i
o
n

o
f

O
c
e
a
n
i
c

M
i
x
e
d

L
a
y
e
r

A
c
i
d
i
f
i
e
d µ
25
2.5.6 Precipitate Nuclear Exchange
In recent years, several new nations have acquired the capability to produce nuclear weapons, and
some of them have proceeded to develop such weapons. Although efforts are underway to prevent further
proliferation of nuclear weapons technology, it is quite possible that additional nations will succeed in
acquiring it over the next few years. Although possessing the weapons themselves, most of these nations
do not yet have the relatively sophisticated detection systems of the better established nuclear powers. This
raises the concern that an unexpected impact might be mistaken for a nuclear strike. If this scenario were to
occur in a region of high tension, it could trigger a retaliatory nuclear strike.
Although nations such as the United States would probably be able to rapidly distinguish an impact
event from a nuclear weapon detonation, it is entirely possible that this information would attract little
credence in a region of high international tension. With its public demanding retaliatory action against the
supposed perpetrator, a national government might discover that events had acquired a grim and irresist-
ible momentum.
This recent development adds urgency to the need for more thorough identification of threatening
bodies, more accurate tracking of them, and the development of effective protection techniques.
2.6 The Credibility Problem
Considering the research described above, one might ask why the danger posed by asteroids and
comets is not given more public attention. The answer to this question is somewhat complex. First, it is
important to distinguish between impacts from small and large objects. Examination of table 5 illustrates
Figure 18. Mass of water lifted into the atmosphere versus impact energy.
Yield (Mton)
M
a
s
s

(
g
/
c
m

2
)
10
0
10
0
10
2
10
4
10
6
10
8
10
–2
10
–4
10
–6
10
–8
Upper Atmosphere Water Mass
Injected Mass of Water
26
Table 5. Consequences of impact by NEOs of various sizes.
10
the consequences of impacts by objects of various sizes. Large impacts, caused by objects with a diameter
greater than the 500-m to 1-km range, produce effects that are felt across the globe. These impacts have
received some attention in the popular media. Smaller objects are likely to cause devastation on a regional
scale. Although more likely to occur than a larger impact, the danger from these smaller bodies has been
ignored; there are several reasons for this.
First, the concept of severe destruction being caused by a collision with an NEO is very alien to
most of the general public. Nobody knows of anyone who has been killed by a falling meteorite and very
few people have witnessed a fall of any size. Human nature tends to naturally concentrate attention on
dangers that are perceived as present in our everyday lives; i.e., automobile accidents, fires, etc., and
excludes those that are more exotic. With all the threats facing humanity in the early 21st century, the
subconscious decision to avoid facing such a nonobvious threat could even be seen as a defense mecha-
nism against being psychologically overwhelmed.
Yield
(Mt)
Interval
LogT
NEO
Diameter
Crater
Diameter Consequences
<10
10–10
2
10
2
–10
3
10
3
–10
4
10
4
–10
5
10
5
–10
6
10
6
–10
7
10
7
–10
8
10
8
–10
9
>10
10
3.0
3.6
4.2
4.8
5.4
6.0
6.6
7.2

75 m
160 m
350 m
0.7 km
1.7 km
3 km
7 km
16 km

1.5 km
3 km
6 km
12 km
30 km
60 km
125 km
250 km

Upper atmosphere detonation of stones and comets;
only irons (<3%) reach surface.

Irons make craters; stones produce airbursts.
Land impacts destroy city-sized areas; e.g., Washington, DC.

Irons and stones produce groundbursts; comets produce airbursts.
Impact destroys urban areas; e.g., New York City.

Impacts on land produce craters; ocean tsunamis become significant.
Land impact destroys area the size of a small state; e.g., Delaware.
Tsunamis reach oceanic scales, exceed damage from land impacts.
Land impact destroys area the size of a moderate state; e.g., Virginia.
Land impact raises enough dust to affect climate, freeze crop. Ocean
impacts generate hemispheric-scale tsunamis. Global ozone destruction.
Land impact destroys area the size of a large state; e.g., California.
Both land and ocean impacts raise dust, change climate. Impact
ejecta are global, triggering widespread fires. Land impact destroys
area the size of a large nation; e.g., Mexico.
Prolonged climate effects, global configuration, probably mass
extinction. Direct destruction approaches continental scale; e.g.,
United States.
Large mass extinction; e.g., K-T-type event.
Threatens survival of all advanced forms of life.
– – –
27
As mentioned previously, our knowledge of comets and asteroids is relatively recent. The idea of
asteroids and comets colliding with the Earth and causing widespread devastation has had even less time to
take root. Also, although careful searches have recently yielded indirect evidence, direct evidence of
impacts from celestial objects is not readily found in the historical record. However, several historical
incidents could be interpreted as asteroid impacts. Sources as diverse as the Christian Bible and Maori
tribal records from New Zealand make reference to catastrophic events that could have been caused by
large impacts.
In many respects, collisions with asteroids and comets are a bigger threat now than in any other
time in history. This can be argued by considering human population growth over the past century, as
depicted in figure 19.
Figure 19. Predicted world population in the last century. Population
is extrapolated through the middle of this century.
11
At the time of the last collision with a substantial object (Tunguska), the Earth’s human population
numbered between 1 and 2 billion. Today, the Earth’s population numbers above 6 billion and is rising
rapidly. The increased population density means that there are far fewer remote places where an impact
event, like that at Tunguska, could occur without causing significant loss of life. As mentioned in the
10
9
8
7
6
5
4
3
2
1
0
1920 1900 1940 1960 1980 2000 2020 2040
Timeframe
N
u
m
b
e
r

o
f

P
e
o
p
l
e

(
b
i
l
l
i
o
n
s
)
28
section on damage mechanisms, even a small impact can cause a large tsunami that would affect much of
the Earth’s shorelines. In fact, consideration of figure 20 shows that a majority of the Earth’s population
actually lives near these shorelines.
An impacting asteroid would probably create large numbers of charged particles during its travel
through the atmosphere. These particles would probably have a severely detrimental effect on the global
electronic and communications infrastructure. Atmospheric dust loading would decrease light reaching
Earth’s surface, placing viable arable land at a very high premium. Our complex and interdependent tech-
nological society, usually well equipped to deal with an isolated crisis, would probably be very vulnerable
to such a varied and large number of coinciding problems. Even impacts from relatively small objects—
a few tens of meters in diameter—pose a much higher threat than any other time in history.
Figure 20. Human population density graph for all continents (except Antarctica).
Greenland and Iceland are not represented as well as some Pacific islands.
11
1 5 25 50 100 250 500 1,000
Persons Per Square Kilometer
29
2.7 The Torino Scale
During an international conference on NEOs held in Turin in 1999, an attempt was made to estab-
lish a type of “Richter Scale” for categorizing the Earth-impact hazard associated with newly discovered
asteroids and comets. In honor of the meeting venue, this system was named the Torino Scale. The Torino
Scale utilizes numbers that range from zero to 10:
• Zero indicates that an object has a zero or negligibly small chance of collision with the Earth. Zero
also categorizes any object too small to penetrate the Earth’s atmosphere intact.
• Ten indicates that a collision is certain, and the impacting object is so large that it is capable of
precipitating a global climatic disaster.
Categorization on the Torino Scale is based on the placement of a close approach event within a
graphical representation of kinetic energy and collision probability (fig. 21).
Figure 21. Illustration of the various category of threat under the Torino Scale.
Orbital predictions for newly discovered bodies are naturally uncertain. Discovery observations
typically involve measurements over only a short orbital track and so, as a body’s orbit characterization
improves, its Torino number can change. Hopefully, it will always reduce as more information becomes
available.
0
1
2
6
7
5
4
10
9
8 3
Global
Regional
Local
>0.99 10
–2
10
–4
10
–6
10
–8
10
8
10
5
100
1
Events Having
No Likely
Consequences
Events Meriting
Careful
Monitoring
Events
Meriting
Concern
Threatening
Events
Certain
Collisions
No Consequences
Collision Probability
K
i
n
e
t
i
c

E
n
e
r
g
y

(
M
t
o
n
)
(5 km)
(1 km)
(100 m)
(20 m)
30
An object that is capable of making multiple close approaches to the Earth will have a separate
Torino Scale value associated with each approach. An object may be summarized by the single highest
value that it attains on the Torino Scale; no fractional or decimal values are used. The various categories
within the scale are shown in figure 22. It should be noted that the Torino Scale has yet to achieve any
wide-scale usage. It is mentioned here for completeness only.
Figure 22. Categories within the Torino Scale.
The likelihood of a collision is zero, or well below
the chance that a random object of the same size
will strike the Earth within the next few decades.
This designation also applies to any small object
that, in the event of a collision, is unlikely to reach
the Earth's surface intact.
The chance of collision is extremely unlikely, about
the same as a random object of the same size
striking the Earth within the next few decades.
A somewhat close, but not unusual, encounter.
Collision is very unlikely.
A close encounter, with 1% or greater chance of a
collision capable of causing localized devastation.
A close encounter, with 1% or greater chance of a
collision capable of causing regional devastation.
A close encounter, with a significant threat of a
collision capable of causing regional devastation.
A close encounter, with significant threat of a
collision capable of causing a global catastrophe.
A close encounter, with an extremely significant threat
of a collision capable of causing a global catastrophe.
A collision capable of causing localized destruction.
Such events occur somewhere on Earth between
once per 50 yr and once per 1,000 yr.
A collision capable of causing regional devastation.
Such events occur between once per 1,000 yr
and once per 100,000 yr.
A collision capable of causing a global climatic
catastrophe. Such events occur once per
100,000 yr, or less often.
TORINO SCALE
Assessing Asteroid and Comet Impact
Hazard Predictions in the 21st Century
Events Having
No Likely
Consequences
Events Meriting
Careful Monitoring
Events Meriting
Concern
Threatening
Events
Certain
Collisions
0
1
2
3
4
5
6
7
8
9
10
31
3. MISSION CONFIGURATIONS
In considering how to counter an incoming object, it is important to consider first whether it is
better to push the object out of the way or to break it up into small pieces. Each option offers its own set of
advantages and disadvantages. Additionally, one must consider how to deliver the energy needed to deflect
or fragment the object. Three methods are discussed here: remote station, interception, and rendezvous.
Under the remote station approach, no complex spacecraft would be sent out to the
approaching NEO. Instead, all operations are conducted remotely, probably from the vicinity of the Earth,
with beamed energy or projectiles being used to perform the deflection or fragmentation. A strategy based
upon interception would involve sending spacecraft out on an intercept trajectory with the incoming NEO;
the resulting high velocity impact(s) would accomplish either deflection or fragmentation. Rendezvous-
based techniques are more propulsively demanding, as they require one to dispatch hardware to actually
match orbits with the incoming NEO.
At first consideration, it would seem that the decision over deflection versus fragmentation is
interlinked with the method chosen. An intercepting object would deliver all of its energy at once, tending
to cause fragmentation instead of deflection. However, an incoming NEO could be deflected by a series of
intercepting objects, each imparting enough momentum to slightly perturb its orbit without causing frag-
mentation. Similarly, it may seem improbable to actually rendezvous with an incoming object only to
subsequently break it up. However, if there were a finite amount of time needed for the system to deliver
the fragmentation energy, then rendezvous becomes necessary.
3.1 Deflection Versus Fragmentation
Figure 23 illustrates the concept of deflecting an incoming object away from an orbit that intersects
with the Earth. In the case illustrated, it is anticipated that the deflection mechanism would require a
significant period of the NEO orbit to deliver the energy necessary to perturb its orbit. Figure 23 shows the
commencement of deflection before aphelion; if undeflected, the object would collide with the Earth near
perihelion.
32
When considering strategies based on deflection, it is important to establish what level of perturba-
tion is necessary to consider the Earth as being safe from collision. One might argue that “a miss is good as
a mile,” but some margin of error is necessary when designing a system that would deflect incoming
objects. The literature commonly uses a figure of 3 Earth radii as a minimum safe approach distance for a
deflected object. This value takes into account the uncertainty in astrodynamical constants that affects
trajectory modeling accuracy for the incoming object.
Figure 24 illustrates the concept of fragmentation. At first sight, this might seem the best
approach as the object’s destruction means that it cannot threaten the Earth on a later orbit. In addition,
there is no requirement to deliver the energy to the NEO in a distributed manner; thus, it can be defeated in
one shot. Finally, as recent Hollywood blockbuster movies clearly demonstrate, there is a unique emo-
tional satisfaction to be derived from destroying a life-threatening object in this emphatic manner.
Figure 23. Illustration of deflection method of threat mitigation.
Figure 24. Illustration of fragmentation method of threat mitigation.
Perturbed
Orbit
Collision Point
Earth and Target
at Detection
Sun
Collision Point
Expanding Debris
Earth and Target
at Detection
Sun
Fragmentation Point
33
Despite its immediate tabloid appeal, fragmentation does introduce several issues that, on reflec-
tion, make it less attractive than deflection. It is important to break up the object into relatively small
components. To break up the object into just a few pieces could actually exacerbate the damage to the
Earth, with several distributed impacts occurring instead of one large impact. The fragmented pieces can
“draft” off one another in the atmosphere; i.e., following pieces can travel within the slipstream of a lead-
ing piece and thus reach the ground relatively intact, alleviating burnup. For these reasons, the suggested
fragmentation criteria is that no fragment should have a diameter >10 m. A major problem arises because
asteroids and comets are suspected to have a very heterogeneous composition with significant internal
structural flaws. Energy deposited into such objects cannot be expected to cause uniform fragmentation.
3.2 Remote Station Versus Interception Versus Rendezvous
The three modes considered in this study to deliver deflection or fragmentation energy to the
incoming object are remote station, interception, and rendezvous. The remote station mode is depicted in
figure 25. A station remains in orbit around either the Earth or the Sun. Energy can be delivered in the form
of projectiles fired from the station by a mass driver or by a focused beam of coherent light, such as a solar
lens or a laser. The advantage of such a system is that it remains close to the Earth and is easily maintained
and upgraded. Also, the system can start deflecting the incoming object almost immediately—which might
be months or years—during which an interceptor or rendezvous system would take to reach the object.
However, there are also several disadvantages. Targeting of the beam or stream of projectiles is not
a trivial issue. For instance, targeting is required to within 1.4¥10
–5
to 2.8¥10
–6
arc s for objects between 1 and
5 au, the approximate orbital radii of Earth and Jupiter. Focusing the beam on the object across such vast
distances would also be very challenging. There would be no vehicle in the vicinity of the object that could
accurately assess the effect of the beam. Any such assessment would have to be conducted remotely from
terrestrial- or Earth-orbiting platforms. This need for remote sensing over large distances would make it
more difficult to properly assess the effect on the NEO. Moreover, unless the station is placed in a polar
orbit, the object will almost certainly be eclipsed once per revolution. Polar orbits would require additional
launches to deliver the station into orbit and would result in higher radiation exposure. Finally, a remote
station would only be able to deflect incoming objects radially away from the station. Over a finite time
period, the station and object will move relative to one another, causing the deflection vector to rotate,
thereby wasting some of the beamed energy. Also, one must remember that radial deflection may not be the
most efficient deflection strategy. See section 6 for a discussion of optimal deflection directions.
34
The interception strategy is depicted in figure 26. After NEO detection, the interceptor is
deployed to intersect it later in its orbit. At this point, deflection or fragmentation can commence. In most
cases, the interceptor will have substantial kinetic energy relative to the NEO. Thus, the interception option
allows use of some of the energy initially stored in the outbound propulsion system to be delivered to the
NEO. Interception options tend to be relatively simple, capitalizing on the high kinetic energy that is
naturally available. The propulsive requirements for interception are substantially less than for rendez-
vous. This difference is further discussed in section 6.
Figure 25. Delivering deflection or fragmentation energy by the remote station mode.
Figure 26. Delivering deflection or fragmentation energy by the interception mode.
Collision Point
Projected
Beam/Projectile—
Commencement
of Deflection
or Fragmentation
Earth and Target
at Detection
Sun
Collision Point
Earth and Target
at Interception—
Commencement
of Deflection
or Fragmentation
Earth and Target
at Detection
Sun
I
n
t
e
r
c
e
p
t
i
o
n
T
r
a
j
e
c
t
o
r
y
35
Despite these advantages, the interception strategy also has its problems. The rate of closure
between the interceptor and NEO can be as high as several tens of miles per second. This is an order of
magnitude higher than the closure rate required for the kinetic kill vehicles used in the U.S. Global Missile
Defense (GMD) program. The GMD program has had a mixed success rate in interception tests against
simulated intercontinental ballistic missile (ICBM) warheads. While it is true that an incoming asteroid or
comet will not maneuver to avoid destruction, and at 10 to 1,000 m in size, offers a larger target than a
3- to 5-m ICBM, the very high closure speeds still pose significant problems for guidance and terminal
maneuvering.
Clearly, only a system with multiple interceptors dispatched in sequence, with the later vehicles
capable of adjusting their trajectories, can provide the opportunity to continually sense changes in the NEO
orbit. It should be noted that the interception strategy has the same type of deflection vector limitations as
for the remote station strategy.
The final strategy is that of rendezvous with the incoming NEO, as shown in figure 27. After
detection, the rendezvous system is deployed and matches orbits with the NEO later along its trajectory.
This strategy is the one required for most types of deflection systems; it has several significant advantages.
Targeting the NEO is much less difficult for a vehicle in a parallel orbit. This strategy offers the best
opportunity to continuously evaluate the NEO during deflection or fragmentation operations. The limita-
tions placed on the direction of the deflection vector that were encountered with the other two strategies are
absent for the rendezvous option, allowing deflection in the direction that requires the least energy. Finally,
this strategy has the greatest synergy with resource utilization missions.
Figure 27. Delivering deflection or fragmentation energy by the rendezvous mode.
Perturbed
Orbit
Earth and Target
at Detection
Sun
Collision Point Earth and Target
at Rendezvous—
Commencement
of Deflection
I
n
t
e
r
c
e
p
t
i
o
n
T
r
a
j
e
c
t
o
r
y
36
Of course, the propulsive requirements to rendezvous with an incoming NEO are much higher than
that for interception. Additionally, the response time for a rendezvous system must include both the out-
bound rendezvous time and whatever inbound time is needed for the fragmentation or deflection process to
take place. The rendezvous vehicle may be sufficiently distant from the Earth to make teleoperations
difficult and would thus require significant onboard autonomy in an unknown environment that offers
many opportunities for unexpected effects. Finally, during fragmentation or deflection, the rendezvous
vehicle will probably be exposed to a hazardous environment filled with ejecta from the asteroid. The
vehicle will have to be designed to resist this environment.
37
4. OUTBOUND PROPULSION
For both the interception and rendezvous techniques, neither fragmentation nor deflection can take
place until the necessary system hardware is transported out to the approaching NEO. Some type of out-
bound propulsion system is required to accomplish this. Several propulsion systems are considered in this
study. They were selected for their ability to meet the mission requirements, their level of technological
maturity, and development status.
Each propulsion system has been assessed against a range of qualitative considerations that have
been appropriately weighed by the MSFC study team. Although these considerations are of considerable
merit in comparing the various outbound propulsion systems, being qualitative, they do not easily lend
themselves to a numerical analysis. Instead, they are classified as high, medium, and low. When presented
diagrammatically, they are indicated by the use of three colors—green, yellow, and red—to indicate
decreasing levels of favor.
Each consideration is categorized as being either first or second order. First-order considerations
include thrust level, scalability, long-term readiness, and compactness. High thrust levels are favorable
because they reduce outbound trip times and gravity losses. Scalability is the parameter that denotes the
ability of a proposed propulsion system to fulfill a range of propulsive needs, and to thus handle a variety
of threat sizes. Since the vehicle will probably have to be constructed and kept until a threat is detected,
long-term storage at a high state of readiness is a major issue. For similar reasons, compactness is also
important. These considerations are all regarded as being of first order, as they directly affect the ability of
the proposed system to meet the mission requirements.
Second-order considerations include usefulness of the system as a weapon, perceived safety, syn-
ergy with other NASA missions, and cost. The propulsion options under consideration are—of necessity—
all high-energy systems that could cause extensive damage if misused. However, the category that assesses
a system’s usefulness as a weapon must consider whether use of the system as a weapon will outperform
existing weapon systems.
NASA always considers safety to be an issue of primary importance. However, some high-energy
systems are considered to be more threatening than others; i.e., nuclear systems. In these cases, extra effort
must be expended to overcome public opinion obstacles, usually generated by the lack of public under-
standing and consequent mistrust of these specific technologies. The benefits of synergy with other NASA
missions need no further explanation. For present purposes, the synergy consideration is divided into manned
and robotic exploration missions as well as missions in which the use of asteroid and comet resources is a
primary goal. Finally, development and deployment costs constitute the final consideration. These consid-
erations are all regarded as being of second order as they do not affect the ability to meet the immediate
mission objectives, but they do address political and economic issues involved in the deployment of a
system.
38
4.1 Staged Chemical
Chemical systems are the most highly developed propulsion technologies currently available. They
also offer the lowest performance of all the options considered for this project. In fact, chemical systems
would not be able to handle many of the propulsive requirements for these missions without staging. For
this reason only the high-performance liquid oxygen (lox)/liquid hydrogen (LH
2
) propellant combination
is considered. A two-stage lox/LH
2
vehicle is shown in figure 28. In the analysis that follows, a value of
465 is used for the specific impulse (I
sp
) for each stage.
Figure 28. Two-stage lox/LH
2
vehicle. Image produced by INTROS.
12
The analysis method used in this project is covered in detail elsewhere
13
and is only summarized
here. Staging calculations requires knowledge of both the inert and propellant masses for each stage as
well as the I
sp
of each stage. The inert mass fraction is instrumental in calculating these masses; it is
defined as
e
n
s
p s
m
m m
n
n n
=
+
.
(1)
The inert mass fraction for the nth stage is the structural mass of the nth stage, m
s
n
divided by the
structural and propellant mass for the nth stage, m
p
n
.
Predicting the inert mass fraction requires knowledge of historical vehicle designs. Curve fits using
these historical data can then be developed (see fig. 29). Here, the inert mass fractions of historical launch
vehicle stages listed in table 6 have been plotted against the amount of propellant contained in each stage.
The line plotted represents a regression fit of the data. This regression line is given by
e
n p
m =
( )
- 0 422 0 0579
10
. log . ,
(2)
where propellant mass is in lbm. The Shuttle’s external tank was not used to calculate this regression as its
mass does not include a propulsion system. If included, this discrepancy would unfairly bias the results.
First Stage Second Stage Inbound System
39
Figure 29. Regression curve fit of lox/LH
2
launch vehicle stages.
0.25
0.2
0.15
0.1
0.05
0
10,000 100,000 1,000,000
Propellant Mass (Ibm)
I
n
e
r
t

M
a
s
s

F
r
a
c
t
i
o
n
Lox/LH
2
Lox/LH
2
Regress
+2Sy
–2Sy
40
Table 6. Data on in-service and historical launch vehicles that have lox-/LH
2
-powered stages.
Vehicle
Length
(in) Propellants Stage
1
M
i
1
(lb)
M
f
2
(lb)
Thrust
3
(lbf)
I
sp
3
(s)
Dia
4
(in)
Span
4
(in)
Titan
(USA)
0
1
2
3
498,758
256,999
64,348
35,843
74,512
12,000
5,849
5,800
1,314,999
526,000
101,999
29,500
238
250
316
444
122
122
122
122
122
122
122
122
1,019.7
878
311
378
Solid
N
2
O
4
/Aerozin
e-50
6
N
2
O
4
/Aerozin
e-50
6
Lox/LH
2
1
Vehicle families use multiples of particular stages to achieve a variety of performance requirements. Zero and first stages
especially are frequently used in numbers to boost performance. The data listed represents one stage only; i.e., zero stage
on the STS represents one RSRM, not two. Zero stage indicates strap-on boosters.
2
Neither initial or final mass includes payload.
3
Thrust and I
sp
represent sea level values for all stages except the last. Vacuum thrust and I
sp
are listed for the final stage.
4
Dia is the diameter of the vehicle fuselage. Span is the total span of the vehicle including fins and wings, if any.
5
There are no engines on the Space Shuttle external tank.
6
Aerozine-50 is a 1:1 mixture of UDMH and N
2
H
4
.
Ariane 5
(ESA)
0
1
2
593,043
374,785
20,679
74,957
33,069
4,806
1,455,049
241,670
6,173
259
310
324
118.1
212.6
157.5
118.1
212.6
157.5
1,220.5
1,181.1
129.9
Solid
Lox/LH
2
N
2
O
4
/MMH
Atlas IIIb
(USA)
1
2
432,300
50,618
30,200
4,696
932,669
44,584
311
451
122
122
118.1
122
1,141.7
460.6
Lox/RP–1
Lox/LH
2
Atlas V
Centaur III
1
2
673,300
50,620
46,060
4,700
860,200
22,300
311.3
450.5
150
120
150
120
1,278
462
Lox/RP–1
Lox/LH
2
Delta III
(USA)
0
1
2
42,609
230,112
42,060
5,031
15,040
5,459
141,250
244,096
24,736
273
254
462
47.2
94.5
94.5
47.2
157.5
157.5
578.7
787.4
346.5
Solid
Lox/RP–1
Lox/LH
2
Delta IV 1
2
499,000
67,700
9,000
7,700
650,000
24,750
365
462.4
200.4
157.2
200.4
157.2
1,606.8
474
Lox/LH
2
Lox/LH
2
H-1
(Japan)
0
1
2
3
9,753
189,156
23,369
4,850
1,532
9,700
3,968
794
58,206
194,844
23,149
17,416
232
253
450
291
31.5
94.5
98.4
51.2
31.5
94.5
98.4
51.2
236.2
866.1
405.5
90.6
Solid
Lox/RP–1
Lox/LH
2
Solid
H-IIA
(Japan)
1
2
250,500
43,200
30,000
6,600
191,000
30,800
338
447
157.2
157.2
157.2
157.2
1,464
362.4
Lox/LH
2
Lox/LH
2
Long March
3D
(China)
0
1
2
3
90,389
394,627
87,193
45,415
6,614
19,842
8,818
6,614
183,508
734,033
186,817
35,274
259
259
260
440
90.6
133.9
133.9
118.1
90.6
275.6
133.9
118.1
629.9
909.4
409.4
346.5
N
2
O
4
/UMDH
N
2
O
4
/UMDH
N
2
O
4
/UMDH
Lox/LH
2
2
1
3
1
Saturn II
Saturn IB
Saturn IVB
Saturn 1C
1,079,800
987,140
261,400
4,872,000
86,000
91,520
28,380
288,000
1,150,000
1,640,000
200,000
7,760,000
425
232
426
264
396
256.8
260.4
396
396
256.8
260.4
396
978
963.6
711.6
1,656
Lox/LH
2
Lox/RP–1
Lox/LH
2
Lox/RP–1
STS (USA)
RSRM
ET
Orbiter
0
1
2
1,299,998
1,655,615
218,958
190,001
65,984
218,515
2,589,796
0
1,536,411
337
363
455
145.7
342.5
192.9
200.8
342.5
937
1,515.7
1,846.5
1,464.6
Solid
Lox/LH
2
Lox/LH
2
41
The energy required to deliver the threat mitigation system is defined as DV. DV is the difference
between the velocity required for the final orbit and the velocity in the initial orbit. This requirement is
defined by the trajectory and is discussed further in section 6. Each stage produces a portion of the total DV
requirement. There is an optimal distribution between the stages that is defined by
DV g I
g I
g I
sp
sp
n sp n
N
n
n
n
=
+
Â
=
0
0
0 1
1 a
ae
,
(3)
where g
0
is the gravitational constant at the Earth’s surface, 9.8066 m/s
2
, I
sp
n
is the specific impulse of the
nth stage, and a is a Lagrange multiplier.
These values can be broken into stage masses using the following equation:
DV g I
M m
M m m
sp
i pay
i p pay n
N
n
n n
n n n
=
+
- +
Ê
Ë
Á
Á
ˆ
¯
˜
˜
Â
=
0
1
ln .
(4)
M
i
n

is the stage mass or the sum of M
s
n
and m
p
n
, and m
pay
n
is the mass of everything above the nth stage.
Thus, for the final stage, m
pay
n
is the payload or the mass of the threat mitigation system. For the (n–1)th
stage, the mass is the payload mass together with the total mass of the nth stage. Finally,
MR
g I
g I
n
sp
n sp
n
n
=
+ a
ae
0
1
0
(5)
l
e
n
n n
n
MR
MR
=
-
-
1
1
,
(6)
where MR
n
and l
n
are, respectively, the mass ratio and payload fraction for the nth stage. Payload fraction
is also defined in a manner similar to inert mass fraction:
l
n
pay
p s
m
m m
n
n n
=
+
.
(7)
Finally, mass ratio is defined as
MR
M
M
n
i
f
n
n
= .
(8)
42
The method for solving these equations is as follows: First, the inert mass fractions are estimated
for each stage. Our code used a value of 0.2 for each stage. Then, equation (3) can be solved for the
Lagrange multiplier. Next, equations (6) and (5) can be solved in series. Equations (7) and (8) can then be
used to derive propellant and inert masses. The propellant masses are used in equation (2) to calculate new
inert mass fractions. This calculational scheme is repeated until convergence is achieved. Our code also
executed this loop for a five-stage system down to a single stage to determine the least number of stages
necessary to meet the DV requirements.
The Lagrange multiplier method is usually of limited use, as it does not handle DV requirements
that include losses, such as gravitational or drag losses. However, these losses are minimal for the types of
system envisioned, where the vehicle would be deployed in low-Earth orbit (LEO). That raises an opera-
tional concern; lox/LH
2
systems are not considered viable candidates for applications where long storage
times are necessary. The systems envisioned would be applied in a situation where the vehicle would be
assembled in advance and then placed in a parking orbit. Only when a threat is detected would the system
be activated. In this type of application, the propellant tanks will be exposed to temperature cycling that
will probably exacerbate the normal problems of cryogenic propellant storage. However, the data used to
predict inert mass fractions were for launch vehicles, which must handle high stresses during ascent and
produce an initial thrust-to-weight ratio of 0.8–0.9 for upper stages and 1.2–1.5 for first stages. Both of
these conditions are alleviated for in-space vehicles, which will result in significant mass savings. There-
fore, for this level of analysis, it is assumed that these mass savings will counteract the additional mass
necessary for the active and passive thermal protection systems necessary to retain cryogenic propellants
indefinitely.
Table 7 lists the qualitative considerations for chemical propulsion. The thrust level and scalability
for chemical systems are excellent. However, the leak possibility for cryogenic propellants and the thermal
protection requirements combine to indicate that long-term readiness will be difficult. The relatively low
performance of chemical systems also makes them massive and the low density of LH
2
means that they
will be physically bulky.
Table 7. Qualitative considerations for outbound propulsion using chemically-powered rockets.
Thrust level
Scalability
Long-term readiness
Compactness
High
High
Low
Low
Low
Medium
Usefulness as weapon
Perceived safety
Synergy with other NASA missions
Manned missions
Robotic missions
Resource utilization missions
Costs
Development
Deployment
Low
High
High
Medium
Medium
First-Order Qualitative Considerations
Second-Order Qualitative Considerations
43
Chemical propulsion does not lend itself for use of these systems as a weapon. Due to their exten-
sive use over the last half century, they are perceived to be very safe by the general public. Chemical
propulsion has strong applicability for manned missions where low trip times are important. For robotic
and resource utilization missions, the level of potential synergy for chemical propulsion is considered to be
medium. Development costs should be low for chemical systems due to the significant heritage and
well-understood technologies of these systems, but deployment costs are higher due to the size and main-
tainability issues.
4.2 Nuclear Thermal Rocket
The nuclear rocket option was included to give a high thrust option, similar to the chemical option
but with improved performance. The I
sp
used for nuclear thermal rocket (NTR) propulsion systems is 850.
Although almost twice as high as that for the chemical option, it is not enough to preclude the necessity
of staging.
NTR engines weigh a considerable amount more than comparable chemical engines because of the
required reactor and shielding mass. Since no NTR vehicles have been developed, the historical approach
that was employed to calculate masses for chemically powered stages cannot be used. Instead, data from
previous NTR engine development programs have been used, and some basic assumptions to apply it to
the same type of analysis that was used for chemical stages have been made. A schematic of the NTR is
shown in figure 30.
Propellant Tank
Pump
Reactor
Core
Nozzle
Figure 30. Schematic of a nuclear thermal rocket.
14
Table 8 lists data from the Nuclear Engine for Rocket Vehicle Applications (NERVA) program.
These data show a linear relationship between the total engine mass (m
reac
) and thermal power produced
by the reactor (P
jet
):
P m
jet reac
= - 0 127 53 8 . .

. (9)
44
This relationship is shown in figure 31.
Small engine
XE
NERVA
Phoebus-2A
2,550
7,700
12,300
41,679
367
1,140
1,570
5,320
Mass
(kg)
Thermal Power
(MW)
Table 8. Data on nuclear rocket engines developed under the NERVA program.
Figure 31. Regression curve fit of NERVA program-developed engines.
The power produced by the rocket jet is related to the I
sp
and thrust produced by the rocket by the
following equation:
P g I F
jet sp
=
1
2
0
.
(10)
For this analysis, it is assumed that the initial thrust-to-weight ratio for an NTR stage is 0.8. This is based
to some extent on upperstage data from launch vehicles, but it represents a conservative estimate. Using
this value, the power can then be related to the vehicle mass as follows:
P g I M
jet sp i
n
n N
n
= Â
=
0 8
2
0
.
,
(11)
0 10,000 20,000 30,000 40,000 50,000
0
1,000
2,000
3,000
4,000
5,000
6,000
NTR
NTR Regress
+2Sy
–2Sy
Total Mass (kg)
T
h
e
r
m
a
l

P
o
w
e
r

(
M
W
)
45
where the summation is intended to calculate the cumulative mass of the nth stage and all of the stages
above it. The reactor mass can then be found by the solution of equation (9).
It would be convenient if this analysis could be conducted in the same way as for the chemical
propulsion system from section 4.1. In this analysis, the reactor mass was divided by the stage mass and
included with half of the inert mass fraction to yield an NTR inert mass fraction:
e
e
( )
( )
. NTR
n
reac
i
n
m
M
chem
n
n
= +
2
(12)
This value can then be incorporated in the same analysis scheme as was used for chemical systems.
The justification for this value is as follows. The reactor mass term accounts for the added mass associated
with NTR engines. The chemical inert mass fraction accounts for the tanks, piping, thrust structure, and
ancillary components necessary to make up a stage. It is divided by 2 because the NTR vehicle uses only
one propellant, LH
2
, instead of two. This assumption is optimistic because the LH
2
tank is much larger and
heavier than the lox tank in a chemical vehicle. Additionally, an on-orbit system maintained indefinitely
will require active and passive thermal protection, as did the chemical systems. However, the chemical
inert mass fraction still contains the mass of the propulsion system too, which is superfluous here. There-
fore, for the current level of analysis, it is assumed that these inconsistencies cancel each other.
Table 9 lists the qualitative considerations for NTRs. Like chemical systems, thrust level and
scalability is believed to be high. Also, like chemical systems, maintenance of LH
2
propellant will make
long-term readiness difficult. However, this system will be more compact than an alternative chemical
system.
Table 9. Qualitative considerations for outbound propulsion using nuclear thermal rockets.
First-Order Qualitative Considerations
Second-Order Qualitative Considerations
Thrust level
Scalability
Long-term readiness
Compactness
High
High
Low
Medium
Medium
Low
Usefulness as weapon
Perceived safety
Synergy with other NASA missions
Manned missions
Robotic missions
Resource utilization missions
Costs
Development
Deployment
Low
Low
High
Medium
High
46
An NTR system would have little usefulness as a weapon. However, as a nuclear system, its
perceived safety to the general public is rather low. This system offers high synergy for manned and
resource utilization missions. However, these systems are most efficient at larger sizes; thus, they will be
less effective for smaller robotic missions. Development costs are complicated by the fact that testing
nuclear systems introduces a number of environmental issues. Deployment costs should be favorable due
to their similarity with chemical systems, but with greater compactness.
4.3 Nuclear Pulse
The nuclear pulse concept was first considered only as an asteroid deflection technique. However,
transport of the deflection system out to the NEO needs to be accomplished as swiftly as possible, and
among propulsion techniques using known technology, external pulsed plasma propulsion (EPPP)
or nuclear-pulsed systems are the best possible performers. EPPP utilizes not only the fission energy liber-
ated in a nuclear device but is also substantially enhanced by a fusion energy release. Performance
approaches that of a fusion-driven spacecraft, but without the additional challenges inherent in fusion
technology.
15
As the fission fuel fragments ejected are extremely massive nuclei, thrust is also consider-
ably higher than for pure fusion concepts.
There are several concepts that have been proposed for EPPP, including the standard pusher plate,
16
rotating cable pusher,
17
pusher plate,
18
magnetic field,
19,20
and large lightweight sail/spinnaker (Medusa
concept) suggested by Solem.
21
All of these momentum transfer mechanisms (MTMs) utilize the same
nuclear detonation energy source. Each couples the tremendous burst of high-velocity particles to the
spacecraft by spreading the intense shock over a longer and more tolerable time period.
In this preliminary study, the most conventional—a term loosely applied to pulsed nuclear
rockets—approach was taken: the pusher plate and shock absorber configuration. The specific geometry
and scaling model was primarily derived from the original Air Force Program, ORION (fig. 32).
22
This
classified program began in 1958 and ended in 1965, at a cost of about $8 to $10 million—significant
research funding for those years. This program achieved tremendous technical success, but failed to main-
tain political support for a number of reasons. Significant data have recently been declassified and served
as the basis for the generic vehicle calculations presented here. In 1999, a smaller NASA study, Project
Gabriel, assessed the ORION concept, as applied to a smaller vehicle design, using current materials and
technologies (fig. 33).
23,24
A significant portion of that work was also used in this study. The pulse unit
designs are extremely general in nature due to the sensitive nature of the technology involved. However,
the results appear to be realistic and well within current state of the art for such devices.
47
Figure 32. U.S. Air Force ORION spacecraft (1964).
Figure 33. NASA Gabriel spacecraft (1999).
Powered Flight
Crew Station
(Shielded)
Crew
Accommodations
Structural Spine
Basic Structure
Includes Pulse-Unit
Delivery System
Secondary
Shock Absorber
Primary
Shock Absorber
Pusher Plate
Pulse-Unit
Ejection Path
Ejected Pulse Unit
Point of Detonation
Standoff
Distance
Basic Nuclear
Pulse Propulsion
Module
Propellant
Magazines
Crew Module
48
The vehicle mass estimator uses geometry and density parameters that are based on rough scaling
of the ORION vehicle. A 30-percent mass contingency factor was applied across the entire estimated
vehicle mass. The pusher plate was assumed to be constructed from solid titanium; the primary shock
absorber was a three-tiered carbon structure and the secondary shock absorber consisted of a set of long,
gas strut type systems. A spreadsheet analysis was developed, which included gross estimates on pulse unit
volume, packing fraction, and number of levels required. Although an effort was made to ensure enough
spacing to preclude a thermal reactor critical geometry being created out of the pulse unit storage, no
calculation or neutronic simulation was used to check for this issue.
Evaluation of the first-order qualitative considerations (table 10) clearly illustrates the outstanding
technical merit of this system. It is the only system that has all high (green) ratings for the primary param-
eters. It also has excellent synergy with many other NASA missions. The robotic mission category is
labeled a medium, since nuclear pulse has less potential application for smaller, near-term probes, but it
receives high marks for interstellar type missions, sample returns, multiplanet tours, and large-scale
robotic exploration.
Table 10. Qualitative considerations for outbound propulsion using nuclear pulse.
The second-order qualitative considerations, which are negative, result from concerns over secu-
rity of nuclear material and technical knowledge. The high cost is primarily generated by the significant
need to regulate and secure fissionable material and prevent nuclear weapon proliferation. The fissionable
material is readily available, and using it in peaceful space applications actually achieves cost savings in
other areas. There are tons of highly enriched uranium and plutonium available from decommissioned
Cold War weapons, both in the United States and in Russia. This material is presently being stored at a high
cost to both nations and reprocessing it for burial will be extremely expensive. Furthermore, no matter
what is done to this material, it will always be possible, although difficult, for someone to extract and
reprocess it for weapons. Only in space is it safe from theft. When used as rocket propellant, it is rendered
completely unusable in any future application. Not only is the exhaust material spread out into space, but
much of the material will escape the solar system. The other factors responsible for the high cost result
First-Order Qualitative Considerations
Second-Order Qualitative Considerations
Thrust level
Scalability
Long-term readiness
Compactness
High
High
High
High
High
High
Usefulness as weapon
Perceived safety
Synergy with other NASA missions
Manned missions
Robotic missions
Resource utilization missions
Costs
Development
Deployment
Medium
Low
High
Medium
High
49
from the large size of the vehicle and also from testing issues. Pulse unit tests will require underground
testing as well as several deep-space tests. The EPPP vehicle will require a monolithic pusher plate of
5–10 m (8 m was used in the final design developed here), some shock absorber development and testing,
and a very fast pulse unit dispenser system. The initial launch costs could also be high, although compa-
rable to other options considered in this study.
Other negative considerations also arise from concerns over security. Perceived safety is assessed
as low, since this type of propulsion system is similar to a weapon system or could even be used as one.
Although a pulse unit would be destructive if accidentally or purposely detonated on Earth, it would cause
rather limited damage, since the devices are very small and not designed to generate the weapon effects of
an atomic bomb. Note that the device will be only meters away from the spacecraft, and its power is held
to a minimum, since many hundreds of these “pulses” must be endured. The arming can be made fail-safe
and only possible in space/onboard the spacecraft.
Nations that have large quantities of fissionable material can easily build large weapons and those
who would have only small amounts, such as a terrorist-supporting state, will find it nearly impossible to
assemble, much less design and fabricate, the many extremely precise mechanisms that are required to
make small amounts attain supercriticality. As far as security and safety are concerned, the historical record
is favorable. No nuclear weapon has ever been lost or accidentally detonated, even after accidents in which
a device has been inadvertently dropped from an aircraft. This record has been maintained for over 50 yr
with thousands of weapons being moved around the world. It would seem very simple to build and store
the devices for only several months at one U.S. location before shipment into space. It is apparent that all
of these many mitigating points are not likely to counteract the likely perceived political risk. Like the
original ORION program, this concept suffers from a basic human fear of ourselves, and it will be difficult
to overcome this, even for purposes of asteroid defense.
The calculations were based on two external inputs: mission DV and payload mass. The mission DV
was based on the trajectory analysis and is discussed in other sections. The payload mass is based on the
nuclear pulse deflection option, it being logical to assume that if approval to build and launch the EPPP
vehicle (fig. 34) was obtained, the asteroid threat mitigation could then be accomplished with the same
type of technology. This conclusion is not absolutely certain, as there may be some chance of using another
deflection scheme, or more likely, that a chemical or electric system would be used for outbound propul-
sion, but carrying a payload of nuclear devices designed to deflect the asteroid.
50
The calculations then proceed with two basic subsystem designs: (1) The vehicle’s MTM, which is
the pusher plate and the shock absorber system, and (2) the pulse unit design. From these two independent
calculations, and using the required mission DV, an iterative solution is employed to determine the propel-
lant required for the mission (the propellant being the number of individual nuclear devices or pulse units
needed). Also, there is structure mass and associated volume that scales with the number of pulse units
required.
The MTM was based on simple geometry and density considerations. The plate was a simple disk
shape, although the real design would have a special taper profile as developed during the ORION pro-
gram, with thickness and diameter specified and a mass calculated using a density, assuming a homog-
enous material. In this point design, the diameter was 8 m with a thickness of 3 cm. The material selected
was titanium, which has a mass density of ª4,500 kg/m
3
.
The primary shock absorber was a three-tier block of carbon microfiber springs. The original ORION
program specified gas-filled toroidal chambers, whose mass was similar to that of the carbon material, but
which was not as safe or reliable. Again, a simple geometry and density approach was used. Each tier was
0.5 m thick and the first; i.e., that connected to the metallic pusher plate, was cylindrical in shape with the
same diameter as the pusher plate. The next two tiers were cone shaped rather than cylindrical. On one
side, each has the same diameter as the preceding tier, and on the other side, the diameter is three-quarters
of that value. The volume is calculated for each section and the density is used to determine the mass. Since
the shock absorber is envisioned to be leaf spring or coil structured, a large percentage of the volume was
considered to consist of voids. For the carbon material with an assumed density of 1,600 kg/m
3
, a carbon
fraction of 10 percent was specified. This was assumed without any rigorous analysis, since the material
manufacturing process and actual property values are presently based on speculation.
Figure 34. EPPP concept vehicle.
Nuclear Shape
Charge Device
Pusher Plate
Chemical
Propulsion
Stage
Structure
Dispensing
Mechanisms
Secondary Shock Absorber System
Three-Tier
Shock Absorber
System
10 m
8 m
Nuclear Pulse
Unit Storage
51
The secondary shock absorber system was chosen to consist of gas tube struts, similar to the ORION
engineering, but modeled as a simple hollow cylinder. Again, simple geometry and density estimates were
made instead of using masses taken from the ORION study. Although the ORION numbers were based on
relatively detailed engineering drawings, the simple geometry involved made it easy to develop a paramet-
ric calculation, which was used to explore a wide trade space area; i.e., the effect of changes in the number
and size of the shocks could be readily investigated. The material selected was titanium with a mass den-
sity of ª4,500 kg/m
3
and a wall thickness of 2 cm. Each of the six tubes was 0.5 m in diameter and 10 m in
length. The sum of the three major components that make up the MTM system gave the total vehicle dry
mass. A 30-percent mass contingency factor was applied to the total vehicle mass. The operation of the
shock absorber system is illustrated figure 35.
Figure 35. Shock absorber operations.
The pulse unit performance parameters were based upon the amount of energy released per nucleus
fissioned, the mass of U235, and the inert mass. The inert mass includes everything else which comprises
the complete device: impurities in the U235 fuel, high explosives, arming circuitry, columniation structure,
channel filler, or low atomic number propellant and casing/mounts. The energy released per U235 nucleus
fissioned is ª185 MeV. This accounts for the thermal energy of the main fragments only and not the
neutrons, product decays, neutrinos, or prompt gamma rays. This value was used to determine the theoreti-
cal average energy release per kilogram of fuel (7.592¥10
13
m
2
/s
2
), using Avogadro’s number and a con-
version factor. An estimated “burnup” fraction (percent of U235 nuclei actually split in the reaction) was
assumed and the total energy released by the pulse unit was then determined; i.e., burnup fraction times
mass of fuel in the pulse unit (kg) times the energy per kilogram. The chemical explosive energy required
1. Uncompressed 2. Fully Compressed 3. Ready for Next Pulse
52
to start the nuclear reaction was neglected. The total fission energy was then assumed to be absorbed
evenly into the total pulse unit mass and the average velocity of the particles determined from the equation:
E mv =1 2
2
/ .
(13)
I
sp
was determined from the average particle velocity times an effectivity factor, divided by the
gravitational constant; i.e., I
sp
=v¥F
E
/g, where g=9.807 m/s
2
). The effectivity factor (F
E
) in the I
sp
equation
accounts for the collimation factor; i.e., plasma burst preferential direction, pusher plate diameter, and
standoff distance (distance from pulse unit detonation to the center of the plate). The effectivity factor
equation was derived by Thane Reynold.
16
From this I
sp
, a total energy yield was determined and was
expressed as an equivalent number of kilotons of tri-nitro-toluene (TNT).
The last spreadsheet calculation determines the number of pulse units needed to be carried as pro-
pellant using a spreadsheet goal seek iteration routine. An initial estimate of the propellant mass is made.
This estimate is multiplied by a generic tankage fraction, assumed to be 15 percent. This accounts for all
the storage, mounting, and dispensing mechanism hardware needed to handle the pulse units. The mass of
the MTM and payload are then added to obtain the start mass of the spacecraft. Using the ideal rocket
equation,
M M e
f i
V
g I
sp
=
-
D
0
,
(14)
a final mass is generated for the spacecraft at the end of the mission. The difference between that and the
start mass is the propellant used. Dividing that quantity by the mass of one complete pulse unit (mass of
U235 fuel and inert mass) established the number of pulse units (rounding up to the next whole value). A
5-percent contingency factor was added to account for misfires, trajectory errors, and other performance
losses. With the final number of whole pulse units determined, the actual total mass of propellant is deter-
mined and compared to the initial estimate. If the estimate was not sufficiently close, a new value was
automatically generated by the computer; i.e., the new value is halfway between the initial estimate and the
calculated value, and the calculation repeated. The conversion criterion was an error better than 1¥10
15
between the two values.
Other calculations were performed to estimate the size of the propellant magazines. The diameter
of the storage area was held to half that of the plate, and the pulse units were estimated by the density of
U235, which has a mass density of 19,000 kg/m
3
, and inert material; i.e., low atomic number with a mass
density assumed to average 1,500 kg/m
3
. A 10-percent void fraction was assumed to determine the gross
volume. Using a length-to-width ratio of 2, the size and shape was ascertained for a single pulse unit.
Required “floor space” was found by assuming a square storage geometry; i.e., packing cylinders, and
adding an additional 25-percent fluff factor. The floor area required was divided by the area available
based on the one-half plate diameter criteria to estimate the number of storage levels. Finally, a level height
of twice the pulse unit length was assigned and the total propellant volume was determined. This was used
to ascertain whether the propellant could be reasonably launched into LEO by conventional chemical
rockets.
53
The EPPP calculation approach was simplistic, but nonetheless is considered to be realistic. Due to
the sensitive nature of any details concerning nuclear devices and their effects, the inputs were intention-
ally conservative and were obtained from open literature sources. No inference should be made regarding
any parameters relating to the true composition, size, geometry, or efficiency of any real nuclear device.
Fusion energy, as well as the use of plutonium fuel, was not considered. If employed, they would be
expected to enhance performance, although they each have their own drawbacks, most notably, their natu-
ral radioactive decay, cost, and availability. The vehicle was also designed using simple geometric and
density-based calculations. No estimate of the effects of erosion, radiation, or shock was made for the
pusher plate design. The 30-percent mass margin was considered adequate, but not necessarily conserva-
tive, for such a preliminary design. Because of the conservative assumptions used in pulse unit design, the
MTM was not additionally burdened with a higher mass margin.
Many small trade studies were conducted in order to gain an understanding of how the entire
system would function. The vehicle has excellent ability to absorb mass growth; a larger vehicle actually
yields improved performance. Surprisingly, the final mission design was not very sensitive to the plate
diameter as long as the plate diameter-to-standoff distance ratio was not altered significantly. The
acceptable vehicle trade space was found to be very large. This gives confidence in the practicality of such
a propulsion system, despite the large uncertainty in the design parameter inputs. Even doubling the pay-
load mass increases the vehicle gross mass by <20 percent or ª50 more pulse units. In general, pulse unit
yields were in the 3 to 4 kton range (present devices are normally in the many megaton range) and the
range of I
sp
was 2,500 to 5,000 s. The general conclusion from the analysis is that this technology has
excellent performance capability and that there are no known technical barriers.
4.4 Solar Sail
The solar sail offers unique capabilities for rendezvous with an incoming object. Interception is
also a possibility but it would not be a first choice for threat mitigation systems depending upon kinetic
energy. Solar sails are capable of substantial inclination changes that are difficult to achieve with other
propulsion systems. Sufficiently large sails are also capable of moving sizable payloads over long dis-
tances. Also, since most NEOs are expected to occupy orbits within that of Jupiter, a solar sail would
remain close enough to the Sun to maintain significant propulsive capability. The analysis method
discussed here follows that given elsewhere.
25,26
There are several solar sail configurations available. The three most popular are square sails, disk
sails, and heliogyros. The difference between these types of sails relates mostly to consideration of struc-
tural design. Square sails use a square sail sheet supported by booms and lines. The heliogyro uses long,
narrow sail blades that are each connected to a central hub. The blades are kept taut through rotation of the
entire sail, including both the blades and the hub. A disk sail attempts to incorporate advantages of both
square sail and heliogyro. It has a circular sail sheet that is kept taut through rotation. Previous studies have
suggested that, in most cases, the square sail is slightly more attractive than the other options. The analysis
presented here is based upon a square sail design (fig. 36).
54
The important dimensions for a square solar sail are illustrated in figure 37. Each side of the sail has
a length (l). The total sail area (A
sail
) is equal to the square of the sail length. The sail will normally be
oriented at an angle of attack (a) relative to the plane perpendicular to the orbital radius vector. The sail
will project an area (A) normal to the incoming luminous flux. The centerline angle (f) is the angle
between the sail normal and force vector. Similarly, the cone angle (q) is the angle between the force vector
and the incident luminous flux.
Figure 36. Artists concept of a billowing, square solar sail.
25
Figure 37. Schematic and dimensions for a square solar sail.
I
F
F
t
F
n
θ
φ
α
Tip Vanes
Control Boom
Payload Envelope
Spars
Sail Film
Stays
To Sun
55
The first task in solar sail design is to model the luminous flux emanating from the Sun. The Sun
should be modeled as a finite disk out to distances as far as 10 solar radii. The solar light pressure can be
calculated as a function of distance from the Sun using the following equation:
P r
L
cR
R
r
s
s
s
( )
= - -
Ê
Ë
Á
Á
ˆ
¯
˜
˜
È
Î
Í
Í
Í
˘
˚
˙
˙
˙
6
1 1
2
2
2
3
2
p
,
(15)
where R
s
is the radius of the Sun, r is the radius of the sail from the center of the Sun, c is the speed of light,
and L
s
is the solar luminosity.
The sail converts this pressure into thrust through absorption or reflection. Spectral reflection is the
preferred method as it is the most efficient. However, 100-percent reflection is not easily achieved.
Absorption of the incident light is less attractive because of the resulting thermal loads that must be radi-
ated away from the reverse side of the sail. This emission process also produces its own radiation pressure,
which partially cancels the motive force. The normal and tangential forces due to absorption are given as
F PA
an
= cos
2
a
(16)
F PA
at
= cos sin , a a
(17)
where A is the cross-sectional area presented to the Sun. The tangential and normal forces due to reflection
are calculated using
F PA r s B s r
rn f
= ¢ ¢ + - ¢
( )
¢
( )
cos cos
2
1 a a
(18)
F PAr s
rt
= ¢ ¢cos sin , a a
(19)
where r¢ and s¢ are the reflection coefficient and the spectral reflection coefficient, respectively. Also, B
f
is
the front non-Lambertian coefficient used to model the nonspectral reflection. The normal force due to
emission is given as
F PA r
B B
e
f f b b
f b
= - ¢
( )
-
+
1
e e
e e
a cos ,
(20)
where B
b
is the non-Lambertian coefficient for the back side of the sail. Additionally, e
f
and e
b
are the
emission coefficients for the front and back, respectively. The values used in this study for these optical
coefficients can be found in table 11. These values can be traced back to a Jet Propulsion Laboratory (JPL)
solar sail model developed during the 1970’s to study a proposed Halley’s Comet rendezvous mission.
56
The total sail normal and tangential forces can be determined by summing the above forces:
F F F F
n an rn e
= + +
(21)
F F F
t at rt
= + .
(22)
The total force and centerline angle are determined using
2 2
t n
F F F + =
(23)
f =
-
tan .
1
F
F
t
n
(24)
Finally, the sail temperature can be calculated by accounting for the thermal flux due to absorption
and emission:
T
r cP
f b
=
- ¢
( )
¢ +
( )
È
Î
Í
Í
˘
˚
˙
˙
1
1
4
cos
,
a
s e e
(25)
where

¢ s

is the Stefan-Boltzmann constant.
The sail can be allowed to billow under the luminous pressure and therefore will not present a plane
surface. The cone angle under billowing can be determined by
q a f = - ,
(26)
Table 11. Optical, billowing force and other parameters used in solar sail analysis.
25
Variable
B
f
B
b
f
b


C
1
C
2
C
3
Front non-Lambertian coefficient
Back non-Lambertian coefficient
Front emissivity coefficient
Back emissivity coefficient
Reflection coefficient
Specular reflection coefficient
Force coefficient 1
Force coefficient 2
Force coefficient 3
Sail loading parameter
Name Ideal Value Study Value
2/3
2/3
0
0
1
1
0.5
0.5
0
Low as possible
0.79
0.55
0.05
0.55
0.88
0.94
0.349
0.662
–0.011
6 gm/m
2
σ
ε
ε
57
and the total force adjusted for billowing is calculated using
F F C C C
norm
= + +
( )
1 2 3
2 4 cos cos . q q
(27)
The coefficients C
1
, C
2
, and C
3
are billowing force coefficients that were also determined during JPL’s
Halley’s Comet study. These coefficients and the optical parameters are not universally applicable but are
used here as a first approximation to take account of billowing and optical physics effects.
The mass of the sail can be calculated using
M A m
i pay
= + s .
(28)
There are several figures of merit that are typically used to compare sail performance. First, the sail
lightness number (b) is a characterization of the sail’s acceleration compared to that due to the local gravi-
tational force imposed by the Sun:
b
p s
=
L
GM c
s
s
2
.
(29)
Here, G is the universal gravitational constant, M
s
is the mass of the Sun, and
s
L
is the solar luminosity.
Figure 38 illustrates the relevance of lightness number to sail orbital capabilities. For a b < 1/2, the
sail travels on an elliptical trajectory around the Sun. For a b of between 1/2 and 1, the sail follows a
hyperbolic escape trajectory. When b = 1, the force exerted on the sail exactly negates the gravitational
force from the Sun. For values of b > 1, the sail trajectory is still hyperbolic, but with a thrust higher than
the local gravitational force.
58
The sail’s efficiency is a measure of its ability to convert the incident luminous flux into a propul-
sive force; it is given by
h =
F
A P
norm
sail
2
.
(30)
This value is a constant for any heliocentric distance or any sail area.
The characteristic acceleration is the acceleration that the sail will experience when it is normal to
the Sun and at a distance of 1 au (the average radius of the Earth’s orbit around the Sun). It is given by
a
PA
M
i
0
2
=
h
.
(31)
The sensitivities to several of these figures of merit are given by the following equations:
D D a
a
m
m
pay
s
s
s
0
0
1
1
=
-
+
s
s
(32)
Figure 38. Sail trajectories relative to lightness number.
25
Sun
r
V
β
β
β
β
β
>1
=1
=0.5
=0
0.5< <1
β 0< <0.5
Deploy Sail
59
a
a
m
m
m
m
s
pay
pay
pay
0
0
1
1
D
D
=
-
+
(33)
a
a
m
m
A
A
pay
s
0
0
1
1
D
D
=
+
.
(34)
These are the sensitivities to sail density, payload mass, and sail area, respectively.
Table 12 lists the qualitative considerations for solar sail outbound propulsion. Solar sails produce
very low thrust. As even the least demanding missions require large sails, there is limited potential for
scaling up the design. The sail does not have to be unfurled until a threat is identified; therefore, it can be
maintained indefinitely in a very compact state.
Table 12. Qualitative considerations for outbound propulsion using solar sails.
It is difficult to conceive of any way in which a solar sail could be used to cause harm. Sails have
the ability to achieve a variety of mission types, as long as the total payload to be delivered is small. This
makes it difficult to conceive manned missions using solar sails. The deployment of a furled sail presents
few complications, although the unfurling process is not trivial. Sail development cost estimates are com-
plicated by the difficulty in measuring the low propulsion levels on a ground-based facility.
Solar sails have unique abilities to change inclination and achieve non-Keplerian orbits. For these
reasons, solar sails have proven in this study to have surprising capabilities to rendezvous with the varied
orbits of potentially incoming objects. Intercept capabilities are less impressive, as sails do not accelerate
quickly. However, the sail should be considered a strong contender for the outbound leg of any rendezvous
concept.
First-Order Qualitative Considerations
Second-Order Qualitative Considerations
Thrust level
Scalability
Long-term readiness
Compactness
Low
Low
High
High
Medium
Low
Usefulness as weapon
Perceived safety
Synergy with other NASA missions
Manned missions
Robotic missions
Resource utilization missions
Costs
Development
Deployment
Low
High
Low
High
High
60
4.5 Solar Collector
The solar collector is closely related to the solar sail; the principle is shown in figure 39. A solar sail
with a curved configuration is used to focus incident light onto a secondary collector. The secondary
collector directs the light outward so as to generate thrust. The curved sail centerline remains parallel to the
vehicle Sun radius vector so that the sail generates the maximum possible thrust. This configuration avoids
the loss of thrust experienced by a conventional solar sail when attempting to direct its thrust vector in an
optimal direction. The focused light can also be effectively used to redirect incoming asteroids and comets,
as will be discussed in section 5.
Figure 39. Solar collector configuration.
As shown in figure 39, the curved sail is assumed to be hemispherical, with the secondary
collector located at the center of the projected sphere. Although this configuration is susceptible to chro-
matic aberration, this is not a concern for this application. Additionally, a hemispherical shape would limit
the direction in which the secondary mirror could focus to the plane perpendicular to the collector centerline.
Thus, figure 39 illustrates a sail with a projected hemispherical area, but with an actual area that is less.
Accounting for the additional area of the hemisphere gives a conservative analysis. This is clear because it
is obvious from the figure that the additional cross-sectional area capturing light is not significant. More
efficient parabolic configurations may be found, but they would require ray tracing or other advanced
calculations that are beyond the scope of this project. The captured light is a function of the cross-sectional
area of the hemisphere. The sail mass is calculated using the hemispherical surface area and the sail-
loading factor found in table 11. Use of these loading factors is somewhat optimistic, as a curved solar
collector will require more structural support to maintain its shape than would a flat solar sail. However,
this optimistic assumption is expected to negate the pessimistic hemispherical shape assumption. The solar
collector uses the same optical parameters listed for the solar sail in table 11.
Secondary
Collector
Projected
Area
Payload
Envelope
Spars
Sail Film
Stays
To Sun
F
r
F
i
l
61
Design of the secondary collector is of particular importance for this concept. If the collector is
made large, then the amount of sail area that it blocks may be prohibitive. The curved sail focus is found
to be
27
¢ =
+
l
l r
1
2 1
,
(35)
where l is the diameter of the curved sail and r is the orbital radius of the sail from the Sun. For all realistic
situations, l/r can be treated as being zero. For a 100-m-diameter sail, equation (35) yields a focus of 50 m,
which gives the location of the secondary collector in figure 39. The magnification is shown to be
M
l
r
=
¢
.
(36)
The curved sail essentially produces an image of the Sun. The image size is
r M R
c s
= ¥ ,
(37)
where R
s
is the radius of the Sun, 6.96¥10
10
m. For the sail >1 au, the magnification is 3.34¥10
–10
and the
image size is 0.233 m. Therefore, the collector size need not be large and its shadow will be insignificant
compared to the curved sail area projected normal to the Sun.
Table 13 lists the qualitative considerations for the solar collector. The solar collector is expected to
have better thrust levels and scalability than the solar sail because of its more efficient use of the incident
solar radiation. Note that the configuration cannot be folded as easily as can a solar sail. The collector has
considerations similar to those of the solar sail in all other respects.
Table 13. Qualitative considerations for outbound propulsion using solar collectors.
First-Order Qualitative Considerations
Second-Order Qualitative Considerations
Thrust level
Scalability
Long-term readiness
Compactness
Medium
Medium
High
Medium
Medium
Low
Usefulness as weapon
Perceived safety
Synergy with other NASA missions
Manned missions
Robotic missions
Resource utilization missions
Costs
Development
Deployment
Low
High
Low
High
High
62
5. THREAT MITIGATION
5.1 Nuclear Fragmentation
The equations used to model the catastrophic fragmentation of a near-Earth solid body asteroid are
based on the work of Thomas J. Ahrens, California Institute of Technology, and Alan W. Harris, JPL.
28
These equations are based on the assumption that an explosive device is placed deep enough below the
asteroid’s surface to produce near-optimum fragmentation. The location for optimum fragmentation is
generally considered to be the geometric center of the target object.
In reference 28, starting with equation (65), p. 920, one can write
v Hr ( ) , cm/ s = ¢
(38)
where r¢ is the idealized radius of the spherical target body and H is a parameter which characterizes the
hardness of the asteroidal material. For hard igneous terrestrial rocks, H=5.72¥10
10
s
–1
. For soft terrestrial
rock; e.g., sandstone, H=2.90¥10
10
s
–1
.
An expression for shock wave internal energy per unit mass can also be written:
E v = 0 5
2
. ,
(39)
where v is the shock-induced particle velocity and E is the energy. Fragmentation takes place when E=E
fracture
=

ª10
–7
erg/gm (energy density needed to break a 10-m object in two).
Finally,
¢ = r
r
W
1 3 /
,
(40)
where r is the actual radius (in meters) and W is the required blast yield of the device. Combining these
three equations gives an equation for the required blast yield as a function of asteroid radius in units of
megatons of TNT:
W
E r
H
=
Ê
Ë
Á
Á
ˆ
¯
˜
˜
2 2 10
3
3 2 /
,
(41)
where the density of the target object is assumed to be 2 g/cm
3
.
63
Figure 40 shows the blast yield required to fracture an asteroid into fragments smaller than 10 m in
size. Although open to debate, it is generally assumed that fragments of this size would be much less likely
than the original body to survive entry through Earth’s atmosphere. Even if any fragments did reach the
ground, the impact of these relatively small objects, spread over a large area, would be less damaging from
a global point of view than from a single massive asteroid strike.
Figure 40. Blast yield—explosive placed at center of body—required
for fragmentation as a function of asteroid radius.
These equations are admittedly somewhat ideal because they assume that:
(1) The asteroid is a perfectly spherical homogeneous structure.
(2) The explosive charge is placed at the exact geometric center.
(3) The explosion fractures the target body into pieces no larger than 10 m in diameter.
However idealized these assumptions may be, they do permit one to estimate the explosive power required
for fragmentation to within an order of magnitude.
To gain a better understanding of the mass of the explosive payload to be delivered to the target
body, data on existing nuclear warheads were tabulated using open literature sources in table 14 and graphed
in figure 41.
29
As shown in the tabulated data, there are no existing devices that could catastrophically
fragment an asteroid >2 km in diameter; this assumes that all of the ideal conditions listed above are
satisfied.
10,000
1,000
100
10
1
0.1
0.01
0.001
0.0001
10 100 1,000 10,000
Asteroid Radius (m)
B
l
a
s
t

Y
i
e
l
d

(
M
t
o
n
)
Hard Igneous Rock
Soft Rock
64
The problem of exactly how the explosive device would be placed at the geometric center of the
target body has not been addressed in this study. One idea considered is to utilize the same technology that
is found in the “long-rod bunker buster” types of ordinance that the U.S. military employs against hardened
underground facilities. This idea has the advantage of not requiring a DV breaking maneuver to rendezvous
with and soft land on the target. Instead, the outbound kinetic energy is utilized to bury the device to the
optimum depth. Again, using open literature sources, the physical characteristics of the BLU–113 penetrator
bunker buster are listed in table 15.
30
Figure 41. Device mass versus explosive yield.
Table 14. Nuclear device masses.
Device Mass (Ibm)
E
x
p
l
o
s
i
v
e

Y
i
e
l
d

(
k
t
o
n
)
10
100
1,000
10,000
100,000
100 1,000 10,000 100,000
y = 3,161.6 ln(x) – 8,481.3
Designation
Mass
(lbm)
Mk–1
Mk–3
Mk–4
Mk–5
Mk–6
Mk–7
Mk–8
Mk–11
Mk–12
Mk–14
Mk–15
Mk–16
Mk–17
Yield
(kton)
16
49
32
120
160
61
30
30
14
7,000
3,900
8,000
15,000
8,900
10,300
10,900
3,125
8,500
1,700
3,280
3,500
1,200
31,000
7,600
42,000
42,000
Designation
Mass
(lbm)
Mk–18
Mk–21
Mk–24
Mk–28
Mk–36
Mk–39
Mk–41
Mk–43
Mk–53
Mk–57
Mk–61
Mk–83
Yield
(kton)
500
5,000
15,000
1,100
10,000
4,000
25,000
1,000
9,000
20
340
1,200
8,600
17,700
42,000
2,320
17,700
6,750
10,670
2,125
8,900
510
716
2,400
65
These characteristics, especially the explosive mass-to-overall mass ratio, were used to estimate a
representative total mass for the explosive payload that must be delivered to the target body (table 16).
Now, the depth these first-generation penetration weapons can achieve has been explained by a simple rule
of thumb. “For typical values for steel and concrete, we expect an upper bound to the penetration depth to
be roughly 10 times the missile length, or about 100 ft (30 m) for a 10 ft (3 m) missile. In actual practice the
impact velocity and penetration depth must be well below this to ensure the contents are not severely
damaged.”
31
It is assumed that the explosive device can be successfully delivered kinetically to the center
of a 200-m-diameter asteroid; anything larger may require the use of some sort of drilling or auger device.
Table 15. BLU–113 penetrator characteristics.
Table 16. Qualitative considerations for threat mitigation using nuclear fragmentation.
Structure
Length
Diameter
Explosive
Overall Mass
Penetration of concrete
Fuse
Weapon system
Thick high-grade steel
153 in
14.5 in
630 lbm (285 kg)
4,400 lbm (2,000 kg)
20 ft (≈6 m)
FMU–143 series
GBU–28
First-Order Qualitative Considerations
Second-Order Qualitative Considerations
Susceptibility to dust cloud
Ability to handle target rotation
Requires landing on target
Usefulness on fragmented body
Swarm option
Low
High
Maybe
Low
Medium
High
Medium
Usefulness as weapon
Perceived safety
Synergy with other NASA missions
Manned missions
Robotic missions
Resource utilization missions
Costs
Development
Deployment
High
Low
Low
Low
Low
66
5.2 Nuclear Deflection
Nuclear deflection of an asteroid or comet would probably be accomplished by using a nuclear
device to produce a highly intense radiation burst. This is done while the device is still a significant dis-
tance from the target object. A relatively thin layer of the body’s surface absorbs the intense, high-energy
electromagnetic radiation—mainly hard x rays or gamma rays—and is vaporized. The intense heating
blows the surface layer off and the ejected mass imparts a reactive impulse to the body.
There are other ways in which a nuclear blast might be used to deflect an asteroid. Near-surface
blasts will break up the surface layer and blow large pieces of physical debris into space. Once again, a
reaction would be produced and act to deflect the body. Unfortunately, energy losses in the fracture process
and the absence of a preferred direction of motion for the detached pieces both serve to limit performance.
Another option would be to place a quantity of some low atomic number material between the
nuclear device and the asteroid. This material would then absorb the radiation and blast energy from the
device, would vaporize, and the resulting debris would impact the asteroid. Although possibly effective if
used on a solid asteroidal body, this technique would be less effective on a comet or dusty and/or soft
asteroid.
A deep subsurface blast could also be used, but the nuclear device must somehow be buried or be
designed to survive the high gravity loads of a deep-penetrating projectile. Unfortunately, such a technique
would still dissipate a majority of the available energy in the fracture process. One could envision a situa-
tion where only a few large fragments are produced or where the pieces have very little separation velocity
and would all still strike the Earth. Even worse, some of the pieces could draft behind others during atmo-
spheric entry, causing as much or even more impact damage on the Earth’s surface than one large piece.
The first process described above—and used in the following analysis—is similar to a laser abla-
tion process. In fact, the energy deposition, absorption, and subsequent plasma expulsion are based on the
same physical principles and governed by the same mathematical equations as for laser ablation. The
asteroid deflection process for nuclear pulse/EPPP is accomplished with a more conventional nuclear
device design and nominally requires two to ten separate devices to be successful. The primary reason for
this is that nuclear energy is efficiently liberated during very high yield detonations. The energy deposited
is then transferred to the in situ propellant—surface material from the target body itself—to produce a very
efficient propulsive technique.
First-order qualitative considerations all received the highest score for this technology. Table 17
lists the rankings as all green. A dusty target body surface or dust clouds do not affect the radiation energy
from the blast. The impulse may be slightly reduced, depending on the particular conditions, but much of
the energy will still be deposited on the surface and will produce thrust. Even the dust cloud itself, should
it absorb a great deal of the energy, will generate large pressures on one side of the body and so assist in the
desired deflection. A tremendous advantage to such an intense flash of radiation is that target rotation is
essentially inconsequential. Like a strobe light that seems to stop a moving object for an instant, the entire
thrusting event is over in milliseconds and the momentum imparted to the object is in one direction. Precise
targeting or landing on the body is not needed and simplifies the mission requirements considerably. Frag-
mented asteroids or comets could be dealt with very well, since each pulse would affect the trajectories of
67
Table 17. Qualitative considerations for threat mitigation using nuclear deflection.
all objects simultaneously. Depending on the circumstances, there are several advantages that could be
capitalized upon so as to maximize the effectiveness of the nuclear pulses. Finally, the swam option is also
every effective with this technology as, after each pulse, an assessment can be made as to the effectiveness
and implications of the event before the next one is sent. One large device might be used to obtain the
maximum efficiency of the fission fuel, but this would not be a critical factor to mission success or cost.
Many plasma pulses will allow precise control, reduce risk, and be more flexible in target engagements.
The second-order qualitative considerations have a mixed assessment ranking. The technique’s
usefulness as a weapon is obviously high, but this point is a little misleading. The very fissile material that
would be used in an asteroid deflection system currently resides in nuclear weapons around the world.
Removing this material from these weapons, and instead, using it for planetary defense, must surely pro-
duce a net benefit. Preexistence of the fissile material is also one of the reasons why the cost is not high for
this option. In fact, besides the expense for nuclear safeguards, which might be accounted as an expense
already being carried by other government agencies, particularly if excess weapon stockpiles are used, this
should be a low-cost option. Perceived safety is a clear problem. However, in reality, nuclear weapons of
this caliber have been safely maintained for over 50 yr without incident. A space-based planetary defense
system would be far more secure, and a ground-based system would be no more difficult or expensive to
maintain than the present silo-based ICBM systems already maintained by several countries. Synergy with
other NASA missions is high, as the physics and many of the specific technology challenges would be
applicable to an EPPP; e.g., nuclear pulsed propulsion, vehicles capable of interplanetary travel with
acceptable trip times for human missions, or for deep-space robotic missions. Of course, extraterrestrial
resource utilization would be greatly facilitated if this technique were available to deliver asteroids or
comets to strategic locations in the solar system.
First-Order Qualitative Considerations
Second-Order Qualitative Considerations
Susceptibility to dust cloud
Ability to handle target rotation
Requires landing on target
Usefulness on fragmented body
Swarm option
Low
High
No
High
High
Medium
Medium
Usefulness as weapon
Perceived safety
Synergy with other NASA missions
Manned missions
Robotic missions
Resource utilization missions
Costs
Development
Deployment
High
Low
High
Medium
High
68
There are two primary hardware components to the nuclear deflection option: the nuclear device
and a small rocket delivery system. For this assessment, the estimate for the nuclear devices is only based
upon the fission of U235. Tritium cores for fusion enhancement and plutonium fuel were not considered
but would be significant performance enhancers. The use of only uranium fuel is helpful because it is
essentially nonradioactive and can be easily handled and stored for long periods of time. The performance
for this application is excellent and little is gained by going to hydrogen bomb performance. Although, it
may be easier to utilize existing and well-proven devices rather than build new ones, even if the design
used a successful early-generation uranium mechanism.
The calculational process begins using inputs for the asteroid mass and its average density. Some
nuclear device characteristics are defined: nominal yield, collimation factor, burst half-angle, burnup frac-
tion, and inert material fraction, along with the standard properties of U235 (185-MeV fission fragment
energy and uncompressed density of 19,100 kg/m
3
). The ideal potential energy in the fuel
(J/kg) is found from the energy release in megaelectron volts multiplied by Avogadro’s number and then
divided by the molecular weight—provided the proper unit conversions are also used. From this value, the
desired yield, and the burnup fraction, it is straightforward to determine the required mass of U235. The
total device mass is then the sum of the fuel and inert mass, where the inert mass is based on the estimated
inert mass fraction (F
inert
) given in equation (42):
M
M
inert
F
F
inert
inert
=
¥
-
U235
100
100
1
( )
( )
.
(42)
It should be noted that this analysis only assumes the kinetic energy from a pure fission reaction of
U235, which is converted to intense thermal electromagnetic energy in the hard x-ray or gamma-ray wave-
lengths. The analysis neglects the neutron energy and prompt gamma ray produced, which would improve
performance. Finally, a simple pulse unit volume is determined by assuming that the inert material (aver-
age) density is simply one-fourth of the density of the fuel, with an additional 10-percent fluff factor on the
entire device.
The planetary body size is determined by assuming a roughly spherical shape and using the mass
and density to estimate a radius. The asteroid is considered to be a hemisphere, illuminated at the optimum
pulse unit distance. The standoff distance—perpendicular distance from detonation point to body’s sur-
face—is selected so that, at the device’s cone half-angle, one-half of the exposed surface is irradiated; i.e.,
one-eighth of the total spherical asteroid surface. The geometric relationships are illustrated in figure 42
and the standoff distance is calculated using equation (43). The average energy deposited on the surface is
based on the volumetric ratio of the radiation cone and the total spherical volume surrounding the detona-
tion (eq. (44)), with a conversion factor of 4.186¥10
12
to convert energy from kilotons to Joules, and
divided by the surface area (pIR
2
) on which the energy is deposited. To account for the losses that might be
involved; i.e., surface reflection, scattering, dust particle interference, reradiation losses to deep space,
etc., only 25 percent of the energy just determined was regarded as available to provide thrust:
69
D
IR
IR
r
standoff
, where = =
tan( ) a
2
2
(43)
E
F E D IR
D IR
avg
collimation yield standoff
standoff
. =
( )
( )( )
1 3
4 3
3
3 2
p
p p
(44)
The depth of energy penetration was estimated at 20 cm, and this volume of surface material was
used to estimate the in situ propellant available. The average velocity of the propellant leaving the body
after one nuclear radiation pulse is calculated from the kinetic energy equation (E = 1/2 mV
2
) using the
mass of the propellant (derived from the disk volume—for the 20-cm-deep region—multiplied by the
average density assumed for the planetoid) and the energy as determined above. Not all of this material
will be ejected in such a way as to contribute momentum to the planetoid in the desired direction. There-
fore, a further efficiency factor of 20 percent was applied before determining the effective I
sp
of the
technique.
The calculation continues with the deflection DV for the body being used to calculate the total
propellant needed at the I
sp
predicted for the pulse unit interaction. The standard exponential rocket equa-
tion is used (eq. (45)), and the total number of pulses required to accomplish the deflection is found by
dividing this result by the propellant mass previously calculated for a single nuclear device. Note the value
is rounded up to the next whole number and no contingency nuclear pulse units were assumed to be carried
in this initial analysis:
M M e
f i
V
gI
sp
=
-
D
.
(45)
Figure 42. Geometric position of the pulse unit to the planetary body and cone half-angle definition.
Cone Half Angle
IR
r
70
A rocket delivery system was assumed to provide standoff distance from the spacecraft (assumed to
be an ORION or Gabriel type derivative as shown in fig. 34) and to provide some placement control of the
fission device. Figure 43 illustrates the operation of the EPPP spacecraft and rocket delivery system. The
calculations for this component of the system were nonrigorous and merely serve to estimate the mass and
volume that would be needed. A 1-km/s DV was assumed and the rocket equation used to estimate the
rocket propellant required to carry the nuclear pulse unit to the asteroid. Table 18 shows the basic assump-
tions made to estimate the rocket size. The chemical propulsion system mass includes the tank and other
subsystems that scale with propellant load. (It is recognized that this is not the most accurate methodology,
but within the framework of this study, it is an acceptable expedient.) The tank volume was estimated by
assuming all the propellant was hydrogen with a density of 71 kg/m
3
and by specifying a 2-m-diameter
cylinder tank.
Figure 43. Nuclear pulse rocket delivery system sketch.
Kickout
Motion
≈1.25 m
≈4 m
Motion Through
Secondary “Strut”
Shock Sleeve
Rocket Fires and Plume
Misses EPPP Spacecraft
This End Faces
Asteroid When
Detonated
Separation Distance
Would be Several km
1. Initial Intense Nuclear Blast
2. Plasma Shock Cone Directed Toward Asteroid
3. Plasma Dust From Asteroid
Primary “Carbon” Shock
Absorber (Three Tier)
(Drawing not to scale)
1
2
3
71
The final part of the analysis confirmed that the fission pulsed detonation methodology is very
insensitive to many of the input parameters and can easily move most threatening bodies safely away from
the Earth. The system’s qualitative parameters ranked very high, and even the secondary considerations
that were unfavorable appear to have some mitigating arguments. The uranium fission technology
assumed is well established, perhaps even to the point of being outdated, and the general estimates
prepared were all conservative in nature. The plasma interaction with the planetoid body surface was not
modeled, but expected propulsive results were established with significant efficiency degradation factors
taken into account. The need for a second rocket delivery system with an arbitrary 1-km/s DV seems
prudent for protection of the carrier spacecraft. Using an all-chemical system to deliver the nuclear devices
from the Earth seems to be impracticable because the intercept time would be exceedingly long, although
this trade space was not explored in detail.
5.3 Solar Sails
Solar sails are found in the literature as nonnuclear options for deflection. Sails were initially con-
sidered as possible deflection mechanisms. Using calculations shown in section 4, the overall performance
of a solar sail can be determined where an asteroid or comet has somehow been connected to it and thus
constitutes part of the payload.
An optimistic calculation of the overall DV imparted to the vehicle plus object is determined by
using
DV a t = ¥
0
, (46)
where a
0
is the characteristic acceleration, defined in the outbound solar sail description, and t is the time
during which the sail is in operation. DV versus time can then be plotted for several sail areal densities and
sizes. Figure 44 shows the plot of an object that is assumed to be a 10-m-diameter carbonaceous chondrite
asteroid. It is obvious from the figure that the sail areal density is insignificant for this application. The
payload mass of the asteroid dominates the sail mass. The lines indicate sail areas of 10
4
, 10
5
, 10
6
, and
10
7
m
2
, respectively. These areas translate to side lengths for a square sail of 100, 316, 1,000, and 3,162 m,
respectively. It is expected that the deflection requirements for an incoming asteroid with <1 yr to deflect
will be around 1–10 m/s. Therefore, all four size sails are able to deflect a 10-m object.
Table 18. Chemical rocket assumptions.
Specific impulse
Chemical propulsion system mass
Nuclear device mass structure contingency
Integration structure and contingency factor
Pulse unit volume contingency factor
Nuclear device rocket volume contingency
Volume integration structure and contingency factor
450
500
15
30
30
15
20
s
kg
Percent
Percent
Percent
Percent
Percent
Characteristic Value Units
72
Of course, a 10-m object is of little concern. From the above, DV versus time can be plotted for
asteroids of four different diameters: 10, 100, 1,000, and 10,000 m (fig. 45). From here, all three larger sails
are able to handle deflection requirements for a 100-m object without an undue amount of time
(1 yr). The largest sail might be able to deflect a 1-km object but would require 5 yr or more. Finally,
deflection of a 10-km object seems outside the capabilities for any reasonably sized sail. From these
results, using this option will necessitate the use of very large sails and will still not provide the capability
of deflecting the largest asteroids and comets.
Figure 44. Deflection DV imposed on 10-m-diameter asteroid.
100 m/8.7 gm/m
2
316 m/8.7 gm/m
2
1,000 m/8.7 gm/m
2
3,162 m/8.7 gm/m
2
100 m/6.0 gm/m
2
316 m/6.0 gm/m
2
1,000 m/6.0 gm/m
2
3,162 m/6.0 gm/m
2
100 m/3.3 gm/m
2
316 m/3.3 gm/m
2
1,000 m/3.3 gm/m
2
3,162 m/3.3 gm/m
2
100 m/0.6 gm/m
2
316 m/0.6 gm/m
2
1,000 m/0.6 gm/m
2
3,162 m/0.6 gm/m
2
100,000
10,000
1,000
100
10
1
0.1
10 100 1,000 10,000
Time (days)
V

(
m
/
s
)

73
Unfortunately, there are several complications that further exacerbate the sail’s inability to deflect
the most threatening objects. As indicated in table 19, any clouds of dust around the object would degrade
the sail material. Most concepts envision the sail somehow encircling the object with a line or net. The
force from the sail would then be transmitted through the line to the object. Obviously, this idea would
have great difficulty if the object were rotating. The sail could be “rolled up” onto the asteroid or comet.
For similar reasons, the sail would not be able to affect more than one fragment of a comet at a time.
Finally, the sizes indicated for these objects are such that it is difficult to see how more than one sail could
operate at a time.
Figure 45. Deflection DV imposed by 6 gm/m
2
solar sail with varying areas
on varying diameter asteroids.
Table 19. Qualitative considerations for threat mitigation using a solar sail.
10
1×10
–10
1×10
–9
1×10
–8
1×10
–7
1×10
–6
1×10
–5
1×10
–4
1×10
–3
1×10
–2
1×10
–1
1
10
100
100 1,000 10,000
Time (days)
V

(
m
/
s
)

100-m-Diameter Asteroid
1,000-m-Diameter Asteroid
10,000-m-Diameter Asteroid
Solid – 100-m-Side Square Sail
Dash – 316-m-Side Square Sail
Dot – 1,000-m-Side Square Sail
Dash-Dot – 3,162-m-Side Square Sail
First-Order Qualitative Considerations
Second-Order Qualitative Considerations
Susceptibility to dust cloud
Ability to handle target rotation
Requires landing on target
Usefulness on fragmented body
Swarm option
Medium
Low
Yes
Low
Low
Medium
Low
Usefulness as weapon
Perceived safety
Synergy with other NASA missions
Manned missions
Robotic missions
Resource utilization missions
Costs
Development
Deployment
Low
High
Low
High
High
74
5.4 Solar Collector
The solar collector is another nonnuclear option considered for use as a threat mitigation technique.
It also offers synergy by using the same technique for both the outbound and inbound legs of a rendezvous-
deflection mission. For deflection, the collected light from the curved sail is directed onto the asteroid by
the secondary mirror, shown schematically in figure 46. It is noted that this concept excludes the propul-
sion system-needed stationkeeping. The added weight of such a system would substantially degrade the
performance of the solar collector on the outbound leg and is most probably superfluous in view of the
forces experienced on the inbound leg.
Figure 46. Schematic of solar collector.
The beam intensity incident on the object is assumed to be equal to that incident on the collector.
Therefore, no losses are included for absorption or the effect of a noncolumninated beam. This is a little
optimistic, but losses must be strongly avoided or the design temperatures required for the collector and
mirror will become untenable. The beam intensity incident on the object is then given by
I r P r c ( ) = ( ) ¥ .
(47)
The pressure as a function of radius was defined in section 5.3 and c is simply the speed of light in a
vacuum. In all respects, the analysis for the collector is the same as in section 5.3. All that is required is to
determine the force imposed on the asteroid or comet by the incident beam. The following discussion deals
with this.
32
The amount of mass ejected from the object is calculated as
˙ , m
I r A
H
v
e
sail
vap
e
=
( )
+
2
2
(48)
Station-Keeping
Propulsion System
Primary Mirror,
1- to 10-km Diameter
Secondary
Mirror
Sunlight
Vaporized
Rock Jet
Asteroid
Deflection Velocity
Heated Spot,
10–100 m
75
where H
vap
is the heat of vaporization (3 MJ/kg for water ice) and v
e
is the velocity of the ejecta. As a first
approximation, the ejecta is assumed to come out at its sonic velocity (ª1 km/s). From this, the force
exerted on the object by the ejecta can be determined:
F m v
e e
= ¢ b ˙ ,
(49)
where ¢ b is a gas expansion factor—assumed to be 0.5—which represents a hemispherical expansion.
For the purpose of this study, the orbital radius of the vehicle from the Sun was assumed to be
1.5 au. This is an assumption based on the average distance during the time period when the collector is
acting on the object. Additionally, 500 kg of payload was assumed to account for avionics and other
systems. The power of the incident beam at 1.5 au is nearly 5 GW for a 100-m-diameter collector.
Obviously, a secondary mirror reflecting this much power will be a major design challenge.
A parametric of DV imposed on an asteroid or comet similar to that for the solar sail above can be
created (fig. 47). Here, a solar collector with a diameter of 100 m is acting on asteroids with diameters of
10, 100, 1,000, and 10,000 m, respectively. The collector can obviously handle asteroids of up to 100 m
without a problem. The intensity of the beam focused on the asteroid will scale with the square of the
diameter of the solar sail. Thus, there is the expectation that a 1-km solar collector may be able to deflect
asteroids of 1-km diameter. There is no obvious path for solar collectors to deflect asteroids of larger
diameter.
Figure 47. DV imposed on varying asteroid sizes by 100-m-diameter solar collector.
1,000,000
100,000
10,000
1,000
100
10
1
10 100 1,000 10,000
0.1
0.01
0.001
0.0001
0.00001
0.000001
10 m
100 m
1,000 m
10,000 m
Time (days)
V

(
m
/
s
)

76
Any dust or ice clouds surrounding the object will degrade the beam from the solar collector. It will
be difficult to match rotation with a rotating object. In this case, the incident beam will strike not a single
site, but instead, a band around the rotating object. The need to constantly heat new portions of the object—
or at least portions not heated since the previous revolution—will most likely result in a lower temperature
and therefore slower ejecta. This will result in less thrust than could be achieved on a nonrotation body. For
a fragmented body, although the beam can be directed on different components, only one can be pushed at
a time. Finally, due to the size of the solar collector, it would be difficult for more than one to operate on an
incoming object at a time. See table 20 for solar collector qualitative considerations.
Table 20. Qualitative considerations for threat mitigation using a solar collector.
5.5 Magnetic Flux Compression
Of the many asteroid deflection approaches researched for this TP, none used electromagnetic forces.
By their very nature, electromagnetic forces could offer an advantage over more conventional explosive or
kinetic impact forces. With explosives and kinetic impacts, the momentum is applied through either a
pressure force from the highly random kinetic thermal energy in a gas or at a localized impact area. Both of
these approaches loose a significant amount of energy by increasing the temperature of the asteroid surface
and possibly breaking the asteroid into pieces. Many researchers agree that breaking the object into pieces
could make the problem more difficult and would be an undesirable effect of any deflection technique,
except in the case of complete pulverization. With electromagnetic forces, a large amount of energy can be
converted into a directed Lorentz force without producing a lot of heat or fracturing the asteroid into
pieces. A rapidly changing magnetic field could be used to generate large countercurrents on the surface of
the asteroid. These surface countercurrents would in turn produce a secondary magnetic field that would
repel away from the original magnetic field and exert a repulsive force on the asteroid surface.
Any conductor in the presence of a changing magnetic field will produce an electric current propor-
tional to the magnitude of the magnetic field and its rate of change. This current is due to the induced
electric field produced in the space that the conductor occupies, and is defined by Lenz’s Law:
First-Order Qualitative Considerations
Second-Order Qualitative Considerations
Mitigation by dust cloud
Ability to handle target rotation
Requires landing on target
Usefulness on fragmented body
Swarm option
Medium
Medium
No
Medium
Low
Medium
Low
Usefulness as weapon
Perceived safety
Synergy with other NASA missions
Manned missions
Robotic missions
Resource utilization missions
Costs
Development
Deployment
Low
High
Low
High
High
77
e f = d dt
B
,
(50)
where e is the induced electromagnetic field, f
B
is the magnetic flux, and t is time. Knowing the magnitude
of the electromagnetic field and the electrical resistance of the conductor (R), this resulting current can be
calculated using Ohm’s Law:
I R =e .
(51)
Electrical resistance can be further defined by the resistivity parameter (r):
R L A = r .
(52)
Resistivity has units of ohms/meter; L is the length of the conductor, and A is the cross-sectional area of the
conducting path. Resistivity has a specific value for different materials and also changes with material
temperature. The inverse of resistivity is known as conductivity (s). Knowing the geometry and material
of a certain conductor, one can calculate with some certainty the electrical resistance.
Magnetic flux is simply defined as the product of the magnetic flux density (B) and the surface area
(A) that the flux lines pass through:
f
B
B A = ¥ .
(53)
If a particle of charge (q) moves with velocity (v) in the presence of both a magnetic field (B) and an
electric field (E), it will experience a force (F) defined by the Lorentz equation:
F q E v B = + ¥ ( ) .
(54)
If the electric field and magnetic field are perpendicular, the force will be applied in a direction perpendicu-
lar to both fields. For the case where charge is traveling through a conductor, the force is defined by the
following equation:
F V J B = ¥
( )
.
(55)
For a conductor with volume (V) and a current density (J) passing through it, both the flowing electrons
and fixed ions in the conductor will experience a force (F) in the same direction due to the interaction of the
electric field and the magnetic field (B). The magnitude of the force is dependent on the strength of the
magnetic field, the rate at which electrical current flows, and the conductivity of the conductor; i.e., the
poorer the conductor, the smaller the resulting force.
Since most asteroids are composed of carbonaceous chondrite materials, which are poor conduc-
tors, it may be necessary to use a target disk with a good conductivity value. This disk would mate up to the
asteroid and serve as a pusher plate. Unlike a pusher plate used for pulsed nuclear or conventional
78
explosives, this disk could be a relatively thin, lightweight metallic foil of sufficient area to be effective.
The reason that this target disk can be much thinner and less massive than an explosion pusher plate is the
difference between the methods by which the force is produced. With the pusher plate, the force is applied
via the shock wave that is produced from the large amount of heat energy released from the explosives.
Without the pusher plate, the explosives would ablate a surface layer off the asteroid and the pressure
would be applied to a more localized area that may not be able to withstand the resulting stress. This could
lead to cracks or pitting. The thick, strong pusher plate acts as a shield similar to a bulletproof vest. It has
sufficient area to spread out the force and sufficient thickness to withstand the localized stress of the
impact/explosion. With the target foil, the force is produced by a Lorentz reaction between currents and
magnetic fields. This force is more evenly distributed through the entire target (like a sail) and the amount
of heat flux released into the surface is greatly reduced. Energy is not wasted in vaporizing or demolishing
the asteroid material. The thin foil could be efficiently packaged and only deployed once the asteroid is
encountered. The target disk need only be thick enough to absorb the magnetic flux. Depending on the rate
at which the flux changes, the disk could be very thin. The thickness necessary to absorb the flux is calcu-
lated using the following equation:
d psm = t
pulse 0
. (56)
Skin depth (d) is a function of the pulse time (t
pulse
), the conductivity (s), and the permeability of free
space (m
0
). As long as the foil thickness is greater than d, the magnetic energy will be coupled completely
into the foil. The faster the pulse, the thinner the foil can be.
Since most magnetic fields are curved, there will be a limit to the diameter of the target disk beyond
which the field lines will reverse direction and produce forces acting in the wrong direction, as shown in
figure 48. This would be very undesirable. The disk diameter is dependent on the distance from the mag-
netic field source to the target disk. This diameter must be optimized, since the critical disk diameter
increases as the source moves farther from the asteroid (more surface area for the flux to be coupled to and
easier to achieve) and decreases as the source gets closer toward the asteroid (stronger field strength, but
may require more difficult maneuvering).
Figure 48. Critical disk diameter.
B
B
B
B
Force
79
A common source of magnetic flux is a solenoid (coil) connected to a power supply. A solenoid
produces axisymmetric, curved, nonuniform magnetic field lines with a similar field line profile to that of
a bar magnet. The curved field lines may be divided into two component vectors—one normal to the target
surface and one parallel to the target surface. The component normal to the target is responsible for induc-
ing a current. The component parallel to the target surface is responsible for producing the Lorentz force.
To accurately sum up the force produced by each varying field line, the Biot-Savart Law can be used:
dB
I d r
r
=
¥ m
p
0
2
4
l
.
(57)
Using Biot-Savart, the magnetic field strength (flux density) (B) is calculated from knowledge of the coil
path vector

( ) l and the distance from the coil to the target (r). By summing up the length of the conductor
path, the total field strength in a given region of space can be calculated. This magnetic field has energy
proportional to the volume within the solenoid and the magnitude of the magnetic field:
E V
B
=
2
0
2m
.
(58)
Using equations (57) and (58), a parametric FORTRAN program was written to calculate the force that
would be produced by an arbitrary solenoid design. If a 1,000-turn solenoid with a radius and length of
1 m, located at a standoff distance of 100 m, was supplied a current of 1 MA ramped up over a period of
10 ms, it would exert a force of over 9 million lb on a 100-m-radius target disk. That results in a very
sizeable impulse. Although not impossible to produce, the logistics are not favorable. The inductance value
of the solenoid in question is ª3.95 H. This is roughly six orders of magnitude larger than most solenoids
used in common electrical components. To force 390 kJ of magnetic energy into the solenoid in 10 ms
would require over 4 billion V of electricity at a power of over 39 MW. This current and voltage level are
not impossible to achieve, but they would require a very large power supply. Since anything more than one
shot on the coil would be impractical (one would need a restoring force on the coil), it is not an economic
solution. Current technology lightweight, high-power, disposable power supplies; i.e., batteries, capaci-
tors, magnetic flux compression generators, etc., are incapable of providing this performance. Originally,
this system was envisioned to use a magnetic flux compression generator.
A magnetic flux compression generator is an ingenious method for converting explosive energy
into electrical energy by compressing a magnetic field. A solenoid containing a magnetic field has an
energy associated with the volume of the solenoid and the magnetic field strength. From equation (58),
it can be seen that the magnetic field strength plays a major role in the amount of energy within the sole-
noid. To increase the magnetic field, the current running through the coil must be increased. Another way
of doing this is by compressing the field lines within the solenoid. By reducing the volume of the solenoid,
the field lines are compressed and their strength is increased. Although the volume is
decreased, since the energy is related to the square of the magnetic field strength, the energy is greatly
increased. A convenient way of decreasing the volume is by placing a conducting sleeve filled with explo-
sives along the centerline of the solenoid as shown in figure 49. This conducting sleeve is referred to as the
armature of the flux compression generator and the surrounding coil is referred to as the stator. The stator
80
Figure 49. Magnetic flux compression generato.
33
is energized with a current pulse, usually supplied by a capacitor bank, and a strong magnetic field is
created within. An instant later, the explosives within the armature are detonated from one side. This
explosion propagates along the armature, expanding the conducting sleeve as it goes. Because the armature
sleeve is conducting as it moves through the magnetic flux within the solenoid, a countercurrent is
induced. This countercurrent creates a magnetic field that repels the solenoid field and compresses it
into the decreasing volume caused by the explosion. The magnetic field and resulting current at the end of
the process are extremely large.
Using this method, currents in the mega-ampere range can be produced for a very brief period of
time, usually on the order of nanoseconds. The flux compression generator can be coupled to a load or can
be uncoupled to produce a very large electromagnetic pulse (EMP). The latter configuration is sometimes
referred to as an E-bomb, since its EMP is often great enough to destroy nearby electrical equipment. The
uncoupled configuration is shown in figure 50.
Figure 50. E-bomb magnetic flux compression generator.
33
High Explosive
Stator Winding
Armature Cylinder
81
Unfortunately, magnetic flux compression generators provide peak powers in nanoseconds instead
of milliseconds and their energy outputs are inadequate for driving the coil proposed for the asteroid
deflection system. There is, however, another aspect of the magnetic flux compression generator that may
have been overlooked at the beginning of this study and has not subsequently been explored. Although a
flux compression generator would never be able to produce an EMP great enough to reproduce the perfor-
mance of the proposed coil, another device may be able to do so.
Nuclear weapons detonated at high altitudes in the Earth’s atmosphere produce a gigantic EMP.
This was demonstrated by accident when a nuclear test conducted high in the atmosphere created an EMP
that completely blacked out Hawaii and other parts of the United States. This EMP traveled down to the
ground, disrupting and destroying electrical devices, batteries, generators, and wires. Since an EMP is
simply a large traveling magnetic flux, it can couple with conductors in the same way in which radio waves
couple with antennas. The major difference between the magnetic flux compression generator and a nuclear
bomb is the way in which the EMP is produced.
When a nuclear weapon is exploded, a large amount of radiation in the form of gamma rays and
x rays are released. These high-energy photons can be absorbed by elements with low atomic numbers,
such as nitrogen and oxygen—two major components of the Earth’s atmosphere. When these high-energy
gamma photons strike an atom, they can knock loose electrons and send them flying away at great speed.
This phenomenon is named the Compton Effect after its discoverer, A.H. Compton. These moving elec-
trons serve as a large electrical current and produce a very large magnetic field for a brief period of time.
This huge EMP can be coupled into conducting materials as electrical current. In the case of high-altitude
nuclear tests, the atmosphere acts as a transformer, converting electromagnetic radiation (gamma rays) into
lower frequency electromagnetic waves (radio waves). The amount of energy is conserved—if one ignores
loss mechanisms—but its form is changed.
Perhaps a device could be constructed to use the compact and abundant energy released during a
nuclear explosion to create an EMP suitable to drive currents in the foil target disk described earlier. To the
author’s knowledge, this idea has not yet been explored.
Except for the nuclear configuration—not yet explored—the magnetic flux driver concept is rela-
tively safe and simple. It could be constructed from existing technology rather inexpensively. This concept
could be very effective on fragmented and rotating bodies that have a high content of ferromagnetic mate-
rials, or an existing magnetic field. Nonmagnetic bodies would require use of a target disk, which would be
less effective on rotating and fragmented bodies. Dust should not be a problem, since it would not block
electromagnetic radio waves. The nature of this option would almost mandate that a swarm approach be
adopted. See table 21 for quantitative considerations.
5.6 Mass Driver
The term mass driver is used to describe a variety of electromagnetic acceleration systems; all use
electrical power to accelerate a bucket containing an inert mass to very high speed. The bucket is designed
either without a top cover or with one that can be opened in flight, so that as it decelerates, the contents
leaves the bucket and continues on at high speed.
82
Although there are several different types of mass drivers,
34
all make use of magnetic attraction
and repulsion between the moving (bucket) and the stationary parts of the device. A distinct type is distin-
guished by the specific shape and location of the magnets.
Although initially proposed
34
for space launch applications, mass drivers are also applicable to
planetary body maneuvering. A special deployment vessel would transport the entire mass driver system
out to the asteroid and install it on the surface. The system consists of the mass driver itself along with its
electrical power source, either a nuclear reactor or a solar array, and its thermal control system. The basic
rationale behind the use of a mass driver in this application is that it would make use of the asteroid surface
material as reaction mass; i.e., expellant. As this material is already present, it is essentially free propellant,
so even a system with a modest exhaust velocity; i.e., expellant speed, could offer good overall performance.
Two options are available regarding the provision of reaction mass:
(1) The deployment team could mine the entire stock of reaction mass before departure and place it
into a suitable storage and dispensing system from which it could be fed to the mass driver as needed.
(2) An automated mining system could be deployed on the asteroid to operate continuously,
together with the mass driver, after the deployment team has departed.
Option (2) is clearly more complex and raises additional questions over reliability and mainte-
nance.
The intent of this study was to establish the general utility of mass driver systems for planetary
maneuvering, not to determine which specific design offers the best performance, although this might be a
suitable subject for a follow-on study. Accordingly, a relatively straightforward design was chosen to
Table 21. Qualitative considerations for threat mitigation using magnetic flux compression.
First-Order Qualitative Considerations
Second-Order Qualitative Considerations
Sensitivity to dust cloud
Ability to handle target rotation
Requires landing on target
Usefulness on fragmented body
Swarm option
Low
Low
Impact
Medium
High
Medium
Low
Usefulness as weapon
Perceived safety
Synergy with other NASA missions
Manned missions
Robotic missions
Resource utilization missions
Costs
Development
Deployment
Low
High
Low
Low
High
83
represent the entire spectrum of designs. The principal concern was to select a design that was relatively
straightforward to model. This turned out to be the design with the greatest geometrical simplicity and
symmetry—the coil gun.
The coil gun uses a series of equally spaced stationary solenoidal electromagnets (drive coils) to
accelerate a bucket carrying a number of smaller but identically spaced solenoids (bucket coils). The con-
cept is illustrated in figure 51.
Figure 51. Coil gun conceptual design.
In the case shown above, which is that modeled, the bucket is surrounded by and attached to four
current-carrying superconducting coils and is accelerated along the axis of the drive coils. Current flowing
in each drive coil produces a magnetic field, a component that exerts an axially directed motive force on
the bucket coils. The relative current flow directions of an arbitrary bucket coil and the nearest pair of drive
coils is shown in figure 52.
Figure 52. Relative current flow directions of bucket and immediately adjacent drive coils.
Bucket Coil Bucket Drive Coil
Direction
of Motion
S
x
Direction
of Motion
Drive
Coil
n
Drive
Coil
n+1
Bucket
Coil
84
Coils with currents flowing in the same sense attract; those flowing in opposite senses repel. Thus,
in the above example, the bucket coil is being repelled by the drive coil, which it has just passed (number n)
and is being attracted to the next one (n+1).
Clearly, for this concept to work, each drive coil must reverse its current flow direction as a bucket
coil passes through it, thus changing an attractive force for an approaching coil into a repulsive force for a
receding coil. The current profile for one drive coil is shown in figure 53. The drive coil current is shown
as a function of the bucket coil position (S being the distance between adjacent drive coils). When the
bucket coil is at –S; i.e., just passing through the center of the preceding drive coil, the drive coil current
begins to increase, producing an attractive force. This current reaches a maximum when the bucket coil is
at –S/2; i.e., halfway between the preceding drive coil and the coil under consideration, and subsequently
declines. It reaches zero as the bucket coil arrives and subsequently goes through the negative portion of its
cycle, repelling the receding bucket coil.
Figure 53. Relative current flow directions of bucket and nearest drive coils.
The design proposed here has four adjacent bucket coils separated by a distance S; i.e., identical to
the spacing between adjacent drive coils. This means that, as bucket coil No. 1 passes through a drive coil,
bucket coil No. 2 passes through the preceding drive coil. This, taken together with the drive coil current
profile shown above, dictates that successive bucket coils must have their current flowing in opposite
directions; i.e., exactly half a cycle out of phase.
The current directions for both drive and bucket coils at any instant is illustrated in figure 54. Note
that the current flowing in each bucket coil does not change its flow direction, although its magnitude
varies—due to induction—during the transit between adjacent drive coils.
Drive Coil
Current
–S
–S/2
S/2 S
x
85
To minimize unnecessary energy losses, it is anticipated that only those drive coils that immedi-
ately surround the bucket, together with a small number ahead and behind, will be energized at any instant.
A minimum of five drive coils must be energized in order to cover the bucket itself. This can be seen
clearly in figure 54. In reality, the number energized will depend upon the speed with which the bucket is
moving and limitations imposed by current-switching technology.
It should be noted that although a bucket coil receives an accelerating force from the closest pair of
drive coils, it experiences alternately retarding and attractive forces from each more distant pair. Fortu-
nately, these more distant coils exert smaller forces than the nearest ones, illustrated in figure 55.
Figure 54. Relative current flow directions of bucket and surrounding drive coils.
Figure 55. Relative current flow directions of bucket and nearby drive coils.
Bucket
Bucket Coil
(Arrow Indicates Direction
of Current Flow)
Drive Coil
(Arrow Indicates Direction
of Current Flow)
Direction
of Motion
Drive Coils Which Accelerate Bucket Coil
Drive Coils Which Retard Bucket Coil
Direction
of Motion
86
When the bucket and its contents have been accelerated to their maximum speed, the bucket is then
decelerated by a set of braking coils. As the bucket decelerates, the nondecelerated expellant separates and
continues at the maximum speed until it exits the mass driver. This is illustrated in figure 56, which shows
three different times during an operational cycle.
Figure 56. Mass driver operation.
The braking process is essentially the same as the acceleration process. The braking coils are of the
same design as the drive coils. The coils are successively energized as the bucket coils pass them. The only
difference is that during braking, the braking coils are energized exactly half a period out of phase with
their drive equivalents. This means that the forces are in the reverse direction to those
encountered during the drive section. Figure 57 shows the disposition of currents during braking. This
should be compared with figure 54, the equivalent diagram for the drive coil portion of the mass driver.
Figure 57. Relative current flow directions of bucket and braking coils.
Drive Coils
Bucket Accelerating
Bucket Acceleration Ends—Deceleration Commences
Braking Coils
Bucket Decelerates—Contents Emerges
and Continues at Maximum Speed
Bucket Braking Coil
Bucket Coil
Direction
of Motion
87
In addition to providing a means of decelerating the bucket, the braking coils also provide a means
of converting its kinetic energy back into electrical energy, thus making it available for use during the next
cycle. Just as the drive coil circuits lose electrical energy as they accelerate the bucket, so the braking coil
circuits gain electrical energy as they decelerate the bucket. A simple schematic of the entire mass driver
system is given in figure 58.
Figure 58. Mass driver system schematic.
The mass driver unit and return rail provide a complete closed circuit around which the bucket
travels. After rapid acceleration and deceleration along the mass driver (coil gun) itself, the bucket moves
at a relatively slow speed along the return rail. It halts at the expellant loading system in order to take on a
fresh load of reaction mass and then proceeds on to the start of the mass driver.
While the loading process is taking place, the bucket is placed in contact with a cold plate whose
function is to extract the waste heat generated within the bucket coils during the previous cycle.
Although superconducting coils are used throughout the design, there is still some heat generated during
the acceleration and deceleration processes that must be extracted to prevent the superconductor from
rising to its critical temperature. The cooling system must also remove waste heat from the superconduct-
ing stationary coils, both drive and braking, between successive operational cycles. Connections between
the thermal management system and both the mass driver and the bucket coil cooling system are shown
above. An array of thermal radiators disposes of the waste heat.
Bucket Return Rail
(Arrows Indicate Direction
of Bucket Travel)
Expellant Storage
and Loading System
Bucket Coil
Cooling System
Power Line
Thermal Radiators
Mass Driver
Coolant Line
Thermal
Management
System
Electrical
Power
Management
System
Solar Array
88
There appear to be only two viable power source options currently available: a nuclear fission
reactor and a solar array. For bodies sufficiently distant from the Sun, only the nuclear option is likely to be
practical; however, for the purpose of this exercise, a solar array has been selected. This choice has been
made for several reasons:
(1) Mass and performance modeling for a solar array are somewhat simpler than for a fission
reactor, making it a more pragmatic choice regarding the analytical effort involved.
(2) A solar array power system currently poses fewer concerns over reliability, an important con-
cern given the likely need to operate the mass driver, untended, over a long period of time.
(3) A solar power system is capable of operating even after sustaining damage. Parts of the array,
which are nonfunctional, can be isolated and the system can continue to work, albeit at a lower power level.
(4) There would likely be some safety concerns associated with the placement of a fission reactor
on a body; i.e., at least initially, on a collision course with the Earth. Obviously, if the mass driver system
were to perform its job correctly, there would be no problems. But in the event, say, of a malfunction in the
mass driver system, it might be necessary to fragment the body. If so, the presence of a nuclear reactor on
its surface could significantly complicate the operation.
Against these considerations, it must be noted that the useful electrical power available per unit
array surface area will be very low. Apart from being further from the Sun than is normal for solar power
application, there will be additional causes of performance degradation. The array will almost certainly
have to be simply laid on the surface of the asteroid and is unlikely to be normal to the incident
radiation. If the asteroid is rotating, which is almost certain to be the case, additional losses will result.
Finally, the process of preparing the expellant, regardless of whether it is done by continuous mining or by
deployment vessel, will likely produce a large amount of dust and debris. The very low accelerations, both
gravitational and of the body itself, will permit this dust and debris to remain in the vicinity of the mass
driver for a significant time, allowing some of it to settle on the array, further reducing its effectiveness.
Some details of the simple model used for the solar array are given in appendix B.
For the purpose of this study, it is assumed that expellant is prepared by the deployment vessel
before its departure. The only other option—continuous automated mining—requires a detailed study in its
own right. The problems of operating mechanically complex equipment, such as that needed to extract and
render asteroidal material, in an ultra-low-gravity environment, are beyond the scope of the current work.
Although it is not the authors’ intent to prejudge the results of this study, the following scenario is
proposed for the application of mass drivers to planetary body maneuvering. A crewed deployment vessel
conducts an extended mission in the vicinity of the asteroid belt. Its targets are a small number of asteroids
whose orbits pose a threat to the Earth in a time period of, say, 10 to 50 yr into the future. At each of these
asteroids, a mass driver is deployed and mining equipment, carried on the deployment vessel, is used to
prepare a stock of reaction mass. At each asteroid, the mass driver commences operation and the deploy-
ment vessel departs for its next target. If necessary, subsequent visits can be arranged to conduct any
needed maintenance and repair. Additional reaction mass can also be produced and stored during these
visits.
89
The mass driver system deployed under this scenario is shown in figures 59 and 60. Figure 59
shows the entire system, which is dominated by the large solar array and thermal radiators. Figure 60
identifies the actual mass driver (coil gun) and associated equipment.
Figure 59. Mass driver system view.
Figure 60. Main components of the mass driver.
Thermal Radiators
Solar Panels
Mass Driver
Mass Driver System
Mass Driver System
Thermal Radiators
Solar Panels
Mass Driver
Major Components
of the Mass
Driver System
Cooling
System
Radiator
Panel
Bucket Loading
and Cooling Facility
Solar Panel
Cable From
Solar Panels
Power Management
and Conditioning System Power
Cable
Coolant
Line
M
a
s
s

D
r
iv
e
r
Bucket
Return Rail
B
u
c
k
e
t
R
e
t
u
r
n
R
a
il
Bucket
Expellant
Storage
Facility
90
Figures 61–66 illustrate system operations. Together, they show the system in operation through a
complete cycle—the discharge of one bucket load of expellant.
Figure 61. System following discharge—bucket returns for reloading.
Figure 62. Bucket about to be reloaded and cooled prior to next discharge.
Bucket Returns Slowly
to Bucket Loading
and Cooling Facility
Bucket Enters
Bucket Loading
and Cooling Facility
91
Figure 63. Loaded bucket about to enter mass driver.
Figure 64. Bucket containing expellant under acceleration within mass driver.
Loaded Bucket
Returns to Mass
Driver Entrance
92
Figure 65. Discharge in progress—bucket is decelerating while expellant mass
exits mass driver at high speed.
Figure 66. Following discharge—decelerated bucket exits mass driver and joins return system.
Expellant Exits
Mass Driver at
High Speed
Bucket Slowly
Exits Mass Driver
and Attaches
to Return System
93
While there may be additional situations under which mass driver systems could be employed for
planetary maneuvering, it is the above scenario that formed the basis of the analysis carried out during the
present study. A simple performance and mass model for the mass driver and its principal supporting
systems is presented in appendix C.
Table 22 lists the qualitative considerations that apply to the use of mass drivers for planetary body
maneuvering. It is easier to conceive application of this technique to asteroid maneuvering than to comet
maneuvering. A comet nucleus is likely to be both fragmented and dusty. A mass driver clearly cannot
move more than a single body and will probably suffer some damage if operated in a dusty environment
due to secondary impacts. Even if installed on a single asteroid in a nondusty environment, rotation will
reduce the mass driver’s efficiency. It will only be able to operate when oriented in the proper direction. To
improve the technique’s efficiency, two mass drivers could be deployed at appropriate “balanced” loca-
tions. These could first be used as necessary to counteract the asteroid rotation and then used together to
provide a torque-free net force.
Table 22. Qualitative considerations for threat mitigation using the mass driver.
The mass driver has a relatively high inherent safety rating against these potential complexities;
this is largely due to our selection of a solar array to provide primary power. Use of a nuclear reactor,
although in many ways more appropriate for the outer solar system, would reduce the safety rating. High-
speed projectiles could cause significant damage, so a mass driver has some intrinsic weapons capability,
but it would clearly not be a weapon of choice in any realistic situation.
5.7 Kinetic Deflection
Although potentially threatening planetary bodies will all be very massive—even a
30-m-diameter sphere of water ice would have a mass of ª15,000 mT—in many cases, only a very
slight orbital perturbation would be necessary to render them harmless. The energy required to carry out
such perturbations, while still large, does not necessarily have to be supplied by an explosive event, whether
First-Order Qualitative Considerations
Second-Order Qualitative Considerations
Sensitivity to dust cloud
Ability to handle target rotation
Requires landing on target
Usefulness on fragmented body
Swarm option
Medium
Medium
Yes
Low
Medium
High
High
Usefulness as weapon
Perceived safety
Synergy with other NASA missions
Manned missions
Robotic missions
Resource utilization missions
Costs
Development
Deployment
Medium
High
Low
Low
High
94
conventional or nuclear. Even a relatively small interceptor craft, if accelerated to a sufficiently high
velocity, possesses a formidable amount of kinetic energy that may be adequate to produce the desired
deflection.
This raises the possibility of using a relatively small, inert interceptor body, raised to high speed, as
a deflection tool. This technique offers the attraction that, once the interceptor’s onboard propulsion sys-
tem has exhausted its propellant, the remaining dry mass still fulfills an important function by virtue of the
kinetic energy it carries.
There are clearly major issues relating to guidance of the interceptor, particularly during its termi-
nal mission phase. With closing speeds perhaps in excess of 30 km/s, maneuvering capabilities will be
limited and reaction times critical. For the purpose of this discussion, issues of guidance and terminal
maneuvering will not be addressed. Clearly, some of the technologies being developed for terrestrial ballis-
tic missile defense are likely to be applicable. Clearly, this subject is not appropriate for an open report.
This section contains the derivation of a simple model for use in evaluating the effectiveness of the
kinetic deflection technique. The model is comprised of two distinct parts: (1) A largely analytical impact
and momentum exchange model,
35
and (2) the issue of planetary body fragmentation.
36
Figure 67 shows a general situation in which an interceptor is about to impact a planetary body. The
velocity vectors of the body and the interceptor can always define a plane, so without loss of generality, the
analysis can be presented in two dimensions. A modified inelastic model is given below.
Although basically inelastic, allowance is made for the ejection of debris from the collision site.
Figure 67. Interception geometry.
Planetary Body
Interceptor
Planetary Body
Velocity Vector
x
y
θ
95
An x-y coordinate system is established as shown and then a momentum conservation equation is
established as follows.
For motion along the x axis,
M u M v M M M u u P
i i i ej x ej NEO NEO
- = + -
( )
- ( ) + cos cos , q q D
(59)
where
M
NEO
= mass of the NEO before impact of interceptor
u = initial speed of NEO
M
i
= mass of interceptor
v
i
= initial speed of interceptor
q = angle between NEO and interceptor velocity vectors (see fig. 67)
M
ej
= mass of material ejected due to impact
Du
x
= change in x component of NEO velocity due to impact
P
ej
= net momentum of ejected material.
Note that the impact and ejection processes are assumed to follow the simple pattern shown in figure 68.
Figure 68. Impact and ejection geometry.
The ejecta material is assumed to emerge from the impact site in the form of a uniform cone; in
other words, it all emerges at the same angle (j) to, and evenly distributed around, the vertical axis.
Assuming that
(60)
Ejecta Cone
ϕ
M M M
NEO i ej
>> ,
,
96
one can rewrite equation (59) to give
Du
M v
M
P
M v
x
i i
ej
i i
= +
Ê
Ë
Á
ˆ
¯
˜
NEO
1 cos . q
(61)
For motion along the y axis,
M v M M M u P
i i i ej y ej
sin sin , q q = + -
( )
-
NEO
D
(62)
and hence,
Du
M v
M
P
M v
y
i i
ej
i i
= +
Ê
Ë
Á
ˆ
¯
˜
NEO
1 sin . q
(63)
Equations (61) and (63) constitute the momentum conservation part of this model. Note that if one
neglects the ejecta momentum terms, the model becomes completely inelastic.
If one denotes the cumulative mass of ejecta traveling faster than v by M(>v), then
35
M v R
Y
v
c
> ( ) = 0 05
3
2
. . r
r
(64)
This equation is based on data obtained from experiments that measured the ejecta produced by impact into
materials of differing hardness. The various quantities are defined as follows:
R
c
= crater radius
r = NEO material mass density
v = ejecta speed
Y = NEO material strength.
Defining v
min
as the minimum speed of ejected material, the total mass of ejecta can clearly be
written as
M R
Y
v
ej c
= 0 05
3
2
. .
min
r
r
(65)
This expression can be used to eliminate Y and R
c
from equation (64) to give
97
M v
v
v
M
ej
> ( ) =
Ê
Ë
Á
ˆ
¯
˜
min
.
2
(66)
Now, consider the small element of ejecta mass which emerges with speed in the range v Æ v + dv. This
mass is given by
M v M v v
v
v
M
v
v v
M
ej ej
> ( ) - > + ( ) =
Ê
Ë
Á
ˆ
¯
˜
-
+
Ê
Ë
Á
ˆ
¯
˜
d
d
min min
,
2 2
(67)
which can be expanded, retaining only first-order terms in dv, to give
d d M v
v
v
M v
ej
>
( )
= 2
2
3
min
.
(68)
Noting our assumption that the ejecta is assumed to all emerge at the same angle (see fig. 68), the following
expression can be written for the vertical component of momentum (dp) of this small portion of ejecta:
d
d
j p
v M v
v
v
ej
=
2
2
3
min
cos .
(69)
This enables one to derive the vertical component of momentum for the entire ejecta mass (P
ej
) as follows:
P
v M
v
dv
ej
ej
v
=

Ú
2
2
2
min
cos
,
min
j
(70)
and hence,
P v M
ej ej
= 2
min
cos . j
(71)
Next, an equation that relates the mass of ejecta to the radius of the crater is used.
35
It is derived on the
assumption—based on empirical evidence—that the crater is one half of an oblate spheroid with a depth
of 0.4 R
c
:
M R
R
ej c
c
=
4
3
0 4
2
2
p
r
.
,
(72)
which can be rewritten as
M R
ej c
= 0 8378
3
. , r (73)
98
from which
R
M
c
ej 3
1 1937
=
.
.
r
(74)
M
ej
can be eliminated from this expression using equation (65) to give
v
Y
min
.
. , =
Ê
Ë
Á
ˆ
¯
˜
0 2443
0 5
r
(75)
and hence equation (71) can be rewritten as
P
Y
M
ej ej
=
Ê
Ë
Á
ˆ
¯
˜
0 4886
0 5
. cos .
.
r
j
(76)
One final equation is also available.
35
This was originally derived using data from a large
number of laboratory cratering experiments and links the total mass of ejecta to the relative speed of
collision (V) and interceptor mass as follows:
M M
V
Y
ej i
i
=
Ê
Ë
Á
ˆ
¯
˜
0 458
2
0 709
. ;
.
r
r
r
(77)
thus, M
ej
can be eliminated from equation (76) and rewritten as
P
M v
V
v
V
Y
ej
i i i i
=
Ê
Ë
Á
Á
ˆ
¯
˜
˜
0 2238
2
0 209
. cos .
.
r
r
r
j
(78)
This is the final quantity required to solve equations (61) and (63).
Note that the relative speed of collision is given by
V v u
i
= + cos . q
(79)
Equation (78) can be used to eliminate (P
ej
/M
i
¥v
i
) from equations (61) and (63), which can then be
used to determine the deflection that results from interceptor impact.
For this method of deflection, the question of fragmentation also needs to be considered. For a
relatively low-energy collision, although some material will be ejected from the impact site, the planetary
body will remain intact. At larger collisional energies, however, there is the possibility that the body may
99
split into a number of fragments as a result of the impact. It is important to be able to understand and model
the fragmentation process, as it may determine the usefulness of kinetic deflection as a threat mitigation
technique. An incoming body that is unexpectedly fragmented by a kinetic deflection interceptor may
subsequently prove very difficult to deal with if some of the fragments are very massive and still pose a
threat.
There is clearly very limited direct experimental data to draw on. Studies have been conducted to
determine the effect of kinetic energy on various NEO analogue materials.
36
The mechanics of collision
and fragmentation are reasonably well understood for impact speeds of up to ª8 km/s. Understanding is
less clear for impact speeds in the 8 to 15 km/s range. At speeds >15 km/s, there is very poor understanding,
which is unfortunate, given that this is the likely regime in which a kinetic deflection interceptor would
operate.
In general terms, it is understood that the initial impact produces a “hydrodynamically induced”
crater; in other words, the cratering process takes place while the solid material of the body has been
rendered fluid-like by the impact energy. The impact produces, on a timescale of between microseconds
and milliseconds, internal shock waves that propagate through the target on a timescale of between tens
and thousands of milliseconds. Depending upon the impact energy and the constitution of the target body,
the final outcome will range from velocity deflection to complete and catastrophic fragmentation.
The critical parameter in determining whether the target will fragment is the collisional energy per
unit mass of the target body. This is written as E
P
/M
T
, where E
P
is the collisional energy and M
T
is the
target mass. The following approximate criteria has been established for catastrophic fragmentation;
36
i.e.,
where the target is rendered into a large number of fragments, each much smaller than the original body:
• For E
P
/M
T
< 0.5 J/gm, the target is considered to be cratered but intact.
• For E
P
/M
T
> 0.5 J/gm, the target is considered to be catastrophically fragmented.
Some empirical data
36
are available that relate the collisional energy to the mass of the largest
fragment produced by the collision. The data are reproduced in table 23 for two different target materials,
showing the relative size of the largest fragment (M
L
/M
T
) at various values of (E
P
/M
T
). Note that M
L
denotes the mass of the largest fragment.
Table 23. Relative size of largest fragment at various collisional energies.
Target Material
Rock (basalt)
Ice
E
P
/M
T
(J/gm)
E
P
/M
T
(J/gm)
E
P
/M
T
(J/gm) M
L
/M
T
M
L
/M
T
M
L
/M
T
0.07
0.01
0.9
0.9
3
0.05
0.1
0.1
0.01
0.01
10
0.6
100
For facilitating computations, these data can be roughly curve fitted as shown in figure 69.
Figure 69. Fragmentation data and curve fit results.
The curve fit results can be summarized as follows:
ln ln ,
M
M
A
E
M
B
L
T
P
T
Ê
Ë
Á
ˆ
¯
˜
=
Ê
Ë
Á
ˆ
¯
˜
+
(80)
where
A = –0.8299 for basalt; –1.0843 for ice.
B = –2.1324 for basalt; –5.2697 for ice.
In conclusion, some indications
36
are noted that current analytical and semiempirical models may
underestimate the momentum change resulting from a high-velocity impact. Chemical reactions, taking
place in the fluid phase after impact, are probably responsible for this discrepancy. Hence, the above data
and model should be viewed as very preliminary. This area requires more work, including, if possible,
actual impact tests on a variety of planetary bodies.
–5
–5
–4
–3
–2
–1
0
1
–4 –3 –2 –1 0 1 2 3
Impact Kinetic Energy/Planetary Body Mass (ln)
L
a
r
g
e
s
t

F
r
a
g
m
e
n
t

M
a
s
s
/
P
l
a
n
e
t

B
o
d
y

M
a
s
s

(
l
n
)
Basalt
Ice
Linear (Basalt)
Linear (Ice)
y=–1.0843x – 5.2697
y=–0.8299x – 2.1324
101
Table 24. Qualitative considerations for threat mitigation using kinetic deflection.
Table 24 lists the qualitative considerations that affect the use of kinetic deflection. Small quantities
of dust surrounding the target body would not pose a major problem provided the interceptor is designed
appropriately. However, larger pieces of debris could pose a serious problem, destroying it before it
impacts on the target. Unlike some of the other techniques, kinetic deflection works well with a rotating
body. In fact, a rotating target is in some respects more desirable than a nonrotating target as it will be
easier to identify the location of its center of mass.
First-Order Qualitative Considerations
Second-Order Qualitative Considerations
Sensitivity to dust cloud
Ability to handle target rotation
Requires landing on target
Usefulness on fragmented body
Swarm option
Low
High
No
Medium
High
Low
Low
Usefulness as weapon
Perceived safety
Synergy with other NASA missions
Manned missions
Robotic missions
Resource utilization missions
Costs
Development
Deployment
High
High
Low
Low
High
Perceived safety is judged to be high because of the lack of any sort of nuclear element to this
technique. Technology developed for ballistic missile interception will probably be directly applicable
here, which should reduce development costs. Note that the closing speed for this application may be an
order of magnitude higher than for missile interception. Although potentially usable as a weapon itself, a
kinetic deflection system would be more likely to be employed in a defensive capacity than in an offensive
one.
In the present application, it is not anticipated that a single interceptor would be employed, but
rather a stream of them, each capable of some terminal-phase maneuvering. The later interceptors in
a stream will have been launched long before the initial interceptors impact, so they must be able to
maneuver in order to respond to changes in the body’s trajectory.
102
6. TRAJECTORY MODELING
6.1 Outbound
The outbound trajectory was modeled simply using two-body orbital mechanics and impulsive
thrust assumptions. These assumptions are not accurate for the continuous thrust propulsion systems con-
sidered, but the use of a more accurate integrating trajectory optimization program would have required
more time than was available for this project. These inaccuracies must be considered in any follow-on
study.
The outbound trajectory is solved as a Gauss problem. Two points in space are known, as well as
the desired transfer time between them. The asteroid’s position at interception or rendezvous is calculated
by assuming an Earth impact position and traveling backward along the asteroid’s orbit to the desired
arrival time, which is given as the number of days before impact. In this study, the asteroid’s orbit is
initialized such that it will impact the Earth at a 45∞ angle on the heliocentric-ecliptic plane. The spacecraft’s
position at departure is the same as the Earth’s at that time. The Earth’s position is calculated by moving the
Earth backward from the impact point by the number of days equal to the asteroid arrival time—given as
the number of days before impact—plus the desired outbound trajectory flight time.
The Gauss problem formulation used in this study is taken from the literature.
37
The universal
variables solution method allows the trajectory to be any type of conic section: an ellipse, a parabola, or a
hyperbola. Two DVs are calculated: (1) Must be applied to depart Earth’s orbit and send the vehicle on the
trajectory that will intercept the asteroid in the desired flight time, and (2) applied upon arrival at the
asteroid. This DV places the vehicle in the asteroid’s orbit and allows for rendezvous with the asteroid.
Interceptor missions like kinetic deflection use only the first DV, since impact with the asteroid is
desired. Other missions that require close asteroid operations must perform both maneuvers.
The calculations were implemented in a Mircosoft
®
Excel workbook and then “wrapped” into the
integrated design environment, ModelCenter
®
. All outbound trajectories are designed to intercept or ren-
dezvous with a baseline NEO that is defined in section 6.2. Example results for missions to the baseline
asteroid are given in figure 70. The DVs for both rendezvous and interception are given. These results
correspond to a total mission time of 3,600 days. Total mission time is defined as the sum of the outbound
flight time and the asteroid arrival time (days before impact), or the time between launch of the system and
the asteroid’s predicted collision with the Earth.
103
6.2 Inbound
The inbound trajectory modeling software determines the minimum impulsive DV required to make
an incoming planetary body miss Earth by some specified distance. In order to allow the inclusion of other
objects, such as Jupiter, during later studies, the software numerically integrates the equations of motion of
the Sun-Earth-planetary body system. This program, Planetary Body Intercept (PBI), iterates over the
search space until a DV is found that is a minimum and also causes the planetary body to miss the surface
of Earth by 3 Earth radii.
To determine minimum DV requirements, it was necessary to find a planetary body that would
definitely collide with Earth, preferably dead center. However, due to the uncertainty in the orbital deter-
mination of the NEOs, it was decided not to conduct a lengthy search of the NEO catalog. Instead, a
fictitious body was created in the following way. First, the orbital elements of 444 known potentially
hazardous asteroids (PHAs) were examined in order to establish those elements that might apply to an
“average” PHA. Next, the PHA database was searched for one asteroid that came close to this average—at
least in terms of orbital size, eccentricity, and inclination. The resulting candidate asteroid was 1999JT6.
Its orbital elements were then modified slightly, so as to force a collision with Earth. For this purpose, the
Earth was placed 45∞ from the x axis of the Heliocentric-Ecliptic coordinate system at the time of the
hypothetical impact, as illustrated in figure 71. The original and modified elements of 1999JT6 are given in
table 25. The modified asteroid orbit is plotted in figure 72.
Figure 70. Outbound trajectory DVs for 3,600-day total mission duration.
0
20
40
60
80
100
120
–800 –300 –1,300 –1,800 –2,300 –2,800 –3,300 –3,800
Asteroid Arrival Time (Days Before Earth Impact)
V

R
e
q
u
i
r
e
d

t
o

P
e
r
f
o
r
m

O
u
t
b
o
u
n
d

T
r
a
j
e
c
t
o
r
y

(
k
m
/
s
)

Rendezvous
Intercept
104
Table 25. Original and modified orbital elements of 1999JT6.
Figure 71. Illustration showing a typical NEO orbit. The velocity of the
planetary body at impact for this case is (–40,0,0)
T
km/s,
parallel to the x axis of the Heliocentric-Ecliptic system.
A
steroid Orbit
y
x
Sun
E
a
r
t
h

s

O
r
b
it
Name
Semimajor axis (km)
Eccentricity
Inclination*
Longitude of ascending node*
Argument of perihelion*
1999JT6
319280491.2
0.579033744
9.568048
79.06928886
38.8692997
M1999JT6
319285502.2
0.5791277
11.47182
45.009263
41.840244
Original Orbital Elements Modified
*All angles are in degrees.
105
To avoid confusion with the real asteroid 1999JT6, this fictitious asteroid has been named M1999JT6.
This step was taken largely to avoid the possibility of readers getting the impression that 1999JT6 is indeed
on a collision course with the Earth. Size and composition of the fictitious asteroid are not specified, since
these characteristics are varied during the analysis. During execution of the PBI program, the user specifies
only the mass and velocity vector of an arbitrary planetary body, which is initially placed at the center of
the Earth. Then, the position of the body is adjusted so that it collides with the surface of the Earth nearly
dead center. For M1999JT6, the mass is varied, but the impact velocity vector is always as given:
V I J K = - + + 35 65 18 5 9 0 . . . . km/s
(81)
This impact velocity matches the modified orbital elements above, which are determined when the asteroid
is at a distance from Earth equal to twice the radius of the Earth’s sphere of influence.
The user also specifies the number of days before impact when the deflection DV is to be applied.
Backward numerical integration of the three-body equations (eq. (82)) positions both the Earth and plan-
etary body at their proper location for this specified time. The equations are written in the Heliocentric-
Ecliptic coordinate system, with the bodies assumed to be point masses:
Figure 72. M1999JT6 orbit plot.
Impact
Asteroid
J
u
p
i
t
e
r

O
r
b
i
t
–6 –5 –4 –3 –2 –1 0 1 2 3 4 5 6
6
5
4
3
2
1
0
–1
–2
–3
–4
–5
–6
Earth
M
a
rs Orb
i
t
106
˙˙
/
/
/
/
/
/
/
r
r
r
r
r
r
r
E S
E S
E S
E S PB
PB E
PB E
PB S
PB S
G M M
GM = -
+
( )
+ -
È
Î
Í
Í
˘
˚
˙
˙
3 3 3
˙˙ .
/
/
/
/
/
/
/
r
r
r
r
r
r
r
PB S
PB S
PB S
B S E
PB E
PB E
E S
E S
G M M
GM = -
+ ( )
- +
È
Î
Í
Í
˘
˚
˙
˙
3 3 3
(82)
At this point, the iterative process of determining the magnitude of the impulsive DV begins. The
direction of the impulse is determined by the user from a list of possible commands shown in table 26.
Each maneuver is made from the perspective of the planetary body and its orbit, not from the perspective
of Earth’s orbit. The program makes an initial guess for the magnitude of the DV and integrates forward
until the planetary body reaches its point of closest approach to the Earth. If the minimum distance between
Earth and the planetary body is not within some user-specified tolerance, a new DV magnitude is chosen
and the process repeated. In practice, with a tolerance of 0.01 Earth radii, the program usually converges to
a solution after three to five iterations. Some cases, however, require many more iterations, or may not
converge at all, as is the case when the planetary body is so close to Earth that no ACCEL maneuver can
result in a miss. In these cases, the program prints a warning message and reports the magnitude of the DV
to be 1¥10
10
km/s.
Table 26. Explanation of the different maneuvers available for use in the program PBI.
Different maneuvers usually result in different DV requirements, and a maneuver that is best in one
situation is not always the best in others. For example, for long-period comets coming in at very high
speeds, UP or DOWN may result in the minimum DV requirement when impact is only a few days away, as
shown in figure 73. Note that the velocities shown in figures 73–77 are those of the planetary body at
impact. However, figure 74 shows that DECEL may result in the minimum DV when the impact is from 50
to 150 days away. The situation is different for the typical asteroid, which would be moving much slower
than a long-period comet. Figure 75 shows that the UP maneuver no longer is the best option when the
planetary body with an impact velocity of 35 km/s is only a few days away; now the best option is to
decelerate the object. With more time available, the preferred option changes from DECEL to OUTSIDE,
ACCEL*
DECEL*
UP
DOWN
INSIDE
Increase the magnitude of the velocity; do not change direction
Decrease the magnitude of the velocity; do not change direction
Rotate the velocity vector up; do not change the magnitude of the velocity
Rotate the velocity vector down; do not change the magnitude of the velocity
Rotate the velocity vector toward the inside of the body’s orbit, or toward
the Sun, whether or not the orbit is direct or retrograde; do not change
the magnitude of the velocity
Rotate the velocity vector toward the outside of the body’s orbit, or toward
the Sun, whether or not the orbit is direct or retrograde; do not change
the magnitude of the velocity
OUTSIDE
Maneuver Description
*Only ACCEL and DECEL change the magnitude of the velocity vector.
107
Figure 73. Required impulsive DV for 42 km/s velocity for various maneuvers to avoid
collision with Earth, showing the benefit of the UP maneuver when impact
is only a few days away.
Figure 74. Required impulsive DV for 42 km/s velocity for various maneuvers to
avoid collision with Earth, showing the benefit of the DECEL and
OUTSIDE maneuvers when impact is several weeks away.
as illustrated in figure 76. Overall, however, the ACCEL maneuver is found to be the most efficient maneu-
ver for rendezvous times of ª300 days or more, as illustrated in figure 77.
INSIDE
OUTSIDE
UP
DOWN
ACCEL
DECEL
V=(–41.36, 0, 7.29)
T
–20 –18 –16 –14 –12 –10 –8 –6 –4 –2
0
0.05
0.1
0.15
0.2
0.25
I
m
p
u
l
s
i
v
e




V

(
k
m
/
s
)

Days Before Impact
INSIDE
OUTSIDE
UP
DOWN
ACCEL
DECEL
V=(–41.36, 0, 7.29)
T
–150 –140 –130 –120 –110 –100 –90 –80 –70 –60 –50
0.004
0.002
0
0.006
0.008
0.014
0.012
0.01
0.016
0.018
0.02
I
m
p
u
l
s
i
v
e




V

(
k
m
/
s
)

Days Before Impact
108
Figure 75. Required impulsive DV for 35 km/s velocity for various maneuvers
to avoid collision with Earth, showing the benefit of the DECEL
maneuver when impact is only a few days away.
Figure 76. Required impulsive DV for various maneuvers to avoid collision
with Earth for planetary body with velocity of 35 km/s.
INSIDE
OUTSIDE
UP
DOWN
ACCEL
DECEL
V=(–34.47, 0, 6.078)
T
–20 –18 –16 –14 –12 –10 –8 –6 –4 –2
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
I
m
p
u
l
s
i
v
e




V

(
k
m
/
s
)

Days Before Impact
INSIDE
OUTSIDE
UP
DOWN
ACCEL
DECEL
V=(–34.47, 0, 6.078)
T
–150 –140 –130 –120 –110 –100 –90 –80 –60 –70 –50
0
0.005
0.01
0.015
0.02
0.025
I
m
p
u
l
s
i
v
e




V

(
k
m
/
s
)

Days Before Impact
109
By limiting the number of possible maneuvers to six rather than performing a global search, com-
putation speed is increased significantly. In many cases, for example, the optimum trajectory may be a
combination of UP and DECEL. However, comparing the DV requirements for a range of successful trajec-
tories gives insight into which combinations could possibly result in a lower impulse requirement. The
goal of manipulating the planetary body’s trajectory is to miss the moving Earth, not to rendezvous with it.
Therefore, maneuvers that delay the arrival of the planetary body generally allow Earth to move out of the
way before the body arrives.
One must realize that the quest for the minimum DV did not include the object’s disposal; i.e., never
threatening Earth again, only its avoidance. By only requiring the body to miss Earth on this one occasion
does not guarantee that it will miss Earth during some future encounter. A maneuver that removes the body
completely from the list of threatening objects is clearly the best option and should be considered in future
studies.
Figure 77. Required impulsive DV for various maneuvers to avoid collision with
Earth for planetary body with velocity of 35 km/s (long lead time).
INSIDE
OUTSIDE
UP
DOWN
ACCEL
DECEL
V=(–34.47, 0, 6.078)
T
–1,000 –900 –800 –700 –600 –500 –400 –300
0.0020
0.0018
0.0016
0.0014
0.0012
0.0010
0.0008
0.0006
0.0004
0.0002
0
I
m
p
u
l
s
i
v
e




V

(
k
m
/
s
)

Days Before Impact
110
7. THREAT PARAMETRIC
Thanks to the pioneering work of researchers like Eugene Shoemaker,
38
the threat and conse-
quences of an impact with a planetary body are now more appreciated, and better understood, than was the
case 30 yr ago. One only has to consider the literature in references 39–44 and many others to see evidence
of increased efforts to understand the threat that Earth faces every day. Although our understanding of the
impact threat is still incomplete, it is far ahead of our understanding of the consequences of the most likely
impacts. Attempting to predict the number of people killed over the next decade, century, or millennium
due to impacts of certain sizes of planetary bodies is a highly speculative endeavor. Recorded impacts in
the developed world are rare and so do not constitute a statistically significant database. This means that
researchers must resort to the use of theoretical models
41
to estimate the number of deaths resulting from
the impact of a planetary body of a certain size, velocity, and type.
Our program, PEOPLE, estimates the number of people saved over the next century if all planetary
bodies of a given type; i.e., chondrite, long-period comet, etc., having kinetic energy less than or equal to
some given value, can be successfully deflected. The number of fatalities prevented is based on the work
done by Shoemaker,
38
Chapman and Morrison,
39
and Lewis
41
using estimates of both the impact
frequency and the number of deaths due to impacts of a certain energy. Over 10,000 runs of a modified
version of John Lewis’s Monte Carlo simulation program were used to generate data for the average num-
ber of deaths due to each type of object (table 27). (The main modification to Lewis’s simulation code
allowed the tallying of deaths due to different types of objects, enabling the study team to focus efforts on
the types of objects that would most likely cause the most deaths over the next century.)
Table 27. Types of planetary bodies examined in the Monte Carlo
simulation and their average contribution to the total number
of deaths over the next century.
Type
Total Deaths
(%)
Chondrite
Achondrite
Iron
Mesosiderite
Pallasite
Comet, short-period
Comet, long-period
60
5
5
<1
<1
6
24
111
The program PEOPLE also determines which parameter—mass or velocity—should be increased
to counter the largest portion of the threat. This determination is based on velocity distributions from
Lewis
41
and Chesley,
43
and size distributions from Gold
40
and Ivezic et al.
44
(Recent data from Brown
et al.
45
indicates that the frequency for Tunguska-sized events may occur only every 1,000 yr as opposed to
previous estimates of every 200 to 300 yr, and that the size distribution of the smaller asteroids may need
to be reassessed.) A static human population of 6 billion is assumed here, as it is throughout most of the
literature. Causes of death include tsunamis, blast waves, firestorms, and direct impacts.
The first step in PEOPLE is to use the mass (m) of the planetary body (kilograms), its impact
velocity (V) relative to Earth (m/s), and its type (see table 27), and determine the equivalent energy yield in
megatons of TNT using the following equation:
Mton =
¥
mV
PB E /
.
.
2
15
8 37 10
(83)
Next, the number of deaths (F ) per year, on average, due to all impacts having this energy or less is
determined. These equations were derived using data taken from Chapman and Morrison,
39
Lewis,
41
and
Gold.
40
The equations are split into four categories based on the planetary body’s equivalent energy yield:
Mton £ 2.5 F = 0.0236 Mton
5.525
2.5 < Mton £ 200,000 F = 164 – 180 Mton
–0.1333
200,000 < Mton £ 1¥10
8
F = 3763 – 4.16¥10
6
Mton
–0.577
Mton > 1¥10
8
F = 3763 – 4.16¥10
6
Mton
–0.577
(if the user wishes; this is an extrapolation). (84)
PEOPLE also calculates the number of fatalities ( f ) that would have been expected on average for
this energy of impact. This is not the cumulative number of fatalities as described above, but is based on the
following equations:
Mton £ 2.5 f = 5.433 Mton
6.325
2.5 < Mton £ 200,000 f = 1000 Mton
2/3
200,000 < Mton £ 1¥10
8
f = 1¥10
8
Mton
0.223
Mton > 1¥10
8
f = 6¥10
9
, the world population. (85)
Caution must be used when applying these equations and estimates. For example, the overwhelm-
ingly dominant event each year is a small one, which would result in only a few fatalities. However, while
a catastrophic event with impact energy of 1¥10
8
Mton could kill several billion people, its likelihood is
only 1 in 100 million.
39,43
Therefore, this type of impact will result in <100 fatalities per year on average,
but when it does occur, the outcome will clearly be catastrophic. Similarly, while their likelihood is only
1 in 10,000, impactors with energies of 1¥10
5
Mton statistically cause ª3,000 fatalities per year, the largest
number by far. Impacts of this size are on the threshold of being globally catastrophic events, large enough
112
to do massive damage throughout the world, but too small to destroy humanity. They are also frequent
enough to result in a high average number of fatalities per year.
39
Finally, if the deflection system used to successfully change the course of the object has some
excess energy, the program PEOPLE determines which planetary body parameter—velocity or mass—
should be increased in order to defeat the maximum portion of the total threat. These equations are based
on data reported by Lewis
41
and Gold
40
on the velocity and size distributions of NEAs and comets. First,
the spherical diameter of the object is determined by
D
m
=
Ê
Ë
Á
ˆ
¯
˜
6
1
3
pr
,
(86)
where the density depends of the type of object, as listed in table 27. The rate of change of cumulative
average fatalities with respect to mass is then calculated from


=
-
F
m
D 37 67
2 42241
. .
.
(87)
Finally, the program estimates the rate of change in fatalities with respect to velocity, using the
appropriate form of equation (88). This calculation depends on the type of object (long-period comet,
short-period comet, or asteroid) and its velocity relative to the Earth at impact:
• Long-period comet:
18 . 21 18 . 11 < £V
2
) 18 . 11 ( 000075 . 0 - =


V
V
F
35 18 . 21 < £V 0075 . 0 =


V
F
50 35 < £V 045 . 0 0015 . 0 - =


V
V
F
86 . 66 50 < £V


=
F
V
0 03 .
86 . 71 86 . 66 < £V 3814 . 0 005256 . 0 + - =


V
V
F
113
• Short-period comet:
34 14 < £V 072 . 0 ) 24 ( 00072 . 0
2
+ - - =


V
V
F
86 . 66 34 < £V 0 =


V
F
86 . 71 86 . 66 < £V 008 . 0 =


V
F
• Asteroid:
86 . 71 18 . 11 < £V
79 . 0
) 18 . 11 ( 1023 . 0
-
- =


V
V
F
. (88)
The velocity distribution of asteroids is based on an equation slightly different from the one used by
Lewis,
41
but it seems to fit the data slightly better.
PEOPLE was incorporated into a model in ModelCenter to determine the number of deaths that
could, on average, be prevented over the next century if all incoming asteroids of a particular energy or less
could be successfully deflected. The results are graphed in figure 78. The threshold asteroid size for a
globally catastrophic event is evident, although the exact size of the asteroid that defines the threshold is
unclear, and is strongly influenced by type.
114
The number of fatalities is very sensitive to the impact location for the smaller asteroids. The
average number of fatalities from the impact of a 100-m chondrite is ª10,000 people, but that number
could increase to the millions if the impact occurred in or near a major city or a densely populated coast-
line. Therefore, the relatively small average number of fatalities per century is little reason to conclude that
defending Earth from these sizes of impacts is unnecessary.
In summary, the average number of deaths per impact and the cumulative number of deaths per
century are based on simulations. These simulations contain both statistical data and estimates gleaned
from experience with high-energy detonations in both populated and unpopulated areas. This results in a
high uncertainty in the estimate of fatalities due to an asteroid impact. An uncertainty analysis of the
estimated fatalities was not completed due to time constraints. However, it is clear that the standard devia-
tion in the estimate of facilities for most sizes of impactors is much larger than the mean. Therefore, the
estimate of average fatalities per century should be used with caution, and is only a guide to illuminate the
direction in which the design of a planetary defense system should proceed in order to prevent the most
fatalities on average.
Figure 78. Average deaths from single asteroid impact versus size,
and the average total number of deaths prevented if all
impacts of equal or less energy can be avoided.
10 100 1,000 10,000
100
1,000
10,000
100,000
1,000,000 10
10
10
9
10
8
10
7
10
6
10
5
10
4
10
3
10
2
10
1
10
0
Chondrite Diameter (m)
D
e
a
t
h

p
e
r

C
e
n
t
u
r
y

f
o
r

A
l
l

I
m
p
a
c
t
o
f

E
q
u
a
l

o
r

L
e
s
s

E
n
e
r
g
y
A
v
e
r
a
g
e

D
e
a
t
h
s

F
r
o
m

S
i
n
g
l
e

I
m
p
a
c
t
Average Deaths From
Single Impact
Average Deaths Prevented
per Century if all Impacts
of Equal or Less Energy
can be Defeated
115
8. PARAMETRIC RESULTS
8.1 Integrated Analysis
The original intent of this project was to evaluate the ability of various combinations of technolo-
gies to defeat the entire threat posed by NEOs. Obviously, this is a very complicated problem. Potential
impactors come in all shapes and sizes, and their orbits vary greatly. To further complicate matters, a large
number of technologies for use in threat mitigation have been examined. Assessing all possible technology
combinations would be prohibitive.
The original intent of this project was to select several technologies based on our understanding
and experience, and to test the ability of each to defeat the total threat. Figure 79 illustrates the proposed
analysis process. Starting with an assumption for the total system mass and the total mission time
allowed, the analysis process then divides, based on the type of threat mitigation concept being
considered—remote station, interception, or deflection. The remote station analysis path assumes both an
incoming asteroid mass and a velocity vector. Running the inbound parametric defines the DV
required to deflect the asteroid, given its size and velocity vector at point of impact, and also the allowable
mission time. Running the remote station tool, based on the DV to be delivered to the object and the
allowable mission time, computes the required remote station mass. If this mass is not equal to the allow-
able mass for the system assumed at the beginning, then the analysis path returns to assume a new asteroid
mass and velocity vector. New DVs and remote station masses are then computed. The new asteroid mass
and velocity vector is selected using the threat assessment tool to maximize the percentage of the total
threat that can be defeated for the assumed total system mass. After closure, the threat assessment tool is
run again in order to compute the total threat that is defeated. The total threat is quantified by the percent-
age of people saved by deployment of the system over a given time period divided by the number of people
expected to die from impact of an NEO over the same period. By running through this process several
times—assuming new total system masses and mission times on each occasion—yields a parametric model
of the total threat defeated as a function of total system mass for lines of constant mission time.
116
The interception branch differs from that of the remote station in that the inner iterative loop deter-
mines the asteroid size and velocity that can be deflected. Here, there is an additional problem in optimiz-
ing the amount of the total mission mass allocated to the outbound propulsion against the amount allocated
to the interception system. First, an interceptor mass is assumed and the size of the resulting object frag-
ments is estimated. Development of an atmospheric entry code that would model the burnup of these
fragments was initiated, but this has not yet been completed. It is intended that this model would
include the effects of drafting; i.e., later objects following in the wake of the earlier objects. If the
fragments survived reentry, then a larger interception mass is assumed and a new iteration ensues. Other-
wise, the calculation proceeds to calculate the DV required and then the mass of the outbound propulsion
stage needed to take the interceptor to the incoming object. This outer loop iterates until the maximum
threatening object that can be mitigated for the total allowable mission mass is found, again using the threat
Remote Station
Assume
M
a
, V
imp
Assume
M
a
, V
imp
Assume
M
a
, V
imp
Assume
M
t
, t
T
Inbound
Trajectory
Parametric
Inbound
Trajectory
Parametric
Outbound
Trajectory
Parametric
Outbound
Trajectory
Parametric
Outbound
Propulsion
Parametric
Outbound
Propulsion
Parametric
Asteroid
Dismantle
Parametric
Debris
Reentry
Parametric
Survive
Reentry?
Stationary
Deflector
Parametric
No
No
No
Yes
Yes
Yes
Yes
No
No
Yes Threat
Parametric
Interception Deflection
Start
Finish
Cumulative
Damage
Percent
Guess: t
i
t
i
Optimal?
∆V
i
∆V
o
∆V
o
∆V
i
M
s
M
s
= M
T
M
b
M
deb
M
i
M
o
M
o
M
b
+ M
o
= M
T
M
i
+ M
o
= M
T
Inbound
Propulsion
Parametric
Figure 79. Proposed analysis process for assessing total amount of threat mitigated.
117
tool. As before, a parametric model, giving the percentage of the total threat defeated as a function of both
total system mass and mission time, can be generated.
The rendezvous branch in figure 79 is the most complicated one of all. Here, the most threatening
object must be found using techniques similar to those employed for the interception branch. However, in
this case, an allowed inbound trip time is also assumed. The analysis process then runs through the tools
that calculate the required inbound DV and mass, as well as the outbound DV and mass. Then, the total mass
can be calculated by summing the inbound and outbound masses and comparing the result with the
assumed total system mass. Even after closure of this inner loop, the interception point, defined by the
allowed inbound interception time, may not be optimal; therefore, another loop is used to find the optimal
interception point. After closure of this loop, the process goes into the threat parametric to find the total
percentage of the threat that is defeated. Again, a parametric model of total threat defeated as a function of
system mass and mission time is generated.
There are still two assumptions built into the above analysis scheme:
(1) Although the distribution of object mass and velocity is taken into account, the possible
distribution in composition is not. That composition is believed to be a secondary factor in performance,
although not in damage caused. In addition, because asteroid and comet compositions are so poorly under-
stood, it was decided that the whole issue of composition would not be addressed for this initial study. Note
that the inclination distribution was taken into account; it is imposed by a nonzero z component in the
incoming object velocity vector at impact.
(2) The deflection study allowed no time for setup after asteroid rendezvous. None of the options
considered required a significant amount of time after rendezvous for these operations, except for the mass
driver, so it was assumed that the required deflection DV was imposed instantaneously upon rendezvous.
Finally, the threat parametric has several implied assumptions defined in section 7.
Unfortunately, these ambitious analysis goals were not completed in the time available. As will be
shown below, a parametric model of total system mass was derived for several architecture options, assum-
ing a standard set of orbital elements for the incoming object. The parametrics were derived using a
process similar to the inner loops shown in figure 79. The architecture options considered and performance
of these options are described in section 8.2.
8.2 Architecture Options
In the time available for this study, several architecture options were considered—either suggested
in the literature or which appeared promising. It must be emphasized that this selection does not constitute
a full list of possible options for threat mitigation. In particular, note that a remote station architecture
option was not considered. The list of cases that were considered and their mission configuration are
shown in table 28. Note that each case or mission scenario was assembled in a unique project file within the
integrated design environment ModelCenter.
118
8.3 Parametric Performance
8.3.1 Staged Chemical/Mass Driver
The basic design process for this scenario is as follows. The inbound trajectory program is initiated
by specifying how many days before impact the deflection is to take place. One output of this tool is the DV
required to deflect the asteroid. This value is used in the mass driver sizing tool. The asteroid’s position at
the time of deflection is another output. With this information and the desired time of flight, the outbound
trajectory tool determines the DV requirement to rendezvous with the asteroid. The mass driver’s total mass
is used as the payload mass for a staged chemical rocket that performs the DV maneuvers required for the
outbound trajectory. The final result is the total system mass, which includes the mass driver and the rocket
required to deliver it to the asteroid. The optimal mission from an energy standpoint is that which requires
the lowest mass. This requires a balance between the conflicting goals of minimizing the deflection DV and
the outbound trajectory DV.
Figure 80, a partial screen shot from the ModelCenter program, shows this basic design process.
The first component, Main, is a simple script. The primary reason for creating Main is to collect all of the
inputs used in the parametric studies into a single component. This frees the user from needing to search
through the extensive input lists of the other programs. It also performs some simple algebraic manipula-
tion necessary to convert units and calculate inputs for subsequent programs.
Table 28. Architecture options considered in this study.
Outbound System Inbound System
Remote Station Versus
Interception Versus
Rendezvous
Deflection Versus
Fragmentation
Staged chemical
Staged chemical
Staged chemical
Nuclear pulse
Solar collector
Mass driver
Kinetic deflector
Nuclear deflection
Nuclear deflection
Solar collector
Rendezvous
Interception
Interception and rendezvous
Rendezvous
Rendezvous
Deflection
Deflection
Deflection
Deflection
Deflection
119
The next component, PBI, is the inbound trajectory tool, which was documented in section 6.2
of this TP. With the asteroid specified and the integration parameters set, the only changing input to this
program is the number of days before impact that the deflection is to occur, and the optimal deflection
direction. The first of these inputs is a user input that comes from the Main program discussed above. The
second input is determined by looking at each deflection direction and finding the best option for the
missions considered. Figure 81 shows the deflection DV for each deflection direction over the range of
100 days to over 10 yr before impact. In the magnified image, it is apparent that for any deflection that
occurs more than 600 days before impact, the ACCEL maneuver requires the least DV. It is also interesting
to note that the DV required oscillates with each asteroid period, and it is obvious for the ACCEL and
DECEL options that each oscillation diminishes in magnitude.
Figure 80. Staged chemical/mass driver model.
Main
PBI
MassDriver
MassDriverDVCalc
GoalSeek
UnitConversion1
Unit
Conversion2
PBO_v4
People_v3
stagedChemical3
120
Figure 81. (a) Optimal deflection direction, (b) optimal deflection
direction—detailed view, and (c) optimal deflection direction
—detailed view—minimal deflection DV.
–0.0005
0
0.0005
0.001
0.0015
0.002
0.0025
0.003
0.0035
0.004
500 –500 –1,500 –2,500 –3,500 –4,500
ACCEL
DECEL
UP
DOWN
INSIDE
OUTSIDE
Maneuver Type
Asteroid Arrival Time (Days Before Earth Impact)
V

(
k
m
/
s
)

Invalid Point
(b)
0
0.00005
0.00015
0.0001
0.0002
0.0003
0.00025
0.0004
0.00035
0.00045
0.0005
500 –500 –1,500 –2,500 –3,500 –4,500
Asteroid Arrival Time (Days Before Earth Impact)
V

(
k
m
/
s
)

ACCEL
DECEL
INSIDE
OUTSIDE
Maneuver Type
(c)
–0.01
0
0.01
0.02
0.03
0.04
0.05
0.06
500 –500 –1,500 –2,500 –3,500 –4,500
ACCEL
DECEL
UP
DOWN
INSIDE
OUTSIDE
Maneuver Type
Asteroid Arrival Time (Days Before Earth Impact)
V

(
k
m
/
s
)

Invalid Point
(a)
Asteroid M1999JT6 Deflection V
(Asteroid's Period is 1,100 Days)


121
Once the deflection DV is known, the mass driver can be sized. The mass driver analysis
program, MassDriver, is not set up to use a DV for sizing, so an iteration loop is introduced.
MassDriverDVCalc is a script that converts the total impulse provided by the mass driver into a DV and
then compares it with the DV required to deflect the asteroid. The optimizer, GoalSeek, changes the amount
of asteroid mass ejected by the mass driver until the two DVs are equal. There is a simple unit conversion
in the loop that converts ejected asteroid mass, or total expellant mass, from metric tons to kilograms.
When the mass driver produces the correct DV, its total system mass value is passed to the staged chemical
tool, where it becomes the payload mass for the outbound vehicle. One interesting feature about the mass
driver is that its total system mass remains relatively constant regardless of the deflection DV required. This
is because no expellant mass needs to be carried out with the mass driver; the asteroid provides it all. There
is only one term in the mass driver sizing relationship that scales with DV—the Expellant Storage facility
mass. This facility houses the expellant mass prior to ejection, and for this study, its mass is equal to
0.01 times the total amount of mass to be ejected. From figure 82 one can see that this change in component
mass results in a minimal shift in the total system mass.
Figure 82. Variation of mass driver total system mass with required asteroid
deflection DV for a 50-m-diameter chondrite.
The outbound trajectory code, Planetary Body—Outbound (PBO), is executed once the asteroid’s
position at rendezvous is passed from PBI. The outbound flight time is calculated and passed from the
Main component. The required DV to rendezvous with the asteroid is output, and is passed to the staged
chemical tool.
Deflection V (km/s)
Mass Driver Mass (t)

4×10
–4
3.5×10
–4
3×10
–4
2.5×10
–4
2×10
–4
1.5×10
–4
1×10
–4
5×10
–5
0
0 –500 –1,000 –1,500 –2,000 –2,500 –3,000 –3,500 –4,000
75
74
73
72
71
70
69
68
67
Asteroid Arrival Time (Days Before Earth Impact)
V

t
o

D
e
f
l
e
c
t

(
k
m
/
s
)

M
a
s
s

D
r
i
v
e
r

M
a
s
s

(
t
)
122
The staged chemical tool uses the payload mass given by the mass driver component and the DV
from the outbound trajectory code, PBO, to size the vehicle that departs from Earth. For this scenario, only
the lox/LH
2
propellant combination is considered, with an assumed vacuum I
sp
of 465 s. Also, the vehicle
was allowed five stages.
With all of the necessary components appropriately linked in a ModelCenter project file, the user
can input a total mission duration and a rendezvous time, and then calculate the total mass required for the
mission for a given asteroid composition and diameter. In this manner, the parametric analysis was
performed for this mission scenario. The total mission duration, which is defined as the number of days
between the vehicle’s departure from Earth and the impending Earth-asteroid collision, was varied from
1,500 days to 10 yr, in increments of 25 days. The point at which the vehicle makes its rendezvous with the
asteroid was varied from 500 to 3,500 days before impact, also in increments of 25 days. The small incre-
ment value was necessary to ensure that optimal orbital transfers were captured. Within these ranges, the
outbound flight time could be as short as 25 days or as long as 3,100 days. Figures 83 and 84 show the
resulting initial spacecraft mass at Earth departure.
Figure 83. Staged chemical/mass driver vehicle mass at Earth departure.
1500 1525 1550
1575 1600 1625
1650 1675 1700
1725 1750 1775
1800 1825 1850
1875 1900 1925
1950 1975 2000
2025 2050 2075
2100 2125 2150
2175 2200 2225
2250 2275 2300
2325 2350 2375
2400 2425 2450
2475 2500 2525
2550 2575 2600
2625 2650 2675
2700 2725 2750
2775 2800 2825
2850 2875 2900
2925 2950 2975
3000 3025 3050
3075 3100 3125
3150 3175 3200
3225 3250 3275
3300 3325 3350
3375 3400 3425
3450 3475 3500
3525 3550 3575
3600
450,000
400,000
350,000
300,000
250,000
200,000
150,000
100,000
50,000
0
–800 –1,300 –1,800 –2,300 –2,800 –3,300 –3,800
Asteroid Arrival Time (Days Before Earth Impact)
S
p
a
c
e
c
r
a
f
t

M
a
s
s

(
t
)

a
t

S
O
I
Total Mission Duration
123
The minimum mass solution for this scenario is 6,850 t. This mission is launched 2,900 days before
the asteroid’s collision with the Earth, and arrives at the asteroid 500 days after launch. The trajectory is
shown in figure 85.
Figure 84. Staged chemical/mass driver vehicle mass at Earth departure (expanded view).
Figure 85. Optimal staged chemical/mass driver mission.
1500 1525 1550
1575 1600 1625
1650 1675 1700
1725 1750 1775
1800 1825 1850
1875 1900 1925
1950 1975 2000
2025 2050 2075
2100 2125 2150
2175 2200 2225
2250 2275 2300
2325 2350 2375
2400 2425 2450
2475 2500 2525
2550 2575 2600
2625 2650 2675
2700 2725 2750
2775 2800 2825
2850 2875 2900
2925 2950 2975
3000 3025 3050
3075 3100 3125
3150 3175 3200
3225 3250 3275
3300 3325 3350
3375 3400 3425
3450 3475 3500
3525 3550 3575
3600
18,000
20,000
16,000
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
–800 –1,300 –1,800 –2,300 –2,800 –3,300 –3,800
Asteroid Arrival Time (Days Before Earth Impact)
S
p
a
c
e
c
r
a
f
t

M
a
s
s

(
t
)

a
t

S
O
I
Minimum Mass Solution
Total Mission Duration
–3 –2 –1.5 –1 –0.5 0 0.5 1.5 2.5 1 2 3 –2.5
–3.5
–3
–2.5
–2
–1.5
–0.5
0
0.5
1
1.5
2
2.5
3
3.5
–1
Sun
Earth
Earth perihelion
Asteroid
Asteroid perihelion
Asteroid asc. node
Transfer trajectory
Start transfer
Finish transfer
Asteroid at start
Earth at finish
Impact position
124
Once the optimal mission solution was found, the asteroid diameter was varied to determine its
effect on mission mass. The results show that this scenario can be used to deflect chondrite asteroids up to
80 m in diameter before the staged chemical’s performance becomes inadequate. Because the mass driver
system mass changed very little, the overall vehicle mass remained <7,000 t. The resulting vehicle masses
are given in figure 86. Mass driver deployed mass and total operating time required to deflect asteroids of
50 to 1,000 m in diameter are given in figures 87 and 88. While the staged chemical propulsion system is
not capable of delivering systems for the larger asteroids, these numbers may be used in future analysis
with a different outbound propulsion system.
Figure 86. Staged chemical/mass driver vehicle mass versus chondrite asteroid diameter.
Figure 87. Mass driver deployed mass and total operating time versus chondrite
asteroid diameter (50–100 m).
10,000
9,000
8,000
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
0 20 40 60 80 100 120
Asteroid Diameter (m)
S
p
a
c
e
c
r
a
f
t

M
a
s
s

(
t
)

a
t

S
O
I
100
95
90
85
80
75
70
65
60
55
50
50
45
40
35
30
25
20
15
10
5
0
40 50 60 70 80 90 100 110
Asteroid Diameter (m)
O
p
e
r
a
t
i
n
g

T
i
m
e

(
d
a
y
s
)
M
a
s
s

D
r
i
v
e
r

D
e
p
l
o
y
e
d

M
a
s
s

(
t
)
Deployed Mass
Operating Time
125
8.3.2 Staged Chemical/Nuclear Deflection
The cases considered in this scenario included both intercept and rendezvous with the asteroid. The
overall layout is illustrated in figure 89. The main difference between this model and that for the staged
chemical/mass driver option (which was considered in sec. 8.3.1) is the replacement of the mass driver
with the nuclear deflection tool called explosion. Starting with the asteroid diameter and type—obtained
from the main script—and the required asteroid DV from PBI, explosion calculates the size and number of
nuclear blasts necessary to deflect the asteroid. The payload mass, which is just the total mass of the
nuclear devices, is passed to the staged chemical tool, stagedChemical3. The staged chemical tool then
sizes the chemical rocket that is required to deliver the nuclear devices from Earth’s sphere of influence to
the asteroid.
Figure 88. Mass driver deployed mass and total operating time versus
chondrite asteroid diameter (100–1,000 m).
Figure 89. Diagram of the ModelCenter setup for the staged chemical/nuclear deflection option.
Main
PBI
km2m Explosion Ibm2kg
stagedChemical3
PBO_v3
2,500
2,250
2,000
1,750
1,500
1,250
1,000
750
500
250
0
0 100 200 300 400 500 600 700 800 900 1,000 1,100
50,000
45,000
40,000
35,000
30,000
25,000
20,000
15,000
10,000
5,000
0
Deployed Mass
Operating Time
Asteroid Diameter (m)
O
p
e
r
a
t
i
n
g

T
i
m
e

(
d
a
y
s
)
M
a
s
s

D
r
i
v
e
r

D
e
p
l
o
y
e
d

M
a
s
s

(
m
t
)
126
Various values of total mission time and rendezvous time are examined to determine those regions
in which the global minimum total system mass resides. Results of these runs for the intercept and rendez-
vous cases are illustrated in figures 90 and 91. Once the regions of interest are identified, the Optimization
component of ModelCenter is used to find the minimum total system mass. This is accomplished by iterat-
ing over both the total mission time and the rendezvous time. With so many peaks and valleys in the total
system mass graphs in figures 90 and 91, it is necessary to restrict the search to specific regions before the
optimization tool can be used. Otherwise, the iteration process is likely to converge on nonoptimal
solutions. Through a thorough set of analysis runs, the minimum total system mass required to defeat a
100-m-diameter chondrite was found to be 847 kg for the intercept case (fig. 92) and 5,620 kg for the
rendezvous case (fig. 93). For the intercept case, the optimum rendezvous and total mission times are
910 days before impact and 1,509 days, respectively. For the rendezvous case, the optimum rendezvous
and total mission times are 132 days before impact and 1,075 days, respectively. These results are some-
what surprising, given the large staged chemical system that was required for the mass driver option, but
they do illustrate the benefits of using nuclear energy to deflect the asteroid. The trajectories for the
optimum solutions are shown in figures 94 and 95.
Figure 90. Total system mass for the staged chemical/nuclear blast option versus
total mission time for various rendezvous times. Here, the staged chemical
system does not match the asteroid’s orbit at encounter.

–100
–200
–500
–850
–900
–950
–1,000
–1,050
–1,400
Rendezvous Time
(Days Before Impact)
10
9
8
7
6
5
4
3
2
1
0
500 1,000 1,500 2,000 2,500
Total Mission Time (Days)
S
p
a
c
e
c
r
a
f
t

M
a
s
s

(
t
)

a
t

S
O
I
Intercept
127
Figure 91. Total system mass for the staged chemical/nuclear blast option versus
total mission time for various rendezvous times (zoomed). Here, the
staged chemical system matches the asteroid’s orbit at encounter.
Figure 92. Minimum total system mass for the staged chemical/nuclear
blast option, showing the optimum rendezvous and total mission
times for intercept.

20
18
16
14
12
10
8
6
4
2
0
500 1,000 1,500 2,000 2,500
Total Mission Time (Days)
Rendezvous
S
p
a
c
e
c
r
a
f
t

M
a
s
s

(
t
)

a
t

S
O
I
–132
–100
–150
–175
–200
–250
–300
–400
–500
–1,000
–1,400
Rendezvous Time
(Days Before Impact)

2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1,470 1,480 1,490 1,500 1,510 1,520 1,530 1,540 1,550
Total Mission Time (Days)
Intercept
S
p
a
c
e
c
r
a
f
t

M
a
s
s

(
t
)

a
t

S
O
I
–200
–100
–500
–850
–900
–950
–1,000
–1,050
–1,400
–910
Rendezvous Time
(Days Before Impact)
–910/1,509
847 kg
128
Figure 93. Minimum total system mass for the staged chemical/nuclear blast option,
showing the optimum rendezvous and total mission times for rendezvous.
Figure 94. Optimum intercept trajectory for the staged chemical/nuclear deflection option.
–132
–100
–150
–150
–200
–250
–300
–400
–500
–1,000
–1,400
10
9.5
9
8.5
8
7.5
7
6.5
6
5.5
5
S
p
a
c
e
c
r
a
f
t

M
a
s
s

(
t
)

a
t

S
O
I
1,050 1,060 1,070 1,080 1,090 1,100
Total Mission Time (Days)
–132/1,075
5,620 kg
Rendezvous Time
(Days Before Impact)
Rendezvous
–3 –2.5 –1.5 –0.5
–2.5
–1.5
–3
–2
–0.5
0.5
1.5
2.5
3
2
1
0
–1
0 0.5 1.5 1 2 2.5 3 –2 –1
Sun
Earth
Earth perihelion
Asteroid
Asteroid perihelion
Asteroid asc. node
Transfer trajectory
Start transfer
Finish transfer
Asteroid at start
Earth at finish
Impact position
129
The optimum rendezvous and total mission time vary little with changing asteroid mass; the opti-
mum times are nearly the same for both 100- and 1,000-m-diameter chondrites. With this information, it is
relatively easy to determine the required size of the staged chemical rocket versus asteroid size. The results
are plotted in figures 96 and 97. With the rendezvous option, the total system mass increases rapidly and
quickly exceeds the 1,000-t limit assumed for this study. However, total system mass for the intercept case
is much smaller and is less sensitive to changing asteroid mass. Given the total system mass constraint of
1,000 t, the largest diameter M1999JT6 chondrite that this system can defeat has a diameter of
9,000 m for the intercept case and 1,000 m for the rendezvous case.
Figure 95. Optimum rendezvous trajectory for the staged chemical/nuclear deflection option.
–3 –2.5 –1.5 –0.5
–2.5
–1.5
–3
–2
–0.5
0.5
1.5
2.5
3
2
1
0
–1
0 0.5 1.5 1 2 2.5 3 –2 –1
Sun
Earth
Earth perihelion
Asteroid
Asteroid perihelion
Asteroid asc. node
Transfer trajectory
Start transfer
Finish transfer
Asteroid at start
Earth at finish
Impact position
130
Figure 96. Minimum total system mass for the staged chemical/nuclear blast
option versus chondrite diameter for both intercept and rendezvous.
Figure 97. Minimum total system mass for the staged chemical/nuclear blast
option versus chondrite diameter for the smaller chondrites, showing
both intercept and rendezvous.
30,000
25,000
30,000
20,000
15,000
10,000
0
0 2,000 4,000 6,000 8,000 10,000
T
o
t
a
l

S
y
s
t
e
m

M
a
s
s

(
t
)
Chondrite Asteroid Diameter (m)
Intercept
Rendezvous
1,000-t limit
Lack of data points is because the
nuclear deflection tool does not
interpolate between the tabular data
used to estimate the size of nuclear
blast required.
1,000
900
800
700
600
500
400
300
200
100
0
0 1,000 2,000 3,000 4,000 5,000
T
o
t
a
l

S
y
s
t
e
m

M
a
s
s

(
t
)
Chondrite Asteroid Diameter (m)
Intercept
Rendezvous
1,000-t limit
Lack of data points is because the
nuclear deflection tool does not
interpolate between the tabular data
used to estimate the size of nuclear
blast required.
131
8.3.3 Staged Chemical/Kinetic Deflection
The staged chemical/kinetic deflection option consists of a chemical rocket that delivers a massive
projectile to the asteroid. This projectile impacts the asteroid, nudging it off of its collision course with the
Earth. The process flow for the analysis of this scenario in ModelCenter is given in figure 98.
Figure 98. Staged chemical/kinetic deflection model.
The process is similar to that of the mass driver option, but there are a few exceptions. First, the
diagram contains two instances of the inbound trajectory code, PBI and PBI_dv. The reason for this appar-
ent duplication is that the asteroid’s deflection direction is actually determined by the outbound trajectory
of the projectile. Unlike the mass driver option, this vehicle impacts the asteroid; it does not rendezvous
with it. Hence, the outbound trajectory, which determines the impact velocity, has a significant influence
on the subsequent deflection. The two instances of PBI are required because the outbound trajectory code
cannot run until the asteroid’s position at interception is known. So, PBI runs first to determine the asteroid’s
position. This feeds into PBO, which gives PBI_dv the deflection direction. PBI_dv then produces the DV
required to deflect the asteroid. Like the mass driver scenario, there is an iteration loop to ensure that the
deflection DV produced is equal to that required. The projectile mass is varied until the momentum neces-
sary to just achieve the desired DV is imparted to the asteroid. The resulting projectile mass is then input to
the staged chemical tool as its payload mass. The staged chemical tool then provides the overall vehicle
mass required to complete the mission. This time the number of stages is allowed to vary between one and
five. All stages use the lox/LH
2
propellant combination, delivering an I
sp
of 465 s.
Main PBI
PBO_v5
PBI_dv
stagedChemical5b
KineticDeflection DV_check
ImpactorMassOptimizer
132
The initial parametric study was performed for a 50-m-diameter asteroid of chondrite composition.
Total mission time was varied from 150 to 1,200 days, and interception with the asteroid occurred between
100 and 1,125 days before impact. With a limited amount of time to complete this study, it was decided to
limit the trade space for this case to nearer term missions only. The justification for this limitation comes
from the fact that a simple kinetic deflection system, placed on an impact trajectory by a staged chemical
propulsion system, is the simplest type of mitigation option that can be envisioned. It can be activated at
relatively short notice—certainly much sooner than the more complex deflection options; e.g., the mass
driver—and largely uses existing technology. In short, this is the option that is most readily available to
counter near-term threats with limited reaction times. It should be noted, however, that even in this limited
trade space, the asteroid arrival times considered encompass one full period of the asteroid (ª1,100 days).
One would expect that extending the trade-space to consider arriving at the asteroid 2,000 or 3,000 days
before Earth impact would result in similar trends to those shown in figure 81. The required deflection DV
should oscillate, diminishing in magnitude with each period, which would result in lower total mission
masses than those presented here. See figure 99 for resulting vehicle masses for the missions
considered.
150CF 175CF
200CF 225CF
250CF 275CF
300CF 325CF
350CF 375CF
400CF 425CF
450CF 475CF
500CF 525
525CF 550
550CF 575
575CF 600
600CF 625
625CF 650
650CF 675
675CF 700
700CF 725
725CF 750
750CF 775
775CF 800
800CF 825
825CF 850
850CF 875
875CF 900
900CF 925
925CF 950
950CF 975
975CF 1000
1000CF 1025
1025CF 1050
1050CF 1075
1075CF 1100
1100CF 1125
1125CF 1150
1150CF 1175
1175CF 1200
1200CF
7,000,000
6,000,000
5,000,000
4,000,000
3,000,000
2,000,000
1,000,000
0
200 –200 –400 –600 –800 –1,000 –1,200 0
Asteroid Arrival Time (Days Before Earth Impact)
S
p
a
c
e
c
r
a
f
t

M
a
s
s

(
t
)

a
t

S
O
I
Total Mission Duration
(a)
133
Figure 99. (a) Staged chemical/kinetic deflection vehicle mass at Earth departure,
(b) detailed view, and (c) detailed view—minimum mass solution.
One additional complication associated with the kinetic deflection option is the possibility that the
impact might cause the asteroid to fragment. Small impacts will produce only craters. Larger impacts will
produce larger craters; however, there is some threshold impact size above that which the asteroid will
actually break—a process referred to as catastrophic fragmentation. This was discussed in
section 5.7. In figure 99(c), missions that result in catastrophic fragmentation are denoted by a CF after the
series label.
150CF 175CF
200CF 225CF
250CF 275CF
300CF 325CF
350CF 375CF
400CF 425CF
450CF 475CF
500CF 525
525CF 550
550CF 575
575CF 600
600CF 625
625CF 650
650CF 675
675CF 700
700CF 725
725CF 750
750CF 775
775CF 800
800CF 825
825CF 850
850CF 875
875CF 900
900CF 925
925CF 950
950CF 975
975CF 1000
1000CF 1025
1025CF 1050
1050CF 1075
1075CF 1100
1100CF 1125
1125CF 1150
1150CF 1175
1175CF 1200
1200CF
1,000
900
800
700
600
500
400
300
200
100
0
200 –200 –400 –600 –800 –1,000 –1,200 0
Asteroid Arrival Time (Days Before Earth Impact)
S
p
a
c
e
c
r
a
f
t

M
a
s
s

(
t
)

a
t

S
O
I
Total Mission Duration
150CF 175CF
200CF 225CF
250CF 275CF
300CF 325CF
350CF 375CF
400CF 425CF
450CF 475CF
500CF 525
525CF 550
550CF 575
575CF 600
600CF 625
625CF 650
650CF 675
675CF 700
700CF 725
725CF 750
750CF 775
775CF 800
800CF 825
825CF 850
850CF 875
875CF 900
900CF 925
925CF 950
950CF 975
975CF 1000
1000CF 1025
1025CF 1050
1050CF 1075
1075CF 1100
1100CF 1125
1125CF 1150
1150CF 1175
1175CF 1200
1200CF
100
90
80
70
60
50
40
30
20
10
0
200 –200 –400 –600 –800 –1,000 –1,200 0
Asteroid Arrival Time (Days Before Earth Impact)
S
p
a
c
e
c
r
a
f
t

M
a
s
s

(
t
)

a
t

S
O
I
Minimum Mass Solution
Total Mission Duration
(b)
(c)
134
The minimum mass mission is 11,853 kg. It corresponds to a 1,025-day total mission duration—
defined as the time from Earth launch to possible Earth-asteroid collision—and impacts the asteroid at
800 days before Earth impact, or 225 days after the vehicle departs Earth. This mission causes only cratering,
not complete fragmentation. It requires four chemical stages to provide the necessary DV for the outbound
trajectory (plotted in fig. 100).
Figure 100. Optimal staged chemical/kinetic deflection mission.
The interceptor mass for this mission is 668 kg and the dry mass of the final lox/LH
2
stage is
259 kg. As the system is currently conceived, the final vehicle stage would jettison the projectile just
before impact. The stage would then perform some small DV maneuver that would allow it to monitor the
projectile’s impact with the asteroid and transmit useful data back to Earth. Such data might be needed to
plan future impacts on either the same target asteroid or some other similar body in the future. If the final
stage were to remain attached to the projectile, it would serve to increase the energy and momentum
available at impact. This might allow a reduction in the projectile mass, thus making the entire system
smaller. This option has not been considered under the present study, but might be addressed in future
work.
Once the optimal mission was determined, the asteroid size was again varied to determine its effect
on the vehicle mass. As a result of this investigation, determination was made that chondrite asteroids up to
400 m in diameter can be successfully deflected; however, the resulting vehicle mass is very high for an
asteroid of this size. Figure 101 shows the required spacecraft mass to deflect chondrite asteroids of up to
400 m in diameter. Figure 102 shows the projectile masses required to deflect asteroids of 50–1,000 m in
diameter.
–4 –3 –2 –1
–6
–5
–4
–3
–2
–1
0
1
2
3
4
5
6
0 1 2 3 4 Sun
Earth
Earth perihelion
Asteroid
Asteroid perihelion
Asteroid asc. node
Transfer trajectory
Start transfer
Finish transfer
Asteroid at start
Earth at finish
Impact position
135
Figure 101. Staged chemical/kinetic deflection vehicle mass versus chondrite asteroid diameter.
Figure 102. Projectile mass versus chondrite asteroid diameter.
3,500
3,000
2,500
2,000
1,500
1,000
500
0
0 200 400 600 800 1,000 1,200
Asteroid Diameter (m)
S
p
a
c
e
c
r
a
f
t

M
a
s
s

(
t
)

a
t

S
O
I
6,000
5,000
4,000
3,000
1,000
2,000
0
0 200 400 600 800 1,000 1,200
Asteroid Diameter (m)
P
r
o
j
e
c
t
i
l
e


M
a
s
s

(
t
)
136
8.3.4 Nuclear Pulse
Rendezvous with the asteroid or comet is required for the nuclear pulse option, which operates as
both the outbound propulsion system and as the deflection device. The overall layout is presented in figure
103. Once again, the main script is merely a central location in which to input parameters; it passes the
asteroid size and type to inboundNP, which is the tool that determines the number of nuclear pulses
required to impart the required DV to the asteroid. As with the other models, this DV is determined by PBI.
Sizing of the nuclear pulse system for the outbound journey is done by PBMExtPulseMC, which takes the
mass from inboundNP; this mass is the payload that must be carried to the asteroid.
Figure 103. Diagram of the ModelCenter setup for the nuclear pulse option.
Various values of total mission time and rendezvous time were examined to determine the regions
in which to concentrate the search for the global minimum total system mass. This is the same approach
that was used for the staged chemical/nuclear deflection option (see fig. 90 for an example). The result of
these runs is illustrated in figure 104. For the rendezvous case, the rendezvous and total mission times are
1,200 and 2,170 days, respectively. The optimum trajectory is illustrated in figure 105.
Main
FixMass
PBI inboundNP kg2mT
PBO_v3 PBMExtPulseMC
137
As with the staged chemical/nuclear deflection option, the optimum rendezvous and total mission
time varies little with changing asteroid mass. Therefore, it is relatively easy to determine the required size
of the nuclear pulse system as a function of asteroid size. The results are plotted in figures 106 and 107.
Figure 105. Optimum rendezvous trajectory for the nuclear pulse option.
Figure 104. Minimum total system mass for the nuclear pulse option, showing
the optimum rendezvous and total mission times.
–200
–100
–500
–800
–1,000
–1,100
–1,200
–1,300
–1,400

0
10
20
30
40
50
60
70
80
90
100
2,000 2,100 2,200 2,300 2,400 2,500
Total Mission Time (Days)
S
p
a
c
e
c
r
a
f
t

M
a
s
s

(
t
)

a
t

S
O
I
Rendezvous Time
(Days Before Impact)
Min Found at –1,200/2,170
Mass = 29.7 t Note: Several Other Rendezvous
Times not Included for Clarity
–3 –2.5 –2 –1 –0.5 –1.5
–3
–2.5
–2
–1.5
–1
0
–0.5
0.5
1
1.5
2
3
2.5
0 0.5 1 1.5 3 2.5 2
Sun
Earth
Earth perihelion
Asteroid
Asteroid perihelion
Asteroid asc. node
Transfer trajectory
Start transfer
Finish transfer
Asteroid at start
Earth at finish
Impact position
138
Given the total system mass constraint of 1,000 t, the largest diameter M1999JT6 chondrite that this
system can defeat would have a diameter of 9,000 m.
T
o
t
a
l

S
y
s
t
e
m

M
a
s
s

(
t
)
25
30
35
40
45
1,000 800 600 400 200 0
Chondrite Diameter (m)
Figure 106. Minimum total system mass for the nuclear pulse option versus chondrite diameter.
Figure 107. Minimum total system mass for the nuclear pulse option versus chondrite
diameter for the smaller chondrites.
0
200
400
600
800
1,000
1,200
1,400
0 2,000 4,000 6,000 8,000 10,000
1,000-t Limit
Chondrite Diameter (m)
T
o
t
a
l

S
y
s
t
e
m

M
a
s
s

(
t
)
139
8.3.5 Solar Collector
Operation of the solar collector system was analyzed in a somewhat different manner than were the
other systems. In this case, there is no payload that the outbound system must deliver, other than some
avionics components, the RCS system, and some other minor components. For the solar collector, the
inbound and outbound systems are the same. Also, unlike the other systems, the minimum total system
mass continued to decrease as the total mission time increased. Therefore, rather than seeking a global
minimum, the analysis instead located the minimum total system mass for two specific total mission times:
3 and 10 yr. In fact, these times were not exactly 3 and 10 yr, but were allowed to fluctuate by up to
100 days or so about the nominal values. This analysis method helps to illustrate the benefit of very long
mission times with this system.
Picking an arbitrary rendezvous and total trip time usually results in either the outbound journey or
the inbound journey dominating the solar collector size requirement; this does not yield an efficient solu-
tion. It seems logical that there must be some combination of rendezvous time and total trip time that
would result in the collector being just large enough for both the outbound and inbound journeys. These
points should yield the minimum total system mass, since the collector is optimally designed for both
stages of the mission. Figure 108 shows the difference between inbound and outbound mass
requirements for various total trip times. Negative values indicate that the inbound portion of the journey,
which is the asteroid deflection portion, dominates the required solar collector size. For clarity, this plot
does not include all of the total mission times that were examined, but it does show that for some values of
total trip time, the inbound and outbound solar collector sizes can match at more than one rendezvous time.
In such cases, however, the total system for one of the solutions was always significantly less than the
others.
Figure 108. Plot of the difference between required outbound and inbound solar
collector sizes. Negative values mean that the inbound requirement
dominates.
–0.8
–1
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
–500 –400 –300 –200 –100 0
Rendezvous Time (Days Before Impact)
D
i
f
f
e
r
e
n
c
e

i
n

T
o
t
a
l

S
y
s
t
e
m

M
a
s
s

(
t
)
Total Mission
Time (days)
1,000
1,500
2,000
2,500
3,000
140
The points at which the inbound and outbound size requirements are equal nearly always corre-
spond to the points at which the total system mass is a minimum for a specific total mission time. In those
instances where this is not the case, the total system mass is within a few kilograms of the minimum value.
A plot of these locations for various total mission times, such as figure 109, reveals the cyclic nature of the
total system mass; this is much like the other systems considered in this study. The plot also shows that the
local minimum total system mass continues decreasing as the total mission time increases. Analyses far
beyond the 10-yr limit for total mission time resulted in a continued decrease of the minimum total system
mass. However, the objective of this analysis was not to find the global minimum total system mass, but
rather to determine the minimum mass for two total mission times. As stated earlier, these two total mission
times are around 10 and 3 yr—with some slight fluctuation to allow the ModelCenter optimizer to find the
local minima.
0
–50
–100
–150
–200
–250
–300
–350
–400
–450
–500
3,500
3,000
2,500
2,000
1,500
1,000
500
0
0 1,000 2,000 3,000 4,000
Total Mission Time (Days)
T
o
t
a
l

S
y
s
t
e
m

M
a
s
s

(
t
)
R
e
n
d
e
z
v
o
u
s

T
i
m
e

(
D
a
y
s

B
e
f
o
r
e

I
m
p
a
c
t
)
Rendezvous
System Mass
Figure 109. Combinations of total mission time and rendezvous time where
inbound and outbound required solar sail sizes are equal,
and the associated total system mass, for 500-m-diameter chondrite.
The relatively simple ModelCenter model is presented in figure 110. All tools in the figure have
been described in previous sections, except for ssc, which is the solar collector sizing tool. As with the
other models, PBI determines the asteroid deflection requirement, based on the rendezvous time, and
PBO_v3 determines the outbound trajectory requirement based on the outbound time. The solar collector
tool gives both the inbound and outbound required system masses as output; the minimum required system
mass, which is the larger of the inbound and outbound requirements; dimensions of the solar collector;
force on the asteroid; acceleration of the asteroid; and some additional data.
141
Analysis of the solar collector option commenced near the end of the study. With limited time
available for additional tool development, PBI and PBO_v3 were used to determine total DV requirements.
Since PBI and PBO_v3 both determine impulsive DV requirements, the use of these values for analysis of
the solar collector tool required the careful application of some conservative estimates. The force imparted
on the asteroid by the solar collector was determined at a heliocentric distance of 1.5 au and assumed to
remain constant. This constant value, coupled with the inbound flight time, determined the inbound DV
requirement. This allowed the determination of the required solar collector size, provided the DV require-
ment for asteroid deflection was available. This analysis method does not accurately determine the
required solar collector size, but it does highlight the trends in system performance for various total mis-
sion times, rendezvous times, and asteroid sizes.
For the staged chemical/nuclear pulse options, the optimum total mission and rendezvous times
were very insensitive to asteroid size. This is not the case with the solar collector option: the optimum
points vary with asteroid size and the rendezvous time changes considerably. For example, the minimum
total system mass to deflect a 100-m-diameter chondrite occurs at a rendezvous time of 112.5 days before
impact, with a total mission time of 3,636 days. But, the minimum for a 1,000-m-diameter chondrite
occurs at a rendezvous time of 824 days before impact, with a total mission time of 3,711 days. Since
inbound and outbound times are determined by asteroid mass, the size and type of an incoming asteroid
would have to be determined quite accurately before the solar collector system was ever launched from
Earth. Alternatively, a fairly large performance margin would have to be built into the system; i.e., a much
larger than optimal solar collector would be required. For the rendezvous case, the optimum trajectory for
the 100-m-diameter chondrite is illustrated in figure 111.
Figure 110. Diagram of the ModelCenter setup for the solar collector option.
Main
PBI
ssc
PBO–v3
142
After extensive analysis, the required total system mass versus asteroid size for a chondrite was
determined for two cases: total mission time of ª3 yr and total mission time of ª10 yr. These results are
plotted in figures 112 and 113.
Figure 111. Optimum rendezvous trajectory for the solar collector option
for a 100-m-diameter chondrite.
Figure 112. Minimum total system mass and size for the solar collector option
versus chondrite diameter.
–3 –2.5 –2 –1 –0.5 –1.5
–3
–2.5
–2
–1.5
–1
0
–0.5
0.5
1
1.5
2
3
2.5
0 0.5 1 1.5 3 2.5 2
Sun
Earth
Earth perihelion
Asteroid
Asteroid perihelion
Asteroid asc. node
Transfer trajectory
Start transfer
Finish transfer
Asteroid at start
Earth at finish
Impact position
180,000
160,000
140,000
120,000
100,000
80,000
60,000
40,000
20,000
0
0 2,000 4,000 6,000 8,000 10,000
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
Asteroid Diameter (m)
S
o
l
a
r

C
o
l
l
e
c
t
o
r

D
i
a
m
e
t
e
r

(
m
)
T
o
t
a
l

S
y
s
t
e
m

M
a
s
s

(
k
g
)

a
t

E
a
r
t
h

S
O
I
Mass (≈3 yr)
Mass (≈10 yr)
Sail Size (≈10 yr)
Sail Size (≈3 yr)
143
Solar collector size, rather than total system mass, is the limiting factor for this option. Even
for a 10-km-diameter chondrite, the total system mass is well within the 1,000-t limit imposed on the
system mass. However, the diameter of the solar collector is a remarkable 6 km. Unfurling and controlling
a collector of this size presents significant technical challenges. Despite these problems, it is clear that
the solar collector could still be effective in deflecting an incoming asteroid.
Figure 113. Minimum total system mass and size for the solar collector option
versus chondrite diameter for the smaller chondrites.
200
400
600
800
1,000
1,200
1,400
1,600
1,800
2,000
0
0 250 500 750 1,000 1,250 1,500 1,750 2,000
1,000
900
800
700
600
500
400
300
200
100
0
Asteroid Diameter (m)
S
o
l
a
r

C
o
l
l
e
c
t
o
r

D
i
a
m
e
t
e
r

(
m
)
T
o
t
a
l

S
y
s
t
e
m

M
a
s
s

(
k
g
)

a
t

E
a
r
t
h

S
O
I
Mass (≈3 yr)
Mass (≈10 yr)
Sail Size (≈10 yr)
Sail Size (≈3 yr)
144
9. CONCLUSIONS AND RECOMMENDATIONS
A wide range of potential mitigation techniques by which threatening planetary bodies could be
either deflected or fragmented have been modeled in this study. Consideration was also given to a range of
transportation methods by which the mitigation hardware could be moved out to an approaching body for
either rendezvous or impact. Several possible combinations of mitigation techniques and transportation
options have been analyzed in detail.
Although there is much that still needs to be done, conclusions and recommendations are given in
sections 9.1 through 9.6.
9.1 Public Awareness
Despite the best efforts of Hollywood, the level of public awareness of this threat is still not high.
Compared with other comparable threats, planetary body collision is still viewed as being a matter of
science fiction rather than one of scientific fact.
While not advocating steps that could lead to hysteria and panic, the facts about this problem
should be properly presented to the general public so as to raise public understanding of the threat and the
ways in which it can be mitigated. Only when in full possession of the facts can the voting public make an
informed decision about what steps should be taken.
9.2 Statistical Problem
The lack of attention given to this threat is in part due to a statistical problem. The chance of a
significant-sized object striking the Earth is fairly low; such collisions might take place perhaps only once
or twice per century. This has led to the danger being downgraded when compared with other threats
to public safety, particularly those relating to acts of terrorism. However, the probability of an impact
taking place cannot be considered in isolation; proper account must also be taken of the likely
consequences of such a collision. Even the impact of a relatively small body would probably be very
severe with fatalities in the millions, wide-scale destruction, and a recovery time possibly extending
over decades.
To obtain proper assessment of the danger, some appropriate parameters, such as the expected
number of fatalities over a period of, say, a decade or a century, must be considered. Table 29 shows the
chances of death by a variety of causes for a typical resident of the United States. It is interesting to note
that the probability of dying due to a planetary body impact is about the same as that of dying due to an
aircraft crash. As table 29 shows, this approach presents an altogether more worrying perspective on the
danger.
145
9.3 Funding of Future Work
While a number of NEO search activities are currently underway, most are proceeding with very
limited funding. In some countries, government agencies have declined to provide funding and, as
recounted, actually scorned the detection efforts. A strong recommendation was made that funding for
these efforts be increased. In particular, sufficient high-quality instruments must be made available to
conduct an all-latitude observation program with the aim of cataloging the entire NEO population.
Although funding is limited for NEO surveys and searches, at least it is nonzero. By comparison,
the study of mitigation techniques is—with the notable exception of this present effort—almost totally
unfunded. Equally important, research into new mitigation techniques is nonexistent, except in those cases
where the technology is under study for some other application.
A strong recommendation is made that a coherent study of mitigation techniques as well as their
likely effectiveness, cost, and deployment times, be undertaken in the very near future. This study, which
would represent an enlarged follow-on to this work, should involve and call upon all of NASA’s consider-
able resources, as well as those of the DOD, the Armed Forces, and other government agencies;
e.g., the National Oceanic and Atmospheric Agency, the Federal Emergency Management Agency, etc.
International collaboration and funding should also be actively sought.
9.4 Development and Deployment of Mitigation Systems
The technical work undertaken in this study shows clearly that, although the mitigation challenge is
formidable, it is not beyond our capabilities, provided preparations are begun well in advance of an impact.
Despite the impression given by Hollywood, it is not practical to wait until a specific threat is identified
Table 29. Causes of death and associated probabilities for a U.S. resident.
45
Cause of Death Chance
Motor vehicle accident
Homicide
Fire
Firearms accident
Electrocution
Asteroid/comet impact
Passenger aircraft crash
Flood
Tornado
Venomous bite or sting
Fireworks accident
Food poisoning by botulism
Drinking water with EPA limit of tricholoethylene
1 in 100
1 in 300
1 in 800
1 in 2,500
1 in 5,000
1 in 20,000
1 in 20,000
1 in 30,000
1 in 60,000
1 in 100,000
1 in 1,000,000
1 in 3,000,000
1 in 10,000,000
146
before starting work on the mitigation system. Systems engineering, system deployment, and in some
cases, technology development, will take several years.
Most strongly recommended is that, following an appropriate study phase, a development program
be initiated immediately, with a view to deploying an operational system as soon as possible. It is already
clear to us that a first-generation protection system will not be able to counter all possible threats; however,
it should be able to defeat those most likely to occur. At the outset of Project Apollo, it was said that, while
the United States could not guarantee to come first in the race to the Moon, failure to act would guarantee
that she would come last. In the same way, it might be said that, while success in protecting the Earth
against a cosmic impact cannot be guaranteed, failure to act will, in the long run, guarantee a major catas-
trophe of regional, if not global, proportions.
9.5 Accomplishments
It was not the intent of this study to select a particular technical option for recommendation as
a threat mitigation system. Instead, our intent was to study the various options, in several cases, using
improved and updated modeling techniques. It was also our aim to categorize these options into different
mission configurations and to propose a method for comparing the large number of possible combinations
of mitigation options and mission configurations. It was also our intent to recommend future work.
Several new tools were created during the course of this project. None should be regarded as
a finished product and all would benefit from further development and refinement. As an example, the
outbound trajectory tool is designed to give a first approximation of the required DV, using high thrust
calculation methods. Similarly, the inbound tool takes a velocity vector at the point of impact and
integrates the trajectory backward in time until the object is well outside the Earth’s sphere of influence. It
then integrates forward, after a deflection DV has been applied to the object, so as to determine the resulting
miss distance from the Earth. The program iterates until a specified closest approach to the Earth has been
achieved. Both these tools would benefit from the use of more accurate, although more calculationally
intensive, techniques.
Numerous outbound propulsion systems and threat mitigation options were considered and mod-
eled using several tools that were created by combining some basic principles of physics with engineering
data available in the open literature. These tools yielded first approximations for the performance and mass
of each technical option.
The data and tools available in the literature have been built on to create a threat assessment tool
that calculates the percentage of the total threat that can be defeated using a given mitigation system.
A procedure for comparing all these technologies will be put into place in the future. In so doing,
identification of mission categories for these technologies have been made and future analyses simplified
by developing a procedure that deals with each category instead of attempting to deal with each individual
technology combination separately.
147
9.6 Assessment of Mitigation Options
Although it was not the purpose of this study to select mitigation options, a preliminary assessment
is possible. Table 30 summarizes the capability of each major system option.
Table 30. Summation of parametric results for mitigation concepts.
The mass driver was coupled with the staged chemical system to offer a non-nuclear threat mitiga-
tion option. However, the relatively massive mass driver system coupled with the least efficient stored-
chemical system yielded unacceptably high initial masses. The mass driver could have been coupled with
the solar sail/collector but would result in sail sides/diameters in the tens of kilometers. However, the mass
driver is an attractive option for moving asteroids with the ultimate purpose of resource utilization. Thus,
this concept should be carried forward in further studies, perhaps with effort expended to reduce the
required mass driver system mass.
By comparison, the combination of a nuclear blast system and a staged chemical outbound propul-
sion system also offered excellent performance. Once again, it was the staged chemical propulsion system
that limited the system performance.
System Maneuver
Time Before Impact
(days)/Outbound
Travel Time (days)
Total System Mass
at SOI (t)
for Different
Asteroid Diameters
(m)
Maximum Diameter
of Asteroid*(m)/
Total System Mass
at Earth SOI (t)
Staged chemical/
mass driver
Staged chemical/
nuclear deflection
Staged chemical/
kinetic deflection
Nuclear pulse
Solar collector
Rendezvous
Rendezvous
Intercept
Intercept
Rendezvous
Rendezvous
(≈3 yr)
Rendezvous
(≈10 yr)
2,900/2,400 NA NA NA
100 1,000 10,000
50/6,849
80/6,918
73.8 NA NA
29.7 41.8 1,240
0.637 1.07 167
0.550 0.636 34.6
0.847 8.27 1,300 9,000/1,000
9,000/1,000
1,000/1,000
260/1,000
5.62 568 87,800
1,509/910
1,025/800
2,170/1,200
1,076/65**
3,635/115**
1,075/132
§
§
* Maximum was constrained to a total system mass at Earth SOI of 1,000 t.
** Times are for 100-m-diameter chondrite. Rendezvous times are greater for larger asteroids, although total missions times change little.

§
The solar collector system is limited more by solar collector size than by total system mass. See figure 112.
148
A kinetic deflection vehicle, carried out by a staged chemical system, is theoretically capable of
deflecting large asteroidal bodies. However, the interceptor vehicle mass required increases rapidly with
asteroid size. Nonetheless, deflection of a 100-m-diameter asteroid is possible.
The nuclear pulse option performs well because of its use of the same, very effective technique for
both outbound travel and deflection. Of all the options considered during this study, nuclear pulse offers
the best prospect of providing an effective mitigation technique using existing and near-term
technology.
The solar collector system showed itself to be capable, but only at the expense of a very large sail
area and the consequent operational problems. As with the nuclear pulse option, it has the advantage of
using a single unified system for both outbound propulsion and deflection.
Each of these options may well find some application in the future, but our initial results indicate
that the nuclear pulse option offers the best defensive capabilities in the near term. This is by no means a
recommendation but merely an observation based on the data at hand. Due to the level of fidelity and
extensive assumptions that have been forced to be made in this limited study, it is recommend that all
options discussed here, as well as other options suggested elsewhere, be carried forward into a higher
fidelity analysis.
9.7 Future Work
A large amount of future work has been identified. All of our tools would benefit from more
detailed analysis procedures. Many of the assumptions made during the development of our technology
tools are in need of refinement. Our trajectory tools would benefit greatly from the ability to model
continuous thrust propulsion systems. Our threat assessment tool requires more research into the available data
on the asteroid and comet population. As a minor example of this, note that there are suggestions in the
literature that cometary rings, such as the Leonid ring, may have nonuniform densities along their cir-
cumference. Since the Earth passes through such rings on a yearly basis, there would be a synodic period
on which the Earth would cross these higher density areas, yielding a higher probability of
impact. Our threat assessment tool also requires further research into the consequences of an impact.
There are several other mitigation options that were not studied because of resource and time limi-
tations; two of these are of particular interest and merit some mention:
(1) Laser ablation is used as either a remote station or as a rendezvous option. This technique would
allow deflection in a manner similar to that of the solar collector with a beam of high-energy coherent light
being directed at the incoming asteroid or comet.
(2) This second new option involves firing inert masses from a mass driver located in Earth orbit,
perhaps at a Lagrange point. This would combine our mass driver and kinetic deflection/fragmentation
tools and would represent another remote station option.
149
Finally, a method to combine the quantitative results from this analysis needs to be established and
the qualitative issues outlined for each technology in the outbound propulsion and threat mitigation
sections in order to compare architectures.
After completion of the more advanced tools above, including the atmospheric reentry tool
described earlier for fragmentation options, the overall threat assessment flow chart could begin, as
described in figure 79.
9.8 Summary Conclusion
The threat posed by NEOs should be taken very seriously. It is well within humanity’s ability to
effectively defend itself against this threat. Development of the necessary technologies would also offer
considerable synergy with NASA’s other missions aimed at understanding the universe and exploring
space. The planetary defense mission is also one for which NASA is uniquely suited and could potentially
offer the Agency a goal that both fires the public imagination and creates a sense of urgency comparable to
that during the Apollo program in the 1960’s. The goal is to persuade those in positions of authority to
continue the efforts presented here.
150
APPENDIX A—CURRENT NEAR-EARTH OBJECT SEARCH PROGRAMS
A.1 SpaceWatch
The SpaceWatch program is run by The University of Arizona’s Lunar and Planetary Laboratory,
established in 1980. Its primary goal is to explore the various populations of small objects in the solar
system, and to study statistical data for asteroids and comets so as to understand the dynamical evolution of
the solar system.
CCD systems scan the Centaur, Trojan, Main Belt, Trans-Neptunian, and Earth-approaching aster-
oid populations. The principal instruments, located on Kitt Peak, are the Steward Observatory 0.9-m
SpaceWatch telescope and the SpaceWatch 1.8-m telescope. SpaceWatch is a pioneer in the use of CCDs
and automation for asteroid and comet detection.
SpaceWatch currently has the distinction of having detected the smallest known asteroid—
1993 KA2, which is about 4 to 9 m in diameter—and has also observed the closest known approach of an
asteroid to the Earth—1994 XM1, which approached to a distance of ª105,000 km.
SpaceWatch continues to detect some 20 to 30 new NEAs per year.
A.2 Spaceguard
Spaceguard is an international association, established in 1996, to promote and coordinate activi-
ties for the discovery, monitoring, and orbital calculation of NEOs. It is intended to promote study
activities at theoretical, observational, and experimental levels of the physical and mineralogical character-
istics of the minor bodies of the solar system, with particular attention to NEOs. It is also intended to
promote and coordinate a ground network—Spaceguard system—backed up by a satellite network for
discovery, observation, and astrometric and physical studies.
The Spaceguard system is a collection of observatories engaged in NEO observations. There are
currently more than 70 observatories registered, located worldwide. Wide ranges of instruments are in use.
Note that Spaceguard is a coordinating body and that the technology available and effort
expended vary widely between the various participating observatories.
A.3 Lincoln Near-Earth Asteroid Research
Lincoln Near-Earth Asteroid Research (LINEAR) is a Massachusetts Institute of Technology
Lincoln Laboratory program funded by the U.S. Air Force and NASA. Its goal is to demonstrate the appli-
cation of technology, originally developed for the surveillance of Earth-orbiting satellites, to the problem
of detecting and cataloging NEOs that threaten the Earth.
151
Equipment consists of a pair of 1-m-diameter, ground-based, electro-optical deep-space surveil-
lance (GEODSS) telescopes at Lincoln Laboratory’s Experimental Test Site on the White Sands Missile
Range, Socorro, NM. The telescopes are equipped with Lincoln Laboratory-developed CCD electro-
optical detectors and collected data are processed on site to generate observations.
Survey results as of April 2002 are as follows:
Number of observations to minor planet center 7,416,832
Number of asteroid detections 1,127,759
Number of new designations 157,920
Number of confirmed NEOs 951
Number of confirmed comets 82
A.4 Near-Earth Asteroid Tracking
Near-Earth asteroid tracking (NEAT) observatory is an autonomous celestial observatory devel-
oped by JPL and funded by NASA to study asteroids and comets. It is based upon a specially designed
CCD camera.
The principal investigator is Dr. Eleanor F. Helin; co-investigators are Dr. Steven H. Pravdo and
Dr. David Rabinowitz.
NEAT is comprised of two autonomous observing systems at the Maui Space Surveillance Site
(MSSS)—NEAT/MSSS and at the Palomar Observatory—NEAT/Palomar. At both sites, the NEAT
cameras use 1.2-m (48-in) telescopes to find NEOs, NEAs, and comets.
Nine new NEAs were discovered during July 2002: three Amors with one >1 km; five Apollos,
including one PHA and one >1 km, and one Aten.
A.5 Lowell Observatory Near-Earth Object Search
The Lowell Observatory near-Earth object search (LONEOS) system can scan the entire sky every
month, accessible from Flagstaff, AZ. It uses a 0.6-m Schmidt telescope and a CCD detector. It has been in
operation since March 1998. The first new discovery was made on June 18, 1998, and is able to record
objects to a magnitude limit near V=19.3, or ª100,000 times fainter than can be seen with the naked eye.
As of August 2001, LONEOS had submitted more than 1 million asteroid observations to the Inter-
national Astronomical Union Minor Planet Center. It is estimated that, after 10 yr of full-time operation,
LONEOS could discover 500 of the 1 km or larger NEOs and perhaps twice as many smaller NEOs, thus
substantially increasing our knowledge of these bodies.
The asteroid discovery summary as of July 9, 2002, includes 10 Aten, 56 Apollo, 55 Amor, and
6 comets, for a total of 137 asteroids.
152
APPENDIX B—SOLAR ARRAY CALCULATIONS
The solar array forms an important part of the proposed mass driver system. The large distances
anticipated between target asteroids and the Sun, coupled with the generally unfavorable incidence condi-
tions, threaten to make the array one of the most massive system elements. Two distinct methods of sizing
the array have been identified.
B.1 Method I
The electrical power (P) can be written as
P FA = ε ,
(89)
where
F = solar flux (W/m
2
)
A = array area (m
2
)
ε = efficiency.
If σ = mass per unit area, then
A
M
=
σ
,
(90)
where M = array mass.
Thus,
σ
ε
=
FM
P
,
(91)
and hence,
σ
ε
=






F
P
M
.
(92)
For Earth orbit, F = 1,300 W/m
2
. For current generation arrays, ε = 0.15 is achievable, and can be
bettered. Reference 46 (p. 333) gives (P/M) = 14 to 47 W/kg. This implies σ = 4 to 14.
153
B.2 Method II
47
Achievable specific power = 130 W/kg (assumed achievable at Earth with ideal array orientation
with respect to the Sun). Let R
E
= Earth’s orbital radius and R
A
= asteroid orbital radius.
Hence, achievable specific power at asteroid = 130¥(R
E
/R
A
)
2
W/kg. Again, it is assumed that the
array is ideally oriented with respect to the Sun.
Introducing an additional degradation factor (a) to take account of (1) nonideal orientation,
(2) possible asteroid rotation, (3) dust obscuration of array, …, gives
Specific power
W
kg
=
Ê
Ë
Á
ˆ
¯
˜
È
Î
Í
˘
˚
˙
130
2
a
R
R
E
A
.
(93)
154
APPENDIX C—MASS DRIVER
Appendix C contains details of a simple model for the key components of the mass driver system.
C.1 Model of the Forces on a Bucket Coil Due to the Nearby Drive Coils
As explained in section 5.7, although a bucket coil receives an accelerating force from the closest
pair of drive coils, it experiences alternately retarding and attractive forces from each more distant pair
(fig. 114). Of course, the more distant pairs of coils produce lower forces than do the nearer coils. For the
force calculations in this analysis, only the effect of the first two pairs of coils will be considered. The
nearest pair provides the major motive force; the next pair provides the major retarding force. By limiting
consideration to these four coils, one essentially conducts a conservative analysis. This is because all
subsequent drive coils—in theory, stretching out to infinity in both directions—can also be grouped into
sets of four coils; the nearest and next-nearest sets are shown below.
Direction
of Motion
Nearest Set of Four Coils—
Included in Analysis
Next Nearest Set of Four Coils—
Not Included in Analysis
Figure 114. Drive coils included in and omitted from the analysis.
If one examines the coils which comprise the next-nearest set of four; i.e., the first coils to be
excluded in the analysis, it is immediately apparent that the nearer pair exerts a motive force and the
slightly more distant pair exerts a retarding force on the bucket coil. Hence, the net force from this set of
four coils will be a motive one. The same argument can be applied to all subsequent; i.e., more distant, sets
of four coils. Each produces a net motive force on the bucket coil, although declining in magnitude as
distance increases. Hence, by their exclusion, the analysis neglects a portion of the overall motive force.
155
The analysis is conducted by considering the motion of a single bucket coil between two adjacent
drive coils. The results can be multiplied to also include the effect of the remaining three-bucket coils.
Figure 115 shows the bucket coil and surrounding four drive coils.
Figure 115. Drive coils included in the analysis.
The force (F) between any two coils (designated as 1 and 2) is given by
F N N I I
M
x
=


1 2 1 2
,
(94)
where
I
1,2
= current in coil 1,2
N
1,2
= number of turns in coil 1,2
M = mutual inductance between one turn of coil 1 and one turn of coil 2.
Equation (94) can be derived from basic magnetic energy considerations for two interacting current-
carrying coils. Appendix D contains the derivation.
The general expression for M can be shown
46
to be as follows:
M x a a
k
k K
k
E ( ) = ( ) -
Ê
Ë
Á
ˆ
¯
˜ -
È
Î
Í
˘
˚
˙
m
0 1 2
0 5
2 2
.
,
(95)
where
Bucket Coil
Drive Coil
n–1
Drive Coil
n
Drive Coil
n+1
Drive Coil
n+2
S S
x
S
Direction
of Motion
156
c = speed of light
a
1,2
= radius of coil 1,2
k
2
= 4×a
1
×a
2
/[x
2
+(a
1
+a
2
)
2
]
x = distance between the two coils
µ
0
= permeability of free space.
The quantities K and E are, respectively, elliptical integrals of the first and second kind and are
given by:
K k
d
k
( ) ·

( )

·
·
ϕ
ϕ
ϕ
ϕ
π
1
2 2
0 5
0
2
sin
.
(96)
and
E k d k ( ) · −
( )

·
·
ϕ ϕ
ϕ
ϕ
π
1
2 2
0 5
0
2
sin .
.
(97)
Hence, for the case under consideration here, the net force acting on the bucket coil due to the four drive
coils is given by
F N N I I
x
M M M M
D B D B n n n n
·


− + + −
( )
+ + − 2 1 1
,
(98)
where
I
D
= current in drive coil
I
B
= current in bucket coil
N
D
= number of turns per drive coil
N
B
= number of turns per bucket coil
M
i
(x) = mutual inductance between one turn of the bucket coil and one turn of drive coil i (i = n – 1, … ,
n + 2).
Note the minus signs in front of the n + 2d and n – 1st coils, which provide retarding forces, and the
plus signs in front of the nth and n + 1st coils, which provide motive forces.
The mutual inductances are given by
M x a a
k
k K k
k
E k
i D B
i
i i
i
i
( ) · ( ) −
|
.

`
,
( ) − ( )

]
]
] µ
0
0 5
2 2
.
,
(99)
157
where
a
D
= radius of drive coil
a
B
= radius of bucket coil
k
n–1
2
= 4¥a
D
¥a
B
/[(S+x)
2
+ (a
D
+ a
B
)
2
]
k
n
2
= 4¥a
D
¥a
B
/[(x +(a
D
+a
B
)
2
]
k
n+1
2
= 4¥a
D
¥a
B
/[(S–x)
2
+ (a
D
+a
B
)
2
]
k
n+2
2
= 4¥a
D
¥a
B
/[({2¥S}–x)
2
+ (a
D
+a
B
)
2
].
To facilitate calculational procedures, the two elliptical integrals, K(k) and E(k), have both been
curve fitted. This was done using raw data
48
accurate to the fourth decimal place. Both curves were fit to
sixth-order polynomial equations. The resulting curves are shown in figure 116 (denoted by Poly(K(m))
and Poly(E(m)), respectively), superimposed upon the raw data. Note that the independent variable in the
graph is denoted by m, which in terms of the quantities given above, is equal to k
2
. The curve fit equations
are displayed in equations (100) and (101).
Figure 116. Elliptical function curve fits—K(m) and E(m) versus m.
K(m)
E(m)
Poly. (K(m))
Poly. (E(m))
0 0.2 0.4 0.6 0.8 1 1.2
4
3.5
3
2.5
2
1.5
1
0.5
0
F
u
n
c
t
i
o
n
Distance From Bucket Coil to Drive Coil (m)
158
Replacing m with k
2
, the equations are as follows:
K k k k k k
2 2
6
2
5
2
4
2
3
86 7180 226 3862 223 7493 102 6237
( )
=
( )
-
( )
+
( )
+
( )
. . . .
+
( )
-
( )
+ 22 0258 1 4009 1 6024
2
2
2
. . . k k
(100)

E k k k k k
2 2
6
2
5
2
4
2
3
1 9341 4 9415 4 8625 2 1863
( )
= -
( )
+
( )
-
( )
+
( )
. . . .

-
( )
-
( )
+ 0 5435 0 3540 1 5701
2
2
2
. . . . k k
(101)
C.2 Drive and Bucket Coil Currents
Although the force equation is fairly straightforward, the calculations are complicated somewhat
by the variations in both drive and bucket coil currents. The drive current can be written as
I I
x
S
D D
=
0
sin ,
p
(102)
where x is the distance between the nth drive coil and the bucket coil and I
D0
is the maximum current in the
drive coil. In what follows, it is assumed that (I
B
/I
D
) < 1.
Although this relatively simple functional dependence can be assumed for I
D
, the bucket current
(I
B
) is more complex. As the bucket carries no power source, the current will vary according to the flux
supplied by the drive coils. The guiding principle is that the net flux through a bucket coil remains con-
stant. When it first enters the mass driver, before any of the drive coils are energized, the total flux (F
1
)
through the front bucket coil is given by
F
1 0 12 0 13 0 14 0
= - + - N LI N M I N M I N M I
B B B B
, (103)
where
L = self-inductance of a bucket coil
M
1j
= mutual inductance between bucket coil 1 and bucket coil j (where j = 2,3,4)
I
0
= initial current in a bucket coil (before any drive coils are energized).
The self-inductance of a bucket coil is given by:
49
L a
a
r
B
B
o
= -
Ê
Ë
Á
ˆ
¯
˜
m
0
8 7
4
ln , (104)
159
where
r
o
= conductor wire radius.
The mutual inductances between bucket coils are given by
M x a
k
k K k
k
E k
ij B
ij
ji ij
ij
ij
( ) = -
Ê
Ë
Á
ˆ
¯
˜
( )
-
( )
È
Î
Í
Í
˘
˚
˙
˙
m
0
2 2
, (105)
where
k
ii+1
2
= k
ii–1
2
= 4¥a
B
2
/[S
2
+(2¥a
B
2
)]
k
ii+2
2
= k
ii–2
2
= 4¥a
B
2
/[(2¥S
2
)+(2¥a
B
2
)]
k
ii+3
2
= k
ii–3
2
= 4¥a
B
2
/[(3¥S
2
)+(2¥a
B
2
)]
and the K and E functions are the same as those given earlier.
Note that the signs of alternate mutual inductances are reversed to account for the current flow
directions in alternate bucket coils, as shown in figure 117.
Figure 117. Bucket coil current directions.
The expression for total flux can be simplified to
F
1 1 0
= N L I
B
, (106)
Coil Numbers
1 2 3 4
Alternating Current Flow Directions in Bucket Coils
160
where L
1
is the total inductance of bucket coil 1, given by
L L M M M
1 12 13 14
= + + + .
(107)
Inspection of figure 117 shows immediately that the four coils will not all have the same total
inductance; i.e., L
1
through L
4
are not all identical. The more centrally located coils will have higher values
than those on the ends. Coils 1 and 4, the two end coils, will both have the same inductance as will coils 2
and 3, the middle coils. In other words, L
1
= L
4
and L
2
= L
3
, but L
1
π L
2
. Inspection shows that the
differences are not large and that, typically, all four coils are within ª5 percent of the average induction. To
avoid having to develop separate induction models for the end and central coils for the purpose of this
model, the average value is used.
When the bucket undergoes acceleration (fig. 118), the total flux through the coil is still F
1
, but this
is now given in terms of the drive coil current and various mutual inductances between drive and bucket
coils by
F
1 1 3 2 1 1
= + - + - +
( )
+ + + -
N L I N I M M M M M
B B D D n n n n n
.
(108)
Figure 118. Bucket and drive coil current directions during acceleration.
Note the signs before each of the mutual inductances. The drive coils that are energized in the same sense
as the bucket coil will add to the flux. Those energized in the opposite sense will subtract from it. Unlike
the motive force equation, which only took into account the n–1, n, n+1, and n+2 drive coils, the effect of
the n+3 coil is also included. This is because it adds to the flux through the bucket coil and hence dimin-
ishes its current. The lower current means that the motive force is reduced. This is in keeping with the
conservative intent throughout this analysis. The n–2 drive coil is not included because it serves to increase
the bucket coil current.
Bucket Coil
Drive Coil
n–1
Drive Coil
n–2
Drive Coil
n
Drive Coil
n+1
Drive Coil
n+2
Drive Coil
n+3
S S S S
x
S
Direction
of Motion
161
Eliminating Φ
1
gives
I I N I
M M M M M
N L
B D D
n n n n n
B
= −
− + − +
+ + + −
0
3 2 1 1
1
(109)
as the bucket coil current. As explained earlier, at any instant, all four-bucket coils have current of the same
magnitude, but with flow direction alternating between successive coils.
C.3 Analysis of Bucket Kinetic Energy and Acceleration
The total mechanical work done on the bucket coil as it moves from the nth to the n+1st drive coil;
i.e., from x=0 to x=S, is given by
∆E Fdx
x
x S
= ∫
=
=
0
.
(110)
Although there are now expressions for I
D
, I
B
, and the various mutual inductances, analytical inte-
gration of the force equation is impractical. A numerical integration, using the following simple difference
equation, is conducted instead:
δ δ δ KE F x N N I I M M M M
D B D B n n n n
= = − + + −
( )
+ + − 2 1 1
,
(111)
where δKE is the incremental increase in bucket kinetic energy between x and x + δx. This can be
rewritten as
δ δ δ KE x x x N N I x I x M x x M x
D B D B n n
→ +
( )
=
( ) ( )
− +
( )

( ) [ ] {
+ + 2 2
+ + ( ) − ( ) [ ]
+ + ( ) − ( ) [ ]
+ +
M x x M x M x x M x
n n n n 1 1
δ δ
− + ( ) − ( ) [ ] }
− −
M x x M x
n n 1 1
δ ,
(112)
where
I x I
x
S
D D
( ) =
0
sin
π (113)
and
I x I N I
M x M x M x M x M x
N L
B D D
n n n n n
B
( ) = −
( ) − ( ) + ( ) − ( ) + ( )
+ + + −
0
3 2 1 1
1
.
(114)
162
These three equations can be integrated in a stepwise fashion to determine the total kinetic energy increase
that the bucket experiences due to a single one of its coils as it traverses from x=0 to x=S. This is then
multiplied by 4 to account for all the bucket coils, thus giving the total increase in bucket kinetic energy,
denoted by DKE.
Note that the model is completely independent of the bucket speed. This means that DKE is inde-
pendent of the bucket speed. For a design such as this, with equally spaced drive (and bucket) coils, this
means that the same increase in kinetic energy is experienced between each pair of drive coils. Thus, one
can write the final bucket speed (v), at the end of the accelerating portion of the mass driver, as
v N
KE
M
DC
B
ª 2
D
,
(115)
where N
DC
is the total number of drive coils. The approximate equality symbol is used in the equation
because the model employed here clearly does not apply at the very start of the mass driver, when there are
no drive coils behind the bucket. Similarly, it does not apply at the very end of the acceleration portion
when there are no more accelerating coils ahead of the bucket. However, provided there is a sufficiently
large number of drive coils, say at least 20, the above expression should suffice.
Note that although the bucket acceleration is not constant as it moves between two adjacent drive
coils, the force profile climbs from zero at x=0, to maximize at approximately x=S/2, and then subse-
quently declines. The average acceleration between any two adjacent drive coils is the same all the way
along the coil gun if one neglects end effects.
This can be seen as follows:
dv
dt
dx
dt
dv
dx
v
dv
dx
d v
dx
= = =
( )
2
2 /
,
(116)
where x is the distance along the mass driver traveled by the bucket. If one denotes the mean acceleration
between two drive coils by ·dv/dtÒ, then between the ith and i+1th drive coils,
dv
dt
v v
S
i i
=
-
+
1
2
1
2 2
,
(117)
and thus
dv
dt
KE M
S
B
=
1
2
D /
,
(118)
which is a constant for all i. Hence, the average acceleration can be treated as being constant.
163
C.4 Drive Coils
Drive coil current has been expressed as a function of the distance (x) traveled by the bucket
between two drive coils:
I x I
x
S
D D
( )
=
0
sin .
p
(119)
In practice, of course, the current must be expressed as a function of time.
If one considers a single drive coil, the ideal current versus time profile would be as depicted in
figure 119. By following a square profile, the drive coil is always at the maximum value to attract or repel
a bucket coil. In practice, a square profile will probably not be achievable and so, instead, assume a simple
oscillating inductance-capacitance-resistance (LCR) circuit. If the effects of resistance are neglected, the
ideal sinusoidal current versus time profile is as shown in figure 120.
Figure 119. Ideal current versus time profile for a single drive coil.
Drive
Coil
Current
Before Bucket
Arrives
Current = 0
After Bucket
Leaves
Current=0
Current at
Maximum Positive
Value for Transit of
Bucket Coil No. 1
Current at
Maximum Positive
Value for Transit of
Bucket Coil No. 2
Current at
Maximum Positive
Value for Transit of
Bucket Coil No. 4
Current at
Maximum Positive
Value for Transit of
Bucket Coil No. 3
Time
164
In theory, this type of current profile can be produced relatively easily by incorporating the drive
coil into the type of circuit shown in figure 121.
Switch
Capacitor
DC
Power
Drive Coil
Figure 121. Oscillating drive circuit—charging capacitor.
The capacitor is charged from a DC power source (the switch is shown in fig. 121 in the charging
position) and is then discharged to produce the required coil current, as shown in figure 122.
DC
Power
Figure 122. Oscillating drive circuit—discharging capacitor.
Figure 120. Sinusoidal current versus time profile for a single drive coil.
Drive
Coil
Current
Before Bucket
Arrives
Current = 0
After Bucket
Leaves
Current=0
Positive Current
for Transit of
Bucket Coil No. 1
Negative Current
for Transit of
Bucket Coil No. 2
Negative Current
for Transit of
Bucket Coil No. 4
Positive Current
for Transit of
Bucket Coil No. 3
Time
165
The LCR of the circuit are selected to give the required oscillation frequency, which will be differ-
ent for each drive coil—earlier coils having a lower frequency than later coils. Resistance and other
energy-loss mechanisms will result in a decaying current profile, as shown in figure 123.
Figure 123. Real current versus time profile for a single drive coil.
Following the transit of one bucket, the circuit is switched back to the DC power source and the
capacitor is charged again in preparation for the next bucket. Note that, although adequate for this simple
model, the above drive coil circuit will need to be modified for practical use. This is because, unless (I
B
/I
D
)
«1, the drive current’s sinusoidal profile for a simple circuit will be disrupted by the intrusion of bucket coil
flux. In practice, to prevent this from happening, the drive coil electrical circuit will probably need to be
fairly complex, with multiple capacitors being discharged in a carefully timed sequence as the bucket coil
passes.
Drive coil circuit oscillation frequencies will vary along the length of the mass driver. As a coil’s
frequency is set by the time taken for the bucket to travel the standard intercoil distance (S), the earlier coils
will have a lower frequency than the later coils. This means that each coil’s circuit must be designed to
produce the frequency (n) that is appropriate for the bucket’s speed when it passes the coil:
n =
v
S
i
,
(120)
where v
i
is the mean speed of the bucket as it traverses the distance 2¥S, centered on the ith bucket.
This frequency-specific aspect of the drive coil design has two implications:
(1) A given drive coil interacts with each of the four bucket coils in succession. Its drive circuit
should ideally be able to change its frequency slightly as each bucket coil passes. This is because the
bucket speed will increase slightly as successive coils pass a given drive coil, in principle, necessitating a
slightly higher frequency. The possibility of modifying drive circuit design to facilitate this should be
investigated.
Drive
Coil
Current
Before Bucket
Arrives
Current = 0
After Bucket
Leaves
Current=0
Negative Current
for Transit of
Bucket Coil No. 2
Negative Current
for Transit of
Bucket Coil No. 4
Positive Current
for Transit of
Bucket Coil No. 1
Positive Current
for Transit of
Bucket Coil No. 3
Time
166
(2) The total mass of the loaded bucket must not vary significantly between “shots” of the mass
driver. This means that the mass of expellant added to the bucket must be accurately metered. Use of a
bucket whose empty mass is significantly greater than the expellant mass it carries would clearly help
reduce this sensitivity. An alternate approach, provided that some variation in drive coil frequency were
possible, would be to weigh the bucket by some means after it has been loaded with expellant. This might
be done by vibrating the bucket and measuring either its amplitude or response frequency.
C.5 Braking Coils
Analysis of the braking process is conducted using exactly the same equations as for the accelera-
tion process. The sole difference is that all of the stationary coil currents are reversed.
C.6 Bucket Design
The bucket is assumed to be of cylindrical shape. As shown in figure 124, it consists of an inner
structural layer surrounded by a dewer that contains liquid nitrogen (LN
2
) and the superconducting coils.
Figure 124. Bucket conceptual design.
The bucket shape is determined by the ratio of the internal height to the internal radius,
h =
h
r
;
(121)
thus, the internal volume (V), available to hold expellant, is given by
V r =hp
3
. (122)
Dewer
(With Insulation)
Bucket Structure
r
h
Bucket Coil LN
2
Reservoir
167
The radius is given in terms of the volume and h by
r
V
=
Ê
Ë
Á
ˆ
¯
˜
hp
1
3
.
(123)
The size of the LN
2
reservoir is dictated by the total heat dissipated from the bucket coils during
one acceleration-deceleration cycle. The LN
2
will warm slightly during the acceleration-deceleration pro-
cess due to the fact that bucket coil currents will vary due to inductive interaction with the stationary coils.
If the total amount of heat generated is denoted by DQ and the allowable LN
2
temperature
extremes are T
L
and T
H
, then, assuming that the thermal capacity of the coils may be neglected compared
to that of the LN
2
,
DQ T T M C
H L
= - ( )
LN LN
2 2
,
(124)
where M
LN
2
is the mass of LN
2
and C
LN
2
is specific heat.
Hence, the minimum mass of LN
2
required is
M
Q
T T C
H L
LN
LN
2
2
=
-
( )
D
.
(125)
The upper temperature limit is set by the critical temperature of the superconducting coils; i.e., the
temperature above which their superconducting properties decline. The large number of so-called high-
temperature superconductors now available, with critical temperatures at or above that of LN
2
at
1 atm, suggest that the normal LN
2
boiling point (77.4 K) as a good value for T
H
.
To minimize the mass of the bucket, it is clearly desirable to minimize the mass of LN
2
that must be
carried. However, there are limits to the value of T
L
that are achievable. Some explanation of how the entire
system works is appropriate here. Having completed its deceleration, the bucket LN
2
reservoir will be near
its maximum temperature (T
H
). The temperature will rise by an additional small amount as it makes its way
via the return leg of the driver, back to the expellant loading hopper. This is simply because the ambient
temperature on the asteroid surface is likely to be 50–100 K higher than that of the LN
2
reservoir. Note that
the bucket design must ensure that the reservoir temperature never exceeds T
H
.
When the bucket has returned to the hopper, it will be placed into thermal contact with a cold plate
at a sufficiently low temperature and for a sufficiently long time period to restore its LN
2
reservoir to T
L
.
For reference, the entire bucket return concept, including reservoir cooling and expellant loading, is shown
in figure 125.
168
The flow of heat from the bucket LN
2
reservoir to the cold plate can be written as
dQ
dt
Kr T T
CP
= -
( )
2
, (126)
where K is a constant, r is the bucket external radius, which will probably determine the contact area with
the cold plate, T is the instantaneous temperature of the bucket LN
2
, and T
CP
is the temperature of the cold
plate. It is assumed that the cold plate is attached to a thermal reservoir and that its temperature may be
taken as a constant.
For a given DQ, the above equation simply shows that the time which needs to be spent in contact
with the cold plate decreases with increasing contact area; i.e., with r
2
, and also decreases with decreasing
cold plate temperature (T
CP
).
This equation can be solved to give the time (t) that the bucket must spend in contact with the cold
plate in order to have its reservoir temperature lowered from T
H
to T
L
:
t = -
-
-
Ê
Ë
Á
ˆ
¯
˜
H
Kr
T T
T T
C L CP
H CP
2
ln , (127)
where H
C
is the total thermal capacity of bucket coils and LN
2
.
Figure 125. Bucket handling through a complete cycle.
Bucket Rotates and Fully
Engages With Frame
Bucket Emerges From
Final Braking Coils
and Partially Engages
With Handling Frame
Bucket Coils
Engage With
Acceleration Coils;
Frame Releases
and Returns to the Bucket
Pickup Point
Frame Moves
Bucket to Start
of Mass Driver
Bucket Engages
With Hopper and
With Cold Plate
Cold Plate
Expellant
Frame Moves
Bucket on Return Path
Cycle of Operation
For Bucket
Hopper
169
The amount of heat dissipated in the bucket coils will, in some way, be proportional to the total time
taken to accelerate and decelerate the bucket. Heat dissipation in the drive and braking coils is handled
somewhat differently.
The current in a drive coil is given by
I x I
x
S
D D
( ) =
Ê
Ë
Á
ˆ
¯
˜
0
sin .
p
(128)
If R is the resistance of the coil, then the total heat dissipated as the drive circuit oscillates through
one cycle is given by
Heat =
=
=
Ú
I Rdt
t
t
2
0
2p
w
; (129)
hence,
Heat = ( )
=
=
Ú
I R t dt
D
t
t
0
2 2
0
2
sin . w
p
w
(130)
Therefore,
Heat =
=
=
Ú
I
R
xdx
D
x
x
0
2 2
0
2
w
p
sin ; (131)
thus,
Heat =
=
=
Ú
2
0
2 2
0
I
R
xdx
D
x
x
w
p
sin . (132)
Hence,
Heat = 2
2
0
2
I
R
D
w
p

= p
w
I
R
D0
2
.
(133)
170
Now in practice, each drive coil will probably oscillate through about four complete cycles.
Thus,
Total heat dissipated = 4
0
2
p
w
I
R
D
.
(134)
C.7 Interference Between Adjacent Drive Coils
As outlined earlier, it is anticipated that the drive coils could each be part of a simple inductance-
capacitance circuit, tuned to the frequency appropriate to each coil’s location along the length of the mass
driver. This raises the question of how adjacent coils, which will have current flowing in opposite direc-
tions (fig. 126) will affect each other. This is a significant question given that, to avoid lengthening the
bucket, which must be at least four coil spacings long, the coils must be relatively close to each other. The
spacing would almost certainly need to be less than the stationary coil diameter.
Before considering the case of two interacting coils, it is useful to consider first a single isolated
coil. For simplicity, the effects of resistance are neglected; they would be very minor anyway for supercon-
ducting coils.
The discharging capacitor circuit is as shown in figure 126.
As the capacitor begins to discharge, the increasing current induces an increasing magnetic flux
(F) through the circuit (fig. 126). The increasing flux induces a back emf, acting against the direction of
the increasing current:
Figure 126. Discharge of isolated drive coil.
Back emf =
d
dt
F
.
(135)
Magnetic
Field
I
+Q
–Q
V
1
V
2
171
From simple capacitor theory, one can see that
V V
Q
C
2 1
= - ,
(136)
where C denotes the capacitance of the capacitor. From consideration of the back emf,
V V
d
dt
2 1
= -
F
;
(137)
thus,
d
dt
Q
C
F
= .
(138)
The capacitor charge and the discharge current are related via
I
dQ
dt
= - ;
(139)
so,
d
dt
I
C
2
2
F
= - .
(140)
The flux can be expressed in terms of the current and the circuit inductance (L) via
F= LI .
(141)
To simplify the equations, the number of turns in the coil is assumed to have been taken into
account in deriving the L value. This gives
d I
dt
L
CI
2
2
1
0 + = ,
(142)
which is the equation for simple harmonic motion and can be solved to give continuously oscillating
solutions of the form
I I e
i t
=
0
w
,
(143)
172
where I
0
is the amplitude of the current oscillation and w its angular frequency, given by
w =
Ê
Ë
Á
ˆ
¯
˜
1
0 5
LC
.
.
(144)
Now consider the case of two adjacent coils with instantaneous AC flow directions, as shown
in figure 127.
Figure 127. Current flow directions of two adjacent drive coils.
The interaction between these two coils can be described as follows. The total flux through coils 1
and 2 can be written as:
F
1 1 1 2
= + L I MI
(145)
and
F
2 2 2 1
= + L I MI ,
(146)
where
F
1,2
= total flux through coils 1 and 2
L
1,2
= self-inductance of coils 1 and 2
I
1,2
= current in coils 1 and 2
M = mutual inductance of coils 1 and 2, which, as the coils are both fixed, remains a constant.
To simplify the equations, the number of turns in each coil is assumed to have been taken into
account in the L and M values.
Coil 1 Coil 2
173
As for the case of the single coil, equation (147) can be written as
d
dt
I
C
2
1
2
1
F
= - ,
(147)
from which F
1
can be eliminated to give
L
d I
dt
M
d I
dt
I
C
1
2
1
2
2
2
2
1
0 + + = .
(148)
Similarly, for circuit 2,
L
d I
dt
M
d I
dt
I
C
2
2
2
2
2
1
2
2
0 + + = .
(149)
Writing
I I e
i t
1 10
=
w
(150)
and
I I e
i t
2 20
=
w
,
(151)
one can derive
I
C
L M I
1
2 2
1
1
-
È
Î
Í
˘
˚
˙
= w w
(152)
and
I
C
L M I
2
2 2
2
1
-
È
Î
Í
˘
˚
˙
= w w .
(153)
These can be solved to give
1
2
2
2
2
C
L M -
È
Î
Í
˘
˚
˙
=
( )
w w ;
(154)
174
hence,
w
2
1
=
±
( ) C L M
.
(155)
Substituting into one of the earlier equations relating the two currents gives
I I
1 2
= ± .
(156)
Hence, when the two currents flow in opposite directions, which is the situation here for adjacent drive
coils, the minus sign applies and so
w
2
1
=
-
( ) C L M
.
(157)
This means that the frequency is shifted upward when compared to an isolated coil that had
w
2
1
=
CL
.
(158)
In addition to the frequency shift, the current will differ from that in an isolated coil. This can be
seen by taking the expression
I I e
i t
1 10
=
w
(159)
and integrating with respect to time to get an expression for the charge on the capacitor (Q
1
):
Q
I
i
e
i t
1
10
=
w
w
,
(160)
where I
10
is the maximum current. Hence, the maximum charge on the capacitor (Q
10
) is given by
Q
I
i
10
10
=
w
.
(161)
This means that, for a given maximum capacitor charge, the maximum current (I
10
) is given by
I i Q
10 10
= w .
(162)
175
Hence, for an isolated coil, the maximum current will be given by
I
Q
CL
10
10
0 5
=
( )
.
,
(163)
whereas for an interacting coil,
I
Q
C L M
10
10
0 5
=
- ( ) ( )
.
.
(164)
Hence, the discharge current is increased by a factor of
CL
C L M M
CL
- ( ) ( )
=
-
Ê
Ë
Á
ˆ
¯
˜
0 5 0 5
1
1
. .
(165)
because of the interaction between the two coils. If the interaction is modeled between a coil and those with
counter-running current on either side, the factor will be increased further to
1
1
2
0 5
-
Ê
Ë
Á
ˆ
¯
˜
M
CL
.
.
(166)
Note that, as the bucket proceeds along the mass driver, the drive coils are activated as they are
required. Now from the above expression, it can be seen that the frequency of a coil will be somewhat
lower when it is the endmost activated coil. As soon as it has energized neighbors on both sides, its
frequency will increase. To avoid having this frequency shift take place while the bucket is very close, it
may be necessary to activate each coil about a half period in advance and deactivate it about a half period
after the ideal start and stop points.
176
C.8 Effect of Bucket Coil Motion on Stationary Coil Circuit Operation
A stationary coil can be represented as an L-C circuit as shown in figure 128.
Switch
Capacitor
DC Power
System
Stationary Coil
Figure 128. Simple model of a drive coil circuit.
For present purposes, it is not important to distinguish between the case of a drive coil and that of
a braking coil. In the former case, the DC power system provides power to the stationary coil circuit; in the
latter case, the DC power system accepts power from it.
Consider the situation where a bucket coil interacts with the stationary coil. As before,
d
dt
I
C
2
2
F
= -
(167)
for the stationary coil. In this case, the total flux is given by
F= + ¢ LI MI ,
(168)
whereas before, L is the stationary coil self-inductance. The mutual inductance between stationary and
bucket coils is denoted by M. As the bucket coil is in motion, this quantity does not remain constant. The
current in the bucket coil circuit is denoted by I¢. Note that, if the bucket coil current flows in the same
direction as that of the stationary coil, I¢ > 0.
Hence,
I
C
d
dt
LI MI + + ¢
( )
=
2
2
0 .
(169)
Now for the bucket coil, one can write
¢ = ¢ ¢ + F L I MI ;
(170)
177
thus,
¢ =
¢ -
¢
I
MI
L
F (171)
This enables one to eliminate I¢ from the main equation to give
I
C
d
dt
LI M
MI
L
+ +
¢ -
¢
Ê
Ë
Á
ˆ
¯
˜
=
2
2
0
F
,
(172)
which can be rewritten as
I
C
d
dt
L
M
L
I M
L
+ -
¢
Ê
Ë
Á
Á
ˆ
¯
˜
˜
+
¢
¢
Ê
Ë
Á
Á
ˆ
¯
˜
˜
=
2
2
2
0
F
.
(173)
This can be expanded to give
L
M
L
d I
dt
dI
dt
d
dt
L
M
L
I
C
d
dt
L
M
L
d
dt
M
L
-
¢
È
Î
Í
Í
˘
˚
˙
˙
+ -
¢
È
Î
Í
Í
˘
˚
˙
˙
+ + -
¢
Ê
Ë
Á
Á
ˆ
¯
˜
˜
È
Î
Í
Í
˘
˚
˙
˙
= -
¢
¢
Ê
Ë
Á
ˆ
¯
˜
2 2
2
2 2
2
2 2
2
2
1 F
.
(174)
Now L and L¢ are both constants. M depends upon the distance between the two coils, which is itself a
function of time (t), and F¢ is a constant; hence, the equation can be rewritten as
f t
d I
dt
f t
dI
dt
f t I f t
1
2
2
2 3 4
( ) + ( ) + ( ) = ( ) ,
(175)
where
f t L
M
L
1
2
( ) = -
¢
f t
L
d M
dt
2
2
2
( ) = -
¢
( )
178
f t
C L
d M
dt
3
2 2
2
1 1
( ) = -
¢
( )
f t
L
d M
dt
4
2
2
( ) . = -
¢
¢
F
Now, sinusoidal solutions with changing amplitude can be expressed as
I I e
i t
=
+ ( )
0
a w
.
(176)
These solutions will only emerge from the complimentary equation corresponding to equation (169):
f t
d I
dt
f t
dI
dt
f t I
1
2
2
2 3
0
( )
+
( )
+
( )
= .
(177)
Solving this for solutions of the type
I Ae
t
=
l
,
(178)
where l is a constant, one obtains
f t f t f t
1
2
2 3
0 ( ) + ( ) + ( ) = l l ;
(179)
hence,
l =
- ( ) ± ( ) - ( ) ( )
( )
f t f t f t f t
f t
2 2
2
1 3
1
4
2
.
(180)
Now in order that the oscillating solutions be obtained, it is necessary that the inequality
4
1 3 2
2
f t f t f t ( ) ( ) > ( )
(181)
be satisfied. If this is not met, then the stationary coil current will not oscillate and the mass driver, as
designed here, will not function properly. For present purposes, simply assume that satisfying this inequal-
ity is a challenge for the detailed design.
179
If this condition is satisfied, then
l =
-
( )
( )
±
( ) ( )
-
( )
( )
f t
f t
i f t f t f t
f t
2
1
1 3 2
2
1
2
4
2
,
(182)
which is of the form
l a w = ± i ,
(183)
with
a = -
( )
( )
=
¢
( )
-
¢
Ê
Ë
Á
Á
ˆ
¯
˜
˜
f t
f t
L
d M
dt
L
M
L
2
3
2
2
2
2
2
.
(184)
Clearly, the sign of the above expression in a particular situation will determine whether the
stationary coil current amplitude is increasing or decreasing. Before proceeding further, it is useful to
consider the sign of the denominator in equation (184). First note that
F= + ¢ LI M
(185)
and
¢ = ¢ ¢ + F L I MI ,
(186)
which means that
F
F
-
¢
¢
= -
¢
Ê
Ë
Á
Á
ˆ
¯
˜
˜
M
L
I L
M
L
2
;
(187)
hence,
L
M
L
M
L
I
-
¢
=
-
¢
¢
2 F
F
.
(188)
180
Now, from the relative sizes of the drive and bucket coils as well as their relative currents, it can be
said that
F F > ¢
and
M

<1 .
(189)
Hence, it follows that
L
M
L
-
¢
>
2
0 .
(190)
Thus far, whether the stationary coil is a drive or a braking coil has not been specified. Also, it has
not been stated whether the bucket coil is approaching or receding from the stationary coil.
C.9 Drive Coil
The following scenarios are for the bucket coil as it approaches and recedes from the drive coil:
• Bucket coil approaches drive coil: In this case, dM/dt > 0; hence, d(M
2
)/dt > 0. This means
that a > 0, and the current amplitude increases.
• Bucket coil recedes from drive coil: In this case, dM/dt < 0; hence, d(M
2
)/dt < 0. This means
that a < 0, and the current amplitude decreases.
To summarize, the drive coil current amplitude increases as the bucket coil approaches and
decreases as it recedes. At first sight, this seems to imply a symmetry in the process, and might lead one to
conclude that the drive coil’s initial and final states are identical. Closer consideration shows that this is not
the case. The magnitude of d(M
2
)/dt is set by the bucket speed. The faster the bucket travels, the greater
will be |d(M
2
)/dt|. Now, as the bucket is accelerating throughout, its speed of recession will be greater than
its speed of approach. Consider two positions equidistant from and on either side of the drive coil, as
shown in figure 129.
181
At the first (left-hand) bucket coil position, d(M
2
)/dt is positive. At the second (right-hand) bucket
coil position, d(M
2
)/dt is negative and of greater magnitude than it was at the first position. Thus, although
these two positions are symmetric about the drive coil, the decrease in drive current amplitude at the
second position exceeds the increase at the first position. This means that, after the bucket coil has passed,
the drive coil current amplitude will be lower than it was initially. Hence, energy has been extracted from
the drive coil circuit, despite the fact that resistance is neglected, simply by virtue of the bucket
acceleration.
C.10 Braking Coil
The following scenarios are for the bucket coil as it approaches and recedes from the braking coil:
• Bucket coil approaches braking coil: In this case, dM/dt > 0; hence, d(M
2
)/dt > 0. This means
that a > 0, and the current amplitude increases.
• Bucket coil recedes from braking coil: In this case, dM/dt < 0; hence, d(M
2
)/dt < 0. This means
that a < 0, and the current amplitude decreases.
These results are identical to those for the drive coil. The difference is that, here the bucket is
decelerating, so the amplitude increases while the bucket approaches and exceeds its decrease while the
bucket recedes. Hence, energy is supplied to the braking coil circuit by virtue of the bucket deceleration.
Note that, although the bucket coil current rises and falls because of induction, the total magnetic
flux through a bucket coil remains constant throughout the acceleration and deceleration processes. Hence,
once again, neglecting resistance, the initial and final bucket coil currents are the same and there is no
change in energy.
Figure 129. The effect of bucket acceleration.
Drive Coil
Bucket Coil
Approaching
at Speed V
Bucket Coil
Receding
at Speed V+δV
x x Direction
of Motion
182
To summarize what is shown above, even with a traditional zero-resistance model, which for an
isolated circuit just produces an undamped sinusoidal oscillation, energy is lost from a drive coil circuit
and supplied to a braking coil circuit simply due to the bucket acceleration and deceleration.
C.11 Self-Induced Coil Stresses
A current-carrying coil loop will experience a radially outward force due to the interaction of its
current with its self-generated magnetic field. The geometry is depicted in figure 130.
Figure 130. Self-induced magnetic field and resulting force.
The radial force on the entire coil (F) is best calculated by considering the change in energy associ-
ated with a slight increase in coil radius.
The total magnetic energy (U) of a system of N circuits is given by
U I
i i
i
N
= Â
=
1
2
1
F ,
(191)
where I
i
is the current through circuit i and F
i
is the flux through circuit i.
In this case, each circuit is a turn of the coil; hence, all are identical. This means that
U NI =
1
2
F ,
(192)
where I and F are the current and flux through each circuit, respectively.
Coil Center
B Field
B Field
Radially Outward Force on Loop Element
Force
Current
Coil Loop With Clockwise-Directed Current
and Self-Induced B Fields
(Directed Into Plane of Figure)
183
Now, the flux through one turn; i.e., one circuit, of the coil is due to the combined effects of the
fields produced by each of the N turns. Hence,
F= NLI ,
(193)
where L is the self-inductance of a single turn; thus,
U LN I =
1
2
2 2
.
(194)
Therefore,
F I N
dL
dR
=
1
2
2 2
,
(195)
where R the radius of the coil.
The self-inductance of a single turn is given by equation (104):
L R
R
r
=
Ê
Ë
Á
ˆ
¯
˜
-
È
Î
Í
Í
˘
˚
˙
˙
m
0
0
8 7
4
ln ; (196)
hence,
dL
dR
R
r
=
Ê
Ë
Á
ˆ
¯
˜
-
È
Î
Í
Í
˘
˚
˙
˙
m
0
0
8 3
4
ln . (197)
Thus, the total radial force on the entire coil is given by
F I N
R
r
=
Ê
Ë
Á
ˆ
¯
˜
-
È
Î
Í
Í
˘
˚
˙
˙
1
2
8 3
4
2 2
0
0
m ln . (198)
The radial force per unit length of hoop per coil (f) is given by
f
F
RN
=
2p
;
(199)
184
thus,
f
R
I N
R
r
=
Ê
Ë
Á
ˆ
¯
˜
-
È
Î
Í
Í
˘
˚
˙
˙
1
4
8 3
4
2
0
0
p
m ln . (200)
The tension (T) within a loop is thus given by
T R f = ;
(201)
so,
T I N
R
r
=
Ê
Ë
Á
ˆ
¯
˜
-
È
Î
Í
Í
˘
˚
˙
˙
1
4
8 3
4
2
0
0
p
m ln . (202)
185
APPENDIX D—COIL FORCE MODEL
Consider an interacting set (N) of current-carrying circuits, each with inductance, capacitance, and
resistance, and each subject to an externally imposed voltage. For circuit k, one can write
V V R I
d
dt
k
ext
k k k
k
+
( )
- - =
F
0 ,
(203)
where
V
k
ext
= voltage applied to circuit k from external sources
V
k
= voltage across the capacitor of circuit k
R
k
= resistance of circuit k
I
k
= current in circuit k
F
k
= total magnetic flux through circuit k, from both the circuit itself and all the other N–1 circuits.
The capacitance emf can be related to the current by
I C
dV
dt
k k
k
= - ,
(204)
where C
k
is the total capacitance of circuit k.
The total available power in the entire system (W) is given by
W V I R I
k
ext
k
k
k k
k
= Â -Â
2
,
(205)
where the sum over all values of k is denoted by .
k
Â
Equation (203) can be used to eliminate V
k
ext
from equation (205) to give
W I
d
dt
I V
k
k
k
k k
k
= Â -Â
F
,
(206)
186
which can be rewritten, using equation (204) to give
W I
d
dt
d
dt
C V
k
k
k
k k
k
= Â + Â
F 1
2
2
,
(207)
where it is assumed that none of the capacitances vary with time, even though the circuits may move
relative to one another. In view of the ultimate application of this model, this assumption may seem ques-
tionable; however, it is only the fixed stationary circuits that have large capacitors built into their
circuits. The bucket coils will have relatively low capacitances.
Noting that the total electrostatic energy (U
e
) is given by
U C V
e k k
k
= Â
1
2
2
.
(208)
This gives
W I
d
dt
dU
dt
k
k
k
e
= Â +
F
.
(209)
Now, in general, the quantity W can be decomposed as follows:
W
d U U
dt
W
e m
mech
=
+
( )
+ ,
(210)
where the total magnetic energy, denoted by U
m
and W
mech
, represents the mechanical work done in
moving the circuits about.
Combining equations (209) and (210) gives
dU
dt
W I
d
dt
m
mech k
k
k
+ = Â
F
.
(211)
Next, note that the total magnetic flux through a circuit is comprised of contributions due to its own
inductance as well as that of each of the remaining circuits. Hence,
F
k ki i
i
L I = Â ,
(212)
where L
ki
denotes the mutual inductance on circuit k due to circuit i.
187
Hence, equation (211) can be written as
dU
dt
W I
d L I
dt
m
mech k
ki i
i k
+ =
( )
  .
(213)
Now, if one considers a restricted situation in which all the circuits are held securely and cannot
engage in any relative motion, then it is clear that:
W
mech
= 0 and dL
dt
ki
= 0
(214)
for all values of k and i.
In this case,
dU
dt
I L
dI
dt
m
k ki
i
i k
= Â Â ,
(215)
from which it follows that
dU
dt
d
dt
I L I
m
k ki i
i k
= Â Â
1
2
.
(216)
This permits one to write the following expression for the magnetic energy. The expression is analogous to
that for electrostatic energy:
U I L I
m k ki i
i k
= Â Â
1
2
.
(217)
Now, consider the more general situation, where the circuits are permitted to engage in relative
motion. Let the instantaneous location of all the circuits be defined by a set of parameters a
r
, where r=1, …, s.
Then, the work done in a movement (da
r
) can be written as
Work done = F da
r r
;
(218)
hence,
W F
da
dt
mech r
r
r
r s
= Â
=
=
1
.
(219)
188
But from equation (213), it is also known that
W I
d L I
dt
dU
dt
mech k
ki i m
i k
=
( )
- Â Â ,
(220)
which can be rewritten using equation (217) as
W I
d L I
dt
d
dt
L I I
mech k
ki i
i k
ki k i
i k
=
( )
- Â Â Â Â
1
2
;
(221)
hence,
W I
d L I
dt
L I
dI
dt
mech k
ki i
i k
ki i
k
i k
=
( )
- Â Â Â Â
1
2
1
2
.
(222)
This can be further simplified to
W I I
dL
dt
mech k i
ki
i k
= Â Â
1
2
.
(223)
Now, L
ki
denotes the mutual inductance on circuit k due to circuit i. Its variation with time is due
solely to the relative motion of the circuits; thus, one can write
dL
dt
dL
da
da
dt
ki ki
r
r
r
= Â ,
(224)
which can be used together with equations (219) and (223) to give
F
da
dt
I I
dL
da
da
dt
r
r
r
k i
ki
r
r
r i k
 =   Â
1
2
.
(225)
From this, it follows that
F I I
dL
da
r k i
ki
r i k
= Â Â
1
2
.
(226)
189
Now, equation (226) is applied to the specific case of two circuits that are in relative motion:
F I
dL
da
I I
dL
da
I I
dL
da
I
dL
da
r
r r r r
= + + +
Ê
Ë
Á
ˆ
¯
˜
1
2
1
2 11
1 2
12
2 1
21
2
2 22
.
(227)
Now, in the case of interest here, with two circular coils in relative motion, there is no distortion of the
coils, so the two self-inductance terms (L
11
and L
22
) are constant. Also, it can be shown from basic energy
considerations that L
ki
= L
ik
for all i and k. Hence,
F I I
dL
da
r
r
=
1 2
12
.
(228)
For two coaxial coils, having N
1
and N
2
turns, respectively, and with motion along only the x axis,
this can be rewritten as
F N N I I
dM
dx
x
=
1 2 1 2
,
(229)
where M is the mutual induction between a single turn of each coil.
190
REFERENCES
1. Adams, R.B.; Alexander, R.A.; Chapman, J.M.; et al.: “Conceptual Design of In-Space Vehicles for
Human Exploration of the Outer Planets,” NASA/TP—2003–212691, Marshall Space Flight Center,
MSFC, AL, November 2003.
2. French, B.M.: Traces of Catastrophe, Lunar and Planetary Institute, pp. 3, 14–15, 1998.
3. Adushkin, V.V.; and Nemchinov, I.V.: “Consequences of Impacts of Cosmic Bodies on the Surface of
the Earth,” Hazards Due to Comets and Asteroids, The University of Arizona Press, Tucson, pp. 721–
778, 1994.
4. “The Barringer Meteorite Crater,” http://www.barringercrater.com, The Cyrus Company, Cited December
12, 2003.
5. Rampino, M.R.; and Haggerty, B.M.: “Extraterrestrial Impacts and Mass Extinctions of Life,” Hazards
Due to Comets and Asteroids, The University of Arizona Press, Tucson, AZ, pp. 827–857, 1994.
6. Levy, D.: Impact Jupiter: The Crash of Comet Shoemaker–Levy 9, Plenum Press, New York, 1995.
7. “The Inner Solar System,” Harvard-Smithsonian Minor Planet Center, http://cfa-www.harvard.edu/
iau/Animations/Inner.gif, Cited March 2003.
8. Hills, J.G.; Nemchinov, I.V.; Popov, S.P.; and Teterev, A.V.: “Tsunami Generated By Small Asteroid
Impacts,” T. Gehrels (ed.), Hazards Due to Comets and Asteroids, The University of Arizona Press,
Tucson, pp. 779–789, 1994.
9. Toon, O.B.; Zahnle, K.; Turco, R.P.; and Covey, C.: “Environmental Perturbations Caused by Asteroid
Impacts,” T. Gehrels (ed.), Hazards Due to Comets and Asteroids, The University of Arizona Press,
Tucson, AZ, pp. 791–826, 1994.
10. Morrison, D.; Chapman, C.R.; and Slovic, P.: “The Impact Hazard,” T. Gehrels (ed.), Hazards Due to
Comets and Asteroids, The University of Arizona Press, Tucson, pp. 59–91, 1994.
11. Center for International Earth Science Information Network, Columbia University, http://www.ciesin.org
Cited March 2003.
12. Lynn, E.: “INTROS: Integrated Rocket Sizing Program,” Marshall Space Flight Center, AL, 2003.
13. Adams, R.; and Straw, A.: “Launch Vehicle Conceptual Design,” Chapter 3 in Space Launch and
Transportation Systems, pp. 79–214, to be published Spring 2004.
191
14. Venetoklis, P.; Gustafson, E.; Maise, G.; and Powell, J.: “Application of Nuclear Propulsion to NEO
Interceptors,” Hazards Due to Comets and Asteroids, The University of Arizona Press, Tucson,
pp. 1089–1110, 1994.
15. Schmidt, G.R.; Bonometti, J.A.; and Irvine, C.A.: “Project Orion and Future Prospects for Nuclear
Pulse Propulsion,” J. Propulsion and Power, Vol. 18, pp. 497–504, May–June 2002.
16. Reynolds, T.W.: “Effective Specific Impulse of External Nuclear Pulse Propulsion Systems,” NASA
Technical Note, NASA TN D–6984, 1972.
17. Cotter, T.P.: “Rotating Cable Pusher for Pulsed-Propulsion Space Vehicle,” Los Alamos Scientific
Laboratory, LA–4666–MS UC–33, Propulsion Systems and Energy Conversion TID–4500, 1971.
18. Nance, J.C.: “Nuclear Pulse Propulsion,” IEEE Transactions on Nuclear Science, February 1965.
19. Martin, A.R., and Bond, A.: “Nuclear Pulse Propulsion: A Historical Review of an Advanced Propulsion
Concept,” J. British Interpl. Soc., Vol. 32, pp. 283–310, 1979.
20. Andrews, J.: and Andrews, D.: “The MagOrion—A Propulsion System for Human Exploration of the
Outer Planets,” American Institute of Physics, New York, pp. 1533–1537, 2000.
21. Solem, J.C.: “Medusa: Nuclear Explosive Propulsion for Interplanetary Travel,” J. British Interpl.
Soc., Vol. 46, pp. 21–26, 1993.
22. Dyson, G.: Project ORION: The True Story of the Atomic Spaceship, Owl Books, 2002.
23. Bonometti, J.A.; Morton, P.J.; and Schmidt, G.R.: “External Pulsed Plasma Propulsion and Its Potential
for the Near Future,” American Institute of Physics, New York, pp. 1236–1241, 2000.
24. Bonometti, J.A.; and Morton, P.J.; “External Pulsed Plasma Propulsion (EPPP) Analysis Maturation,”
AIAA–2000–3610, 2000.
25. McInnes, C.: Solar Sailing, Praxis Publishing Ltd., Chichester, UK, pp. 33–109, 1999.
26. Wright, J.: Space Sailing, Taylor and Francis, Inc, pp. 225–239, 1992.
27. Serway, Physics for Scientists and Engineers, Saunders College Publishing, Philadelphia, 1986.
28. Ahrens, T.J.; and Harris, A.W.: “Deflection and Fragmentation of Near-Earth Asteroids,” Hazards Due
to Comets and Asteroids, The University of Arizona Press, Tucson, pp. 897–924, 1994.
29. “Dan’s History,” http://www.danshistory.com/nuke.shtml Cited April 14, 2003.
30. “Dan’s History,” http://danshistory.com/lgb.shtml Cited April 14, 2003.
192
31. Nelson, R.A.: “Low-Yield Earth Penetrating Nuclear Weapons,” FSA Public Interest Report, p. 4,
January/February 2001.
32. Melosh, H.J.; Nemchinov, I.V.; and Zetzer, Y.I.: “Non-Nuclear Strategies for Deflecting Comets and
Asteroids,” Hazards Due to Comets and Asteroids, The University of Arizona Press, Tucson, p. 1111–
1132, 1994.
33. Harris, T.: “How Stuff Works,” http://science.howstuffworks.com/e-bomb.htm Cited December 2002.
34. O’Neill, G.K.; and O’Leary, B.: “Space-Based Manufacturing From Nonterrestrial Materials,” Progress
in Astronautics and Aeronautics, published by the American Institute of Aeronautics and Astronautics,
Vol. 57, 1977.
35. Ahrens, T.J.; and Harris, A.W.: “Deflection and Fragmentation of Near-Earth Asteroids,” Hazards Due
to Comets and Asteroids, The University of Arizona Press, Tucson, AZ, pp. 897–927, 1994.
36. Tedeschi, W.J.: “Mitigation of the NEO Impact Hazard Using Kinetic Energy,” in Proc. Planetary
Defense Workshop, Lawrence Livermore National Laboratory, Livermore, CA, pp. 313–323, May 22–
26, 1995.
37. Bate, R.R.; Mueller, D.D.; and White, J.E.: Fundamentals of Astrodynamics, Dover Publications, Inc.,
New York, 1971.
38. Shoemaker, E.M.: “Asteroid and Comet Bombardment of the Earth,” Annual Rev. Earth and Planetary
Sci., Vol. 11, pp. 461–494, May 1983.
39. Chapman, C.R.; and Morrison, D.: “Impacts on the Earth by Asteroids and Comets: Assessing the
Hazard,” Nature, Vol. 367, pp. 33–40, 6 January 1994.
40. Gold, R.E.: “SHIELD—A Comprehensive Earth Protection System: A Phase I Report to the NASA
Institute for Advanced Concepts,” Report No. SDO–10974, May 28, 1999.
41. Lewis, J.S.: Comet and Asteroid Impact Hazards on a Populated Earth, Academic Press, 1999.
42. Jeffers, S.V.; Manley, S.P.; Bailey, M.E.; and Asher, D.J.: “Near-Earth Object Velocity Distributions
and Consequences for the Chicxulub Impactor,” Mon. Not. R. Astron. Soc., Vol. 327, p. 327, 2001.
43. Chesley, S.; Chodas, P.; Milani, A.; et al.: “Quantifying the Risk Posed by Potential Earth Impacts,”
Icarus, Vol. 159, pp. 423–432, 2002.
44. Ivezic, Z.; Tabachnik, S.; and Rafikov, R.: “Solar System Objects Observed in the Sloan Digital Sky
Survey Commissioning Data,” Astron. J., Vol. 122, pp. 2749–2784, November 2001.
45. Brown, P.; Spalding, R.E.; ReVelle, D.O.; et al.: “The Flux of Small Near-Earth Objects Colliding with
Earth,” Nature, p. 142, 21 November 2002.
193
46. Reeves, E.I.: “Spacecraft Design and Sizing,” Chapter 10 in Space Mission Analysis and Design, edited
by James R. Wertz and Wiley J. Larson, Microcosm Press, El Segundo, CA, 1999.
47. NASA Glenn Research Center Web-page: http://spacescience.nasa.gov/osstech/solar.htm Cited
December 2002.
48. Abramowitz, M.; and Stegun, I.A.: Handbook of Mathematical Functions, Dover Publications Inc.,
New York, NY, pp. 608–609, 1965.
49. Becker, R.: Electromagnetic Fields and Interactions, Dover Publications Inc., New York, NY, pp. 193–
197, 1982.
REPORT DOCUMENTATION PAGE
Form Approved
OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources,
gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this
collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operation and Reports, 1215 Jefferson
Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188), Washington, DC 20503
1. AGENCY USE ONLY (Leave Blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED
4. TITLE AND SUBTITLE 5. FUNDING NUMBERS
6. AUTHORS
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
REPORT NUMBER
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING
AGENCY REPO NUMBER
11. SUPPLEMENTARY NOTES
12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE
13. ABSTRACT (Maximum 200 words)
14. SUBJECT TERMS 15. NUMBER OF PAGES
16. PRICE CODE
17. SECURITY CLASSIFICATION
OF REPORT
18. SECURITY CLASSIFICATION
OF THIS PAGE
19. SECURITY CLASSIFICATION
OF ABSTRACT
20. LIMITATION OF ABSTRACT
NSN 7540-01-280-5500
Survey of Technologies Relevant to Defense From Near-Earth Objects
Standard Form 298 (Rev. 2-89)
Prescribed by ANSI Std. 239-18
298-102
R.B. Adams, R. Alexander, J. Bonemetti, J. Chapman, S. Fincher,
R. Hopkins, M. Kalkstein, T. Polsgrove, G. Statham,* and S. White*
George C. Marshall Space Flight Center
Marshall Space Flight Center, AL 35812
National Aeronautics and Space Administration
Washington, DC 20546
Prepared by the Advanced Concepts Department, Space Transportation Directorate
*ERC, Inc., 555 Sparkman Drive, Executive Plasma, Suite 1622, Huntsville, AL 35816
Unclassified-Unlimited
Subject Category 18
Availability: NASA CASI 301–621–0390
Several recent near-miss encounters with asteroids and comets have focused attention on the threat of a
catastrophic impact with the Earth. This Technical Publication reviews the historical impact record and
current understanding of the number and location of near-Earth objects (NEOs) to address their impact
probability. Various ongoing projects intended to survey and catalog the NEO population are also reviewed.
Details are given of a Marshall Space Flight Center-led study intended to develop and assess various can-
didate systems for protection of the Earth against NEOs. Details of analytical tools, trajectory tools, and a
tool that was created to model both the undeflected inbound path of an NEO as well as the modified, post-
deflection path are given. A representative selection of these possible options was modeled and evaluated.
It is hoped that this study will raise the level of attention about this very real threat and also demonstrate
that successful defense is both possible and practicable, provided appropriate steps are taken.
asteroids, comets, near-Earth object, bombardment of Earth,
trajectory modeling tools
Unclassified Unclassified Unclassified
222
Unlimited
M–1100
Technical Publication July 2004
NASA/TP—2004–213089

The NASA STI Program Office…in Profile
Since its founding, NASA has been dedicated to the advancement of aeronautics and space science. The NASA Scientific and Technical Information (STI) Program Office plays a key part in helping NASA maintain this important role. The NASA STI Program Office is operated by Langley Research Center, the lead center for NASA’s scientific and technical information. The NASA STI Program Office provides access to the NASA STI Database, the largest collection of aeronautical and space science STI in the world. The Program Office is also NASA’s institutional mechanism for disseminating the results of its research and development activities. These results are published by NASA in the NASA STI Report Series, which includes the following report types: • TECHNICAL PUBLICATION. Reports of completed research or a major significant phase of research that present the results of NASA programs and include extensive data or theoretical analysis. Includes compilations of significant scientific and technical data and information deemed to be of continuing reference value. NASA’s counterpart of peer-reviewed formal professional papers but has less stringent limitations on manuscript length and extent of graphic presentations. • TECHNICAL MEMORANDUM. Scientific and technical findings that are preliminary or of specialized interest, e.g., quick release reports, working papers, and bibliographies that contain minimal annotation. Does not contain extensive analysis. • CONTRACTOR REPORT. Scientific and technical findings by NASA-sponsored contractors and grantees. • CONFERENCE PUBLICATION. Collected papers from scientific and technical conferences, symposia, seminars, or other meetings sponsored or cosponsored by NASA. • SPECIAL PUBLICATION. Scientific, technical, or historical information from NASA programs, projects, and mission, often concerned with subjects having substantial public interest. • TECHNICAL TRANSLATION. English-language translations of foreign scientific and technical material pertinent to NASA’s mission. Specialized services that complement the STI Program Office’s diverse offerings include creating custom thesauri, building customized databases, organizing and publishing research results…even providing videos. For more information about the NASA STI Program Office, see the following: • Access the NASA STI Program Home Page at http://www.sti.nasa.gov • E-mail your question via the Internet to help@sti.nasa.gov • Fax your question to the NASA Access Help Desk at 301–621– 0134 • Telephone the NASA Access Help Desk at 301–621–0390 • Write to: NASA Access Help Desk NASA Center for AeroSpace Information 7121 Standard Drive Hanover, MD 21076–1320 301–621–0390

NASA/TP—2004–213089

Survey of Technologies Relevant to Defense From Near-Earth Objects
R.B. Adams, R. Alexander, J. Bonometti, J. Chapman, S. Fincher, R. Hopkins, M. Kalkstein, and T. Polsgrove Marshall Space Flight Center, Marshall Space Flight Center, Alabama G. Statham and S. White ERC, Inc., Huntsville, Alabama

National Aeronautics and Space Administration Marshall Space Flight Center • MSFC, Alabama 35812

July 2004
i

Acknowledgments
The authors would like to thank Langley Research Center for supporting this worthwhile endeavor through Revolutionary Aerospace System Concepts (RASC) funding, and the Marshall Space Flight Center astronomy group for the technical assistance of Roy Young. .

TRADEMARKS
Trade names and trademarks are used in this report for identification only. This usage does not constitute an official endorsement, either expressed or implied, by the National Aeronautics and Space Administration.

Available from:

NASA Center for AeroSpace Information 7121 Standard Drive Hanover, MD 21076–1320 301–621–0390

National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 703–487–4650

ii

as well as the threat that they pose to the Earth. effective means of defense against the NEO threat are available to us. The scientific community now accepts that these events are just major examples of the continuous ongoing bombardment of the Earth by a wide variety of objects. Section 7 describes the means by which all of the above tools were incorporated into a single “master” design environment and used to search for optimal solutions to the planetary defense problem. This impact created a crater 1 mi in diameter and the resulting shockwave killed every large mammal within a radius of 24 km. In 1908. Many feel that. and evaluating their effectiveness. The dust prevented much of the Sun’s radiation from reaching the Earth’s surface and the resulting Ice Age caused the extinction of ≈50 percent of the existing animal species— including. Approximately 50. Section 6 introduces the trajectory modeling tools used in the study. the exponential growth of world population. This explosive burst—since termed an airburst—occurred above Siberia. this threat can be ignored. Against this it can be argued that. Section 4 describes in detail the propulsion technologies that were considered as candidates to transport defense hardware out to an Earth-bound NEO. by both orbital iii . and outlines our growing understanding of NEO impacts. many unexplained catastrophes in the past may actually be attributable to the impact of asteroids or comets. the dinosaurs. as our knowledge of the solar system in general—and its minor bodies in particular—is very recent. section 2 presents the threat of asteroids and comets. It left a zone of destruction nearly 40 km in diameter around the point of disintegration. Section 5 reviews the actual defense or mitigation technologies that would be used to either fragment or deflect an Earth-bound NEO. most of them fragments of either asteroids or comets in orbit around the Sun. The impact created a crater nearly 300 km in diameter and drove very large quantities of dust and debris into the atmosphere. combined with our increasing technological dependence. conclusions and recommendations from this project are presented in section 9. makes humanity much more vulnerable to the consequences of a near-Earth object (NEO) impact. Methods of categorization for asteroids and comets. near the town of Tunguska. Finally. developing mitigation options. Despite these worrying trends. The airburst took place very close to the altitude at which the maximum amount of ground damage would result. The extreme heat and aerodynamic pressures generated during entry caused the object to disintegrate explosively at an altitude of ≈8 km. Section 8 contains results from parametric analyses that were conducted using the tools and data described in sections 2 through 7. because there appears to have been no loss of life due to cosmic bombardment. Additionally. That is the topic of the Marshall Space Flight Center-led study reported in this Technical Publication (TP). Each section is intended to gradually introduce the reader to the various facets of the problem—quantifying the NEO threat. a small comet or asteroid with a diameter of ≈50 m entered the Earth’s atmosphere over eastern Russia.000 yr ago a stony-iron asteroid with a diameter of ≈150 m struck the Earth in what is now north-central Arizona. After the introduction in section 1. Section 2 gives an overview of our current understanding of asteroids and comets. This TP is divided into nine sections. Section 3 discusses various mission configurations that might be selected as part of a defense strategy.EXECUTIVE SUMMARY Nearly 300 million yr ago a comet with a diameter of ≈10 km struck the Earth near Chicxulub in present-day Mexico. most famously.

The fragmentation option breaks the NEO into pieces. In addition. Grouping these options into categories allows us to develop analysis processes applicable across an entire category. This allows the vehicle to operate on or near to the NEO for an extended duration. delivering its deflection or fragmentation energy over an extended time instead of in one short impulse. The solar sail was considered. An interceptor would travel across the inner solar system and actually impact on the NEO. are introduced. under the rendezvous option. comets. rather than having to establish a different process for each option. deflection of the NEO by purely kinetic means—utilizing an ultra-high velocity inert projectile—was considered. There is some uncertainty over the maximum fragment size that would not threaten the Earth. The deflection option leaves the object largely intact. the mitigation hardware would be transported out to the NEO and would match orbits with it. This categorization approach is new to this study. The immediate physical effects of various types of impact are then described. The solar collector. pulsed magnetic field was considered. this is usually set at 3 Earth radii. a diameter of 10 m was selected here as a first approximation. showed remarkable promise for all but the largest asteroids and comets. describing the development of human knowledge of asteroids. The novel option of deflection by use of a rapidly growing. Finally. The outbound model solved the Gauss problem for high thrust trajectories. For trajectories in the inner solar system requiring large inclination changes. as are the longer-term consequences. the frequency. some thoughts are offered on the current lack of significant public credibility regarding this threat. Finally. mitigation options can be divided into two categories: deflection and fragmentation. location. The use of nuclear devices to deflect the NEO was also considered. At the highest level.parameters and composition. evidence of its previous use was nowhere in the literature. Furthermore. a derivative of the solar sail—the solar collector—was considered. solar sails are highly competitive and so were likewise considered. and sequence of known and suspected impacts on the Earth are presented. each small enough to burn up in the Earth’s atmosphere. Several propulsion systems were considered as candidates to place the mitigation vehicle onto either an intercept or a rendezvous trajectory with the NEO. A multistage liquid oxygen/liquid hydrogen rocket was selected as the baseline option due to its technical maturity. was considered and was modeled in some detail. and the impacts from these bodies. A timeline is then given. This timeline helps to explain why our historical record does not contain more information on possible impacts. A scaled-down derivative of the ORION concept of the 1960’s was also retained—referred to as the nuclear pulse technique. Finally. transported out to and installed on the incoming asteroid. The means by which energy is delivered—to either fragment or deflect an NEO—can also be categorized. A remote station would project a beam or fire a projectile at the incoming NEO. Both the outbound and inbound trajectories were modeled using tools developed under this study. however. the use of a mass driver. Next. The nuclear thermal rocket was retained as an additional option because of its superior specific impulse. but was rejected. Current knowledge of the NEO population within the solar system is then discussed. Finally. sufficient to ensure that it will miss the Earth by a distance greater than or equal to a certain minimum value—in the literature. The inbound model calculated iv . several very interesting concepts for threat mitigation were considered. Fragmentation of the incoming NEO using nuclear devices was assessed. as it was found to be impractical for all but the smallest asteroids. There are many different options that are potentially available to defend against an incoming NEO. but changes its velocity by a small amount.

07 0. for this first attempt at a solution.800 NA Rendezvous Rendezvous (≈3 yr) Rendezvous (≈10 yr) 2.847 5.000 § § * Maximum was constrained to a total system mass at Earth SOI of 1. by the mean number of lives that would be saved. Our threat parametric builds on the results of previous studies in modeling the overall consequences of an impact. We intend to revisit this issue and model the effects of gravitational losses at the earliest opportunity in the future. However.6 9.170/1.636 1. including all options evaluated here.000/1. we integrated our trajectory.849 80/6. these trajectory calculation techniques are only approximate. The next objective was to combine the tools and concepts described above to determine the optimal configurations to defeat the threat.62 73.000 t. § The solar collector system is limited more by solar collector size than by total system mass.900/2.400 100 NA Asteroid Diameters (m) 1. be considered for any future work. Time and funding constraints necessitated a less ambitious approach. Rendezvous times are greater for larger asteroids. and threat mitigation tools to quantify the relationship between required system mass and size of object deflected. ** Times are for 100-m-diameter chondrite.8 8. recognizing our limited modeling capability for this study. However. These results are combined with estimates of the average number of fatalities for an asteroid or comet of a given size and composition and are then used to determine the average number of deaths resulting from this threat over a given time period.000/1.550 41. Our model uses the current knowledge of the asteroid and comet population to execute Monte Carlo simulations to establish the probability of impact.000 Intercept Rendezvous Intercept 1. we strongly recommend a broad spectrum of deflection technologies.637 0. propulsion. as well as qualitative comparisons documented here.025/800 0. Our various concepts for threat mitigation can then be evaluated by the percentage of this threat that each can defeat.635/115** 29.000 1.000/1.200 1.backwards in time from the point of impact with the Earth in order to determine the instantaneous ∆V necessary to achieve the minimum required deflection. Inclusion of the Sun’s gravitational field in these calculations yields significantly different results than the two body approximations found in the literature. Total System Mass at SOI (t) for Different Time Before Impact (days)/Outbound System Staged chemical/ mass driver Staged chemical/ nuclear deflection Staged chemical/ kinetic deflection Nuclear pulse Solar collector Maneuver Rendezvous Travel Time (days) 2. Some of our propulsive techniques do not provide high thrust levels and. and ultimately.075/132 1.300 87. The inbound trajectory calculations accounted for the influence of both the Sun and the Earth.27 568 NA 1.000 NA 10.000 NA Maximum Diameter of Asteroid*(m)/ Total System Mass at Earth SOI (t) 50/6. Assuming a baseline asteroid orbit. v . although total missions times change little. as they neglect gravity losses. Based on these results. we were willing to accept the consequent level of uncertainty in our answers. These results are summarized for the concepts evaluated in the study in the table below. See figure 112. we considered the nuclear pulse option the most viable for the overall threat.509/910 1.076/65** 3.000 260/1.7 0.240 167 34.8 1.918 9.

............................................................................... 5..................................2 Remote Station Versus Interception Versus Rendezvous ....................................................3 5.2 4............................. INTRODUCTION .................................................. Nuclear Pulse ............................................................6 2.................. OUTBOUND PROPULSION ....................................................................................... Measuring the Near-Earth Object Population .................................................................................................................................................................................................................................................................................... Solar Collector .....................................................................5 5...........................................................................2 5........6 5.................................................. Damage Mechanisms ............1 4. 5...............4 5.............................................................................................................7 Categorization of Asteroids and Comets ..........................................................................................1 Outbound ........................................ Magnetic Flux Compression ............... The Credibility Problem ............................................................................. 1 3 3 7 13 16 18 25 29 31 31 33 37 38 43 46 53 60 62 62 66 71 74 76 81 93 102 102 103 3...................................................5 Staged Chemical .......................... Nuclear Thermal Rocket ........... Kinetic Deflection................. THREAT ........................................................................................... Solar Sail ................................................................................................. 4............... 2............................... Solar Sails ......................4 2..................................4 4.................................................................................................................................. 2.............2 2............................................................................................. 6............................................................................... The Torino Scale ................ Earth’s Impact Record ............ 4.............................................................5 2....................................................... 3......................................................................................................................................................................................................................................................................... Nuclear Deflection .....................................................1 2......................................1 5....................................................3 2............. Mass Driver ....... History Related to Near-Earth Objects ............... 6......................................................... 3................ 6...........................................................1 Deflection Versus Fragmentation ... Solar Collector .......................................... MISSION CONFIGURATIONS .......................................... TRAJECTORY MODELING ......................2 Inbound ...................................................................................................................................................................3 4........................................... vii .....................7 Nuclear Fragmentation ....................................................... THREAT MITIGATION ..................TABLE OF CONTENTS 1....................................

....................................... Future Work ........................ Interference Between Adjacent Drive Coils.............2 9...................................... Drive and Bucket Coil Currents ....................6 9........... 110 115 115 117 118 144 144 144 145 145 146 147 148 149 150 150 150 150 151 151 152 152 152 154 154 158 161 163 166 166 170 176 APPENDIX A—CURRENT NEAR-EARTH OBJECT SEARCH PROGRAMS .....................8 Public Awareness ..........................................................................................................1 C................................................................. Effect of Bucket Coil Motion on Stationary Coil Circuit Operation .................. C..........1 9.................. B...... APPENDIX B—SOLAR ARRAY CALCULATIONS ................................................................................................................2 A..................................................................................1 Method I ................................................................................................. Lincoln Near-Earth Asteroid Research .............................................................................................. Near-Earth Asteroid Tracking ............................................................................................................................................................................................................................................. 8....................................................2 C.....................................................1 Integrated Analysis .. APPENDIX C—MASS DRIVER ..... Funding of Future Work ......... 8........3 C.. Spaceguard .............................7 C............ 9........................................... Development and Deployment of Mitigation Systems ................................................................ 9.....................3 A.........7 9.. 8............................................................ Analysis of Bucket Kinetic Energy and Acceleration ......................... Braking Coils .......... Assessment of Mitigation Options ..........4 C.................................................................... Statistical Problem ..........................................4 9....3 Parametric Performance ..........................................................6 C.......................4 A.....2 Architecture Options ...................................5 9.. Accomplishments ...................................................................................................................... Drive Coils ...............................................................................................1 A.........................................................................................................................5 SpaceWatch ..................... CONCLUSIONS AND RECOMMENDATIONS ..................................................................................................................................................... THREAT PARAMETRIC ..........................2 Method II............. B............................................................................................................................... PARAMETRIC RESULTS ................................................................................................................................. 8......................5 C........................................... Summary Conclusion ... Bucket Design ............................ A.................8 Model of the Forces on a Bucket Coil Due to the Nearby Drive Coils ....................................3 9................................................................... Lowell Observatory Near-Earth Object Search .................................. viii ....................................................................................................................................................................................................................................................TABLE OF CONTENTS (Continued) 7........................................................................

.................................................................................................................................................................................................................... C.... 180 181 182 185 190 ix ....................... REFERENCES ...................................................................................................TABLE OF CONTENTS (Continued) C. C.......................11 Self-Induced Coil Stresses ...............................................................................9 Drive Coil ................................10 Braking Coil ....................... APPENDIX D—COIL FORCE MODEL ..................................................................

.. 11.................................... 2002........................................................................................................................................................000 km distance versus initial meteor radius for Fe meteor ............ Location of known impact craters noting diameters as of 1998.......... other asteroids are green... Projected area affected from the Tunguska blast of 1908.. 17 13.... 10 11 12 13 8........... 1862 Apollo........ Orbits of representative short-period comets relative to the outer planets... Several hundred thousand casualties can be expected from such an impact..................... Calculated impact areas from the Barringer meteor ......................... 2.................. Number of known NEAs versus time........LIST OF FIGURES 1... and 1221 Amor relative to Earth and Mars .. 9 7............ 20 14................................................... Arrows depict the location and direction trees were knocked down from the blast ............................ and comets are blue...... 8 6................................................... Alabama........ USA ...................... Organization of RASC FY 2002 activities ...... NEOs are red........... Note the rapid increase in discoveries in recent years due to the use of CCDs and increased interest in the asteroid and comet threat ...................................................................... Aerial view of the Barringer Impact Crater in Arizona.............................. 20 x ... plotted relative to the inner planets..... 10................. Extinction events versus time according to the fossil record ...... 3......................... Age distribution of these craters is also included ... 6 5... Orbits of 2062 Aten............................................... Tunguska impact area superimposed over Madison County............. Deep-water wave height at 1........... USA. Deep-water wave height at 1................................................................... Halley’s Comet represents the Halley class and Comet Crommelin represents the Jupiter class ... 16 12........ 1 4 5 4......... Orbits of representative short..........000 km distance versus initial meteor radius for soft stone meteor .. Location of known minor planets on March 2...........and long-period comets relative to the solar system......................................................... Comet Hale-Bopp represents long-period comets ...................... 9.

...........LIST OF FIGURES (Continued) 15........................................................... U.............................. 24......................................... Delivering deflection or fragmentation energy by the interception mode .......... 25... Mass of water lifted into the atmosphere versus impact energy ... Regression curve fit of NERVA program-developed engines ............................................................................................................................. 16.......................................... 17.. Image produced by INTROS ....................... Greenland and Iceland are not represented as well as some Pacific islands ............... Human population density graph for all continents (except Antarctica)................................................... 22............... 34...................... Blast wave damage versus impact energy ..... 32....................... Regression curve fit of lox/LH2 launch vehicle stages .......... Air Force ORION spacecraft (1964) ... Two-stage lox/LH2 vehicle.........S. 19................ Population is extrapolated through the middle of this century .... 21 23 24 25 27 20......................................... Illustration of deflection method of threat mitigation .................... 30............................................................ EPPP concept vehicle ........... Blast wave damage versus impact energy .............................. 29................................................. 31........................................................................................ xi .............................................................................. 26................................................. 33............................ NASA Gabriel spacecraft (1999) ......................... Delivering deflection or fragmentation energy by the rendezvous mode ......................... Delivering deflection or fragmentation energy by the remote station mode .... 18................ 27.......... Illustration of fragmentation method of threat mitigation ...... 28 29 30 32 32 34 34 35 38 39 43 44 47 47 50 21.............................. Schematic of a nuclear thermal rocket .... Density and optical depth of atmospheric dust versus impact energy ... 23............... Predicted world population in the last century..... 28. Illustration of the various category of threat under the Torino Scale ........................ Categories within the Torino Scale ......................................................

........................... Blast yield—explosive placed at center of body—required for fragmentation as a function of asteroid radius ........................................................... E-bomb magnetic flux compression generator ....... Nuclear pulse rocket delivery system sketch ..... 48.......................................... Deflection ∆V imposed on 10-m-diameter asteroid ........................................................................................... Shock absorber operations ....................... Relative current flow directions of bucket and nearest drive coils ...... 45.................. square solar sail ................................... 36............LIST OF FIGURES (Continued) 35..................... 38....................... 73 74 75 78 80 80 83 83 84 85 46.............................................................................................. 52............................ 51................. 37.................. Critical disk diameter .......................................... 44................ Relative current flow directions of bucket and immediately adjacent drive coils ..................................................................................... 50....... 49....... Schematic and dimensions for a square solar sail ................................. Magnetic flux compression generato ............................................ Sail trajectories relative to lightness number ........................................................... 54...... 40................................. 42........... ∆V imposed on varying asteroid sizes by 100-m-diameter solar collector ........................................ Relative current flow directions of bucket and surrounding drive coils ...................... Coil gun conceptual design ............................................................ 53...................................................................................................................................... Deflection ∆V imposed by 6 gm/m2 solar sail with varying areas on varying diameter asteroids .............................................................................................. 69 70 72 43...... 39....... Schematic of solar collector ................. Solar collector configuration .. Artists concept of a billowing................. 47............................... .................. Geometric position of the pulse unit to the planetary body and cone half-angle definition ..................................................................................................... Device mass versus explosive yield ................................... xii 51 54 54 58 60 63 64 41..............................................................

.. Relative current flow directions of bucket and braking coils .......................................................... Relative current flow directions of bucket and nearby drive coils ............ The velocity of the planetary body at impact for this case is (–40........ 62................. 57.... 68.... 60..............600-day total mission duration ................................................................. Outbound trajectory ∆Vs for 3................. Following discharge—decelerated bucket exits mass driver and joins return system .......... Illustration showing a typical NEO orbit. 92 94 95 100 103 67.......................... 59........................................................................ Fragmentation data and curve fit results ................................................................................ 63........................ Mass driver operation .. 61............................................................................. Main components of the mass driver ........... parallel to the x axis of the Heliocentric-Ecliptic system .. 71........................................................................ Interception geometry ............................................................................................... 70....................... System following discharge—bucket returns for reloading ... Required impulsive ∆V for 42 km/s velocity for various maneuvers to avoid collision with Earth......................... Mass driver system schematic ............................... Impact and ejection geometry .............................. xiii 85 86 86 87 89 89 90 90 91 91 92 66........................ M1999JT6 orbit plot ............................. 73................................................................................. Loaded bucket about to enter mass driver ................ 69.......................................... 107 .................LIST OF FIGURES (Continued) 55........................ 104 105 72......................... 56.............................................................................................0)T km/s.................................................. 64.............. Discharge in progress—bucket is decelerating while expellant mass exits mass driver at high speed .................................................... Mass driver system view ........................................................ showing the benefit of the UP maneuver when impact is only a few days away ........................ Bucket about to be reloaded and cooled prior to next discharge ........... Bucket containing expellant under acceleration within mass driver .................................................... 58.....................0...................................... 65........................

... Required impulsive ∆V for various maneuvers to avoid collision with Earth for planetary body with velocity of 35 km/s (long lead time) ............................ 120 82.........................................000 m) .... Staged chemical/mass driver vehicle mass at Earth departure ............................................................................................................................................. Staged chemical/mass driver vehicle mass versus chondrite asteroid diameter ................. 107 75............................................................... showing the benefit of the DECEL and OUTSIDE maneuvers when impact is several weeks away .......... Average deaths from single asteroid impact versus size........................................................ 108 77......................................................... 121 122 123 123 124 83.............................. Required impulsive ∆V for 42 km/s velocity for various maneuvers to avoid collision with Earth................................LIST OF FIGURES (Continued) 74.... (b) optimal deflection direction—detailed view....................................... Mass driver deployed mass and total operating time versus chondrite asteroid diameter (100–1............... Required impulsive ∆V for various maneuvers to avoid collision with Earth for planetary body with velocity of 35 km/s ................ and (c) optimal deflection direction—detailed view—minimal deflection ∆V ... 86... 125 xiv ....... 114 116 119 79................... 109 78..... 87.................................... 108 76..... Staged chemical/mass driver vehicle mass at Earth departure (expanded view) ......................................................................... Required impulsive ∆V for 35 km/s velocity for various maneuvers to avoid collision with Earth........................................ 85. and the average total number of deaths prevented if all impacts of equal or less energy can be avoided .. 80................ Mass driver deployed mass and total operating time versus chondrite asteroid diameter (50–100 m) .............................. Proposed analysis process for assessing total amount of threat mitigated ........ 84........ 124 88.......... (a) Optimal deflection direction........................... showing the benefit of the DECEL maneuver when impact is only a few days away ............. 81... Optimal staged chemical/mass driver mission ................................. Staged chemical/mass driver model ........................... Variation of mass driver total system mass with required asteroid deflection ∆V for a 50-m-diameter chondrite ...............

.................. 137 137 105............................................... Minimum total system mass for the staged chemical/nuclear blast option versus chondrite diameter for both intercept and rendezvous ......... 103................. showing both intercept and rendezvous .................. 102.............. 128 128 129 94................................ Optimum rendezvous trajectory for the nuclear pulse option . 133 134 135 135 136 100................ showing the optimum rendezvous and total mission times .................................. the staged chemical system matches the asteroid’s orbit at encounter............ ............. 90.. 130 97.. Minimum total system mass for the staged chemical/nuclear blast option.......... Here.......... Staged chemical/kinetic deflection vehicle mass versus chondrite asteroid diameter ............... 101......................................................... Minimum total system mass for the staged chemical/nuclear blast option............ Diagram of the ModelCenter setup for the staged chemical/nuclear deflection option ........................ Optimum intercept trajectory for the staged chemical/nuclear deflection option ......... Total system mass for the staged chemical/nuclear blast option versus total mission time for various rendezvous times......... 95..................... Here....... Optimal staged chemical/kinetic deflection mission ................. 127 93... xv 125 126 91............ Minimum total system mass for the nuclear pulse option........... 130 131 98. 127 92....... and (c) detailed view—minimum mass solution ..... 104................. Optimum intercept trajectory for the staged chemical/nuclear deflection option ............ Staged chemical/kinetic deflection model ..... the staged chemical system does not match the asteroid’s orbit at encounter................ Total system mass for the staged chemical/nuclear blast option versus total mission time for various rendezvous times (zoomed).. ..................................................................... 99............. (a) Staged chemical/kinetic deflection vehicle mass at Earth departure............ Projectile mass versus chondrite asteroid diameter . showing the optimum rendezvous and total mission times for intercept .......................... Diagram of the ModelCenter setup for the nuclear pulse option ............... showing the optimum rendezvous and total mission times for rendezvous ... Minimum total system mass for the staged chemical/nuclear blast option versus chondrite diameter for the smaller chondrites......................LIST OF FIGURES (Continued) 89. (b) detailed view................. 96.. ...

. Combinations of total mission time and rendezvous time where inbound and outbound required solar sail sizes are equal................ 119................................................................................................... Sinusoidal current versus time profile for a single drive coil ..................................................................... 121................... 118.......................................................... Minimum total system mass for the nuclear pulse option versus chondrite diameter ..... 120....... Plot of the difference between required outbound and inbound solar collector sizes..................... 139 109............... Bucket and drive coil current directions during acceleration ................................................................................................................... Oscillating drive circuit—charging capacitor ........ 143 154 155 157 159 160 163 164 164 164 165 114.................................. Oscillating drive circuit—discharging capacitor .. 140 141 110..................................... and the associated total system mass for 500-m-diameter chondrite ... Ideal current versus time profile for a single drive coil .. ..................................... Minimum total system mass and size for the solar collector option versus chondrite diameter for the smaller chondrites ........................ Elliptical function curve fits—K(m) and E(m) versus m ............. Drive coils included in the analysis .............................. 117............................ Drive coils included in and omitted from the analysis .................................................................................... 123................ Minimum total system mass for the nuclear pulse option versus chondrite diameter for the smaller chondrites ............................................ 116................... Negative values mean that the inbound requirement dominates.................................. 122.................................................................................................. xvi 138 138 108................................................... Bucket coil current directions ........ 142 112.......................... 107.................. 142 113........ Diagram of the ModelCenter setup for the solar collector option ..... 111................................................. Optimum rendezvous trajectory for the solar collector option for a 100-m-diameter chondrite ....................... Minimum total system mass and size for the solar collector option versus chondrite diameter ........................... Real current versus time profile for a single drive coil .. ................................ 115.....................................LIST OF FIGURES (Continued) 106..................

............................................................ Bucket handling through a complete cycle ................ Discharge of isolated drive coil .................................. Simple model of a drive coil circuit .. Current flow directions of two adjacent drive coils .................................................. 127..................................... 128.... 130................................... Self-induced magnetic field and resulting force ................................. The effect of bucket acceleration ........................ Bucket conceptual design ........................................................... 125..............LIST OF FIGURES (Continued) 124.............................................................. 126.................................................... 166 168 170 172 176 181 182 xvii ................................................... 129....................................

............................................................................. 6 7 15 4.............. Nuclear device masses .................................... BLU–113 penetrator characteristics ......... 18..... Data on in-service and historical launch vehicles that have lox-/LH2-powered stages ......... xviii ...... Optical................................................................. Timeline of scientific discoveries relevant to humanity’s knowledge of asteroids and comets ........................... Qualitative considerations for threat mitigation using nuclear fragmentation ............................ 18 26 40 42 44 45 48 56 59 61 64 65 65 67 71 5........... 2...... 14.......... 8... 3.. 17................ Qualitative considerations for outbound propulsion using solar sails ........................................ Chemical rocket assumptions .................... 6.......... 13................ the distance between the Earth and the Moon is ≈240... Qualitative considerations for threat mitigation using nuclear deflection ............000 mi.... .............................. 16... Data on nuclear rocket engines developed under the NERVA program ........... Qualitative considerations for outbound propulsion using solar collectors ............ billowing force and other parameters used in solar sail analysis ................................. Recent near misses by comets and asteroids..... 12...... Consequences of impact by NEOs of various sizes . 11...................... 7.. Qualitative considerations for outbound propulsion using nuclear pulse ........... 9............... By comparison......................... Comparable terrestrial events for NEOs of various diameters .......................................................................... Description of asteroid and comet compositions and representative predicted densities ..................LIST OF TABLES 1........................................ 15............................................... Qualitative considerations for outbound propulsion using chemically-powered rockets ....... 10............................ Qualitative considerations for outbound propulsion using nuclear thermal rockets ..

............................. 29... Causes of death and associated probabilities for a U.S..... Architecture options considered in this study .................. 25............................................................................... Summation of parametric results for mitigation concepts ....LIST OF TABLES (Continued) 19.... Types of planetary bodies examined in the Monte Carlo simulation and their average contribution to the total number of deaths over the next century .................................................. 20................................. 27........... 24. 21................. 23........ Qualitative considerations for threat mitigation using magnetic flux compression ... Qualitative considerations for threat mitigation using a solar sail ...................... 22..... Original and modified orbital elements of 1999JT6 ..... Qualitative considerations for threat mitigation using the mass driver ......... Relative size of largest fragment at various collisional energies . 73 76 81 93 99 101 104 106 110 118 145 147 28...................................... 26.......................... xix ......... Qualitative considerations for threat mitigation using a solar collector ..... resident ................ Explanation of the different maneuvers available for use in the program PBI .......... 30.................................................. Qualitative considerations for threat mitigation using kinetic deflection ........

LIST OF ACRONYMS AND SYMBOLS AANEAS au carb CCD CF DOD EMP EPPP Fe GEODSS GMD H H2O HOPE ICBM ICR JPL K-T LaRC LCR Anglo-Australian Near-Earth Asteroid Survey astronomical unit carbonaceous chondrite charged-coupled device catastrophic fragmentation Department of Defense electromagnetic pulse external pulsed plasma propulsion iron ground-based electro-optical deep space surveillance global missile defense hardness parameter for asteroidal material water human outer planet exploration intercontinental ballistic missile inductance capacitance resistor Jet Propulsion Laboratory Cretaceous-Tertiary (extinction event) Langley Research Center inductance-capacitance-resistance xx .

LIST OF ACRONYMS AND SYMBOLS (Continued) LEO LH2 LINEAR LN2 LONEOS lox LP Ma MSFC MSSS MTM NEA NEAT NEO NERVA Ni NO N2O4 NTR PBI PBM low-Earth orbit liquid hydrogen Lincoln near-Earth asteroid research liquid nitrogen Lowell Observatory near-Earth object search liquid oxygen long period (comet) mega annum (million years) Marshall Space Flight Center Maui Space Surveillance Site momentum transfer mechanism near-Earth asteroid near-Earth asteroid tracking near-Earth object nuclear engine for rocket vehicle applications nickel nitrogen oxide nitrogen tetroxide nuclear thermal rocket planetary body intercept planetary body maneuvering

xxi

LIST OF ACRONYMS AND SYMBOLS (Continued)

PBO PHA RASC SOI SP TD30 TNT TP UDMH USA

planetary body—outbound potentially hazardous asteroid revolutionary aerospace systems concept sphere of influence short period (comet) Advanced Concepts Department tri-nitro-toluene Technical Publication unsymmetrical dimethyl hydrazine United States of America

xxii

NOMENCLATURE

A a a0 aB aD B Bb Bf C C1,C2,C3

area—solar array, solar sail, or cross sectional semimajor axis of orbit (astronomical units or kilometers) solar sail characteristic acceleration radius of bucket coil radius of drive coil magnetic flux density; magnetic field non-Lambertian coefficient for backside of sail non-Lambertian coefficient for front side of sail electrical capacitance solar sail force coefficients specific heat of liquid nitrogen speed of light spherical diameter standoff distance elliptical integrals of the second kind; electric field; energy energy of collision orbital eccentricity number of expected average fatalities per year for all impacts of some given energy or less; total radial force acting on a current-carrying coil; solar flux; force; thrust absorptive force normal to solar sail centerline absorptive force tangential to solar sail centerline effectivity factor emissive force tangential to solar sail centerline inert mass fraction solar sail force normal to solar sail centerline

CLN 2
c D Dstandoff E EP e F Fan Fat FE Fe Finert Fn

xxiii

NOMENCLATURE (Continued)

Fnorm Frn Frt Ft f fi G g0 H Hc Hvap h I I0 IB ID ID0 Ij I(r) Isp Isp J K KE k L
n

total force adjusted for billowing reflective force normal to solar sail centerline reflective force tangential to solar sail centerline solar sail force tangential to solar sail centerline number of expected average fatalities per year for an impact of a given energy; radial force per unit length of coil substitute variable (used in bucket coil motion analysis); i = 1,2,3,4 constant of gravitation gravitational constant at Earth’s surface asteroid “hardness” total thermal capacity of bucket coils and their LN2 coolant heat of vaporization bucket internal height orbital inclination (degrees); electrical current initial current current flowing in bucket coil; solar sail front side emission coefficient current flowing in drive coil maximum current flowing in drive coil current flowing in coil j luminous intensity specific impulse specific impulse of the nth stage current density elliptical integrals of the first kind; thermal conduction constant between bucket LN2 reservoir and external cold plate kinetic energy variable used to evaluate elliptical integrals electrical self-inductance; conductor length xxiv

orbital inclination (degrees) mutual inductance between two current-carrying. length. single-turn coils. initial mass mutual inductance between bucket coils i and j initial mass of nth stage mutual inductance between one turn of a bucket coil and one turn of drive coil I when the bucket coil is a distance x from the drive coil mass of largest fragment mass of liquid nitrogen used to cool the bucket superconducting coils mass of near-Earth object mutual inductance between one turn of a bucket coil and one turn of stationary coil n mass ratio of nth stage mass of Sol (the Sun) mass of target NEO mass mass flow rate of ejecta payload mass payload mass of nth stage propellant mass of nth stage NTR reactor mass mass relative to Earth Mi(x) ML M LN 2 MNEO Mn MRn MS MT m & me mpay mpay mp n n mreac m/s xxv .NOMENCLATURE (Continued) Ls l M MB Mbody Mej Mf Mf Mi Mij Mi n n solar luminosity diameter of solar collector or curved sail. mass of solar array. magnification factor for solar collector mass of loaded bucket mass of body mass of material ejected due to the interceptor impact final mass final mass of nth stage mass of interceptor.

NOMENCLATURE (Continued) ms N NB ND NDC Ni P Pej Pjet po Q. idealized radius of target body acceleration of orbiting body wire radius position vector position vector of the Earth relative to the Sun (other objects use similar notation. radius of vehicle from Sun. tension temperature of cold place maximum temperature of liquid nitrogen used to cool bucket superconducting coils xxvi n r′ ˙˙ r ro r rE/S S s′ T Tcp TH . radius of coil asteroid orbital radius crater radius Earth’s orbital radius radius of Sun bucket internal radius. distance from coil to the target reflection coefficient. standard intercoil distance spectral reflection coefficient temperature.q R Ra Rc RE RS r inert mass of nth stage number of circuits in an electrical system. number of turns in coil number of turns per bucket coil number of turns per drive coil total number of drive coils number of turns in coil i electrical power momentum of ejected material jet power total pressure electrical charge resistance of coil. time derivatives use the overdot notation) distance between adjacent stationary coils.

bucket internal volume.NOMENCLATURE (Continued) TL t U Ue UM V VE/S v ve vi vmin W x Y α minimum temperature of liquid nitrogen used to cool bucket superconducting coils time total magnetic energy total electrostatic energy magnetic energy relative speed of collision. vertical component of momentum overall efficiency of solar electric power system. induced electromagnetic field emission coefficient for back emission coefficient for front inert mass fraction β β′ ∆KE ∆Q δ δp ε εb εf εn xxvii . electrical potential velocity vector of the Earth relative to the Sun (other objects use similar notation) velocity velocity of ejecta interceptor speed. total amount of heat generated in mass driver bucket during one operational cycle skin depth coefficient. angle of attack sail lightness number gas expansion factor change in kinetic energy experienced by mass driver bucket as it moves between two adjacent drive coils generated in mass driver bucket during one operational cycle. total amount of heat. mass driver bucket speed as it passes drive coil i minimum speed of ejected material required blast yield distance between bucket and drive coils NEO material strength Lagrange multiplier.

NOMENCLATURE (Continued) η λ λn µ0 ν θ ϕ ρ σ ratio of bucket internal height to internal radius. final mass driver bucket speed mass driver stationary coil current oscillation frequency solar sail cone angle variable of integration used to evaluate elliptical intgrals. efficiency of solar sail variable used in equation solutions payload mass fraction for nth stage permeability of free space ejecta speed. angle between impact and ejecta NEO material mass density. conductivity Stefan-Boltzmann constant time during which bucket must remain in contact with external cold plate in order for its coils to be cooled after one operational cycle magnetic flux total magnetic flux through bucket coil i solar sail centerline angle longitude of the ascending node (degrees) argument of Perihelion (degrees). oscillation frequency of coil current coil path vector σ′ τ Φ Φi φ Ω ω r l Subscripts E PB S denoting the Earth denoting the planetary body denoting the Sun xxviii . resistivity solar array mass per unit area or solar sail loading parameter.

investigate. RASC Group 1: Human/Robotic Exploration Group 2: Human Exploration of the Solar System Beyond Mars Group 3: Aerospace Mobility Studies Group 4: In-Space Remote Sensing Group 5: Tool Development Human outer planet exploration (HOPE)— develop revolutionary aerospace systems concepts for human exploration of the outer planets Planet body maneuvering (PBM)— develop.1 MSFC’s participation in group 4 activities is documented in this Technical Publication (TP).TECHNICAL PUBLICATION SURVEY OF TECHNOLOGIES RELEVANT TO DEFENSE FROM NEAR-EARTH OBJECTS 1. MSFC’s activities were confined to planetary body maneuvering (PBM). led from TD30/Advanced Concepts. and evaluate techniques for maneuvering potentially Earth-threatening planetary bodies (asteroids. 1 . managed from Langley Research Center (LaRC) selected a broad range of projects for the year’s activities. the revolutionary aerospace systems concepts (RASCs) activity. Organization of RASC FY 2002 activities. These projects were organized into five groups as shown in figure 1. comet nuclei) Figure 1. Marshall Space Flight Center (MSFC) participated in two of these groups. however. The work completed for group 2—human outer planet exploration (HOPE)—is documented in NASA/TP—2003–212691. Several projects were funded under group 4. INTRODUCTION In FY 2002.

However.Work under the PBM project was confined to defense of the Earth from collisions from asteroids and comets. Many of the technologies developed for protective maneuvering of planetary bodies are also applicable to maneuvering these bodies for resource utilization. NASA’s unique capabilities may well make it the most uniquely qualified organization in the world to take on the daunting task of protecting the planet from this threat. Asteroids and comets can be maneuvered— in a careful and controlled way—close to the Earth to be mined for structural materials and water. Using the limited knowledge currently available on the solar system’s asteroid and comet population. 2 . It has been suggested that the mission of planetary defense is best suited to the Department of Defense (DOD). and trajectory tools to calculate optimal deflection techniques for use against specific threat scenarios. Finally. role to play. The above mission statement suggests very strongly that the Agency should address this threat. Parametric results could then be generated using the linked propulsion. Despite these other potential applications for maneuvering technologies. this project concentrated solely on planetary defense as the most critical mission—a uniquely suitable one for NASA. This can be seen immediately from the NASA mission statement: To understand and protect our home planet To explore the Universe and search for life To inspire the next generation of explorers … as only NASA can. Research conducted by the TD30 PBM team to understand and categorize the threat of impact by an asteroid or comet is summarized in this TP. This argument is based on the DOD’s extensive experience in the interception of highspeed objects. if not a leading. Analytical tools were developed to model the actual deflection techniques. deflection. these parametric results are presented and a set of conclusions established as to the effectiveness of each deflection method. an analytical tool was developed to estimate the number of human lives that could potentially be lost because of this threat over a specified period of time. The mass of these bodies could also anchor rotating tethers and skyhooks. such bodies could be intentionally targeted to impact Mars or Venus in order to alter rotation speed and/or atmospheric composition. These various tools were then linked with an additional set of tools capable of modeling both inbound and outbound trajectories. the very high energies necessary for deflection of massive planetary bodies combined with the unique problems of operation in interplanetary space suggest that NASA will have a major. Propulsion technologies suitable for reaching the approaching object were then researched and deflection methods investigated. Finally.

and 1221 Amor are all significant because they name their respective categories.2. Apollo. These categorizations are discussed further below. By comparison. low-inclination solar orbits. the asteroids 2062 Aten. Finally. The majority of asteroids have both aphelia and perihelia between the Martian and Jovian orbits. Cometary orbital eccentricities are usually higher than those of asteroids (e > 0. THREAT 2. 1862 Apollo. Cometary aphelia are usually located beyond the orbits of the gas giants. with some even having e > 0.017 au. usually in eccentric.000-km diameter). Asteroids in the Aten class have orbits with apehelia <0. 3 . The most visible and distinctive features of comets are the coma and tail. Most of the mass of a comet is contained within a comparatively tiny central nucleus.35 for most comets. Asteroids are small objects (<1. The orbits of Aten. Comets are ice-rich bodies that become visually prominent when heat from the Sun causes their trapped volatiles to sublimate. They are thought to be largely of rocky and/or metallic composition. Asteroids can also be categorized according to their orbital parameters. and also by dust. produced by the release of these volatile compounds.1 Categorization of Asteroids and Comets Asteroids and comets are both categorized as minor planets.3 au). asteroids in this class could easily be perturbed into a collision trajectory. Apollo asteroids have semimajor axes greater than Earth’s (<1 au) and perihelia <1. Hence. Even though these orbits do not necessarily cross that of the Earth. Amors have orbits between Earth and Mars (1 au < r < 1.9).983 au and semimajor axes less than that of the Earth. now thought likely to have been formed directly from the primordial solar nebula. and Amor class asteroids are illustrated in figure 2. although there are exceptions. This designation is appropriate on the grounds of both size and orbital parameters. An additional level of categorization can also be established on the basis of composition. Each category is usually named after its first representative. comets can be distinguished from asteroids by both orbit and composition. Distinguishing between asteroids and comets is not entirely straightforward.

their orbital periods are <200 yr.Figure 2. Halley’s comets remain within the solar system. Although they can have longer periods. clearly shown by the 76-yr period of the example from which they take their name. Additionally. 1862 Apollo. Short-period comets are believed to originate from the Kuiper belt and generally have inclinations <30∞. The orbits of Halley’s Comet and a representative Jupiter class comet are shown in figure 3. Orbits of 2062 Aten. 4 . Shortperiod comets are further broken down into Halley and Jupiter classes. Comets are generally listed as short or long period. Their orbits also categorize comets. and 1221 Amor relative to Earth and Mars. Jupiter class comets have perihelia within Jupiter’s orbit.

5 . Orbits of representative short-period comets relative to the outer planets. Hale-Bopp. Halley’s Comet represents the Halley class and Comet Crommelin represents the Jupiter class. Long-period comets are believed to come from the Oort cloud.–50 –40 –30 –20 –10 0 10 20 30 40 50 50 Sun Earth 40 Mars 30 20 10 0 –10 –20 –30 –40 –50 Jupiter Saturn Uranus Neptune Pluto Crommelin Halley Figure 3. The orbit of a representative long-period comet. they have periods well above 200 yr and appear to have no preferred orbital inclination. is found in figure 4.

having been deduced from samples found after atmospheric entry.3 Density (g/cm3) 1. Note that chondrites comprise 88 percent of the asteroid population.and long-period comets relative to the solar system.2 7. Comet Hale-Bopp represents long-period comets.1 Asteroids Chondrite Achondrite Iron Stony irons Mesosiderite Pallasite Comets Short period Long period 6 . Density (g/cm3) 3. Description of asteroid and comet compositions and representative predicted densities.6 3.–350 –250 –150 –50 50 150 250 350 350 Sun Earth 250 Mars Jupiter 150 Saturn 50 Uranus Neptune –50 Pluto Crommelin Halley –150 –250 Hale-Bopp –350 Figure 4.4 1. Table 1. minor planets can be organized according to their composition. Orbits of representative short. A list of composition types with approximate densities is given in table 1. Previous studies indicate that ª80 percent of impact-related deaths are likely to be caused by chondrites and short-period comets. These densities are highly conjectural. Finally.9 5.0 4.

Germany Total annual energy released from Earth (seismic. Australia Lake Mein.2. Comparable terrestrial events for NEOs of various diameters. seismic events.500 ton 20. Somewhat perversely. and the like are constantly erasing crater sites.000 Mton 87.7E7 Mton Impact Frequency (Per Myr) 250. these sinkholes contained potable water.) Sudbury.007 Crater Diameter 35 m 75 m 120 m 450 m 1.000 28. but most of the evidence has been destroyed or covered by geological processes and vegetation. 1945) Typical hydrogen bomb explosion (1 Mton) Barringer Meteor Crater. Crater diameters and terrestrial events likely to inflict comparable damage are also listed. The actual number of Earth impacts is thought to be much higher. volcanic. which is widely believed to have initiated an ice age at the boundary between the Cretaceous and Tertiary periods and the consequent extinction of over 50 percent of the then existing species of flora and fauna. the Chicxulub crater was eventually only identified from radar density mapping performed by a petroleum company owned by the Mexican government. Sweden. As an interesting side note. In the case of Chicxulub. Table 2.000 69. At the other end of the size spectrum is the Chicxulub impact. etc. including the dinosaurs. Oasis. Sometimes there is circumstantial evidence that indicates that a major impact crater is present.1 km 5 km 10 km 20 km 31 km 200 km Comparable Terrestrial Event Minimum damaging earthquake (M=5) Largest chemical explosion (Heligoland Fortifications.to 100-km range. Note that even the smallest diameter objects are capable of causing very major damage. Libya Haughton Dome. Arizona. Mexico Figure 5 illustrates the location of 145 known impact craters distributed around the world. South Africa. Canada. 7 . Note that the majority of the craters are <50 Ma (Mega annum.000 Mton 310. without which it may have been impossible for the Spaniards to explore that portion of the continent in the 1700’s. For instance. Rochechouart.000 Mton 8. Siberia. France.5 km 10 km Yield (TNT Equivalent) 500 ton 4. vegetation.9 1. or million years) old.2 Earth’s Impact Record Table 2 lists the impact frequency for progressively larger near-Earth objects (NEOs). Tunguska explosion. Bosumtwi.2 NEO Diameter 2m 4m 6m 23 m 55 m 250 m 500 m 1 km 1.000 ton 1 Mton 11 Mton 1. Vredefort. Goat Paddock.000 2. Norway.4000 Mton 10. Chicxulub. 1947) Atomic bomb explosion (Hiroshima. Ghana.700 540 35 10 2. Canada. craters >100 km are not always simple to find because their effects are so widespread as to not be easily recognized as impact craters. Russia Gardnos. Additionally. a large number of sinkholes were found around the periphery of the impact crater. This evidence supports the theory that wind and water erosion. A 23-m-diameter object can cause destruction equivalent to the nuclear weapon used at Hiroshima at the end of World War II. Japan.4 0. most crater diameters are in the 50. Ries Crater.

8 . but as yet.02–5 km 30 20 10 0 50 150 250 350 Age (Ma) 450 550 >600 5–20 km 20–50 km 50–100 km 100–300 km Figure 5.50 40 Number of Craters Crater Diameter 0. note that few impact structures have been found underwater. There is no obvious reason why these continents would have received a higher impact flux than the others. In addition. Location of known impact craters noting diameters as of 1998. Finally. Many more impacts have been found in Europe. It is expected that cratering is mitigated by the cushioning effect of the oceans and that the erosion rate is higher for submerged craters. undiscovered. Age distribution of these craters is also included.2 Figure 5 also illustrates which parts of the Earth have been subjected to the most thorough search for evidence of extraterrestrial impacts. it is clearly more difficult to find craters in deep water. North America. on the other continents. It seems likely that impact structures exist in equivalent numbers. and Australia than in other regions.

Projected area affected from the Tunguska blast of 1908. Unfortunately. modeling and empirical evidence suggest that the heights at which such explosive blasts are most likely to occur are similar to those determined—by nuclear weapons experts—to cause maximum surface damage. 9 .The last significant impact on Earth was the Tunguska event of 1908. Instead. These directions are mapped in figure 6. This impact is believed to have been caused by a 30. The estimated blast point is illustrated in figure 6.to 60-m object that detonated at a height of ª8 km above the Earth’s surface. they clearly indicate the location of the center of the event.3 Most strikes by large NEOs do not reach the Earth’s surface. N 2 0 S 5 Scale (km) 10 Figure 6. usually causes it to disintegrate explosively. the combination of heat and stress. Investigators who explored the area during a series of expeditions between 1958 and 1965 carefully recorded the direction in which trees had fallen as a result of the blast. which the object experiences as it travels at very high speeds through the atmosphere. Arrows depict the location and direction trees were knocked down from the blast.

such as a large city like New York City or London. Tunguska is an unpopulated area in Russian Siberia. Superimposed over a more densely populated area. the authors’ residence and location of MSFC. USA. the devastation would cause the deaths of several million people. killing the majority of its 250. A similar strike in a populated area would have caused widespread devastation. Several hundred thousand casualties can be expected from such an impact. in the United States of America (USA). Living creatures inside this area are not thought to have survived the event. Fortunately. Tunguska impact area superimposed over Madison County. An impact of this magnitude would devastate the county.000 inhabitants. 10 .Figure 6 indicates the total ground area affected by even this relatively small object. Alabama. Scale 10 km 5 miles Figure 7. Alabama. Consider figure 7—the Tunguska event superimposed over Madison County.

It is ª1 mi in diameter and ª570 ft deep and is believed to have been caused by an object ª150 m in diameter that impacted between 25.Figure 8 depicts an aerial view of the Barringer meteorite crater in Arizona. 11 . Although this damage is less than that expected from the airburst of a non-Ni-Fe object of similar diameter. This object did survive the transit through the atmosphere and physically impacted the Earth’s surface. Aerial view of the Barringer Impact Crater in Arizona. As can be seen. This scenario is characteristic of large nickel (Ni)-iron (Fe) asteroids that can survive the thermal and stress experienced during atmospheric entry.000 and 50. Figure 8. the impact would cause hurricaneforce winds with resulting damage up to 40 km away. Additionally. human injuries and fatalities are expected up to 24 km from the impact point. USA.4 The projected devastation from this impact is shown in figure 9. the destruction is still far from trivial.000 yr ago. USA.

4 Since impacts are concealed or erased due to natural processes. Calculated impact areas from the Barringer meteor. Many of the spikes in this graph coincide closely in time with major known impact craters. 12 . consideration must be given to other types of evidence in order to determine the Earth’s impact history. Other peaks in this graph coincide with possible stratigraphic evidence of impact—material in the geologic strata that could be due to ejecta distributed around the world following an impact.Meteor City Winslow Meteor Crater Fireball Scorches Plants and Animals Out to 10 km West Sunset Mt. Large Animals Killed or Wounded by Pressure Pulse and Air Blast up to 24 km From Impact East Sunset Mt. Figure 10 shows the estimated percentage of all species that were driven to extinction as a function of time. Hurricane-Force Winds up to 40 km Figure 9.

there is evidence of several reversals that cannot be explained by the normal sequence of periodic changes. The Earth’s magnetic field periodically reverses naturally. The Earth’s field could be reversed by a sharp impact of sufficient magnitude.3 History Related to Near-Earth Objects The concept that the Earth is threatened by impacts from space is not universally accepted. Extinction events versus time according to the fossil record. It is entertaining to look at some remarks made through the ages concerning impacts from extraterrestrial objects.5 Possible additional evidence of impacts is provided by the record of occasional geomagnetic field reversals contained in frozen lava flows in the Earth’s crust. however. The first quote is frequently attributed to President Thomas Jefferson: 13 . 2. the mechanism is similar to that which acts when a ferromagnet is struck with a hammer to realign its magnetic field.Large (Diameter >70 km) Dated Crater * Stratigraphic Evidence of Impact Possible Stratigraphic Evidence of Impact 50 59.5 40 Extinction (genera) (%) 30 20 10 0 0 100 200 300 Million Years 400 500 Figure 10.

is a scandal. An effective search program clearly requires the use of telescopes in both the Northern and Southern Hemispheres. Efforts in the Southern Hemisphere were adversely affected by the Australian government’s decision to withdraw funding. Yet such an object is apt … to do a lot of damage. 1807 To be fair. Pale Blue Dot. Teller indicates that under current circumstances there is likely to be little or no warning before a catastrophe occurs. Edward Teller. … Just in dollars it could be billions. LLNL. as ours. Carl Sagan considered the idea of moving asteroids and comets—both for resource utilization and planetary defense—as a potential unifying endeavor for humanity: “Since hazards from asteroids and comets must apply to inhabited planets all over the Galaxy.” —Edward Teller. Some astronomers have strongly suggested that the threat necessitates a larger investment than is currently being made in the business of searching for and categorizing NEOs. to my mind. Additional comments by him in the same time period suggest that he was more open to the audacious theory that meteoroids were of extraterrestrial origin. if there are such.“I could more easily believe two Yankee Professors would lie than that stones would fall from heaven. also commented on the threat posed by asteroids and comets: “Here is the situation that. renowned physicist and developer of the hydrogen bomb. 1995 Teller illustrates the threat represented by NEOs. Additionally. but would require participation of all of humanity. the Australian Minister for Science made the following comment: 14 . We won’t have any indication of it. leave their planets. he believed that the development of systems needed to defeat this threat should not be shouldered by one nation. intelligent beings everywhere will have to unify their home worlds politically.” —Carl Sagan. At the time. This premise will be addressed in section 9. is spaceflight or extinction. Their eventual choice.” —President Thomas Jefferson. it is unclear whether President Jefferson actually said these words. 1994 Here Sagan suggests that planetary defense from NEOs is a strong justification for the continued exploration of space. and move small nearby worlds around. and I think people can understand that it is a scandal: There is a probability of a few percent in the next century of the arrival of a stony asteroid … approximately 100 m in diameter. It is a practical certainty that … it will come completely unannounced. and in lives it might reach millions.

One prominent member of the Australian survey team responded to the Minister’s actions in a direct manner. Anglo-Australian Near-Earth Asteroid Survey (AANEAS). I’m just not convinced that the hype and alarm and even fear-mongering is enough to justify an instant investment. etc. 2002 This action threatened to end all survey efforts in the only participating Southern Hemisphere country.” —Duncan Steel. In considering popular skepticism about the threat of Earth impacts. the Australian survey efforts continue but are funded by sources in the United States and Europe. Timeline of scientific discoveries relevant to humanity’s knowledge of asteroids and comets. “The dinosaurs did not have a space program. Although knowledge of the existence of comets predates written records. has only been widely publicized since 1980. 60 Minutes Interview. Apollos. Kant. The Alvarez theory. Australian Minister for Science. identifying the Chicxulub impact as being responsible for the Cretaceous-Tertiary (K-T) extinction. Chladni suggests extraterrestrial origin of meteorites Giuseppe Piazzi discovers Ceres Reinmuth discovers 1862 Apollo Discovery of Atens. Alvarez theorized asteroid impact for K-T extinction Recorded impact of Comet Shoemaker–Levy 9 into Jupiter 15 . 300 B. et al. 1543 1608 1609 1609 1687 1705 18th C 1794 1801 1932 19th/20th C 1980 1994 Aristarchus theorizes circular orbits around Sun Copernicus proposes Sun-centered system Hans Lippershey invents the telescope Galileo makes first astronomical telescope observations Kepler develops first two laws of planetary motion Newton publishes The Principia Halley reports findings on cometary trajectories Existence of asteroid belt theorized by Bode. Table 3. England The continued efforts to survey the population of NEOs are summarized in appendix A. now Professor. That’s why they died.” —Peter McGauran. it is instructive to consider the development of human knowledge about comets and asteroids. 2002 former member. As can be seen. Today. University of Salford.“I’m not going to be spooked or panicked into spending scarce research dollars on a fruitless attempt to predict the next asteroid. Table 3 contains a list of notable and relevant scientific discoveries.C. most of human history comets were revered (or feared) as omens and not regarded as objects of scientific curiosity. our understanding of asteroids only dates back about two centuries.

16 . Figure 11 illustrates the large population of NEOs around the Earth at any time. documented this historic event in reference 6. The red circles are minor objects that have perihelia <1. does create a misleading impression. plotted relative to the inner planets. since larger objects are easier to detect. the relative location of these objects can be plotted together with the orbits of the inner planets. however. neither population is well understood. 2002. 1994. our knowledge of the known NEO population is biased toward these larger objects. Also. and comets are blue. Unfortunately. Using the known NEO population. other asteroids are green. a cometary impact with Jupiter was recently observed. of course.3 au. This plot is shown in figure 11 and gives the location of all known objects on March 2. It is believed that orbital parameters are currently known for only ª10 percent of the total NEO population.While an impact between a significantly sized NEO and Earth has not been directly recorded. The green circles are minor objects that are not considered candidates for Earth impact. David Levy. Comet Shoemaker–Levy 9 was observed impacting Jupiter by the approaching Galileo spacecraft. On July 16. co-discoverer of the comet. The planets are shown as crosshair circles on their orbits. NEOs are red. Results of the impact were recorded by telescopes all over the world and in Earth orbit. 2. one must develop a proper understanding of the populations of these two types of bodies.4 Measuring the Near-Earth Object Population To determine the overall threat posed by the solar system’s asteroids and comets. in fact. it is overwhelmingly empty. Blue squares represent periodic comets. Figure 11. Location of known minor planets on March 2. the inner solar system—particularly the asteroid belt—appears to be full of NEOs. Due to the finite pixel size. 2002.

have led to a significant increase in the time and resources available to locate these objects. Chamberlin (JPL) 14 January 2002 Figure 12. such searches were conducted by the human study of photographic plates. The number of known near-Earth asteroids (NEAs) is plotted shown as a function time in figure 12. (3) Use of the Internet has facilitated international coordination and data sharing from sky surveys. Before the advent of CCD technology.000 800 600 400 200 0 1980 All NEAs Large NEAs Number 1983 1986 1989 Year 1992 1995 1998 2001 1995 LLNL Planetary Defense Workshop Alan B. The rapid increase in the number of objects identified over recent years can be explained in a number of ways: 1980/01–2001/12 1. and the theory that other Earth impacts have also affected the planet in the past. Note the rapid increase in discoveries in recent years due to the use of CCDs and increased interest in the asteroid and comet threat.Our knowledge of the NEO population has increased significantly in the past few years. (2) The development of charged-coupled devices (CCDs) has computerized the previous manual process of searching for new celestial objects.200 1. 17 .400 1.7 (1) The recent acceptance of the Alvarez hypothesis.800 1. Number of known NEAs versus time.600 1.

The mechanisms can be broadly categorized as either directly or indirectly linked to the impact itself. 28 km/s Will fly by Earth on February 1. some are straightforward while others affect the Earth’s ecosphere in a more indirect manner. This is due to the fact that CCDs enable the detection of smaller objects more than was previously thought possible. not a lot of applicable empirical data upon which to draw.000 mi from Earth 2002 NT7 1. Table 4. In at least one instance. table 4 gives some details. Several movies and television programs have been released in recent years. of the NEO threat. but this is to be expected. the total number of asteroids located has increased more rapidly than the number of large asteroids. several asteroids have passed the Earth at a distance of less than twice the Moon’s orbital radius from the Earth in the past year. over recent years. nitric oxide. Within the directly linked category are the tsunami and the blast wave. giving the general public an appreciation. the detection was made after the asteroid had already passed the point of closest approach.to 400-m diameter January 7. will fly 330. however scientifically distorted. 2001 YB5 300.000 mi from Earth 2002 NY40 800-m diameter August 18. major news outlets have published several articles detailing the Earth’s close encounters with NEOs. To our knowledge.2-km diameter.000 mi.000 mi from Earth 2002 EM7 70-m diameter March 8. the distance between the Earth and the Moon is ª240. and water vapor into the atmosphere. 2019 2. fortunately. Recently. flew 288. There is.Note that. has raised the level of coverage given in the general media to this threat. as well as the possible geopolitical outcome of an impact. Within the indirectly linked category are the effects of releasing large quantities of dust. coupled with the increase in public awareness. 2002. The increase in NEO detection rates. Some of the mechanisms presented here are a little speculative. flew 375. 2002. 2002. By comparison. 18 . Recent near misses by comets and asteroids.5 Damage Mechanisms The potential damage mechanisms resulting from an impact are several and varied.

the cumulative damage over an extended area was greater.000 km distant from the impact point of an Fe meteor of varying size.6. New Guinea. is likely to fragment. The Tunguska meteorite produced an airburst at an altitude of ª8 km while traveling at ª20 km/s.5. such as one comprised largely of stony material. will be able to withstand the ram pressure-induced stresses and will strike the surface—either land or water—largely intact. By these mechanisms. Tsunamis have caused great damage and loss of life in coastal areas.. This force. and speed. Shallower water results in a lower speed. These waves are two-dimensional disturbances whose height diminishes inversely with distance from their point of origin. Although the kinetic energy released locally is clearly greater for a surface impact.5. A body with a more solid internal structure. but at the present high level of detail. A sufficiently friable body. The key to development of a tsunami is the relationships between wave speed and water depth. There will undoubtedly be some difference in the magnitude and extent of a particular mechanism if produced by an airburst rather than by a surface impact. 2.000 km distant from the impact point of a soft stone meteor of varying size8 is shown in figure 13. and among wave energy.e. particularly low-lying regions. killing more than 2. such differences are not important. an incoming body will experience a considerable ram pressure force on its leading surface. particularly one at the optimum altitude. Figure 14 shows the height of a deep-water wave 1. a 30-ft-high tsunami came ashore in Papua. regarding the resulting damage mechanisms.100 people.1 Tsunami Both an oceanic impact and an airburst over the ocean will generate high-energy water waves. This is referred to as a surface impact. a wave of moderate amplitude in the deep ocean can increase in height by an order of magnitude as it comes ashore. is referred to as an airburst. Although the level of destruction immediately below the airburst was less than would have been produced by a surface impact.3 Now. This large-scale disintegration. damage can extend over a wider area from an airburst. while the body is still at some altitude. the potential for causing damage ashore increases proportionately.5. During passage through the atmosphere. In July of 1998.8 19 . in addition to decelerating the body. it is important to discuss the two types of impact events that can occur—air burst and surface impact. no crater was produced. As a wave approaches the continental shelf. The cause of the tsunami was determined to be an underwater landslide that took place more than 2.Before considering the damage mechanisms. The potential energy imparted by the stress forces being suddenly released causes the fragments to fly apart explosively. With increased wave height. both airbursts and surface impacts can give rise to all of those detailed in sections 2. particularly one with a largely metallic composition. i. but slower waves become steeper due to simple energy conservation.000 mi distant. will place it under considerable internal stress. Recent modeling studies indicate that the surface impact of a 400-m-diameter body—at any point in the Atlantic Ocean—would devastate both the American and European/African coastlines with final wave heights in excess of 60 m. height. with some resulting strain.1 through 2.3 The height of deep-water waves 1. the shallower water slows it and increases its height.

000 1.0 km/s vo = 20.2 km/s vo = 15. 10.0 km/s 0.0 km/s vo = 20.000 Figure 14.000 Radius (m) 10.000 Height (m) 100 10 1 vo = 11.0 km/s vo = 22. Deep-water wave height at 1.0 km/s 0.0 km/s vo = 22.2 km/s vo = 15.0 km/s vo = 30.000 10.000 km distance versus initial meteor radius for soft stone meteor. 20 .000 1.0 km/s vo = 25.0 km/s vo = 25.000 Height (m) 100 10 1 vo = 11.10.000 Figure 13. Deep-water wave height at 1.1 10 100 Radius (m) 1.1 100 1.000 km distance versus initial meteor radius for Fe meteor.0 km/s vo = 30.

the following abbreviations are used: SP = short-period comet LP = long-period comet Carb = carbonaceous chondrite. The surface explosion curve is derived on the basis that only 3 percent of the impact energy goes into shock waves. Figure 15 (taken from ref. Blast wave damage versus impact energy.2 Blast Wave Blast waves are produced by both airbursts and surface impacts. defined as the area that experiences an overpressure of 4 psi or higher. 9) consists of a graph showing the total damage area. A 4-psi (27. Area of Earth 108 Area of U. 21 . The destructive potential of the wave is determined by the total explosive energy of either the airburst or the surface impact as well as the altitude at which the explosion occurs.03.944-kPa) overpressure corresponds to hurricane-force winds of 70 mph. i. e = 0.e.5. Within the graph..972-kPa) overpressure will create a severe hazard due to flying debris. The shock is characterized by the peak overpressure. The blast wave consists of a shock wave followed by a substantial wind.S. K-T Damage Area (km2) 107 106 Airburst at Optimum Height 105 Stone 104 Tunguska 103 LP Fe Carb 102 10–1 100 101 102 103 104 105 106 107 108 Impact Energy (Mton) SP Impact-Generated Vapor Cloud Surface Exploration at ε = 0. as a function of the impact energy.2. Even a 2-psi (13.03 Figure 15.

taking into account the settling process—some dust will leave the atmosphere.3 Atmospheric Dust Loading Submicron dust reaching the stratosphere as a result of an impact will adversely affect the regional and global climate as well as disrupting important biochemical mechanisms. based upon the impacting body’s composition.000-Mton nuclear exchange could reduce continental interior temperatures to –25 ˚C within as little as 2 wk.5. Most work to date has concentrated upon the atmospheric dust loading. (3) The effect of the atmospheric dust upon Earth’s radiation balance must be calculated. its kinetic energy. The sequence of events is complex and four distinct steps are required to model the effects of dust loading: (1) The amount of dust lofted by the impact must be derived.9 22 . mass migration. resulting from the use of nuclear weapons. precipitation. Modeling results indicate that a 5. Quantification is difficult without extremely complex numerical modeling. and the nature of the impact. heat transfer from ocean water would probably mitigate somewhat.2. and possibly warfare. It is anticipated that a large impact could cause a drop of several degrees in global temperature and the loss of one or more year’s crop with resulting starvation. social disorder. resulting in minimum temperatures of only about –5 ˚C. (2) The rate at which dust disperses around the globe must be derived. and photosynthesis must be established. (4) The effect of low light levels on surface temperature. In practice. such as photosynthesis.5 Figure 16 shows the density of atmospheric dust and the resulting optical depth as functions of impact energy.

500 K. at which point additional NO would be created directly from atmospheric gases. sulfate. NO forms if the resulting ejecta plume moves through the atmosphere at more than 2 km/s.4 Nitrogen Oxide Production Four significant mechanisms for the production of nitrogen oxides (NOs) during an impact have been identified: (1) The formation of NO in the shock waves that accompany both atmospheric entry and explosive disintegration.5.or carbonaceous-rich impacting bodies may also produce sulfur dioxide and carbon dioxide. Elevated NO levels in the atmosphere would also 23 . 2. (4) In the case of an impact from an extremely large body. further NO will be formed within the subsequent reentry shocks. (2) If a surface crater is created. the hot ejecta dispersed through the atmosphere could briefly bring the local temperature up to ª1. (3) If the ejecta plume leaves the atmosphere. In addition to elevated atmospheric NO levels. Density and optical depth of atmospheric dust versus impact energy.102 Total Mass 100 Submicron Dust K-T Clay Layer 105 Optical Depth of Submicron Dust Mass (gm/cm–2) 103 10–2 101 10–4 10–6 10–1 Pinatubo Optical Depth 10–8 10–3 10–10 100 102 104 Impact Energy (Mton) 10–5 106 108 Figure 16. respectively. The most direct consequence of elevated atmospheric NO levels are the production of acid rain.

Figure 17. The impact of additional water vapor at high altitude is uncertain. Analysis indicates ª150 million kg of water would be vaporized per megaton of impact energy. i. The resulting increase in water levels above the tropopause would give rise to a “greenhouse” effect and would substantially increase Earth’s temperature. volume of NO as a fraction of total air volume. the resulting greenhouse effect could significantly increase global temperatures. Figure 18 shows the mass of water injected into the atmosphere as a function of impact energy.5 Water Vapor Injection Under normal circumstances.e. mainly because the mechanisms by which water leaves the upper atmosphere are not properly understood. this means that impacts as low as 104 Mton are capable of doubling that level. An NO volume mixing ratio. if mixed uniformly throughout the atmosphere.threaten the integrity of the Earth’s ozone layer. taken from Toon et al.. Blast wave damage versus impact energy. 1016 1015 1014 1013 NO Well-Mixed in Atmosphere Removes Ozone UV Screen NO Mostly in Stratosphere or Mesosphere 10–3 10–4 10–5 NO From Global Ejecta: 10–6 2-mm Particles 10–7 NO From Global Ejecta: 200-µm Particles 10–8 10–9 100 101 102 103 104 105 106 Impact Energy (Mton) 107 108 109 10–1 10–2 Fraction of Oceanic Mixed Layer Acidified NO (moles) 1012 1011 1010 109 108 107 NO From Plume at Impact Site Figure 17. as low as 2¥10–7.9 contains a brief summary of the effects of NO on both the ozone layer and oceanic acidity levels. A large oceanic impact would inject large amounts of water into the upper atmosphere. There are two mechanisms by which water can reach the upper atmosphere: direct splash to high altitude and water vaporization.9 24 .5.4 The current water vapor level above the tropopause is approximately 2¥10–4 to 6¥10–4 gm/cm2. would render the current ozone ultraviolet screen ineffective. 2. the upper atmosphere has very low humidity levels.. However.

Although possessing the weapons themselves.6 Precipitate Nuclear Exchange In recent years.Upper Atmosphere Water Mass 100 Injected Mass of Water Mass (g/cm–2) 10–2 10–4 10–6 10–8 100 102 104 Yield (Mton) 106 108 Figure 18. This raises the concern that an unexpected impact might be mistaken for a nuclear strike. Examination of table 5 illustrates 25 . Mass of water lifted into the atmosphere versus impact energy.6 The Credibility Problem Considering the research described above. a national government might discover that events had acquired a grim and irresistible momentum. one might ask why the danger posed by asteroids and comets is not given more public attention. Although efforts are underway to prevent further proliferation of nuclear weapons technology. Although nations such as the United States would probably be able to rapidly distinguish an impact event from a nuclear weapon detonation. more accurate tracking of them. and some of them have proceeded to develop such weapons. and the development of effective protection techniques. most of these nations do not yet have the relatively sophisticated detection systems of the better established nuclear powers. it is important to distinguish between impacts from small and large objects. First. several new nations have acquired the capability to produce nuclear weapons. If this scenario were to occur in a region of high tension. it could trigger a retaliatory nuclear strike. it is quite possible that additional nations will succeed in acquiring it over the next few years. it is entirely possible that this information would attract little credence in a region of high international tension. 2.5. 2. With its public demanding retaliatory action against the supposed perpetrator. The answer to this question is somewhat complex. This recent development adds urgency to the need for more thorough identification of threatening bodies.

. K-T-type event.2 350 m 6 km 104–105 4. e.g. Land impact destroys area the size of a large nation.8 0..e. stones produce airbursts. Smaller objects are likely to cause devastation on a regional scale.. the danger from these smaller bodies has been ignored.. United States. Global ozone destruction. Direct destruction approaches continental scale. Both land and ocean impacts raise dust. Land impacts destroy city-sized areas. Large mass extinction..6 160 m 3 km 103–104 4. produce effects that are felt across the globe. First. Land impact raises enough dust to affect climate. e. Human nature tends to naturally concentrate attention on dangers that are perceived as present in our everyday lives. comets produce airbursts.g. DC. automobile accidents.. Tsunamis reach oceanic scales. probably mass extinction. e.10 Yield (Mt) <10 Interval LogT – 3.2 – 16 km – 250 km – the consequences of impacts by objects of various sizes.g. Prolonged climate effects. change climate. These impacts have received some attention in the popular media. triggering widespread fires. Ocean impacts generate hemispheric-scale tsunamis. Land impact destroys area the size of a moderate state. e. freeze crop. and excludes those that are more exotic.0 NEO Diameter – 75 m Crater Diameter – 1.. i. Land impact destroys area the size of a small state. New York City.Table 5. only irons (<3%) reach surface. Impacts on land produce craters. exceed damage from land impacts. e. Mexico. Threatens survival of all advanced forms of life. caused by objects with a diameter greater than the 500-m to 1-km range..7 km 12 km 105–106 5. 26 . Consequences of impact by NEOs of various sizes. the concept of severe destruction being caused by a collision with an NEO is very alien to most of the general public.0 3 km 60 km 107–108 6. Washington.6 7 km 125 km 108–109 >1010 7.g.7 km 30 km 106–107 6.5 km Consequences Upper atmosphere detonation of stones and comets. Although more likely to occur than a larger impact. Land impact destroys area the size of a large state. the subconscious decision to avoid facing such a nonobvious threat could even be seen as a defense mechanism against being psychologically overwhelmed. Nobody knows of anyone who has been killed by a falling meteorite and very few people have witnessed a fall of any size. California.g. Large impacts. etc. Irons make craters.g. Impact destroys urban areas.. e. Irons and stones produce groundbursts. Virginia. ocean tsunamis become significant.g. fires.. With all the threats facing humanity in the early 21st century.4 1.g. e. Impact ejecta are global. global configuration. there are several reasons for this. 10–102 102–103 3. Delaware. e.

direct evidence of impacts from celestial objects is not readily found in the historical record. Today. Also. as depicted in figure 19. the Earth’s human population numbered between 1 and 2 billion. our knowledge of comets and asteroids is relatively recent. This can be argued by considering human population growth over the past century. The increased population density means that there are far fewer remote places where an impact event. Sources as diverse as the Christian Bible and Maori tribal records from New Zealand make reference to catastrophic events that could have been caused by large impacts.As mentioned previously. 10 9 8 7 6 5 4 3 2 1 0 1900 1920 1940 1960 1980 Timeframe 2000 2020 2040 Number of People (billions) Figure 19.11 At the time of the last collision with a substantial object (Tunguska). although careful searches have recently yielded indirect evidence. collisions with asteroids and comets are a bigger threat now than in any other time in history. As mentioned in the 27 . Predicted world population in the last century. several historical incidents could be interpreted as asteroid impacts. the Earth’s population numbers above 6 billion and is rising rapidly. could occur without causing significant loss of life. In many respects. Population is extrapolated through the middle of this century. like that at Tunguska. However. The idea of asteroids and comets colliding with the Earth and causing widespread devastation has had even less time to take root.

000 Figure 20. An impacting asteroid would probably create large numbers of charged particles during its travel through the atmosphere. These particles would probably have a severely detrimental effect on the global electronic and communications infrastructure. In fact. would probably be very vulnerable to such a varied and large number of coinciding problems.11 28 . Greenland and Iceland are not represented as well as some Pacific islands.section on damage mechanisms. Even impacts from relatively small objects— a few tens of meters in diameter—pose a much higher threat than any other time in history. usually well equipped to deal with an isolated crisis. consideration of figure 20 shows that a majority of the Earth’s population actually lives near these shorelines. 1 5 25 50 100 250 500 Persons Per Square Kilometer 1. Our complex and interdependent technological society. Atmospheric dust loading would decrease light reaching Earth’s surface. placing viable arable land at a very high premium. Human population density graph for all continents (except Antarctica). even a small impact can cause a large tsunami that would affect much of the Earth’s shorelines.

99 8 5 9 100 (100 m) Regional Kinetic Energy (Mton) Local Collision Probability Events Having No Likely Consequences Events Meriting Careful Monitoring Events Meriting Concern Threatening Events Certain Collisions Figure 21.7 The Torino Scale During an international conference on NEOs held in Turin in 1999. Hopefully. In honor of the meeting venue. • Ten indicates that a collision is certain.2. Illustration of the various category of threat under the Torino Scale. an attempt was made to establish a type of “Richter Scale” for categorizing the Earth-impact hazard associated with newly discovered asteroids and comets. and the impacting object is so large that it is capable of precipitating a global climatic disaster. 108 (5 km) 6 7 10 Global 2 105 (1 km) 1 0 4 3 1 (20 m) No Consequences 10–8 10–6 10–4 10–2 >0. 29 . Orbital predictions for newly discovered bodies are naturally uncertain. this system was named the Torino Scale. Discovery observations typically involve measurements over only a short orbital track and so. Categorization on the Torino Scale is based on the placement of a close approach event within a graphical representation of kinetic energy and collision probability (fig. it will always reduce as more information becomes available. its Torino number can change. 21). Zero also categorizes any object too small to penetrate the Earth’s atmosphere intact. as a body’s orbit characterization improves. The Torino Scale utilizes numbers that range from zero to 10: • Zero indicates that an object has a zero or negligibly small chance of collision with the Earth.

000 yr and once per 100. The various categories within the scale are shown in figure 22. A somewhat close. Such events occur somewhere on Earth between once per 50 yr and once per 1. Such events occur once per 100. or well below the chance that a random object of the same size will strike the Earth within the next few decades. It should be noted that the Torino Scale has yet to achieve any wide-scale usage. Collision is very unlikely.An object that is capable of making multiple close approaches to the Earth will have a separate Torino Scale value associated with each approach. with a significant threat of a collision capable of causing regional devastation. with 1% or greater chance of a collision capable of causing regional devastation. A collision capable of causing localized destruction.000 yr.000 yr. is unlikely to reach the Earth's surface intact. with significant threat of a collision capable of causing a global catastrophe. The chance of collision is extremely unlikely. encounter. in the event of a collision. with 1% or greater chance of a collision capable of causing localized devastation. A close encounter. about the same as a random object of the same size striking the Earth within the next few decades. A collision capable of causing regional devastation. 30 . with an extremely significant threat of a collision capable of causing a global catastrophe.000 yr. but not unusual. A close encounter. A close encounter. This designation also applies to any small object that. It is mentioned here for completeness only. An object may be summarized by the single highest value that it attains on the Torino Scale. or less often. TORINO SCALE Assessing Asteroid and Comet Impact Hazard Predictions in the 21st Century Events Having No Likely Consequences 0 The likelihood of a collision is zero. A collision capable of causing a global climatic catastrophe. A close encounter. Categories within the Torino Scale. no fractional or decimal values are used. Events Meriting Careful Monitoring 1 Events Meriting Concern 2 3 4 Threatening Events 5 6 7 Certain Collisions 8 9 10 Figure 22. Such events occur between once per 1. A close encounter.

At first consideration. 31 . Under the remote station approach. no complex spacecraft would be sent out to the approaching NEO. it is important to consider first whether it is better to push the object out of the way or to break it up into small pieces. an incoming NEO could be deflected by a series of intercepting objects. Similarly. probably from the vicinity of the Earth. the resulting high velocity impact(s) would accomplish either deflection or fragmentation. the object would collide with the Earth near perihelion. A strategy based upon interception would involve sending spacecraft out on an intercept trajectory with the incoming NEO. one must consider how to deliver the energy needed to deflect or fragment the object. 3. However. if there were a finite amount of time needed for the system to deliver the fragmentation energy. it would seem that the decision over deflection versus fragmentation is interlinked with the method chosen. An intercepting object would deliver all of its energy at once. then rendezvous becomes necessary. Instead. Additionally. it is anticipated that the deflection mechanism would require a significant period of the NEO orbit to deliver the energy necessary to perturb its orbit. MISSION CONFIGURATIONS In considering how to counter an incoming object. Figure 23 shows the commencement of deflection before aphelion. each imparting enough momentum to slightly perturb its orbit without causing fragmentation. it may seem improbable to actually rendezvous with an incoming object only to subsequently break it up. tending to cause fragmentation instead of deflection. However. all operations are conducted remotely. In the case illustrated. if undeflected. Each option offers its own set of advantages and disadvantages. Rendezvousbased techniques are more propulsively demanding. as they require one to dispatch hardware to actually match orbits with the incoming NEO. Three methods are discussed here: remote station.1 Deflection Versus Fragmentation Figure 23 illustrates the concept of deflecting an incoming object away from an orbit that intersects with the Earth. and rendezvous. with beamed energy or projectiles being used to perform the deflection or fragmentation.3. interception.

as recent Hollywood blockbuster movies clearly demonstrate. there is no requirement to deliver the energy to the NEO in a distributed manner. When considering strategies based on deflection. Finally. Illustration of deflection method of threat mitigation.Collision Point Perturbed Orbit Sun Earth and Target at Detection Figure 23.” but some margin of error is necessary when designing a system that would deflect incoming objects. Collision Point Expanding Debris Fragmentation Point Sun Earth and Target at Detection Figure 24. Illustration of fragmentation method of threat mitigation. In addition. thus. it is important to establish what level of perturbation is necessary to consider the Earth as being safe from collision. it can be defeated in one shot. this might seem the best approach as the object’s destruction means that it cannot threaten the Earth on a later orbit. At first sight. The literature commonly uses a figure of 3 Earth radii as a minimum safe approach distance for a deflected object. This value takes into account the uncertainty in astrodynamical constants that affects trajectory modeling accuracy for the incoming object. there is a unique emotional satisfaction to be derived from destroying a life-threatening object in this emphatic manner. Figure 24 illustrates the concept of fragmentation. One might argue that “a miss is good as a mile. 32 .

the object will almost certainly be eclipsed once per revolution. 3. There would be no vehicle in the vicinity of the object that could accurately assess the effect of the beam. there are also several disadvantages. Finally. and rendezvous. unless the station is placed in a polar orbit. the station and object will move relative to one another. with several distributed impacts occurring instead of one large impact.e.2 Remote Station Versus Interception Versus Rendezvous The three modes considered in this study to deliver deflection or fragmentation energy to the incoming object are remote station. To break up the object into just a few pieces could actually exacerbate the damage to the Earth. Also. one must remember that radial deflection may not be the most efficient deflection strategy. However. make it less attractive than deflection. the suggested fragmentation criteria is that no fragment should have a diameter >10 m. on reflection. For instance. thereby wasting some of the beamed energy. See section 6 for a discussion of optimal deflection directions. A station remains in orbit around either the Earth or the Sun.or Earth-orbiting platforms. interception.Despite its immediate tabloid appeal. such as a solar lens or a laser. Any such assessment would have to be conducted remotely from terrestrial. Energy can be delivered in the form of projectiles fired from the station by a mass driver or by a focused beam of coherent light. The advantage of such a system is that it remains close to the Earth and is easily maintained and upgraded. It is important to break up the object into relatively small components. following pieces can travel within the slipstream of a leading piece and thus reach the ground relatively intact. the approximate orbital radii of Earth and Jupiter. fragmentation does introduce several issues that. Moreover. the system can start deflecting the incoming object almost immediately—which might be months or years—during which an interceptor or rendezvous system would take to reach the object. The remote station mode is depicted in figure 25. Energy deposited into such objects cannot be expected to cause uniform fragmentation. Targeting of the beam or stream of projectiles is not a trivial issue. Also. i.4¥10–5 to 2. causing the deflection vector to rotate. 33 . targeting is required to within 1.8¥10–6 arc s for objects between 1 and 5 au. A major problem arises because asteroids and comets are suspected to have a very heterogeneous composition with significant internal structural flaws. alleviating burnup.. For these reasons. Over a finite time period. Focusing the beam on the object across such vast distances would also be very challenging. This need for remote sensing over large distances would make it more difficult to properly assess the effect on the NEO. a remote station would only be able to deflect incoming objects radially away from the station. Polar orbits would require additional launches to deliver the station into orbit and would result in higher radiation exposure. The fragmented pieces can “draft” off one another in the atmosphere.

The interception strategy is depicted in figure 26. the interceptor will have substantial kinetic energy relative to the NEO. Delivering deflection or fragmentation energy by the remote station mode. 34 tio n Tr Sun aje cto r y . capitalizing on the high kinetic energy that is naturally available. After NEO detection. At this point. Collision Point Earth and Target at Interception— Commencement of Deflection or Fragmentation p ce er nt I Earth and Target at Detection Figure 26. Delivering deflection or fragmentation energy by the interception mode. In most cases. deflection or fragmentation can commence. the interception option allows use of some of the energy initially stored in the outbound propulsion system to be delivered to the NEO. Thus. Interception options tend to be relatively simple.Collision Point Sun Projected Beam/Projectile— Commencement of Deflection or Fragmentation Earth and Target at Detection Figure 25. This difference is further discussed in section 6. The propulsive requirements for interception are substantially less than for rendezvous. the interceptor is deployed to intersect it later in its orbit.

to 5-m ICBM. This strategy offers the best opportunity to continuously evaluate the NEO during deflection or fragmentation operations. Delivering deflection or fragmentation energy by the rendezvous mode. the interception strategy also has its problems. this strategy has the greatest synergy with resource utilization missions. The rate of closure between the interceptor and NEO can be as high as several tens of miles per second. only a system with multiple interceptors dispatched in sequence.S. it has several significant advantages.000 m in size. This is an order of magnitude higher than the closure rate required for the kinetic kill vehicles used in the U. Clearly. Global Missile Defense (GMD) program. The GMD program has had a mixed success rate in interception tests against simulated intercontinental ballistic missile (ICBM) warheads. The limitations placed on the direction of the deflection vector that were encountered with the other two strategies are absent for the rendezvous option. The final strategy is that of rendezvous with the incoming NEO. the rendezvous system is deployed and matches orbits with the NEO later along its trajectory. as shown in figure 27. offers a larger target than a 3.Despite these advantages. the very high closure speeds still pose significant problems for guidance and terminal maneuvering. allowing deflection in the direction that requires the least energy. It should be noted that the interception strategy has the same type of deflection vector limitations as for the remote station strategy. Targeting the NEO is much less difficult for a vehicle in a parallel orbit. can provide the opportunity to continually sense changes in the NEO orbit. This strategy is the one required for most types of deflection systems. pt io nT Sun raj ect ory 35 . After detection. with the later vehicles capable of adjusting their trajectories. Collision Point Perturbed Orbit Earth and Target at Rendezvous— Commencement of Deflection ce ter In Earth and Target at Detection Figure 27. and at 10 to 1. While it is true that an incoming asteroid or comet will not maneuver to avoid destruction. Finally.

during fragmentation or deflection.Of course. the rendezvous vehicle will probably be exposed to a hazardous environment filled with ejecta from the asteroid. The vehicle will have to be designed to resist this environment. 36 . the propulsive requirements to rendezvous with an incoming NEO are much higher than that for interception. Additionally. The rendezvous vehicle may be sufficiently distant from the Earth to make teleoperations difficult and would thus require significant onboard autonomy in an unknown environment that offers many opportunities for unexpected effects. Finally. the response time for a rendezvous system must include both the outbound rendezvous time and whatever inbound time is needed for the fragmentation or deflection process to take place.

neither fragmentation nor deflection can take place until the necessary system hardware is transported out to the approaching NEO. In these cases. some high-energy systems are considered to be more threatening than others. but they do address political and economic issues involved in the deployment of a system. 37 . they are classified as high. They were selected for their ability to meet the mission requirements. Each propulsion system has been assessed against a range of qualitative considerations that have been appropriately weighed by the MSFC study team. the category that assesses a system’s usefulness as a weapon must consider whether use of the system as a weapon will outperform existing weapon systems. development and deployment costs constitute the final consideration. Finally. extra effort must be expended to overcome public opinion obstacles. yellow. and cost. the synergy consideration is divided into manned and robotic exploration missions as well as missions in which the use of asteroid and comet resources is a primary goal. they do not easily lend themselves to a numerical analysis. compactness is also important. These considerations are all regarded as being of second order as they do not affect the ability to meet the immediate mission objectives. and development status. Scalability is the parameter that denotes the ability of a proposed propulsion system to fulfill a range of propulsive needs. Several propulsion systems are considered in this study. First-order considerations include thrust level. they are indicated by the use of three colors—green. long-term readiness. However. i. and compactness. OUTBOUND PROPULSION For both the interception and rendezvous techniques.4. long-term storage at a high state of readiness is a major issue. The propulsion options under consideration are—of necessity— all high-energy systems that could cause extensive damage if misused.e. Second-order considerations include usefulness of the system as a weapon. Although these considerations are of considerable merit in comparing the various outbound propulsion systems. For similar reasons. Since the vehicle will probably have to be constructed and kept until a threat is detected. their level of technological maturity.. usually generated by the lack of public understanding and consequent mistrust of these specific technologies. and to thus handle a variety of threat sizes. These considerations are all regarded as being of first order. For present purposes. When presented diagrammatically. Some type of outbound propulsion system is required to accomplish this. as they directly affect the ability of the proposed system to meet the mission requirements. perceived safety. Each consideration is categorized as being either first or second order. However. NASA always considers safety to be an issue of primary importance. nuclear systems. and red—to indicate decreasing levels of favor. Instead. and low. medium. synergy with other NASA missions. scalability. High thrust levels are favorable because they reduce outbound trip times and gravity losses. The benefits of synergy with other NASA missions need no further explanation. being qualitative.

If included. this discrepancy would unfairly bias the results. They also offer the lowest performance of all the options considered for this project. n Predicting the inert mass fraction requires knowledge of historical vehicle designs.1 Staged Chemical Chemical systems are the most highly developed propulsion technologies currently available. (1) The inert mass fraction for the nth stage is the structural mass of the nth stage.4. a value of 465 is used for the specific impulse (Isp) for each stage. Staging calculations requires knowledge of both the inert and propellant masses for each stage as well as the Isp of each stage. First Stage Second Stage Inbound System Figure 28. Curve fits using these historical data can then be developed (see fig. This regression line is given by en = 0. Here. 29).0579 . ms divided by the n structural and propellant mass for the nth stage. mp . A two-stage lox/LH2 vehicle is shown in figure 28. For this reason only the high-performance liquid oxygen (lox)/liquid hydrogen (LH2) propellant combination is considered. Image produced by INTROS. 38 .12 The analysis method used in this project is covered in detail elsewhere13 and is only summarized here. the inert mass fractions of historical launch vehicle stages listed in table 6 have been plotted against the amount of propellant contained in each stage.0. The line plotted represents a regression fit of the data. The Shuttle’s external tank was not used to calculate this regression as its mass does not include a propulsion system. The inert mass fraction is instrumental in calculating these masses. it is defined as m sn en = m pn + m s n . (2) where propellant mass is in lbm. chemical systems would not be able to handle many of the propulsive requirements for these missions without staging. Two-stage lox/LH2 vehicle.422 log10 ( m p ) . In fact. In the analysis that follows.

0.05 0 10.000. Regression curve fit of lox/LH2 launch vehicle stages. 39 .000 Propellant Mass (Ibm) 1.2 +2Sy –2Sy Inert Mass Fraction 0.1 0.15 0.000 Figure 29.25 Lox/LH2 Lox/LH2 Regress 0.000 100.

8 1.844 23.8 260.300 450.846.656 200.5 98.000 250 101.000 4.9 118.173 324 4.606.6 133.200 311.348 35.7 342.1 405.850 250.817 35.2 31.389 394.031 15.999 64.6 133. zero stage on the STS represents one RSRM.5 120 141. 3Thrust and I represent sea level values for all stages except the last.455.1 396 256.584 451 118. Data on in-service and historical launch vehicles that have lox-/LH2-powered stages.800 86. if any. 6 Aerozine-50 is a 1:1 mixture of UDMH and N2H4. Zero and first stages especially are frequently used in numbers to boost performance.6 122 122 122 122 1.500 444 1 Vehicle families use multiples of particular stages to achieve a variety of performance requirements. i.4 51.5 937 1.5 122 122 Span4 (in) Length (in) Vehicle Ariane 5 (ESA) Atlas IIIb (USA) Atlas V Centaur III Delta III (USA) Delta IV H-1 (Japan) Stage1 0 1 2 1 2 1 2 0 1 2 1 2 0 1 2 3 1 2 0 1 2 3 2 1 3 1 0 1 2 0 1 2 3 Propellants Solid Lox/LH2 N2O4/MMH Lox/RP–1 Lox/LH2 Lox/RP–1 Lox/LH2 Solid Lox/RP–1 Lox/LH2 Lox/LH2 Lox/LH2 Solid Lox/RP–1 Lox/LH2 Solid Lox/LH2 Lox/LH2 N2O4/UMDH 74.200 90.818 6. Zero stage indicates strap-on boosters.040 5.4 51.000 6.140 91.515 1.2 157.1 629.627 87.000 1.679 432.670 310 33.614 19. not two.984 0 218.620 42.000 261.5 860.872.500 43.800 447 183.512 1. sp sp 4 Dia is the diameter of the vehicle fuselage.655.150.614 932.6 1.000 67.300 50.411 498.806 30.001 2.5 650.515.6 150 120 47.532 9..5 1.000 365 200.796 1.508 734.9 133.9 909.141. The data listed represents one stage only.112 42.5 94.206 194.1 157.464.274 259 259 260 440 425 232 426 264 337 363 455 157.096 254 24. Vacuum thrust and I are listed for the final stage.999 316 29.193 45.1 1.842 8.314.3 150 22.299.758 256.700 5.6 275.2 157.6 H-IIA (Japan) Long March 3D (China) 191.2 866.957 1. Mi1 (lb) 593.000 1.618 673.369 4.700 3.5 129.459 9.700 1.760. 5 There are no engines on the Space Shuttle external tank.2 236.2 1.079.700 9. 40 .4 157.849 5.615 65.4 157.8 963.5 192.000 338 30.2 90.600 6.669 311 44.416 232 253 450 291 31.Table 6.4 346.278 462 578.9 118.520 1.800 526.4 409.380 200.5 N2O4/UMDH N2O4/UMDH Lox/LH2 Lox/LH2 Lox/RP–1 Lox/LH2 Lox/RP–1 Solid Lox/LH2 Lox/LH2 Solid N2O4/Aerozin e-506 N2O4/Aerozin e-506 Lox/LH2 Saturn II Saturn IB Saturn IVB Saturn 1C STS (USA) RSRM ET Orbiter Titan (USA) 1.5 94.2 157.609 230.250 273 244.1 1.200 4.6 396 1.2 90.736 462 47.753 189.5 212.4 24.9 118.5 90.e.696 46.049 259 241.000 288.8 474 157.536.033 186.000 5.998 190.060 4.999 238 12.4 1.5 98.300 50.4 711.156 23.6 157.9 122 122 122 122 200.1 212.785 20.7 122 460. 2 Neither initial or final mass includes payload.000 987.415 Mf2 (lb) Thrust3 (lbf) Isp3 (s) Dia4 (in) 118.400 28.640.7 342.019.2 94.060 499.2 58.4 346.7 787.843 396 978 256.000 7.149 17.6 260.464 362.000 7.5 94. Span is the total span of the vehicle including fins and wings.043 374.220.069 6.750 462.181.958 218.968 794 30.7 878 311 378 74.589.5 1.5 157.4 396 145.

These values can be broken into stage masses using the following equation: Ê ˆ N Mi n + m pay n ˜ . DV = Â g0 Ispn ln Á ÁM -m +m ˜ in pn pay n ¯ n =1 Ë (4) Mi is the stage mass or the sum of Ms and mp . This requirement is defined by the trajectory and is discussed further in section 6. DV is the difference between the velocity required for the final orbit and the velocity in the initial orbit.en MRn . For the (n–1)th of n stage.The energy required to deliver the threat mitigation system is defined as DV. for the final stage. respectively. the mass ratio and payload fraction for the nth stage. mpay is the payload or the mass n the threat mitigation system.8066 m/s2. There is an optimal distribution between the stages that is defined by N DV = Â g0 Ispn n =1 ag0 Ispn + 1 aen g0 Ispn . 9. Isp is the specific impulse of the n nth stage. (7) MRn = Mi n M fn . MRn = ag0 Ispn + 1 aen g 0 Ispn (5) ln = 1 .1 (6) where MRn and ln are. (8) 41 . MRn . Payload fraction is also defined in a manner similar to inert mass fraction: m pay n ln = Finally. and a is a Lagrange multiplier. and mpay is the mass of everything above the nth stage. the mass is the payload mass together with the total mass of the nth stage. Each stage produces a portion of the total DV requirement. Finally. (3) where g0 is the gravitational constant at the Earth’s surface. n n n Thus. mass ratio is defined as m pn + m s n .

However. the data used to predict inert mass fractions were for launch vehicles. which will result in significant mass savings. Only when a threat is detected would the system be activated. equations (6) and (5) can be solved in series. these losses are minimal for the types of system envisioned. equation (3) can be solved for the Lagrange multiplier. The propellant masses are used in equation (2) to calculate new inert mass fractions. The relatively low performance of chemical systems also makes them massive and the low density of LH2 means that they will be physically bulky. it is assumed that these mass savings will counteract the additional mass necessary for the active and passive thermal protection systems necessary to retain cryogenic propellants indefinitely. However.2 for each stage. Next. where the vehicle would be deployed in low-Earth orbit (LEO). That raises an operational concern. The systems envisioned would be applied in a situation where the vehicle would be assembled in advance and then placed in a parking orbit. However. The thrust level and scalability for chemical systems are excellent. the inert mass fractions are estimated for each stage. Equations (7) and (8) can then be used to derive propellant and inert masses. This calculational scheme is repeated until convergence is achieved. Our code also executed this loop for a five-stage system down to a single stage to determine the least number of stages necessary to meet the DV requirements. The Lagrange multiplier method is usually of limited use. Table 7. Both of these conditions are alleviated for in-space vehicles. such as gravitational or drag losses. Table 7 lists the qualitative considerations for chemical propulsion.5 for first stages. Qualitative considerations for outbound propulsion using chemically-powered rockets. Then.The method for solving these equations is as follows: First. First-Order Qualitative Considerations Thrust level Scalability Long-term readiness Compactness Second-Order Qualitative Considerations Usefulness as weapon Perceived safety Synergy with other NASA missions Manned missions Robotic missions Resource utilization missions Costs Development Deployment Low High High Medium Medium Low Medium High High Low Low 42 . lox/LH2 systems are not considered viable candidates for applications where long storage times are necessary. as it does not handle DV requirements that include losses. the propellant tanks will be exposed to temperature cycling that will probably exacerbate the normal problems of cryogenic propellant storage. which must handle high stresses during ascent and produce an initial thrust-to-weight ratio of 0.2–1.9 for upper stages and 1. the leak possibility for cryogenic propellants and the thermal protection requirements combine to indicate that long-term readiness will be difficult. Our code used a value of 0. In this type of application. for this level of analysis. Therefore.8–0.

A schematic of the NTR is shown in figure 30.53. and some basic assumptions to apply it to the same type of analysis that was used for chemical stages have been made. the historical approach that was employed to calculate masses for chemically powered stages cannot be used. Instead. Due to their extensive use over the last half century. they are perceived to be very safe by the general public.8 .Chemical propulsion does not lend itself for use of these systems as a weapon. but deployment costs are higher due to the size and maintainability issues. These data show a linear relationship between the total engine mass (mreac) and thermal power produced by the reactor (Pjet): P jet = 0. The Isp used for nuclear thermal rocket (NTR) propulsion systems is 850. data from previous NTR engine development programs have been used. Pump Propellant Tank Reactor Core Nozzle Figure 30. Schematic of a nuclear thermal rocket. the level of potential synergy for chemical propulsion is considered to be medium. NTR engines weigh a considerable amount more than comparable chemical engines because of the required reactor and shielding mass. Although almost twice as high as that for the chemical option. it is not enough to preclude the necessity of staging. For robotic and resource utilization missions.127 mreac . similar to the chemical option but with improved performance. Development costs should be low for chemical systems due to the significant heritage and well-understood technologies of these systems.14 Table 8 lists data from the Nuclear Engine for Rocket Vehicle Applications (NERVA) program. 4. (9) 43 . Since no NTR vehicles have been developed.2 Nuclear Thermal Rocket The nuclear rocket option was included to give a high thrust option. Chemical propulsion has strong applicability for manned missions where low trip times are important.

000 2.000 20.300 41.320 This relationship is shown in figure 31.8. Data on nuclear rocket engines developed under the NERVA program.550 7.700 12. Using this value.000 40. Regression curve fit of NERVA program-developed engines.8 g0 Isp  Mi n . it is assumed that the initial thrust-to-weight ratio for an NTR stage is 0.000 5. 2 (10) For this analysis. Mass (kg) Small engine XE NERVA Phoebus-2A 2.000 50.000 0 0 10.000 3.000 1.Table 8. 6. 2 n=N P jet = (11) 44 . the power can then be related to the vehicle mass as follows: n 0. but it represents a conservative estimate.570 5. The power produced by the rocket jet is related to the Isp and thrust produced by the rocket by the following equation: 1 P jet = g0 Isp F .000 30. This is based to some extent on upperstage data from launch vehicles.140 1.000 Total Mass (kg) NTR NTR Regress +2Sy –2Sy Figure 31.679 Thermal Power (MW) 367 1.000 Thermal Power (MW) 4.

The chemical inert mass fraction accounts for the tanks. Table 9 lists the qualitative considerations for NTRs. However. like chemical systems. Qualitative considerations for outbound propulsion using nuclear thermal rockets. Like chemical systems. for the current level of analysis. piping. However. it is assumed that these inconsistencies cancel each other. 2 (12) This value can then be incorporated in the same analysis scheme as was used for chemical systems. and ancillary components necessary to make up a stage. thrust structure. the reactor mass was divided by the stage mass and included with half of the inert mass fraction to yield an NTR inert mass fraction: mreac n Mi n e(NTR ) n = + e(chem) n . First-Order Qualitative Considerations Thrust level Scalability Long-term readiness Compactness Second-Order Qualitative Considerations Usefulness as weapon Perceived safety Synergy with other NASA missions Manned missions Robotic missions Resource utilization missions Costs Development Deployment Low Low High Medium High Medium Low High High Low Medium 45 . this system will be more compact than an alternative chemical system. the chemical inert mass fraction still contains the mass of the propulsion system too. The justification for this value is as follows. LH2. It would be convenient if this analysis could be conducted in the same way as for the chemical propulsion system from section 4. Also. thrust level and scalability is believed to be high. Table 9. Therefore. Additionally. This assumption is optimistic because the LH2 tank is much larger and heavier than the lox tank in a chemical vehicle. an on-orbit system maintained indefinitely will require active and passive thermal protection. as did the chemical systems.1. The reactor mass can then be found by the solution of equation (9). maintenance of LH2 propellant will make long-term readiness difficult.where the summation is intended to calculate the cumulative mass of the nth stage and all of the stages above it. In this analysis. instead of two. The reactor mass term accounts for the added mass associated with NTR engines. which is superfluous here. It is divided by 2 because the NTR vehicle uses only one propellant.

Performance approaches that of a fusion-driven spacecraft. ORION (fig.17 pusher plate. EPPP utilizes not only the fission energy liberated in a nuclear device but is also substantially enhanced by a fusion energy release. at a cost of about $8 to $10 million—significant research funding for those years.3 Nuclear Pulse The nuclear pulse concept was first considered only as an asteroid deflection technique. the most conventional—a term loosely applied to pulsed nuclear rockets—approach was taken: the pusher plate and shock absorber configuration. but without the additional challenges inherent in fusion technology. including the standard pusher plate. In 1999. but with greater compactness. Significant data have recently been declassified and served as the basis for the generic vehicle calculations presented here.18 magnetic field. However. The pulse unit designs are extremely general in nature due to the sensitive nature of the technology involved. 33).21 All of these momentum transfer mechanisms (MTMs) utilize the same nuclear detonation energy source. Deployment costs should be favorable due to their similarity with chemical systems. Project Gabriel.19. Each couples the tremendous burst of high-velocity particles to the spacecraft by spreading the intense shock over a longer and more tolerable time period. 32). external pulsed plasma propulsion (EPPP) or nuclear-pulsed systems are the best possible performers.An NTR system would have little usefulness as a weapon.20 and large lightweight sail/spinnaker (Medusa concept) suggested by Solem. thus. This program achieved tremendous technical success. This system offers high synergy for manned and resource utilization missions. thrust is also considerably higher than for pure fusion concepts. However. using current materials and technologies (fig. and among propulsion techniques using known technology. but failed to maintain political support for a number of reasons. these systems are most efficient at larger sizes.16 rotating cable pusher. assessed the ORION concept. they will be less effective for smaller robotic missions. a smaller NASA study. as applied to a smaller vehicle design. Development costs are complicated by the fact that testing nuclear systems introduces a number of environmental issues. 46 .24 A significant portion of that work was also used in this study.22 This classified program began in 1958 and ended in 1965. However.15 As the fission fuel fragments ejected are extremely massive nuclei. the results appear to be realistic and well within current state of the art for such devices. transport of the deflection system out to the NEO needs to be accomplished as swiftly as possible.23. 4. In this preliminary study. as a nuclear system. its perceived safety to the general public is rather low. However. There are several concepts that have been proposed for EPPP. The specific geometry and scaling model was primarily derived from the original Air Force Program.

NASA Gabriel spacecraft (1999).S. 47 . U.Powered Flight Crew Station (Shielded) Structural Spine Crew Accommodations Crew Module Propellant Magazines Basic Structure Includes Pulse-Unit Delivery System Secondary Shock Absorber Primary Shock Absorber Pusher Plate Pulse-Unit Ejection Path Basic Nuclear Pulse Propulsion Module Standoff Distance Ejected Pulse Unit Point of Detonation Figure 32. Air Force ORION spacecraft (1964). Figure 33.

A 30-percent mass contingency factor was applied across the entire estimated vehicle mass. Qualitative considerations for outbound propulsion using nuclear pulse. but it receives high marks for interstellar type missions. The fissionable material is readily available. the primary shock absorber was a three-tiered carbon structure and the secondary shock absorber consisted of a set of long. sample returns. A spreadsheet analysis was developed.The vehicle mass estimator uses geometry and density parameters that are based on rough scaling of the ORION vehicle. Although an effort was made to ensure enough spacing to preclude a thermal reactor critical geometry being created out of the pulse unit storage. multiplanet tours. which are negative. Evaluation of the first-order qualitative considerations (table 10) clearly illustrates the outstanding technical merit of this system. it will always be possible. There are tons of highly enriched uranium and plutonium available from decommissioned Cold War weapons. gas strut type systems. result from concerns over security of nuclear material and technical knowledge. and using it in peaceful space applications actually achieves cost savings in other areas. The robotic mission category is labeled a medium. and number of levels required. Table 10. near-term probes. although difficult. Furthermore. no calculation or neutronic simulation was used to check for this issue. which included gross estimates on pulse unit volume. it is rendered completely unusable in any future application. both in the United States and in Russia. packing fraction. Only in space is it safe from theft. The other factors responsible for the high cost result 48 . When used as rocket propellant. It also has excellent synergy with many other NASA missions. It is the only system that has all high (green) ratings for the primary parameters. for someone to extract and reprocess it for weapons. The high cost is primarily generated by the significant need to regulate and secure fissionable material and prevent nuclear weapon proliferation. The pusher plate was assumed to be constructed from solid titanium. since nuclear pulse has less potential application for smaller. This material is presently being stored at a high cost to both nations and reprocessing it for burial will be extremely expensive. but much of the material will escape the solar system. Not only is the exhaust material spread out into space. no matter what is done to this material. First-Order Qualitative Considerations Thrust level Scalability Long-term readiness Compactness Second-Order Qualitative Considerations Usefulness as weapon Perceived safety Synergy with other NASA missions Manned missions Robotic missions Resource utilization missions Costs Development Deployment Medium Low High Medium High High High High High High High The second-order qualitative considerations. and large-scale robotic exploration.

it would cause rather limited damage. and its power is held to a minimum. since many hundreds of these “pulses” must be endured. but carrying a payload of nuclear devices designed to deflect the asteroid. since this type of propulsion system is similar to a weapon system or could even be used as one. much less design and fabricate. location before shipment into space. It would seem very simple to build and store the devices for only several months at one U. such as a terrorist-supporting state. Pulse unit tests will require underground testing as well as several deep-space tests. it being logical to assume that if approval to build and launch the EPPP vehicle (fig. the many extremely precise mechanisms that are required to make small amounts attain supercriticality. the asteroid threat mitigation could then be accomplished with the same type of technology. Nations that have large quantities of fissionable material can easily build large weapons and those who would have only small amounts. this concept suffers from a basic human fear of ourselves. some shock absorber development and testing. 49 . The calculations were based on two external inputs: mission DV and payload mass. the historical record is favorable. that a chemical or electric system would be used for outbound propulsion. The EPPP vehicle will require a monolithic pusher plate of 5–10 m (8 m was used in the final design developed here).from the large size of the vehicle and also from testing issues. although comparable to other options considered in this study. will find it nearly impossible to assemble. Note that the device will be only meters away from the spacecraft. and it will be difficult to overcome this. Like the original ORION program. This conclusion is not absolutely certain. Other negative considerations also arise from concerns over security. The arming can be made fail-safe and only possible in space/onboard the spacecraft. even for purposes of asteroid defense.S. even after accidents in which a device has been inadvertently dropped from an aircraft. 34) was obtained. This record has been maintained for over 50 yr with thousands of weapons being moved around the world. The mission DV was based on the trajectory analysis and is discussed in other sections. and a very fast pulse unit dispenser system. as there may be some chance of using another deflection scheme. No nuclear weapon has ever been lost or accidentally detonated. or more likely. The initial launch costs could also be high. since the devices are very small and not designed to generate the weapon effects of an atomic bomb. Although a pulse unit would be destructive if accidentally or purposely detonated on Earth. Perceived safety is assessed as low. As far as security and safety are concerned. The payload mass is based on the nuclear pulse deflection option. It is apparent that all of these many mitigating points are not likely to counteract the likely perceived political risk.

the diameter was 8 m with a thickness of 3 cm. The material selected was titanium. On one side. but which was not as safe or reliable. and using the required mission DV. Again. The plate was a simple disk shape. a carbon fraction of 10 percent was specified. Each tier was 0. a large percentage of the volume was considered to consist of voids. although the real design would have a special taper profile as developed during the ORION program. the diameter is three-quarters of that value. and on the other side. which has a mass density of ª4.5 m thick and the first. i. and (2) the pulse unit design. was cylindrical in shape with the same diameter as the pusher plate. that connected to the metallic pusher plate. The volume is calculated for each section and the density is used to determine the mass. The calculations then proceed with two basic subsystem designs: (1) The vehicle’s MTM. The primary shock absorber was a three-tier block of carbon microfiber springs. Since the shock absorber is envisioned to be leaf spring or coil structured. whose mass was similar to that of the carbon material. The next two tiers were cone shaped rather than cylindrical. an iterative solution is employed to determine the propellant required for the mission (the propellant being the number of individual nuclear devices or pulse units needed). each has the same diameter as the preceding tier. assuming a homogenous material. a simple geometry and density approach was used. which is the pusher plate and the shock absorber system. Also. 50 . there is structure mass and associated volume that scales with the number of pulse units required.500 kg/m3.e.Pusher Plate Nuclear Shape Charge Device 10 m Nuclear Pulse Unit Storage 8m Chemical Propulsion Stage Structure Dispensing Mechanisms Secondary Shock Absorber System Three-Tier Shock Absorber System Figure 34. For the carbon material with an assumed density of 1. The MTM was based on simple geometry and density considerations. since the material manufacturing process and actual property values are presently based on speculation. EPPP concept vehicle. The original ORION program specified gas-filled toroidal chambers. This was assumed without any rigorous analysis. with thickness and diameter specified and a mass calculated using a density. In this point design. From these two independent calculations..600 kg/m3.

e. simple geometry and density estimates were made instead of using masses taken from the ORION study. Again. The operation of the shock absorber system is illustrated figure 35. i. columniation structure. This accounts for the thermal energy of the main fragments only and not the neutrons. arming circuitry. the mass of U235. high explosives. but modeled as a simple hollow cylinder. The chemical explosive energy required 51 . i. The inert mass includes everything else which comprises the complete device: impurities in the U235 fuel. using Avogadro’s number and a conversion factor. Uncompressed 2.The secondary shock absorber system was chosen to consist of gas tube struts. similar to the ORION engineering.e. The sum of the three major components that make up the MTM system gave the total vehicle dry mass. The pulse unit performance parameters were based upon the amount of energy released per nucleus fissioned. The energy released per U235 nucleus fissioned is ª185 MeV.5 m in diameter and 10 m in length. and the inert mass. An estimated “burnup” fraction (percent of U235 nuclei actually split in the reaction) was assumed and the total energy released by the pulse unit was then determined. or prompt gamma rays. This value was used to determine the theoretical average energy release per kilogram of fuel (7. neutrinos. 1. Fully Compressed 3. Each of the six tubes was 0. A 30-percent mass contingency factor was applied to the total vehicle mass. the simple geometry involved made it easy to develop a parametric calculation. burnup fraction times mass of fuel in the pulse unit (kg) times the energy per kilogram. channel filler. or low atomic number propellant and casing/mounts. Ready for Next Pulse Figure 35.500 kg/m3 and a wall thickness of 2 cm. the effect of changes in the number and size of the shocks could be readily investigated. Although the ORION numbers were based on relatively detailed engineering drawings.592¥1013 m2/s2). which was used to explore a wide trade space area. product decays.. The material selected was titanium with a mass density of ª4.. Shock absorber operations.

e. If the estimate was not sufficiently close. a level height of twice the pulse unit length was assigned and the total propellant volume was determined. 52 .16 From this Isp. A 5-percent contingency factor was added to account for misfires. Other calculations were performed to estimate the size of the propellant magazines. i. low atomic number with a mass density assumed to average 1. the size and shape was ascertained for a single pulse unit. and other performance losses. This accounts for all the storage. the new value is halfway between the initial estimate and the calculated value.e. packing cylinders. plasma burst preferential direction. and standoff distance (distance from pulse unit detonation to the center of the plate).e.. Required “floor space” was found by assuming a square storage geometry. Using the ideal rocket equation. a new value was automatically generated by the computer. With the final number of whole pulse units determined. (14) a final mass is generated for the spacecraft at the end of the mission. The difference between that and the start mass is the propellant used.. mounting. Using a length-to-width ratio of 2. (13) Isp was determined from the average particle velocity times an effectivity factor. The effectivity factor (FE) in the Isp equation accounts for the collimation factor.807 m/s2). a total energy yield was determined and was expressed as an equivalent number of kilotons of tri-nitro-toluene (TNT). which has a mass density of 19. The last spreadsheet calculation determines the number of pulse units needed to be carried as propellant using a spreadsheet goal seek iteration routine. and adding an additional 25-percent fluff factor. The diameter of the storage area was held to half that of the plate. and the calculation repeated. The conversion criterion was an error better than 1¥1015 between the two values. i. and the pulse units were estimated by the density of U235. The total fission energy was then assumed to be absorbed evenly into the total pulse unit mass and the average velocity of the particles determined from the equation: E = 1 / 2 mv 2 . DV g 0 I sp M f = Mi e . Isp=v¥FE/g. An initial estimate of the propellant mass is made. The effectivity factor equation was derived by Thane Reynold.500 kg/m3. The mass of the MTM and payload are then added to obtain the start mass of the spacecraft. assumed to be 15 percent. Dividing that quantity by the mass of one complete pulse unit (mass of U235 fuel and inert mass) established the number of pulse units (rounding up to the next whole value). pusher plate diameter. trajectory errors. and inert material. i.000 kg/m3. and dispensing mechanism hardware needed to handle the pulse units. i. where g=9. i. Finally. A 10-percent void fraction was assumed to determine the gross volume. The floor area required was divided by the area available based on the one-half plate diameter criteria to estimate the number of storage levels.e. This estimate is multiplied by a generic tankage fraction..to start the nuclear reaction was neglected..e.. This was used to ascertain whether the propellant could be reasonably launched into LEO by conventional chemical rockets. divided by the gravitational constant. the actual total mass of propellant is determined and compared to the initial estimate.

The analysis method discussed here follows that given elsewhere. If employed. Due to the sensitive nature of any details concerning nuclear devices and their effects. including both the blades and the hub. the square sail is slightly more attractive than the other options. pulse unit yields were in the 3 to 4 kton range (present devices are normally in the many megaton range) and the range of Isp was 2. the final mission design was not very sensitive to the plate diameter as long as the plate diameter-to-standoff distance ratio was not altered significantly. In general. The difference between these types of sails relates mostly to consideration of structural design. Many small trade studies were conducted in order to gain an understanding of how the entire system would function. Sufficiently large sails are also capable of moving sizable payloads over long distances. Square sails use a square sail sheet supported by booms and lines. The blades are kept taut through rotation of the entire sail. Also. but nonetheless is considered to be realistic. No estimate of the effects of erosion. the inputs were intentionally conservative and were obtained from open literature sources. most notably. The vehicle was also designed using simple geometric and density-based calculations. The vehicle has excellent ability to absorb mass growth. cost.4 Solar Sail The solar sail offers unique capabilities for rendezvous with an incoming object.The EPPP calculation approach was simplistic. The 30-percent mass margin was considered adequate. they would be expected to enhance performance. or shock was made for the pusher plate design. Surprisingly. Because of the conservative assumptions used in pulse unit design. disk sails. Interception is also a possibility but it would not be a first choice for threat mitigation systems depending upon kinetic energy. despite the large uncertainty in the design parameter inputs. The analysis presented here is based upon a square sail design (fig. 4. for such a preliminary design. This gives confidence in the practicality of such a propulsion system. Fusion energy. or efficiency of any real nuclear device.000 s. No inference should be made regarding any parameters relating to the true composition. as well as the use of plutonium fuel.500 to 5. in most cases.25. radiation. It has a circular sail sheet that is kept taut through rotation. narrow sail blades that are each connected to a central hub. the MTM was not additionally burdened with a higher mass margin. but not necessarily conservative. The acceptable vehicle trade space was found to be very large. was not considered. and heliogyros. 36). 53 . and availability. Previous studies have suggested that. Solar sails are capable of substantial inclination changes that are difficult to achieve with other propulsion systems. A disk sail attempts to incorporate advantages of both square sail and heliogyro. a larger vehicle actually yields improved performance. a solar sail would remain close enough to the Sun to maintain significant propulsive capability. although they each have their own drawbacks. geometry. Even doubling the payload mass increases the vehicle gross mass by <20 percent or ª50 more pulse units. The heliogyro uses long. size. their natural radioactive decay.26 There are several solar sail configurations available. The general conclusion from the analysis is that this technology has excellent performance capability and that there are no known technical barriers. The three most popular are square sails. since most NEOs are expected to occupy orbits within that of Jupiter.

The total sail area (Asail) is equal to the square of the sail length. Similarly.Figure 36. square solar sail. The centerline angle (f) is the angle between the sail normal and force vector. Tip Vanes Control Boom Payload Envelope Spars Fn θ α F φ F t I Sail Film Stays To Sun Figure 37. the cone angle (q) is the angle between the force vector and the incident luminous flux. The sail will project an area (A) normal to the incoming luminous flux. Artists concept of a billowing. Each side of the sail has a length (l). 54 .25 The important dimensions for a square solar sail are illustrated in figure 37. Schematic and dimensions for a square solar sail. The sail will normally be oriented at an angle of attack (a) relative to the plane perpendicular to the orbital radius vector.

Spectral reflection is the preferred method as it is the most efficient. which partially cancels the motive force. The sail converts this pressure into thrust through absorption or reflection. The normal and tangential forces due to absorption are given as Fan = PA cos2 a (16) (17) Fat = PA cosa sin a .s¢) r¢ cosa ( ) (18) (19) Frt = PAr¢s¢cosa sin a . where A is the cross-sectional area presented to the Sun. (20) where Bb is the non-Lambertian coefficient for the back side of the sail.r¢) e f B f . However. 55 . ef and eb are the emission coefficients for the front and back.eb Bb e f + eb cosa . The normal force due to emission is given as Fe = PA(1 . 100-percent reflection is not easily achieved. where r¢ and s¢ are the reflection coefficient and the spectral reflection coefficient. The tangential and normal forces due to reflection are calculated using Frn = PA r¢s¢ cos2 a + B f (1 . The Sun should be modeled as a finite disk out to distances as far as 10 solar radii. The solar light pressure can be calculated as a function of distance from the Sun using the following equation: 3˘ È Í Ê R2 ˆ 2 ˙ Ls s P ( r) = Í1 . r is the radius of the sail from the center of the Sun.Á1 . respectively. This emission process also produces its own radiation pressure. Absorption of the incident light is less attractive because of the resulting thermal loads that must be radiated away from the reverse side of the sail.The first task in solar sail design is to model the luminous flux emanating from the Sun. and Ls is the solar luminosity. respectively. Also.2 ˜ ˙ . These values can be traced back to a Jet Propulsion Laboratory (JPL) solar sail model developed during the 1970’s to study a proposed Halley’s Comet rendezvous mission. Additionally. The values used in this study for these optical coefficients can be found in table 11. ˜ Á 2 6pcRs Í Ë r ¯ ˙ Î ˚ (15) where Rs is the radius of the Sun. c is the speed of light. Bf is the front non-Lambertian coefficient used to model the nonspectral reflection.

349 0. billowing force and other parameters used in solar sail analysis. Fn (24) Finally.55 0. Optical.5 0 Low as possible Study Value 0. The cone angle under billowing can be determined by q =a -f . T =Í Í s ¢ e f + eb ˙ Î ˚ ( ) (25) where s ¢ is the Stefan-Boltzmann constant.Table 11.25 Variable Bf Bb εf Name Front non-Lambertian coefficient Back non-Lambertian coefficient Front emissivity coefficient Back emissivity coefficient Reflection coefficient Specular reflection coefficient Force coefficient 1 Force coefficient 2 Force coefficient 3 Sail loading parameter Ideal Value 2/3 2/3 0 0 1 1 0.05 0.88 0. the sail temperature can be calculated by accounting for the thermal flux due to absorption and emission: 1 È 1 .5 0.79 0.011 6 gm/m2 r´ s´ C1 C2 C3 σ εb The total sail normal and tangential forces can be determined by summing the above forces: Fn = Fan + Frn + Fe (21) (22) Ft = Fat + Frt . (26) 56 .662 –0.r¢ cP cosa ˘ 4 ( ) ˙ .94 0. The sail can be allowed to billow under the luminous pressure and therefore will not present a plane surface. The total force and centerline angle are determined using F = Fn2 + Ft 2 (23) F f = tan-1 t .55 0.

the sail lightness number (b) is a characterization of the sail’s acceleration compared to that due to the local gravitational force imposed by the Sun: Ls . the sail trajectory is still hyperbolic. For values of b > 1. Ms is the mass of the Sun.and the total force adjusted for billowing is calculated using Fnorm = F (C1 + C2 cos 2q + C3 cos 4q ) . the force exerted on the sail exactly negates the gravitational force from the Sun. but with a thrust higher than the local gravitational force. C2. For a b < 1/2. 57 . G is the universal gravitational constant. (27) The coefficients C1. and Ls is the solar luminosity. Figure 38 illustrates the relevance of lightness number to sail orbital capabilities. When b = 1. 2pGM scs b= (29) Here. and C3 are billowing force coefficients that were also determined during JPL’s Halley’s Comet study. (28) There are several figures of merit that are typically used to compare sail performance. For a b of between 1/2 and 1. These coefficients and the optical parameters are not universally applicable but are used here as a first approximation to take account of billowing and optical physics effects. First. The mass of the sail can be calculated using Mi = sA + m pay . the sail follows a hyperbolic escape trajectory. the sail travels on an elliptical trajectory around the Sun.

it is given by Fnorm . 2 Asail P h= (30) This value is a constant for any heliocentric distance or any sail area.25 The sail’s efficiency is a measure of its ability to convert the incident luminous flux into a propulsive force. It is given by a0 = 2hPA .5 Sun r V Deploy Sail β =1 β >1 Figure 38.0< β <0.5< β <1 β =0 β =0. Sail trajectories relative to lightness number. Mi (31) The sensitivities to several of these figures of merit are given by the following equations: Da0 = a0 -1 Ds s m pay s s 1+ ms (32) 58 .5 0. The characteristic acceleration is the acceleration that the sail will experience when it is normal to the Sun and at a distance of 1 au (the average radius of the Earth’s orbit around the Sun).

it can be maintained indefinitely in a very compact state. Intercept capabilities are less impressive. although the unfurling process is not trivial. as long as the total payload to be delivered is small. Qualitative considerations for outbound propulsion using solar sails. Sails have the ability to achieve a variety of mission types. The deployment of a furled sail presents few complications. First-Order Qualitative Considerations Thrust level Scalability Long-term readiness Compactness Second-Order Qualitative Considerations Usefulness as weapon Perceived safety Synergy with other NASA missions Manned missions Robotic missions Resource utilization missions Costs Development Deployment Low High Low High High Medium Low Low Low High High It is difficult to conceive of any way in which a solar sail could be used to cause harm. However. The sail does not have to be unfurled until a threat is identified. Solar sails produce very low thrust. Table 12. As even the least demanding missions require large sails. This makes it difficult to conceive manned missions using solar sails. therefore. the sail should be considered a strong contender for the outbound leg of any rendezvous concept. 59 . Sail development cost estimates are complicated by the difficulty in measuring the low propulsion levels on a ground-based facility. Table 12 lists the qualitative considerations for solar sail outbound propulsion. For these reasons. as sails do not accelerate quickly.Dm pay -1 a0 = Da0 1 + ms m pay m pay (33) a0 = Da0 1 DA . and sail area. there is limited potential for scaling up the design. m pay A 1+ ms (34) These are the sensitivities to sail density. payload mass. solar sails have proven in this study to have surprising capabilities to rendezvous with the varied orbits of potentially incoming objects. respectively. Solar sails have unique abilities to change inclination and achieve non-Keplerian orbits.

The captured light is a function of the cross-sectional area of the hemisphere. as a curved solar collector will require more structural support to maintain its shape than would a flat solar sail. Thus. Solar collector configuration. Accounting for the additional area of the hemisphere gives a conservative analysis. Projected Area Secondary Collector Payload Envelope To Sun Fi l Stays Sail Film Spars Fr Figure 39. The secondary collector directs the light outward so as to generate thrust. This configuration avoids the loss of thrust experienced by a conventional solar sail when attempting to direct its thrust vector in an optimal direction. Additionally. this is not a concern for this application. Although this configuration is susceptible to chromatic aberration. this optimistic assumption is expected to negate the pessimistic hemispherical shape assumption. However. More efficient parabolic configurations may be found. the principle is shown in figure 39. the curved sail is assumed to be hemispherical. 60 . The solar collector uses the same optical parameters listed for the solar sail in table 11.4. The curved sail centerline remains parallel to the vehicle Sun radius vector so that the sail generates the maximum possible thrust. Use of these loading factors is somewhat optimistic. but they would require ray tracing or other advanced calculations that are beyond the scope of this project. with the secondary collector located at the center of the projected sphere. as will be discussed in section 5.5 Solar Collector The solar collector is closely related to the solar sail. figure 39 illustrates a sail with a projected hemispherical area. The focused light can also be effectively used to redirect incoming asteroids and comets. A solar sail with a curved configuration is used to focus incident light onto a secondary collector. The sail mass is calculated using the hemispherical surface area and the sailloading factor found in table 11. This is clear because it is obvious from the figure that the additional cross-sectional area capturing light is not significant. As shown in figure 39. a hemispherical shape would limit the direction in which the secondary mirror could focus to the plane perpendicular to the collector centerline. but with an actual area that is less.

Qualitative considerations for outbound propulsion using solar collectors. the collector size need not be large and its shadow will be insignificant compared to the curved sail area projected normal to the Sun. l/r can be treated as being zero.233 m. 6. If the collector is made large.34¥10–10 and the image size is 0. The image size is rc = M ¥ Rs . Note that the configuration cannot be folded as easily as can a solar sail. The curved sail focus is found to be27 l¢ = 1 . (37) where Rs is the radius of the Sun. which gives the location of the secondary collector in figure 39. For the sail >1 au. The collector has considerations similar to those of the solar sail in all other respects. the magnification is 3. For a 100-m-diameter sail. For all realistic situations. Therefore. Table 13 lists the qualitative considerations for the solar collector. The magnification is shown to be M= l¢ . Table 13. r (36) The curved sail essentially produces an image of the Sun. First-Order Qualitative Considerations Thrust level Scalability Long-term readiness Compactness Second-Order Qualitative Considerations Usefulness as weapon Perceived safety Synergy with other NASA missions Manned missions Robotic missions Resource utilization missions Costs Development Deployment Low High Low High High Medium Low Medium Medium High Medium 61 . The solar collector is expected to have better thrust levels and scalability than the solar sail because of its more efficient use of the incident solar radiation.Design of the secondary collector is of particular importance for this concept.96¥1010 m. 2 1 + l r (35) where l is the diameter of the curved sail and r is the orbital radius of the sail from the Sun. equation (35) yields a focus of 50 m. then the amount of sail area that it blocks may be prohibitive.

sandstone. JPL. (40) where r is the actual radius (in meters) and W is the required blast yield of the device. (39) where v is the shock-induced particle velocity and E is the energy. Á ˜ H Ë ¯ (41) where the density of the target object is assumed to be 2 g/cm3. Finally. Combining these three equations gives an equation for the required blast yield as a function of asteroid radius in units of megatons of TNT: Ê 2 E 2 r10 3 ˆ 3 / 2 ˜ W =Á . Fragmentation takes place when E=Efracture = ª10–7 erg/gm (energy density needed to break a 10-m object in two). The location for optimum fragmentation is generally considered to be the geometric center of the target object. and Alan W.g.28 These equations are based on the assumption that an explosive device is placed deep enough below the asteroid’s surface to produce near-optimum fragmentation. For soft terrestrial rock. Harris.5v 2 .5. An expression for shock wave internal energy per unit mass can also be written: E = 0.. THREAT MITIGATION 5. 920.72¥1010 s–1. Ahrens. p.1 Nuclear Fragmentation The equations used to model the catastrophic fragmentation of a near-Earth solid body asteroid are based on the work of Thomas J. H=2. In reference 28. e. (38) where r¢ is the idealized radius of the spherical target body and H is a parameter which characterizes the hardness of the asteroidal material. one can write v(cm / s) = Hr ¢ . California Institute of Technology. starting with equation (65). 62 . H=5. r¢ = r W 1/3 . For hard igneous terrestrial rocks.90¥1010 s–1.

Although open to debate.000 Figure 40. would be less damaging from a global point of view than from a single massive asteroid strike. However idealized these assumptions may be.29 As shown in the tabulated data. Even if any fragments did reach the ground.000 10.001 0. (2) The explosive charge is placed at the exact geometric center. it is generally assumed that fragments of this size would be much less likely than the original body to survive entry through Earth’s atmosphere. (3) The explosion fractures the target body into pieces no larger than 10 m in diameter. Blast yield—explosive placed at center of body—required for fragmentation as a function of asteroid radius.01 0.1 0. To gain a better understanding of the mass of the explosive payload to be delivered to the target body. data on existing nuclear warheads were tabulated using open literature sources in table 14 and graphed in figure 41. there are no existing devices that could catastrophically fragment an asteroid >2 km in diameter. they do permit one to estimate the explosive power required for fragmentation to within an order of magnitude. the impact of these relatively small objects. These equations are admittedly somewhat ideal because they assume that: (1) The asteroid is a perfectly spherical homogeneous structure.000 1.0001 10 100 Asteroid Radius (m) 1. this assumes that all of the ideal conditions listed above are satisfied.000 100 Blast Yield (Mton) Hard Igneous Rock Soft Rock 10 1 0.Figure 40 shows the blast yield required to fracture an asteroid into fragments smaller than 10 m in size. 63 . spread over a large area. 10.

000 Device Mass (Ibm) 10. Device mass versus explosive yield. military employs against hardened underground facilities.500 1.000 9.6 ln(x) – 8.600 17.700 42.750 10. One idea considered is to utilize the same technology that is found in the “long-rod bunker buster” types of ordinance that the U.000 1.000 4. This idea has the advantage of not requiring a DV breaking maneuver to rendezvous with and soft land on the target.500 1.400 Designation Mk–1 Mk–3 Mk–4 Mk–5 Mk–6 Mk–7 Mk–8 Mk–11 Mk–12 Mk–14 Mk–15 Mk–16 Mk–17 Designation Mk–18 Mk–21 Mk–24 Mk–28 Mk–36 Mk–39 Mk–41 Mk–43 Mk–53 Mk–57 Mk–61 Mk–83 100.000 Figure 41.900 510 716 2.30 64 .000 42. The problem of exactly how the explosive device would be placed at the geometric center of the target body has not been addressed in this study.000 15.125 8. using open literature sources.320 17.280 3.000 3.000 100 10 100 1.900 3.700 3.200 Mass (lbm) 8.161.300 10.670 2.000 1.000 y = 3.000 100.000 Mass (lbm) 8.000 1.000 25.700 6.000 20 340 1.900 8.3 Explosive Yield (kton) 10.125 8.000 Yield (kton) 500 5. Again.000 15.000 7.200 31.481. Yield (kton) 16 49 32 120 160 61 30 30 14 7.600 42. the physical characteristics of the BLU–113 penetrator bunker buster are listed in table 15. the outbound kinetic energy is utilized to bury the device to the optimum depth.900 10. Instead.Table 14.100 10.S. Nuclear device masses.000 2.

Structure Length Diameter Explosive Overall Mass Penetration of concrete Fuse Weapon system Thick high-grade steel 153 in 14.Table 15.”31 It is assumed that the explosive device can be successfully delivered kinetically to the center of a 200-m-diameter asteroid.400 lbm (2. Qualitative considerations for threat mitigation using nuclear fragmentation. especially the explosive mass-to-overall mass ratio. Table 16. were used to estimate a representative total mass for the explosive payload that must be delivered to the target body (table 16). the depth these first-generation penetration weapons can achieve has been explained by a simple rule of thumb. In actual practice the impact velocity and penetration depth must be well below this to ensure the contents are not severely damaged.000 kg) 20 ft (≈6 m) FMU–143 series GBU–28 These characteristics. anything larger may require the use of some sort of drilling or auger device. BLU–113 penetrator characteristics. or about 100 ft (30 m) for a 10 ft (3 m) missile. “For typical values for steel and concrete. Now. we expect an upper bound to the penetration depth to be roughly 10 times the missile length. First-Order Qualitative Considerations Susceptibility to dust cloud Ability to handle target rotation Requires landing on target Usefulness on fragmented body Swarm option Second-Order Qualitative Considerations Usefulness as weapon Perceived safety Synergy with other NASA missions Manned missions Robotic missions Resource utilization missions Costs Development Deployment High Low Low Low Low High Medium Low High Maybe Low Medium 65 .5 in 630 lbm (285 kg) 4.

Although possibly effective if used on a solid asteroidal body. this technique would be less effective on a comet or dusty and/or soft asteroid. the energy deposition. Precise targeting or landing on the body is not needed and simplifies the mission requirements considerably. the entire thrusting event is over in milliseconds and the momentum imparted to the object is in one direction. This material would then absorb the radiation and blast energy from the device. The asteroid deflection process for nuclear pulse/EPPP is accomplished with a more conventional nuclear device design and nominally requires two to ten separate devices to be successful. Another option would be to place a quantity of some low atomic number material between the nuclear device and the asteroid. will generate large pressures on one side of the body and so assist in the desired deflection. In fact. and the resulting debris would impact the asteroid. Even the dust cloud itself. and subsequent plasma expulsion are based on the same physical principles and governed by the same mathematical equations as for laser ablation. causing as much or even more impact damage on the Earth’s surface than one large piece. A deep subsurface blast could also be used. One could envision a situation where only a few large fragments are produced or where the pieces have very little separation velocity and would all still strike the Earth. but much of the energy will still be deposited on the surface and will produce thrust. The first process described above—and used in the following analysis—is similar to a laser ablation process. The primary reason for this is that nuclear energy is efficiently liberated during very high yield detonations. The energy deposited is then transferred to the in situ propellant—surface material from the target body itself—to produce a very efficient propulsive technique. should it absorb a great deal of the energy. Unfortunately. A tremendous advantage to such an intense flash of radiation is that target rotation is essentially inconsequential. depending on the particular conditions. Table 17 lists the rankings as all green. but the nuclear device must somehow be buried or be designed to survive the high gravity loads of a deep-penetrating projectile. The intense heating blows the surface layer off and the ejected mass imparts a reactive impulse to the body. a reaction would be produced and act to deflect the body. A dusty target body surface or dust clouds do not affect the radiation energy from the blast. A relatively thin layer of the body’s surface absorbs the intense. This is done while the device is still a significant distance from the target object.5. Even worse. since each pulse would affect the trajectories of 66 . Near-surface blasts will break up the surface layer and blow large pieces of physical debris into space.2 Nuclear Deflection Nuclear deflection of an asteroid or comet would probably be accomplished by using a nuclear device to produce a highly intense radiation burst. absorption. Once again. First-order qualitative considerations all received the highest score for this technology. such a technique would still dissipate a majority of the available energy in the fracture process. some of the pieces could draft behind others during atmospheric entry. The impulse may be slightly reduced. Like a strobe light that seems to stop a moving object for an instant. would vaporize. Fragmented asteroids or comets could be dealt with very well. energy losses in the fracture process and the absence of a preferred direction of motion for the detached pieces both serve to limit performance. high-energy electromagnetic radiation—mainly hard x rays or gamma rays—and is vaporized. Unfortunately. There are other ways in which a nuclear blast might be used to deflect an asteroid.

an assessment can be made as to the effectiveness and implications of the event before the next one is sent. Qualitative considerations for threat mitigation using nuclear deflection. The technique’s usefulness as a weapon is obviously high. but this would not be a critical factor to mission success or cost. as the physics and many of the specific technology challenges would be applicable to an EPPP. must surely produce a net benefit. First-Order Qualitative Considerations Susceptibility to dust cloud Ability to handle target rotation Requires landing on target Usefulness on fragmented body Swarm option Second-Order Qualitative Considerations Usefulness as weapon Perceived safety Synergy with other NASA missions Manned missions Robotic missions Resource utilization missions Costs Development Deployment High Low High Medium High Medium Medium Low High No High High all objects simultaneously. e.Table 17. The very fissile material that would be used in an asteroid deflection system currently resides in nuclear weapons around the world. and be more flexible in target engagements. In fact. Removing this material from these weapons.. and a ground-based system would be no more difficult or expensive to maintain than the present silo-based ICBM systems already maintained by several countries. A space-based planetary defense system would be far more secure. or for deep-space robotic missions. Depending on the circumstances. Many plasma pulses will allow precise control. The second-order qualitative considerations have a mixed assessment ranking. Perceived safety is a clear problem. the swam option is also every effective with this technology as. Of course. using it for planetary defense. Synergy with other NASA missions is high. besides the expense for nuclear safeguards. and instead. Finally. there are several advantages that could be capitalized upon so as to maximize the effectiveness of the nuclear pulses. which might be accounted as an expense already being carried by other government agencies. but this point is a little misleading. after each pulse. reduce risk. this should be a low-cost option. 67 .g. in reality. One large device might be used to obtain the maximum efficiency of the fission fuel. nuclear pulsed propulsion. extraterrestrial resource utilization would be greatly facilitated if this technique were available to deliver asteroids or comets to strategic locations in the solar system. Preexistence of the fissile material is also one of the reasons why the cost is not high for this option. nuclear weapons of this caliber have been safely maintained for over 50 yr without incident. vehicles capable of interplanetary travel with acceptable trip times for human missions. However. particularly if excess weapon stockpiles are used.

The average energy deposited on the surface is based on the volumetric ratio of the radiation cone and the total spherical volume surrounding the detonation (eq.e.( F inert ) 100 . even if the design used a successful early-generation uranium mechanism. one-eighth of the total spherical asteroid surface.100 kg/m 3 ). The analysis neglects the neutron energy and prompt gamma ray produced. The ideal potential energy in the fuel (J/kg) is found from the energy release in megaelectron volts multiplied by Avogadro’s number and then divided by the molecular weight—provided the proper unit conversions are also used.. The geometric relationships are illustrated in figure 42 and the standoff distance is calculated using equation (43). with a conversion factor of 4. The performance for this application is excellent and little is gained by going to hydrogen bomb performance. Tritium cores for fusion enhancement and plutonium fuel were not considered but would be significant performance enhancers. the desired yield. along with the standard properties of U235 (185-MeV fission fragment energy and uncompressed density of 19. i. Some nuclear device characteristics are defined: nominal yield.186¥1012 to convert energy from kilotons to Joules. collimation factor. surface reflection. reradiation losses to deep space.There are two primary hardware components to the nuclear deflection option: the nuclear device and a small rocket delivery system.e. burst half-angle. with an additional 10-percent fluff factor on the entire device. a simple pulse unit volume is determined by assuming that the inert material (average) density is simply one-fourth of the density of the fuel. and divided by the surface area (pIR2) on which the energy is deposited. illuminated at the optimum pulse unit distance. Although. and inert material fraction. (44)). dust particle interference. The total device mass is then the sum of the fuel and inert mass. The asteroid is considered to be a hemisphere. it is straightforward to determine the required mass of U235. where the inert mass is based on the estimated inert mass fraction (Finert) given in equation (42): Minert = M U235 ¥ ( F inert ) 100 1 . The standoff distance—perpendicular distance from detonation point to body’s surface—is selected so that.. one-half of the exposed surface is irradiated. To account for the losses that might be involved. From this value. The planetary body size is determined by assuming a roughly spherical shape and using the mass and density to estimate a radius. (42) It should be noted that this analysis only assumes the kinetic energy from a pure fission reaction of U235. burnup fraction. The calculational process begins using inputs for the asteroid mass and its average density. it may be easier to utilize existing and well-proven devices rather than build new ones. The use of only uranium fuel is helpful because it is essentially nonradioactive and can be easily handled and stored for long periods of time. etc. Finally. which would improve performance. only 25 percent of the energy just determined was regarded as available to provide thrust: 68 . at the device’s cone half-angle. the estimate for the nuclear devices is only based upon the fission of U235. which is converted to intense thermal electromagnetic energy in the hard x-ray or gamma-ray wavelengths. For this assessment. and the burnup fraction.. scattering. i.

and the total number of pulses required to accomplish the deflection is found by dividing this result by the propellant mass previously calculated for a single nuclear device.IR r Cone Half Angle Figure 42. (45)). The average velocity of the propellant leaving the body after one nuclear radiation pulse is calculated from the kinetic energy equation (E = 1/2 mV2) using the mass of the propellant (derived from the disk volume—for the 20-cm-deep region—multiplied by the average density assumed for the planetoid) and the energy as determined above. Dstandoff = 2 IR . The calculation continues with the deflection DV for the body being used to calculate the total propellant needed at the Isp predicted for the pulse unit interaction. a further efficiency factor of 20 percent was applied before determining the effective Isp of the technique. The standard exponential rocket equation is used (eq. (45) 69 . and this volume of surface material was used to estimate the in situ propellant available. Not all of this material will be ejected in such a way as to contribute momentum to the planetoid in the desired direction. where IR = r 2 tan(a ) (43) E avg = Fcollimation E yield 1 3 pDstandoff IR 3 ( 4 3 pD3standoff ( )(pIR ) 2 ). (44) The depth of energy penetration was estimated at 20 cm. Therefore. Note the value is rounded up to the next whole number and no contingency nuclear pulse units were assumed to be carried in this initial analysis: DV gI sp M f = Mi e . Geometric position of the pulse unit to the planetary body and cone half-angle definition.

A rocket delivery system was assumed to provide standoff distance from the spacecraft (assumed to be an ORION or Gabriel type derivative as shown in fig. Plasma Dust From Asteroid Figure 43. A 1-km/s DV was assumed and the rocket equation used to estimate the rocket propellant required to carry the nuclear pulse unit to the asteroid. The chemical propulsion system mass includes the tank and other subsystems that scale with propellant load.25 m ≈4 m Kickout Motion (Drawing not to scale) Rocket Fires and Plume Misses EPPP Spacecraft Motion Through Secondary “Strut” Shock Sleeve This End Faces Asteroid When Detonated Primary “Carbon” Shock Absorber (Three Tier) Separation Distance Would be Several km 1 2 3 1. it is an acceptable expedient. 34) and to provide some placement control of the fission device. Plasma Shock Cone Directed Toward Asteroid 3. ≈1. (It is recognized that this is not the most accurate methodology. The calculations for this component of the system were nonrigorous and merely serve to estimate the mass and volume that would be needed. Figure 43 illustrates the operation of the EPPP spacecraft and rocket delivery system. Table 18 shows the basic assumptions made to estimate the rocket size.) The tank volume was estimated by assuming all the propellant was hydrogen with a density of 71 kg/m3 and by specifying a 2-m-diameter cylinder tank. Nuclear pulse rocket delivery system sketch. Initial Intense Nuclear Blast 2. but within the framework of this study. 70 .

but expected propulsive results were established with significant efficiency degradation factors taken into account. all four size sails are able to deflect a 10-m object.000. 105. The lines indicate sail areas of 104. Characteristic Specific impulse Chemical propulsion system mass Nuclear device mass structure contingency Integration structure and contingency factor Pulse unit volume contingency factor Nuclear device rocket volume contingency Volume integration structure and contingency factor Value 450 500 15 30 30 15 20 Units s kg Percent Percent Percent Percent Percent The final part of the analysis confirmed that the fission pulsed detonation methodology is very insensitive to many of the input parameters and can easily move most threatening bodies safely away from the Earth. The plasma interaction with the planetoid body surface was not modeled. respectively. and 3. Using calculations shown in section 4. Therefore. defined in the outbound solar sail description. although this trade space was not explored in detail. It is expected that the deflection requirements for an incoming asteroid with <1 yr to deflect will be around 1–10 m/s. 316. Using an all-chemical system to deliver the nuclear devices from the Earth seems to be impracticable because the intercept time would be exceedingly long. 106.162 m. respectively. 1. An optimistic calculation of the overall DV imparted to the vehicle plus object is determined by using DV = a0 ¥ t . the overall performance of a solar sail can be determined where an asteroid or comet has somehow been connected to it and thus constitutes part of the payload. and t is the time during which the sail is in operation. The system’s qualitative parameters ranked very high. (46) where a0 is the characteristic acceleration. It is obvious from the figure that the sail areal density is insignificant for this application. and 107 m2. The payload mass of the asteroid dominates the sail mass. DV versus time can then be plotted for several sail areal densities and sizes.3 Solar Sails Solar sails are found in the literature as nonnuclear options for deflection. Sails were initially considered as possible deflection mechanisms. perhaps even to the point of being outdated. 5. These areas translate to side lengths for a square sail of 100. and the general estimates prepared were all conservative in nature. 71 .Table 18. Chemical rocket assumptions. and even the secondary considerations that were unfavorable appear to have some mitigating arguments. The uranium fission technology assumed is well established. The need for a second rocket delivery system with an arbitrary 1-km/s DV seems prudent for protection of the carrier spacecraft. Figure 44 shows the plot of an object that is assumed to be a 10-m-diameter carbonaceous chondrite asteroid.

162 m/6.3 gm/m2 10 316 m/3.6 gm/m2 ∆V (m/s) Figure 44.000 m/0. all three larger sails are able to handle deflection requirements for a 100-m object without an undue amount of time (1 yr).0 gm/m2 316 m/6. Finally.1 10 100 Time (days) 1.0 gm/m2 3.0 gm/m2 100 m/3.000 m/8.000 100 m/8. 100. From the above. The largest sail might be able to deflect a 1-km object but would require 5 yr or more.100. DV versus time can be plotted for asteroids of four different diameters: 10. From these results. 72 .000.6 gm/m2 0.162 m/0.000 316 m/8.6 gm/m2 3. deflection of a 10-km object seems outside the capabilities for any reasonably sized sail. 1. Of course.000 m/3. From here.000 100 m/6.7 gm/m2 1.000 10.000 m/6.000 m (fig. and 10.162 m/8.7 gm/m2 3. Deflection DV imposed on 10-m-diameter asteroid. using this option will necessitate the use of very large sails and will still not provide the capability of deflecting the largest asteroids and comets.0 gm/m2 100 1.162 m/3.7 gm/m2 1.6 gm/m2 316 m/0.000 1. a 10-m object is of little concern.3 gm/m2 1.3 gm/m2 100 m/0.3 gm/m2 1 3. 45).7 gm/m2 10.

Finally. Most concepts envision the sail somehow encircling the object with a line or net. Obviously. For similar reasons.100-m-Diameter Asteroid 1. this idea would have great difficulty if the object were rotating.162-m-Side Square Sail 100 10 1 1×10–1 1×10–2 ∆V (m/s) 1×10–3 1×10–4 1×10–5 1×10–6 1×10–7 1×10–8 1×10–9 1×10–10 10 100 Time (days) 1. any clouds of dust around the object would degrade the sail material.000-m-Diameter Asteroid 10. Qualitative considerations for threat mitigation using a solar sail.000-m-Side Square Sail Dash-Dot – 3. there are several complications that further exacerbate the sail’s inability to deflect the most threatening objects. Table 19. The sail could be “rolled up” onto the asteroid or comet. the sizes indicated for these objects are such that it is difficult to see how more than one sail could operate at a time. Deflection DV imposed by 6 gm/m2 solar sail with varying areas on varying diameter asteroids.000 Figure 45. First-Order Qualitative Considerations Susceptibility to dust cloud Ability to handle target rotation Requires landing on target Usefulness on fragmented body Swarm option Second-Order Qualitative Considerations Usefulness as weapon Perceived safety Synergy with other NASA missions Manned missions Robotic missions Resource utilization missions Costs Development Deployment Low High Low High High Medium Low Medium Low Yes Low Low 73 . As indicated in table 19.000 10.000-m-Diameter Asteroid Solid – 100-m-Side Square Sail Dash – 316-m-Side Square Sail Dot – 1. the sail would not be able to affect more than one fragment of a comet at a time. Unfortunately. The force from the sail would then be transmitted through the line to the object.

1. Schematic of solar collector. the collected light from the curved sail is directed onto the asteroid by the secondary mirror. It is noted that this concept excludes the propulsion system-needed stationkeeping. no losses are included for absorption or the effect of a noncolumninated beam.3. This is a little optimistic. The beam intensity incident on the object is then given by I ( r) = P ( r) ¥ c . Station-Keeping Propulsion System Sunlight Secondary Mirror Vaporized Rock Jet Heated Spot. the analysis for the collector is the same as in section 5.32 The amount of mass ejected from the object is calculated as I ( r) Asail H vap + ˙ me = 2 ve . It also offers synergy by using the same technique for both the outbound and inbound legs of a rendezvousdeflection mission. All that is required is to determine the force imposed on the asteroid or comet by the incident beam. The added weight of such a system would substantially degrade the performance of the solar collector on the outbound leg and is most probably superfluous in view of the forces experienced on the inbound leg. but losses must be strongly avoided or the design temperatures required for the collector and mirror will become untenable. 10–100 m Primary Mirror. For deflection. shown schematically in figure 46.to 10-km Diameter Asteroid Deflection Velocity Figure 46.5. (47) The pressure as a function of radius was defined in section 5. Therefore.3 and c is simply the speed of light in a vacuum. The beam intensity incident on the object is assumed to be equal to that incident on the collector. In all respects.4 Solar Collector The solar collector is another nonnuclear option considered for use as a threat mitigation technique. (48) 2 74 . The following discussion deals with this.

000 m 10 1 0.000 100 ∆V (m/s) 10 m 100 m 1. Here. respectively. 1. a secondary mirror reflecting this much power will be a major design challenge.00001 0.5—which represents a hemispherical expansion. DV imposed on varying asteroid sizes by 100-m-diameter solar collector.1 0. For the purpose of this study. the force exerted on the object by the ejecta can be determined: ˙ F = b ¢meve .000 10. The collector can obviously handle asteroids of up to 100 m without a problem.000. As a first approximation.01 0.000 1.5 au.000001 10 100 1. The intensity of the beam focused on the asteroid will scale with the square of the diameter of the solar sail. 47). there is the expectation that a 1-km solar collector may be able to deflect asteroids of 1-km diameter. 75 . From this. 1. This is an assumption based on the average distance during the time period when the collector is acting on the object.5 au is nearly 5 GW for a 100-m-diameter collector.001 0. Thus. the ejecta is assumed to come out at its sonic velocity (ª1 km/s). and 10. 100.000 Time (days) Figure 47. 500 kg of payload was assumed to account for avionics and other systems. (49) where b ¢ is a gas expansion factor—assumed to be 0. the orbital radius of the vehicle from the Sun was assumed to be 1. A parametric of DV imposed on an asteroid or comet similar to that for the solar sail above can be created (fig. Additionally.000.0001 0. a solar collector with a diameter of 100 m is acting on asteroids with diameters of 10. The power of the incident beam at 1.000 m.000 100. There is no obvious path for solar collectors to deflect asteroids of larger diameter.where Hvap is the heat of vaporization (3 MJ/kg for water ice) and ve is the velocity of the ejecta. Obviously.000 m 10.000 10.

Any dust or ice clouds surrounding the object will degrade the beam from the solar collector. it would be difficult for more than one to operate on an incoming object at a time. The need to constantly heat new portions of the object— or at least portions not heated since the previous revolution—will most likely result in a lower temperature and therefore slower ejecta. but instead. Any conductor in the presence of a changing magnetic field will produce an electric current proportional to the magnitude of the magnetic field and its rate of change. See table 20 for solar collector qualitative considerations. This will result in less thrust than could be achieved on a nonrotation body. For a fragmented body. although the beam can be directed on different components. Qualitative considerations for threat mitigation using a solar collector. With explosives and kinetic impacts. A rapidly changing magnetic field could be used to generate large countercurrents on the surface of the asteroid. none used electromagnetic forces. Table 20. Many researchers agree that breaking the object into pieces could make the problem more difficult and would be an undesirable effect of any deflection technique. due to the size of the solar collector. Finally. and is defined by Lenz’s Law: 76 . In this case. This current is due to the induced electric field produced in the space that the conductor occupies.5 Magnetic Flux Compression Of the many asteroid deflection approaches researched for this TP. First-Order Qualitative Considerations Mitigation by dust cloud Ability to handle target rotation Requires landing on target Usefulness on fragmented body Swarm option Second-Order Qualitative Considerations Usefulness as weapon Perceived safety Synergy with other NASA missions Manned missions Robotic missions Resource utilization missions Costs Development Deployment Low High Low High High Medium Low Medium Medium No Medium Low 5. By their very nature. a band around the rotating object. the momentum is applied through either a pressure force from the highly random kinetic thermal energy in a gas or at a localized impact area. With electromagnetic forces. except in the case of complete pulverization. only one can be pushed at a time. electromagnetic forces could offer an advantage over more conventional explosive or kinetic impact forces. It will be difficult to match rotation with a rotating object. a large amount of energy can be converted into a directed Lorentz force without producing a lot of heat or fracturing the asteroid into pieces. These surface countercurrents would in turn produce a secondary magnetic field that would repel away from the original magnetic field and exert a repulsive force on the asteroid surface. Both of these approaches loose a significant amount of energy by increasing the temperature of the asteroid surface and possibly breaking the asteroid into pieces. the incident beam will strike not a single site.

the force is defined by the following equation: F = V ( J ¥ B) . (53) If a particle of charge (q) moves with velocity (v) in the presence of both a magnetic field (B) and an electric field (E). it may be necessary to use a target disk with a good conductivity value.e. the rate at which electrical current flows. (50) where e is the induced electromagnetic field. one can calculate with some certainty the electrical resistance. and A is the cross-sectional area of the conducting path. both the flowing electrons and fixed ions in the conductor will experience a force (F) in the same direction due to the interaction of the electric field and the magnetic field (B). the smaller the resulting force. (54) If the electric field and magnetic field are perpendicular. Knowing the magnitude of the electromagnetic field and the electrical resistance of the conductor (R). (55) For a conductor with volume (V) and a current density (J) passing through it. (52) Resistivity has units of ohms/meter. fB is the magnetic flux. The inverse of resistivity is known as conductivity (s). Since most asteroids are composed of carbonaceous chondrite materials. Resistivity has a specific value for different materials and also changes with material temperature. and t is time. the poorer the conductor. this resulting current can be calculated using Ohm’s Law: I =e R . This disk would mate up to the asteroid and serve as a pusher plate. i. Knowing the geometry and material of a certain conductor. Electrical resistance can be further defined by the resistivity parameter (r): (51) R = rL A .e = df B dt . it will experience a force (F) defined by the Lorentz equation: F = q( E + v ¥ B) . which are poor conductors. L is the length of the conductor. the force will be applied in a direction perpendicular to both fields. Unlike a pusher plate used for pulsed nuclear or conventional 77 . and the conductivity of the conductor. The magnitude of the force is dependent on the strength of the magnetic field.. For the case where charge is traveling through a conductor. Magnetic flux is simply defined as the product of the magnetic flux density (B) and the surface area (A) that the flux lines pass through: fB = B ¥ A .

this disk could be a relatively thin. The target disk need only be thick enough to absorb the magnetic flux. the thinner the foil can be. The thick. the explosives would ablate a surface layer off the asteroid and the pressure would be applied to a more localized area that may not be able to withstand the resulting stress. B B Force B B Figure 48. strong pusher plate acts as a shield similar to a bulletproof vest. 78 . The disk diameter is dependent on the distance from the magnetic field source to the target disk. The thin foil could be efficiently packaged and only deployed once the asteroid is encountered. but may require more difficult maneuvering). the force is produced by a Lorentz reaction between currents and magnetic fields. Without the pusher plate. the conductivity (s). The thickness necessary to absorb the flux is calculated using the following equation: d = t pulse ps m0 . the force is applied via the shock wave that is produced from the large amount of heat energy released from the explosives.explosives. lightweight metallic foil of sufficient area to be effective. Depending on the rate at which the flux changes. the disk could be very thin. the magnetic energy will be coupled completely into the foil. (56) Skin depth (d) is a function of the pulse time (tpulse). With the pusher plate. The faster the pulse. Since most magnetic fields are curved. since the critical disk diameter increases as the source moves farther from the asteroid (more surface area for the flux to be coupled to and easier to achieve) and decreases as the source gets closer toward the asteroid (stronger field strength. Energy is not wasted in vaporizing or demolishing the asteroid material. and the permeability of free space (m0). With the target foil. It has sufficient area to spread out the force and sufficient thickness to withstand the localized stress of the impact/explosion. As long as the foil thickness is greater than d. as shown in figure 48. Critical disk diameter. This diameter must be optimized. This would be very undesirable. The reason that this target disk can be much thinner and less massive than an explosion pusher plate is the difference between the methods by which the force is produced. This force is more evenly distributed through the entire target (like a sail) and the amount of heat flux released into the surface is greatly reduced. there will be a limit to the diameter of the target disk beyond which the field lines will reverse direction and produce forces acting in the wrong direction. This could lead to cracks or pitting.

To force 390 kJ of magnetic energy into the solenoid in 10 ms would require over 4 billion V of electricity at a power of over 39 MW. This is roughly six orders of magnitude larger than most solenoids used in common electrical components. capacitors. This magnetic field has energy proportional to the volume within the solenoid and the magnitude of the magnetic field: E =V B2 . but they would require a very large power supply. Although not impossible to produce. the current running through the coil must be increased. the field lines are compressed and their strength is increased. A solenoid produces axisymmetric. From equation (58). By reducing the volume of the solenoid. The component parallel to the target surface is responsible for producing the Lorentz force. To increase the magnetic field. a parametric FORTRAN program was written to calculate the force that would be produced by an arbitrary solenoid design. the logistics are not favorable. If a 1. nonuniform magnetic field lines with a similar field line profile to that of a bar magnet. was supplied a current of 1 MA ramped up over a period of 10 ms. this system was envisioned to use a magnetic flux compression generator. A solenoid containing a magnetic field has an energy associated with the volume of the solenoid and the magnetic field strength. Although the volume is decreased. Current technology lightweight. Since anything more than one shot on the coil would be impractical (one would need a restoring force on the coil). To accurately sum up the force produced by each varying field line. A magnetic flux compression generator is an ingenious method for converting explosive energy into electrical energy by compressing a magnetic field. 2m0 (58) Using equations (57) and (58). batteries. etc. A convenient way of decreasing the volume is by placing a conducting sleeve filled with explosives along the centerline of the solenoid as shown in figure 49. That results in a very sizeable impulse. the total field strength in a given region of space can be calculated. the Biot-Savart Law can be used: m I dl ¥ r dB = 0 .A common source of magnetic flux is a solenoid (coil) connected to a power supply. This conducting sleeve is referred to as the armature of the flux compression generator and the surrounding coil is referred to as the stator. high-power.e. This current and voltage level are not impossible to achieve. it is not an economic solution. the energy is greatly increased. 4p r 2 (57) Using Biot-Savart.000-turn solenoid with a radius and length of 1 m. located at a standoff distance of 100 m. curved. it would exert a force of over 9 million lb on a 100-m-radius target disk. By summing up the length of the conductor path. The component normal to the target is responsible for inducing a current. The curved field lines may be divided into two component vectors—one normal to the target surface and one parallel to the target surface. the magnetic field strength (flux density) (B) is calculated from knowledge of the coil path vector (l) and the distance from the coil to the target (r). The inductance value of the solenoid in question is ª3.. magnetic flux compression generators. i.95 H. since the energy is related to the square of the magnetic field strength. Another way of doing this is by compressing the field lines within the solenoid.. it can be seen that the magnetic field strength plays a major role in the amount of energy within the solenoid. are incapable of providing this performance. Originally. disposable power supplies. The stator 79 .

usually on the order of nanoseconds. usually supplied by a capacitor bank. The magnetic field and resulting current at the end of the process are extremely large. a countercurrent is induced. currents in the mega-ampere range can be produced for a very brief period of time. since its EMP is often great enough to destroy nearby electrical equipment. This countercurrent creates a magnetic field that repels the solenoid field and compresses it into the decreasing volume caused by the explosion.33 is energized with a current pulse.33 80 . Because the armature sleeve is conducting as it moves through the magnetic flux within the solenoid. expanding the conducting sleeve as it goes. the explosives within the armature are detonated from one side.Armature Cylinder Stator Winding High Explosive Figure 49. The latter configuration is sometimes referred to as an E-bomb. The flux compression generator can be coupled to a load or can be uncoupled to produce a very large electromagnetic pulse (EMP). Figure 50. This explosion propagates along the armature. Magnetic flux compression generato. Using this method. E-bomb magnetic flux compression generator. The uncoupled configuration is shown in figure 50. and a strong magnetic field is created within. An instant later.

or an existing magnetic field. When these high-energy gamma photons strike an atom.6 Mass Driver The term mass driver is used to describe a variety of electromagnetic acceleration systems. This huge EMP can be coupled into conducting materials as electrical current. These high-energy photons can be absorbed by elements with low atomic numbers. generators. The major difference between the magnetic flux compression generator and a nuclear bomb is the way in which the EMP is produced. This was demonstrated by accident when a nuclear test conducted high in the atmosphere created an EMP that completely blacked out Hawaii and other parts of the United States. converting electromagnetic radiation (gamma rays) into lower frequency electromagnetic waves (radio waves). since it would not block electromagnetic radio waves.H. The bucket is designed either without a top cover or with one that can be opened in flight. These moving electrons serve as a large electrical current and produce a very large magnetic field for a brief period of time. which would be less effective on rotating and fragmented bodies. 5. they can knock loose electrons and send them flying away at great speed. In the case of high-altitude nuclear tests. It could be constructed from existing technology rather inexpensively. all use electrical power to accelerate a bucket containing an inert mass to very high speed. The nature of this option would almost mandate that a swarm approach be adopted. A. Although a flux compression generator would never be able to produce an EMP great enough to reproduce the performance of the proposed coil. magnetic flux compression generators provide peak powers in nanoseconds instead of milliseconds and their energy outputs are inadequate for driving the coil proposed for the asteroid deflection system. When a nuclear weapon is exploded. Nuclear weapons detonated at high altitudes in the Earth’s atmosphere produce a gigantic EMP. Perhaps a device could be constructed to use the compact and abundant energy released during a nuclear explosion to create an EMP suitable to drive currents in the foil target disk described earlier. the contents leaves the bucket and continues on at high speed. Except for the nuclear configuration—not yet explored—the magnetic flux driver concept is relatively safe and simple. another device may be able to do so. however. This concept could be very effective on fragmented and rotating bodies that have a high content of ferromagnetic materials. To the author’s knowledge. disrupting and destroying electrical devices. There is. so that as it decelerates. The amount of energy is conserved—if one ignores loss mechanisms—but its form is changed. another aspect of the magnetic flux compression generator that may have been overlooked at the beginning of this study and has not subsequently been explored. Nonmagnetic bodies would require use of a target disk. this idea has not yet been explored. This phenomenon is named the Compton Effect after its discoverer. such as nitrogen and oxygen—two major components of the Earth’s atmosphere. Since an EMP is simply a large traveling magnetic flux. This EMP traveled down to the ground. Compton.Unfortunately. the atmosphere acts as a transformer. 81 . and wires. batteries. Dust should not be a problem. a large amount of radiation in the form of gamma rays and x rays are released. it can couple with conductors in the same way in which radio waves couple with antennas. See table 21 for quantitative considerations.

Qualitative considerations for threat mitigation using magnetic flux compression. so even a system with a modest exhaust velocity. it is essentially free propellant. could offer good overall performance.. A special deployment vessel would transport the entire mass driver system out to the asteroid and install it on the surface. not to determine which specific design offers the best performance. expellant. Accordingly.34 all make use of magnetic attraction and repulsion between the moving (bucket) and the stationary parts of the device.e. Although initially proposed34 for space launch applications. The system consists of the mass driver itself along with its electrical power source. mass drivers are also applicable to planetary body maneuvering. expellant speed. i. A distinct type is distinguished by the specific shape and location of the magnets. i. after the deployment team has departed.e. First-Order Qualitative Considerations Sensitivity to dust cloud Ability to handle target rotation Requires landing on target Usefulness on fragmented body Swarm option Second-Order Qualitative Considerations Usefulness as weapon Perceived safety Synergy with other NASA missions Manned missions Robotic missions Resource utilization missions Costs Development Deployment Low High Low Low High Medium Low Low Low Impact Medium High Although there are several different types of mass drivers. The basic rationale behind the use of a mass driver in this application is that it would make use of the asteroid surface material as reaction mass. although this might be a suitable subject for a follow-on study. As this material is already present.. (2) An automated mining system could be deployed on the asteroid to operate continuously. together with the mass driver.Table 21. Option (2) is clearly more complex and raises additional questions over reliability and maintenance. a relatively straightforward design was chosen to 82 . The intent of this study was to establish the general utility of mass driver systems for planetary maneuvering. either a nuclear reactor or a solar array. and its thermal control system. Two options are available regarding the provision of reaction mass: (1) The deployment team could mine the entire stock of reaction mass before departure and place it into a suitable storage and dispensing system from which it could be fed to the mass driver as needed.

the bucket is surrounded by and attached to four current-carrying superconducting coils and is accelerated along the axis of the drive coils. In the case shown above.represent the entire spectrum of designs. Current flowing in each drive coil produces a magnetic field. This turned out to be the design with the greatest geometrical simplicity and symmetry—the coil gun. Direction of Motion Bucket Coil Bucket Drive Coil Figure 51. Relative current flow directions of bucket and immediately adjacent drive coils. The relative current flow directions of an arbitrary bucket coil and the nearest pair of drive coils is shown in figure 52. 83 . a component that exerts an axially directed motive force on the bucket coils. which is that modeled. Coil gun conceptual design. Drive Coil n Drive Coil Bucket n+1 Coil x Direction of Motion S Figure 52. The principal concern was to select a design that was relatively straightforward to model. The coil gun uses a series of equally spaced stationary solenoidal electromagnets (drive coils) to accelerate a bucket carrying a number of smaller but identically spaced solenoids (bucket coils). The concept is illustrated in figure 51.

Coils with currents flowing in the same sense attract.. each drive coil must reverse its current flow direction as a bucket coil passes through it.e. The drive coil current is shown as a function of the bucket coil position (S being the distance between adjacent drive coils). This current reaches a maximum when the bucket coil is at –S/2. thus changing an attractive force for an approaching coil into a repulsive force for a receding coil. This means that. halfway between the preceding drive coil and the coil under consideration. i. Note that the current flowing in each bucket coil does not change its flow direction. producing an attractive force. 2 passes through the preceding drive coil. i. repelling the receding bucket coil. the bucket coil is being repelled by the drive coil. The current directions for both drive and bucket coils at any instant is illustrated in figure 54. The design proposed here has four adjacent bucket coils separated by a distance S. This. the drive coil current begins to increase.e. 1 passes through a drive coil. 84 . although its magnitude varies—due to induction—during the transit between adjacent drive coils.e. When the bucket coil is at –S.. identical to the spacing between adjacent drive coils. bucket coil No. exactly half a cycle out of phase. It reaches zero as the bucket coil arrives and subsequently goes through the negative portion of its cycle. i. Relative current flow directions of bucket and nearest drive coils. which it has just passed (number n) and is being attracted to the next one (n+1). taken together with the drive coil current profile shown above. Clearly... and subsequently declines. Drive Coil Current S/2 –S –S/2 S x Figure 53. in the above example. dictates that successive bucket coils must have their current flowing in opposite directions. as bucket coil No. Thus.e. The current profile for one drive coil is shown in figure 53. for this concept to work. i. just passing through the center of the preceding drive coil. those flowing in opposite senses repel.

In reality. 85 . will be energized at any instant. Relative current flow directions of bucket and nearby drive coils.Drive Coil (Arrow Indicates Direction of Current Flow) Direction of Motion Bucket Bucket Coil (Arrow Indicates Direction of Current Flow) Figure 54. To minimize unnecessary energy losses. these more distant coils exert smaller forces than the nearest ones. it is anticipated that only those drive coils that immediately surround the bucket. It should be noted that although a bucket coil receives an accelerating force from the closest pair of drive coils. together with a small number ahead and behind. Relative current flow directions of bucket and surrounding drive coils. it experiences alternately retarding and attractive forces from each more distant pair. A minimum of five drive coils must be energized in order to cover the bucket itself. This can be seen clearly in figure 54. the number energized will depend upon the speed with which the bucket is moving and limitations imposed by current-switching technology. Fortunately. Drive Coils Which Accelerate Bucket Coil Direction of Motion Drive Coils Which Retard Bucket Coil Figure 55. illustrated in figure 55.

Mass driver operation. Bucket Coil Direction of Motion Bucket Braking Coil Figure 57. This should be compared with figure 54. the braking coils are energized exactly half a period out of phase with their drive equivalents. The braking coils are of the same design as the drive coils. which shows three different times during an operational cycle. the nondecelerated expellant separates and continues at the maximum speed until it exits the mass driver. Relative current flow directions of bucket and braking coils. This means that the forces are in the reverse direction to those encountered during the drive section. The only difference is that during braking. The coils are successively energized as the bucket coils pass them. The braking process is essentially the same as the acceleration process. Drive Coils Braking Coils Bucket Accelerating Bucket Acceleration Ends—Deceleration Commences Bucket Decelerates—Contents Emerges and Continues at Maximum Speed Figure 56. This is illustrated in figure 56. Figure 57 shows the disposition of currents during braking. the bucket is then decelerated by a set of braking coils. As the bucket decelerates. the equivalent diagram for the drive coil portion of the mass driver. 86 .When the bucket and its contents have been accelerated to their maximum speed.

In addition to providing a means of decelerating the bucket, the braking coils also provide a means of converting its kinetic energy back into electrical energy, thus making it available for use during the next cycle. Just as the drive coil circuits lose electrical energy as they accelerate the bucket, so the braking coil circuits gain electrical energy as they decelerate the bucket. A simple schematic of the entire mass driver system is given in figure 58.

Bucket Return Rail (Arrows Indicate Direction of Bucket Travel)

Mass Driver

Electrical Power Management System

Solar Array

Expellant Storage and Loading System

Thermal Management System Thermal Radiators

Bucket Coil Cooling System Coolant Line Power Line

Figure 58. Mass driver system schematic. The mass driver unit and return rail provide a complete closed circuit around which the bucket travels. After rapid acceleration and deceleration along the mass driver (coil gun) itself, the bucket moves at a relatively slow speed along the return rail. It halts at the expellant loading system in order to take on a fresh load of reaction mass and then proceeds on to the start of the mass driver. While the loading process is taking place, the bucket is placed in contact with a cold plate whose function is to extract the waste heat generated within the bucket coils during the previous cycle. Although superconducting coils are used throughout the design, there is still some heat generated during the acceleration and deceleration processes that must be extracted to prevent the superconductor from rising to its critical temperature. The cooling system must also remove waste heat from the superconducting stationary coils, both drive and braking, between successive operational cycles. Connections between the thermal management system and both the mass driver and the bucket coil cooling system are shown above. An array of thermal radiators disposes of the waste heat.

87

There appear to be only two viable power source options currently available: a nuclear fission reactor and a solar array. For bodies sufficiently distant from the Sun, only the nuclear option is likely to be practical; however, for the purpose of this exercise, a solar array has been selected. This choice has been made for several reasons: (1) Mass and performance modeling for a solar array are somewhat simpler than for a fission reactor, making it a more pragmatic choice regarding the analytical effort involved. (2) A solar array power system currently poses fewer concerns over reliability, an important concern given the likely need to operate the mass driver, untended, over a long period of time. (3) A solar power system is capable of operating even after sustaining damage. Parts of the array, which are nonfunctional, can be isolated and the system can continue to work, albeit at a lower power level. (4) There would likely be some safety concerns associated with the placement of a fission reactor on a body; i.e., at least initially, on a collision course with the Earth. Obviously, if the mass driver system were to perform its job correctly, there would be no problems. But in the event, say, of a malfunction in the mass driver system, it might be necessary to fragment the body. If so, the presence of a nuclear reactor on its surface could significantly complicate the operation. Against these considerations, it must be noted that the useful electrical power available per unit array surface area will be very low. Apart from being further from the Sun than is normal for solar power application, there will be additional causes of performance degradation. The array will almost certainly have to be simply laid on the surface of the asteroid and is unlikely to be normal to the incident radiation. If the asteroid is rotating, which is almost certain to be the case, additional losses will result. Finally, the process of preparing the expellant, regardless of whether it is done by continuous mining or by deployment vessel, will likely produce a large amount of dust and debris. The very low accelerations, both gravitational and of the body itself, will permit this dust and debris to remain in the vicinity of the mass driver for a significant time, allowing some of it to settle on the array, further reducing its effectiveness. Some details of the simple model used for the solar array are given in appendix B. For the purpose of this study, it is assumed that expellant is prepared by the deployment vessel before its departure. The only other option—continuous automated mining—requires a detailed study in its own right. The problems of operating mechanically complex equipment, such as that needed to extract and render asteroidal material, in an ultra-low-gravity environment, are beyond the scope of the current work. Although it is not the authors’ intent to prejudge the results of this study, the following scenario is proposed for the application of mass drivers to planetary body maneuvering. A crewed deployment vessel conducts an extended mission in the vicinity of the asteroid belt. Its targets are a small number of asteroids whose orbits pose a threat to the Earth in a time period of, say, 10 to 50 yr into the future. At each of these asteroids, a mass driver is deployed and mining equipment, carried on the deployment vessel, is used to prepare a stock of reaction mass. At each asteroid, the mass driver commences operation and the deployment vessel departs for its next target. If necessary, subsequent visits can be arranged to conduct any needed maintenance and repair. Additional reaction mass can also be produced and stored during these visits. 88

The mass driver system deployed under this scenario is shown in figures 59 and 60. Figure 59 shows the entire system, which is dominated by the large solar array and thermal radiators. Figure 60 identifies the actual mass driver (coil gun) and associated equipment.
Mass Driver System

Mass Driver System

Thermal Radiators

Thermal Radiators

Mass Driver
Mass Driver

Solar Panels

Solar Panels

Figure 59. Mass driver system view.
Major Components of the Mass Driver System Bucket Loading and Cooling Facility Expellant Storage Facility Cooling System Radiator Panel Power Management Cable From and Conditioning System Solar Panels Power Cable Coolant Line Bucket Return Rail

Buc Ma ket ss Retu Dri rn R ver ail

Bucket

Solar Panel

Figure 60. Main components of the mass driver.

89

Figures 61–66 illustrate system operations. Together, they show the system in operation through a complete cycle—the discharge of one bucket load of expellant.

Bucket Returns Slowly to Bucket Loading and Cooling Facility

Figure 61. System following discharge—bucket returns for reloading.

Bucket Enters Bucket Loading and Cooling Facility

Figure 62. Bucket about to be reloaded and cooled prior to next discharge.

90

Loaded bucket about to enter mass driver. Bucket containing expellant under acceleration within mass driver. 91 . Figure 64.Loaded Bucket Returns to Mass Driver Entrance Figure 63.

92 . Discharge in progress—bucket is decelerating while expellant mass exits mass driver at high speed. Bucket Slowly Exits Mass Driver and Attaches to Return System Figure 66.Expellant Exits Mass Driver at High Speed Figure 65. Following discharge—decelerated bucket exits mass driver and joins return system.

A simple performance and mass model for the mass driver and its principal supporting systems is presented in appendix C. Table 22. it is the above scenario that formed the basis of the analysis carried out during the present study.While there may be additional situations under which mass driver systems could be employed for planetary maneuvering. but it would clearly not be a weapon of choice in any realistic situation. It is easier to conceive application of this technique to asteroid maneuvering than to comet maneuvering. Table 22 lists the qualitative considerations that apply to the use of mass drivers for planetary body maneuvering. Use of a nuclear reactor. A comet nucleus is likely to be both fragmented and dusty. would reduce the safety rating. two mass drivers could be deployed at appropriate “balanced” locations. only a very slight orbital perturbation would be necessary to render them harmless. does not necessarily have to be supplied by an explosive event.000 mT—in many cases. The energy required to carry out such perturbations. Highspeed projectiles could cause significant damage. To improve the technique’s efficiency.7 Kinetic Deflection Although potentially threatening planetary bodies will all be very massive—even a 30-m-diameter sphere of water ice would have a mass of ª15. so a mass driver has some intrinsic weapons capability. Qualitative considerations for threat mitigation using the mass driver. whether 93 . while still large. It will only be able to operate when oriented in the proper direction. These could first be used as necessary to counteract the asteroid rotation and then used together to provide a torque-free net force. First-Order Qualitative Considerations Sensitivity to dust cloud Ability to handle target rotation Requires landing on target Usefulness on fragmented body Swarm option Second-Order Qualitative Considerations Usefulness as weapon Perceived safety Synergy with other NASA missions Manned missions Robotic missions Resource utilization missions Costs Development Deployment Medium High Low Low High High High Medium Medium Yes Low Medium The mass driver has a relatively high inherent safety rating against these potential complexities. rotation will reduce the mass driver’s efficiency. A mass driver clearly cannot move more than a single body and will probably suffer some damage if operated in a dusty environment due to secondary impacts. Even if installed on a single asteroid in a nondusty environment. 5. this is largely due to our selection of a solar array to provide primary power. although in many ways more appropriate for the outer solar system.

inert interceptor body. This technique offers the attraction that. 94 . Interception geometry. possesses a formidable amount of kinetic energy that may be adequate to produce the desired deflection. if accelerated to a sufficiently high velocity. This section contains the derivation of a simple model for use in evaluating the effectiveness of the kinetic deflection technique. issues of guidance and terminal maneuvering will not be addressed. raised to high speed. so without loss of generality. With closing speeds perhaps in excess of 30 km/s. Clearly. Although basically inelastic. once the interceptor’s onboard propulsion system has exhausted its propellant. The velocity vectors of the body and the interceptor can always define a plane. For the purpose of this discussion. Even a relatively small interceptor craft.35 and (2) the issue of planetary body fragmentation. The model is comprised of two distinct parts: (1) A largely analytical impact and momentum exchange model. There are clearly major issues relating to guidance of the interceptor. Planetary Body x Planetary Body Velocity Vector θ y Interceptor Figure 67. the remaining dry mass still fulfills an important function by virtue of the kinetic energy it carries. as a deflection tool. maneuvering capabilities will be limited and reaction times critical. this subject is not appropriate for an open report. allowance is made for the ejection of debris from the collision site. some of the technologies being developed for terrestrial ballistic missile defense are likely to be applicable. This raises the possibility of using a relatively small.conventional or nuclear. the analysis can be presented in two dimensions. Clearly. particularly during its terminal mission phase.36 Figure 67 shows a general situation in which an interceptor is about to impact a planetary body. A modified inelastic model is given below.

(60) 95 . ( ) (59) where MNEO u Mi vi q Mej Dux Pej = mass of the NEO before impact of interceptor = initial speed of NEO = mass of interceptor = initial speed of interceptor = angle between NEO and interceptor velocity vectors (see fig. the vertical axis. Mej .Mej ( u .Du x ) + Pej cosq . it all emerges at the same angle (j) to. in other words. M NEO u .An x-y coordinate system is established as shown and then a momentum conservation equation is established as follows. Impact and ejection geometry.Mi vi cosq = M NEO + Mi . 67) = mass of material ejected due to impact = change in x component of NEO velocity due to impact = net momentum of ejected material. For motion along the x axis. Ejecta Cone ϕ Figure 68. Note that the impact and ejection processes are assumed to follow the simple pattern shown in figure 68. The ejecta material is assumed to emerge from the impact site in the form of a uniform cone. and evenly distributed around. Assuming that M NEO >> Mi .

If one denotes the cumulative mass of ejecta traveling faster than v by M(>v). The various quantities are defined as follows: Rc r v Y = crater radius = NEO material mass density = ejecta speed = NEO material strength. M NEO Ë Mi vi ¯ Du x = For motion along the y axis. Note that if one neglects the ejecta momentum terms. (61) Mi vi sin q = M NEO + Mi . M NEO Ë Mi vi ¯ Du y = (63) Equations (61) and (63) constitute the momentum conservation part of this model.05 Rc r Y 2 rv min . Pej ˆ M i vi Ê Á1 + ˜ sin q . Defining vmin as the minimum speed of ejected material. (64) This equation is based on data obtained from experiments that measured the ejecta produced by impact into materials of differing hardness. then35 3 M (> v ) = 0.05 Rc r Y rv 2 .Pej sin q . the total mass of ejecta can clearly be written as 3 Mej = 0. (65) This expression can be used to eliminate Y and Rc from equation (64) to give 96 .one can rewrite equation (59) to give Pej ˆ M i vi Ê Á1 + ˜ cosq . the model becomes completely inelastic. ( ) (62) and hence.Mej Du y .

4 Rc r Rc . (69) This enables one to derive the vertical component of momentum for the entire ejecta mass (Pej) as follows: • vmin 2 2 vmin M ej cos j Pej = Ú v2 dv . an equation that relates the mass of ejecta to the radius of the crater is used. Ë v + dv ¯ (67) which can be expanded.Á min ˜ Mej . 3 2 (72) which can be rewritten as Mej = 0. Ë v ¯ (66) Now. 68).M (> v + dv ) = Á min ˜ Mej Ë v ¯ Êv ˆ2 . Pej = 2v min Mej cosj . This mass is given by Êv ˆ2 M (> v ) .8378r R3 . consider the small element of ejecta mass which emerges with speed in the range v Æ v + dv. retaining only first-order terms in dv.35 It is derived on the assumption—based on empirical evidence—that the crater is one half of an oblate spheroid with a depth of 0. (71) Next. the following expression can be written for the vertical component of momentum (dp) of this small portion of ejecta: 2 2v min Mejdv dp = v 3 v cosj . (68) Noting our assumption that the ejecta is assumed to all emerge at the same angle (see fig.4 Rc: Mej = 4p 2 0. to give d M (> v ) = 2 2 v min v3 Mejdv .Êv ˆ2 M (> v ) = Á min ˜ Mej . c (73) 97 . (70) and hence.

Note that the relative speed of collision is given by V = vi + u cosq . For this method of deflection.209 r V Ê rV 2 ˆ ˜ Á = 0.458 Mi Á ri Ë Y ˜ ¯ 0.4886Á ˜ Ë r¯ 0. Mej can be eliminated from equation (76) and rewritten as 0. although some material will be ejected from the impact site.2443Á ˜ Ë r¯ 0.2238 cosj . ri vi Á Y ˜ M i vi ¯ Ë Pej (78) This is the final quantity required to solve equations (61) and (63).1937 Mej r . (74) Mej can be eliminated from this expression using equation (65) to give ÊYˆ vmin = 0. For a relatively low-energy collision. (75) and hence equation (71) can be rewritten as ÊY ˆ Pej = 0. which can then be used to determine the deflection that results from interceptor impact. (77) thus.35 This was originally derived using data from a large number of laboratory cratering experiments and links the total mass of ejecta to the relative speed of collision (V) and interceptor mass as follows: r Ê rV 2 ˆ Mej = 0. (79) Equation (78) can be used to eliminate (Pej /Mi¥vi) from equations (61) and (63). the planetary body will remain intact. the question of fragmentation also needs to be considered.709 . At larger collisional energies. however.from which 3 Rc = 1. there is the possibility that the body may 98 .5 . (76) One final equation is also available.5 M ej cos j .

it is understood that the initial impact produces a “hydrodynamically induced” crater. the final outcome will range from velocity deflection to complete and catastrophic fragmentation.9 0. EP /MT (J/gm) 0.07 0. on a timescale of between microseconds and milliseconds. which is unfortunate. Note that ML denotes the mass of the largest fragment. Some empirical data36 are available that relate the collisional energy to the mass of the largest fragment produced by the collision.1 0. Understanding is less clear for impact speeds in the 8 to 15 km/s range.9 ML /MT 0. Table 23. where EP is the collisional energy and MT is the target mass.36 i. An incoming body that is unexpectedly fragmented by a kinetic deflection interceptor may subsequently prove very difficult to deal with if some of the fragments are very massive and still pose a threat. the target is considered to be catastrophically fragmented. in other words. The critical parameter in determining whether the target will fragment is the collisional energy per unit mass of the target body. Studies have been conducted to determine the effect of kinetic energy on various NEO analogue materials. In general terms.e.5 J/gm. This is written as EP /MT. Depending upon the impact energy and the constitution of the target body. The impact produces.01 0. where the target is rendered into a large number of fragments. there is very poor understanding.5 J/gm. There is clearly very limited direct experimental data to draw on.6 Target Material Rock (basalt) Ice ML /MT 0. each much smaller than the original body: • For EP /MT < 0. internal shock waves that propagate through the target on a timescale of between tens and thousands of milliseconds..05 EP /MT (J/gm) 10 0. At speeds >15 km/s. showing the relative size of the largest fragment (ML/MT) at various values of (EP /MT).split into a number of fragments as a result of the impact. given that this is the likely regime in which a kinetic deflection interceptor would operate. as it may determine the usefulness of kinetic deflection as a threat mitigation technique. The data are reproduced in table 23 for two different target materials. the target is considered to be cratered but intact. It is important to be able to understand and model the fragmentation process.01 99 .01 EP /MT (J/gm) 3 0.36 The mechanics of collision and fragmentation are reasonably well understood for impact speeds of up to ª8 km/s. Relative size of largest fragment at various collisional energies. • For EP /MT > 0. the cratering process takes place while the solid material of the body has been rendered fluid-like by the impact energy. The following approximate criteria has been established for catastrophic fragmentation.1 ML /MT 0.

For facilitating computations.2697 for ice. taking place in the fluid phase after impact. Ë MT ¯ Ë MT ¯ where A = –0. 1 Basalt Ice Linear (Basalt) Linear (Ice) Largest Fragment Mass/Planet Body Mass (ln) 0 –1 –2 –3 –4 y=–1.8299x – 2. The curve fit results can be summarized as follows: ÊM ˆ ÊE ˆ ln Á L ˜ = A ln Á P ˜ + B .0843 for ice. In conclusion. Chemical reactions. are probably responsible for this discrepancy.1324 –5 –5 –4 –3 –2 –1 0 1 2 3 Impact Kinetic Energy/Planetary Body Mass (ln) Figure 69. (80) 100 .8299 for basalt. –5. if possible. actual impact tests on a variety of planetary bodies. including. This area requires more work. some indications36 are noted that current analytical and semiempirical models may underestimate the momentum change resulting from a high-velocity impact. the above data and model should be viewed as very preliminary. Hence.0843x – 5. B = –2. Fragmentation data and curve fit results.2697 y=–0. –1. these data can be roughly curve fitted as shown in figure 69.1324 for basalt.

a rotating target is in some respects more desirable than a nonrotating target as it will be easier to identify the location of its center of mass.Table 24 lists the qualitative considerations that affect the use of kinetic deflection. In the present application. Technology developed for ballistic missile interception will probably be directly applicable here. Note that the closing speed for this application may be an order of magnitude higher than for missile interception. However. destroying it before it impacts on the target. larger pieces of debris could pose a serious problem. In fact. kinetic deflection works well with a rotating body. 101 . The later interceptors in a stream will have been launched long before the initial interceptors impact. a kinetic deflection system would be more likely to be employed in a defensive capacity than in an offensive one. Although potentially usable as a weapon itself. so they must be able to maneuver in order to respond to changes in the body’s trajectory. but rather a stream of them. First-Order Qualitative Considerations Sensitivity to dust cloud Ability to handle target rotation Requires landing on target Usefulness on fragmented body Swarm option Second-Order Qualitative Considerations Usefulness as weapon Perceived safety Synergy with other NASA missions Manned missions Robotic missions Resource utilization missions Costs Development Deployment High High Low Low High Low Low Low High No Medium High Perceived safety is judged to be high because of the lack of any sort of nuclear element to this technique. Unlike some of the other techniques. Qualitative considerations for threat mitigation using kinetic deflection. which should reduce development costs. Small quantities of dust surrounding the target body would not pose a major problem provided the interceptor is designed appropriately. Table 24. it is not anticipated that a single interceptor would be employed. each capable of some terminal-phase maneuvering.

2. These results correspond to a total mission time of 3. The DVs for both rendezvous and interception are given. The calculations were implemented in a Mircosoft® Excel workbook and then “wrapped” into the integrated design environment. a parabola. The spacecraft’s position at departure is the same as the Earth’s at that time. The Gauss problem formulation used in this study is taken from the literature.600 days. Example results for missions to the baseline asteroid are given in figure 70. since impact with the asteroid is desired. which is given as the number of days before impact. The outbound trajectory is solved as a Gauss problem. as well as the desired transfer time between them. Total mission time is defined as the sum of the outbound flight time and the asteroid arrival time (days before impact). These inaccuracies must be considered in any follow-on study. All outbound trajectories are designed to intercept or rendezvous with a baseline NEO that is defined in section 6. These assumptions are not accurate for the continuous thrust propulsion systems considered. In this study.37 The universal variables solution method allows the trajectory to be any type of conic section: an ellipse. or the time between launch of the system and the asteroid’s predicted collision with the Earth. Two points in space are known. 102 . but the use of a more accurate integrating trajectory optimization program would have required more time than was available for this project. Two DVs are calculated: (1) Must be applied to depart Earth’s orbit and send the vehicle on the trajectory that will intercept the asteroid in the desired flight time. Other missions that require close asteroid operations must perform both maneuvers. TRAJECTORY MODELING 6. This DV places the vehicle in the asteroid’s orbit and allows for rendezvous with the asteroid. ModelCenter®.6. Interceptor missions like kinetic deflection use only the first DV. or a hyperbola.1 Outbound The outbound trajectory was modeled simply using two-body orbital mechanics and impulsive thrust assumptions. The Earth’s position is calculated by moving the Earth backward from the impact point by the number of days equal to the asteroid arrival time—given as the number of days before impact—plus the desired outbound trajectory flight time. and (2) applied upon arrival at the asteroid. The asteroid’s position at interception or rendezvous is calculated by assuming an Earth impact position and traveling backward along the asteroid’s orbit to the desired arrival time. the asteroid’s orbit is initialized such that it will impact the Earth at a 45∞ angle on the heliocentric-ecliptic plane.

To determine minimum DV requirements. so as to force a collision with Earth. and inclination. In order to allow the inclusion of other objects. the Earth was placed 45∞ from the x axis of the Heliocentric-Ecliptic coordinate system at the time of the hypothetical impact.600-day total mission duration. Planetary Body Intercept (PBI). For this purpose. the orbital elements of 444 known potentially hazardous asteroids (PHAs) were examined in order to establish those elements that might apply to an “average” PHA. This program.∆V Required to Perform Outbound Trajectory (km/s) 120 Rendezvous Intercept 100 80 60 40 20 0 –300 –800 –1. such as Jupiter. a fictitious body was created in the following way. Instead. iterates over the search space until a DV is found that is a minimum and also causes the planetary body to miss the surface of Earth by 3 Earth radii. The modified asteroid orbit is plotted in figure 72.800 –2. However.300 –1. Outbound trajectory DVs for 3. during later studies. 6. as illustrated in figure 71.300 –3. Next. the software numerically integrates the equations of motion of the Sun-Earth-planetary body system. preferably dead center. 103 . First. the PHA database was searched for one asteroid that came close to this average—at least in terms of orbital size.800 –3. Its orbital elements were then modified slightly. eccentricity. The original and modified elements of 1999JT6 are given in table 25. it was decided not to conduct a lengthy search of the NEO catalog. it was necessary to find a planetary body that would definitely collide with Earth. The resulting candidate asteroid was 1999JT6.2 Inbound The inbound trajectory modeling software determines the minimum impulsive DV required to make an incoming planetary body miss Earth by some specified distance.800 Asteroid Arrival Time (Days Before Earth Impact) Figure 70.300 –2. due to the uncertainty in the orbital determination of the NEOs.

0)T km/s. Original and modified orbital elements of 1999JT6.y Ear th O ’s rbit x Sun Asteroi d Orbit Figure 71.5791277 11.840244 104 .47182 45. Original 1999JT6 319280491.568048 79.2 0.009263 41. The velocity of the planetary body at impact for this case is (–40.579033744 9.06928886 38.2 0.0. Table 25.8692997 Modified M1999JT6 319285502. Orbital Elements Name Semimajor axis (km) Eccentricity Inclination* Longitude of ascending node* Argument of perihelion* *All angles are in degrees. parallel to the x axis of the Heliocentric-Ecliptic system. Illustration showing a typical NEO orbit.

65 I + 18. To avoid confusion with the real asteroid 1999JT6. the user specifies only the mass and velocity vector of an arbitrary planetary body. Then. the position of the body is adjusted so that it collides with the surface of the Earth nearly dead center. Backward numerical integration of the three-body equations (eq. Size and composition of the fictitious asteroid are not specified. which are determined when the asteroid is at a distance from Earth equal to twice the radius of the Earth’s sphere of influence. (81) This impact velocity matches the modified orbital elements above. For M1999JT6. this fictitious asteroid has been named M1999JT6. but the impact velocity vector is always as given: V = -35. the mass is varied. During execution of the PBI program. M1999JT6 orbit plot.0K km/s . which is initially placed at the center of the Earth. since these characteristics are varied during the analysis. The equations are written in the HeliocentricEcliptic coordinate system. The user also specifies the number of days before impact when the deflection DV is to be applied.–6 –5 –4 –3 –2 –1 0 1 2 Jup iter 3 4 5 6 6 5 4 3 Or bit Ma rs Orbit 2 Impact 1 0 –1 Asteroid –2 –3 –4 –5 –6 Earth Figure 72. (82)) positions both the Earth and planetary body at their proper location for this specified time.5 J + 9. This step was taken largely to avoid the possibility of readers getting the impression that 1999JT6 is indeed on a collision course with the Earth. with the bodies assumed to be point masses: 105 .

with a tolerance of 0. and a maneuver that is best in one situation is not always the best in others. Each maneuver is made from the perspective of the planetary body and its orbit. Different maneuvers usually result in different DV requirements. However. The direction of the impulse is determined by the user from a list of possible commands shown in table 26. as shown in figure 73. which would be moving much slower than a long-period comet. With more time available. or toward the Sun. do not change the magnitude of the velocity Rotate the velocity vector toward the outside of the body’s orbit.˙˙E / S = r G( M E + M S ) rE / S 3 È rE / S + GM PB Í Í Î ˘ rPB / S ˙ 3 3 rPB / E rPB / S ˙ ˚ rPB / E ˙˙PB / S = r G( M PB + M S ) rPB / S 3 È ˘ Í rPB / E + rE / S ˙ .GM E Í rPB / E 3 rE / S 3 ˙ Î ˚ (82) At this point. figure 74 shows that DECEL may result in the minimum DV when the impact is from 50 to 150 days away. do not change the magnitude of the velocity OUTSIDE *Only ACCEL and DECEL change the magnitude of the velocity vector. the program usually converges to a solution after three to five iterations. The program makes an initial guess for the magnitude of the DV and integrates forward until the planetary body reaches its point of closest approach to the Earth. If the minimum distance between Earth and the planetary body is not within some user-specified tolerance. UP or DOWN may result in the minimum DV requirement when impact is only a few days away. 106 . Maneuver ACCEL* DECEL* UP DOWN INSIDE Description Increase the magnitude of the velocity. The situation is different for the typical asteroid. or may not converge at all. Some cases. Explanation of the different maneuvers available for use in the program PBI. however. Note that the velocities shown in figures 73–77 are those of the planetary body at impact. as is the case when the planetary body is so close to Earth that no ACCEL maneuver can result in a miss. whether or not the orbit is direct or retrograde.01 Earth radii. now the best option is to decelerate the object. a new DV magnitude is chosen and the process repeated. not from the perspective of Earth’s orbit. rB / S . whether or not the orbit is direct or retrograde. the preferred option changes from DECEL to OUTSIDE. In practice. do not change the magnitude of the velocity Rotate the velocity vector down. do not change direction Rotate the velocity vector up. Figure 75 shows that the UP maneuver no longer is the best option when the planetary body with an impact velocity of 35 km/s is only a few days away. for long-period comets coming in at very high speeds. For example. do not change the magnitude of the velocity Rotate the velocity vector toward the inside of the body’s orbit. or toward the Sun. require many more iterations. do not change direction Decrease the magnitude of the velocity. In these cases. the iterative process of determining the magnitude of the impulsive DV begins. Table 26. the program prints a warning message and reports the magnitude of the DV to be 1¥1010 km/s.

05 0 –20 –18 –16 –14 –12 –10 –8 –6 –4 –2 Days Before Impact Figure 73.002 0 –150 –140 –130 –120 –110 –100 –90 –80 –70 –60 –50 Days Before Impact Figure 74. showing the benefit of the DECEL and OUTSIDE maneuvers when impact is several weeks away.25 INSIDE 0.004 0.2 OUTSIDE UP DOWN 0. V=(–41.02 0. V=(–41. the ACCEL maneuver is found to be the most efficient maneuver for rendezvous times of ª300 days or more.29)T 0.008 0. 107 .1 0. 0.29)T 0. as illustrated in figure 77.018 0. 0. however. Required impulsive DV for 42 km/s velocity for various maneuvers to avoid collision with Earth.36.01 0.014 0.15 ACCEL DECEL Impulsive ∆ V (km/s) 0. showing the benefit of the UP maneuver when impact is only a few days away. 7. Overall.016 INSIDE OUTSIDE UP DOWN ACCEL DECEL Impulsive ∆ V (km/s) 0. 7.012 0.006 0.36. Required impulsive DV for 42 km/s velocity for various maneuvers to avoid collision with Earth.as illustrated in figure 76.

02 Impulsive ∆ V (km/s) 0. showing the benefit of the DECEL maneuver when impact is only a few days away.078)T 0.078)T 0.3 INSIDE OUTSIDE UP DOWN ACCEL DECEL Impulsive ∆ V (km/s) 0. Required impulsive DV for various maneuvers to avoid collision with Earth for planetary body with velocity of 35 km/s.01 0.15 0. Required impulsive DV for 35 km/s velocity for various maneuvers to avoid collision with Earth. 6.V=(–34. 0. 108 .2 0. V=(–34.25 0.35 0.015 0.05 0 –20 –18 –16 –14 –12 –10 Days Before Impact –8 –6 –4 –2 Figure 75.47.1 0.005 0 –150 –140 –130 –120 –110 –100 –90 –80 –70 –60 –50 Days Before Impact Figure 76. 0. 6.025 INSIDE OUTSIDE UP DOWN ACCEL DECEL 0.47.

e.000 –900 –800 –700 –600 –500 –400 –300 Days Before Impact Figure 77. The goal of manipulating the planetary body’s trajectory is to miss the moving Earth. One must realize that the quest for the minimum DV did not include the object’s disposal. the optimum trajectory may be a combination of UP and DECEL. comparing the DV requirements for a range of successful trajectories gives insight into which combinations could possibly result in a lower impulse requirement.078)T 0. for example.V=(–34.0016 INSIDE OUTSIDE UP DOWN ACCEL DECEL Impulsive ∆ V (km/s) 0.0010 0. By limiting the number of possible maneuvers to six rather than performing a global search.0004 0.0002 0 –1.0006 0. In many cases. computation speed is increased significantly.47. only its avoidance.0012 0. never threatening Earth again. A maneuver that removes the body completely from the list of threatening objects is clearly the best option and should be considered in future studies. maneuvers that delay the arrival of the planetary body generally allow Earth to move out of the way before the body arrives. 6. 0.0018 0. However. not to rendezvous with it. Therefore.0008 0. Required impulsive DV for various maneuvers to avoid collision with Earth for planetary body with velocity of 35 km/s (long lead time). By only requiring the body to miss Earth on this one occasion does not guarantee that it will miss Earth during some future encounter. 109 .. i.0014 0.0020 0.

short-period Comet. Total Deaths (%) 60 5 5 <1 <1 6 24 Type Chondrite Achondrite Iron Mesosiderite Pallasite Comet. Types of planetary bodies examined in the Monte Carlo simulation and their average contribution to the total number of deaths over the next century.. estimates the number of people saved over the next century if all planetary bodies of a given type.. century. long-period comet. can be successfully deflected.7. PEOPLE. This means that researchers must resort to the use of theoretical models41 to estimate the number of deaths resulting from the impact of a planetary body of a certain size.000 runs of a modified version of John Lewis’s Monte Carlo simulation program were used to generate data for the average number of deaths due to each type of object (table 27). or millennium due to impacts of certain sizes of planetary bodies is a highly speculative endeavor.39 and Lewis41 using estimates of both the impact frequency and the number of deaths due to impacts of a certain energy. velocity. it is far ahead of our understanding of the consequences of the most likely impacts. and better understood. Over 10. One only has to consider the literature in references 39–44 and many others to see evidence of increased efforts to understand the threat that Earth faces every day. chondrite.38 the threat and consequences of an impact with a planetary body are now more appreciated. long-period 110 . than was the case 30 yr ago. THREAT PARAMETRIC Thanks to the pioneering work of researchers like Eugene Shoemaker. Recorded impacts in the developed world are rare and so do not constitute a statistically significant database. Our program. i. The number of fatalities prevented is based on the work done by Shoemaker. Attempting to predict the number of people killed over the next decade.) Table 27. enabling the study team to focus efforts on the types of objects that would most likely cause the most deaths over the next century. etc. Although our understanding of the impact threat is still incomplete. (The main modification to Lewis’s simulation code allowed the tallying of deaths due to different types of objects.38 Chapman and Morrison.e. and type. having kinetic energy less than or equal to some given value.

43 Therefore.577 (if the user wishes.37 ¥ 10 15 .525 F = 164 – 180 Mton–0. its likelihood is only 1 in 100 million. Similarly.) A static human population of 6 billion is assumed here. For example.5 < Mton £ 200.The program PEOPLE also determines which parameter—mass or velocity—should be increased to counter the largest portion of the threat.45 indicates that the frequency for Tunguska-sized events may occur only every 1. large enough 111 . (85) Caution must be used when applying these equations and estimates.000.40 The equations are split into four categories based on the planetary body’s equivalent energy yield: Mton £ 2.16¥106 Mton–0. its impact velocity (V) relative to Earth (m/s).5 2.325 f = 1000 Mton2/3 f = 1¥108 Mton0. the overwhelmingly dominant event each year is a small one.000 < Mton £ 1¥108 Mton > 1¥108 f = 5. This determination is based on velocity distributions from Lewis41 and Chesley. the largest number by far.39 Lewis.5 2. the world population. but when it does occur. impactors with energies of 1¥105 Mton statistically cause ª3.223 f = 6¥109.433 Mton6. and that the size distribution of the smaller asteroids may need to be reassessed. while their likelihood is only 1 in 10. and direct impacts. This is not the cumulative number of fatalities as described above.000 yr as opposed to previous estimates of every 200 to 300 yr. (83) Next. The first step in PEOPLE is to use the mass (m) of the planetary body (kilograms). the outcome will clearly be catastrophic.41 and Gold.16¥106 Mton–0.43 and size distributions from Gold40 and Ivezic et al. Impacts of this size are on the threshold of being globally catastrophic events. on average.44 (Recent data from Brown et al.000 200. while a catastrophic event with impact energy of 1¥108 Mton could kill several billion people. this type of impact will result in <100 fatalities per year on average.5 < Mton £ 200. as it is throughout most of the literature. the number of deaths (F ) per year. Causes of death include tsunamis.577 F = 3763 – 4.000 < Mton £ 1¥108 Mton > 1¥108 F = 0. firestorms. However. but is based on the following equations: Mton £ 2. and its type (see table 27).39. (84) PEOPLE also calculates the number of fatalities ( f ) that would have been expected on average for this energy of impact. this is an extrapolation). due to all impacts having this energy or less is determined. and determine the equivalent energy yield in megatons of TNT using the following equation: 2 mVPB / E Mton = 8.1333 F = 3763 – 4.0236 Mton5.000 fatalities per year. blast waves. which would result in only a few fatalities. These equations were derived using data taken from Chapman and Morrison.000 200.

42241 .0015V .0.18 21. ∂m (87) Finally.to do massive damage throughout the world.005256V + 0.18 £ V < 21. Ëp r¯ 1 (86) where the density depends of the type of object.0075 ∂V ∂F = 0.03 ∂V ∂F = -0.18) 2 ∂V ∂F = 0. the program estimates the rate of change in fatalities with respect to velocity.67 D-2. They are also frequent enough to result in a high average number of fatalities per year. using the appropriate form of equation (88).3814 ∂V 66.86 £ V < 71. as listed in table 27.18 £ V < 35 35 £ V < 50 50 £ V < 66. the program PEOPLE determines which planetary body parameter—velocity or mass— should be increased in order to defeat the maximum portion of the total threat.045 ∂V 11. First. These equations are based on data reported by Lewis41 and Gold40 on the velocity and size distributions of NEAs and comets.000075(V .39 Finally.86 ∂F = 0.11. or asteroid) and its velocity relative to the Earth at impact: • Long-period comet: ∂F = 0. the spherical diameter of the object is determined by Ê 6m ˆ 3 D=Á ˜ . but too small to destroy humanity.86 112 . if the deflection system used to successfully change the course of the object has some excess energy. short-period comet. The rate of change of cumulative average fatalities with respect to mass is then calculated from ∂F = 37. This calculation depends on the type of object (long-period comet.

00072(V .0. although the exact size of the asteroid that defines the threshold is unclear.072 ∂V ∂F =0 ∂V ∂F = 0.86 £ V < 71.18) .86 66.1023(V . and is strongly influenced by type. on average.• Short-period comet: ∂F = -0.79 . The threshold asteroid size for a globally catastrophic event is evident.86 (88) Lewis. 113 . be prevented over the next century if all incoming asteroids of a particular energy or less could be successfully deflected. PEOPLE was incorporated into a model in ModelCenter to determine the number of deaths that could.41 The velocity distribution of asteroids is based on an equation slightly different from the one used by but it seems to fit the data slightly better. ∂V 11.86 • Asteroid: ∂F = 0.008 ∂V 14 £ V < 34 34 £ V < 66.11. The results are graphed in figure 78.24) 2 + 0.18 £ V < 71.

These simulations contain both statistical data and estimates gleaned from experience with high-energy detonations in both populated and unpopulated areas. This results in a high uncertainty in the estimate of fatalities due to an asteroid impact. the average number of deaths per impact and the cumulative number of deaths per century are based on simulations.000 Average Deaths From Single Impact 108 107 106 105 104 103 102 101 100 10 Death per Century for All Impact of Equal or Less Energy 100. However. In summary.000 1. Therefore.000 10. it is clear that the standard deviation in the estimate of facilities for most sizes of impactors is much larger than the mean. and the average total number of deaths prevented if all impacts of equal or less energy can be avoided.000 Figure 78. the relatively small average number of fatalities per century is little reason to conclude that defending Earth from these sizes of impacts is unnecessary. Average deaths from single asteroid impact versus size.000 100 100 1.000 people. Therefore. the estimate of average fatalities per century should be used with caution. and is only a guide to illuminate the direction in which the design of a planetary defense system should proceed in order to prevent the most fatalities on average.000. 114 . but that number could increase to the millions if the impact occurred in or near a major city or a densely populated coastline.000 Chondrite Diameter (m) 10.1010 109 Average Deaths From Single Impact Average Deaths Prevented per Century if all Impacts of Equal or Less Energy can be Defeated 1. An uncertainty analysis of the estimated fatalities was not completed due to time constraints. The average number of fatalities from the impact of a 100-m chondrite is ª10. The number of fatalities is very sensitive to the impact location for the smaller asteroids.

Figure 79 illustrates the proposed analysis process. or deflection. By running through this process several times—assuming new total system masses and mission times on each occasion—yields a parametric model of the total threat defeated as a function of total system mass for lines of constant mission time. based on the type of threat mitigation concept being considered—remote station. the analysis process then divides.1 Integrated Analysis The original intent of this project was to evaluate the ability of various combinations of technologies to defeat the entire threat posed by NEOs. Potential impactors come in all shapes and sizes.8. The remote station analysis path assumes both an incoming asteroid mass and a velocity vector. and to test the ability of each to defeat the total threat. Running the inbound parametric defines the DV required to deflect the asteroid. To further complicate matters. and also the allowable mission time. The original intent of this project was to select several technologies based on our understanding and experience. a large number of technologies for use in threat mitigation have been examined. then the analysis path returns to assume a new asteroid mass and velocity vector. Running the remote station tool. Obviously. interception. given its size and velocity vector at point of impact. Starting with an assumption for the total system mass and the total mission time allowed. 115 . and their orbits vary greatly. this is a very complicated problem. computes the required remote station mass. PARAMETRIC RESULTS 8. Assessing all possible technology combinations would be prohibitive. New DVs and remote station masses are then computed. If this mass is not equal to the allowable mass for the system assumed at the beginning. the threat assessment tool is run again in order to compute the total threat that is defeated. The total threat is quantified by the percentage of people saved by deployment of the system over a given time period divided by the number of people expected to die from impact of an NEO over the same period. After closure. The new asteroid mass and velocity vector is selected using the threat assessment tool to maximize the percentage of the total threat that can be defeated for the assumed total system mass. based on the DV to be delivered to the object and the allowable mission time.

Proposed analysis process for assessing total amount of threat mitigated.Remote Station Interception Deflection Assume Mt. First. Development of an atmospheric entry code that would model the burnup of these fragments was initiated. the calculation proceeds to calculate the DV required and then the mass of the outbound propulsion stage needed to take the interceptor to the incoming object. then a larger interception mass is assumed and a new iteration ensues. Vimp Asteroid Dismantle Parametric Assume Ma. The interception branch differs from that of the remote station in that the inner iterative loop determines the asteroid size and velocity that can be deflected. tT Assume Ma. Vimp Guess: ti Start Inbound Trajectory Parametric ∆ Vi Stationary Deflector Parametric Inbound Trajectory Parametric No ∆ Vi Inbound Propulsion Parametric Mb Debris Reentry Parametric Mdeb Survive Reentry? Yes Mi Outbound Trajectory Parametric No ∆Vo Outbound Propulsion Parametric Outbound Trajectory Parametric ∆ Vo No Outbound Propulsion Parametric Finish Ms Ms = M T Yes Mo Cumulative Damage Percent Mi + M o = M T Mo Mb + M o = M T No Yes Yes Threat Parametric No ti Optimal? Yes Figure 79.e. This outer loop iterates until the maximum threatening object that can be mitigated for the total allowable mission mass is found.. Otherwise. Here. Vimp Assume Ma. If the fragments survived reentry. It is intended that this model would include the effects of drafting. there is an additional problem in optimizing the amount of the total mission mass allocated to the outbound propulsion against the amount allocated to the interception system. later objects following in the wake of the earlier objects. i. but this has not yet been completed. again using the threat 116 . an interceptor mass is assumed and the size of the resulting object fragments is estimated.

Note that the inclination distribution was taken into account. the total mass can be calculated by summing the inbound and outbound masses and comparing the result with the assumed total system mass. it was decided that the whole issue of composition would not be addressed for this initial study. defined by the allowed inbound interception time. Unfortunately. note that a remote station architecture option was not considered. As before. After closure of this loop. There are still two assumptions built into the above analysis scheme: (1) Although the distribution of object mass and velocity is taken into account. (2) The deflection study allowed no time for setup after asteroid rendezvous. it is imposed by a nonzero z component in the incoming object velocity vector at impact.2. as well as the outbound DV and mass. can be generated. Here. several architecture options were considered—either suggested in the literature or which appeared promising. a parametric model of total threat defeated as a function of system mass and mission time is generated. the process goes into the threat parametric to find the total percentage of the threat that is defeated. 117 .2 Architecture Options In the time available for this study. In addition. because asteroid and comet compositions are so poorly understood. The list of cases that were considered and their mission configuration are shown in table 28. That composition is believed to be a secondary factor in performance. The parametrics were derived using a process similar to the inner loops shown in figure 79. As will be shown below. Then. Finally. giving the percentage of the total threat defeated as a function of both total system mass and mission time. Even after closure of this inner loop. Again. In particular. a parametric model. these ambitious analysis goals were not completed in the time available. None of the options considered required a significant amount of time after rendezvous for these operations. It must be emphasized that this selection does not constitute a full list of possible options for threat mitigation. the interception point. an allowed inbound trip time is also assumed. except for the mass driver. the most threatening object must be found using techniques similar to those employed for the interception branch. The rendezvous branch in figure 79 is the most complicated one of all. the possible distribution in composition is not. a parametric model of total system mass was derived for several architecture options. Note that each case or mission scenario was assembled in a unique project file within the integrated design environment ModelCenter. However. in this case. may not be optimal. another loop is used to find the optimal interception point. assuming a standard set of orbital elements for the incoming object. 8. therefore.tool. The architecture options considered and performance of these options are described in section 8. so it was assumed that the required deflection DV was imposed instantaneously upon rendezvous. although not in damage caused. The analysis process then runs through the tools that calculate the required inbound DV and mass. the threat parametric has several implied assumptions defined in section 7.

the outbound trajectory tool determines the DV requirement to rendezvous with the asteroid. shows this basic design process. With this information and the desired time of flight. a partial screen shot from the ModelCenter program. Remote Station Versus Interception Versus Outbound System Staged chemical Staged chemical Staged chemical Nuclear pulse Solar collector Inbound System Mass driver Kinetic deflector Nuclear deflection Nuclear deflection Solar collector Rendezvous Rendezvous Interception Interception and rendezvous Rendezvous Rendezvous Deflection Versus Fragmentation Deflection Deflection Deflection Deflection Deflection 8. The mass driver’s total mass is used as the payload mass for a staged chemical rocket that performs the DV maneuvers required for the outbound trajectory.3. This value is used in the mass driver sizing tool. Architecture options considered in this study. which includes the mass driver and the rocket required to deliver it to the asteroid. The first component. This requires a balance between the conflicting goals of minimizing the deflection DV and the outbound trajectory DV.3 Parametric Performance 8. Main. The final result is the total system mass. It also performs some simple algebraic manipulation necessary to convert units and calculate inputs for subsequent programs. One output of this tool is the DV required to deflect the asteroid. The asteroid’s position at the time of deflection is another output. The optimal mission from an energy standpoint is that which requires the lowest mass. The inbound trajectory program is initiated by specifying how many days before impact the deflection is to take place.Table 28. Figure 80.1 Staged Chemical/Mass Driver The basic design process for this scenario is as follows. 118 . is a simple script. This frees the user from needing to search through the extensive input lists of the other programs. The primary reason for creating Main is to collect all of the inputs used in the parametric studies into a single component.

the only changing input to this program is the number of days before impact that the deflection is to occur. the ACCEL maneuver requires the least DV. In the magnified image. Figure 81 shows the deflection DV for each deflection direction over the range of 100 days to over 10 yr before impact. Staged chemical/mass driver model. which was documented in section 6. and it is obvious for the ACCEL and DECEL options that each oscillation diminishes in magnitude. 119 .2 of this TP. The first of these inputs is a user input that comes from the Main program discussed above. and the optimal deflection direction. PBI. The next component. It is also interesting to note that the DV required oscillates with each asteroid period. The second input is determined by looking at each deflection direction and finding the best option for the missions considered. it is apparent that for any deflection that occurs more than 600 days before impact. is the inbound trajectory tool. With the asteroid specified and the integration parameters set.Main PBO_v4 PBI Unit Conversion2 MassDriver MassDriverDVCalc People_v3 stagedChemical3 UnitConversion1 GoalSeek Figure 80.

01 (a) 500 –500 –1.100 Days) ∆ V (km/s) 0.500 Asteroid Arrival Time (Days Before Earth Impact) –4.500 –2.0004 ∆ V (km/s) 0.02 0.0002 0.06 0.500 –2. (b) optimal deflection direction—detailed view.500 0.04 Asteroid M1999JT6 Deflection ∆V (Asteroid's Period is ≈ 1.500 –3.0035 0.500 Asteroid Arrival Time (Days Before Earth Impact) –4.004 0.01 0 Invalid Point Maneuver Type ACCEL DECEL UP DOWN INSIDE OUTSIDE –0.00035 0.002 0. 120 .500 0.500 –3.0025 0.500 Asteroid Arrival Time (Days Before Earth Impact) –4.500 –2.00045 0. and (c) optimal deflection direction —detailed view—minimal deflection DV.0003 0.0005 (b) 500 –500 –1.0015 0.500 –3. (a) Optimal deflection direction.500 Figure 81.00025 0.003 Maneuver Type ACCEL DECEL UP DOWN INSIDE OUTSIDE ∆V (km/s) 0.0005 0.00015 0.0005 0 Invalid Point –0.05 0.001 0.0.00005 0 Maneuver Type ACCEL DECEL INSIDE OUTSIDE (c) 500 –500 –1.0001 0.03 0.

The outbound flight time is calculated and passed from the Main component. is not set up to use a DV for sizing. changes the amount of asteroid mass ejected by the mass driver until the two DVs are equal.500 –2.000 –1.5×10–4 2×10–4 1.500 –3. MassDriver. the mass driver can be sized. The outbound trajectory code.01 times the total amount of mass to be ejected. From figure 82 one can see that this change in component mass results in a minimal shift in the total system mass. The required DV to rendezvous with the asteroid is output. 75 4×10–4 Deflection ∆V (km/s) Mass Driver Mass (t) 3. This facility houses the expellant mass prior to ejection. The optimizer. MassDriverDVCalc is a script that converts the total impulse provided by the mass driver into a DV and then compares it with the DV required to deflect the asteroid.500 –4. the asteroid provides it all.000 69 68 67 Asteroid Arrival Time (Days Before Earth Impact) Figure 82. so an iteration loop is introduced.000 –2. its mass is equal to 0. When the mass driver produces the correct DV. from metric tons to kilograms. There is a simple unit conversion in the loop that converts ejected asteroid mass. GoalSeek.000 –3.5×10–4 1×10–4 5×10–5 0 0 –500 –1. The mass driver analysis program. or total expellant mass. Variation of mass driver total system mass with required asteroid deflection DV for a 50-m-diameter chondrite. is executed once the asteroid’s position at rendezvous is passed from PBI. Planetary Body—Outbound (PBO). There is only one term in the mass driver sizing relationship that scales with DV—the Expellant Storage facility mass.5×10–4 3×10–4 74 73 ∆V to Deflect (km/s) Mass Driver Mass (t) 72 71 70 2.Once the deflection DV is known. where it becomes the payload mass for the outbound vehicle. 121 . and is passed to the staged chemical tool. and for this study. its total system mass value is passed to the staged chemical tool. One interesting feature about the mass driver is that its total system mass remains relatively constant regardless of the deflection DV required. This is because no expellant mass needs to be carried out with the mass driver.

the outbound flight time could be as short as 25 days or as long as 3. Within these ranges. the user can input a total mission duration and a rendezvous time. and then calculate the total mass required for the mission for a given asteroid composition and diameter. For this scenario. in increments of 25 days. With all of the necessary components appropriately linked in a ModelCenter project file.000 50. the vehicle was allowed five stages.500 days before impact. The small increment value was necessary to ensure that optimal orbital transfers were captured.000 0 –800 –1.800 –2.800 –3.300 Asteroid Arrival Time (Days Before Earth Impact) 1525 1600 1675 1750 1825 1900 1975 2050 2125 2200 2275 2350 2425 2500 2575 2650 2725 2800 2875 2950 3025 3100 3175 3250 3325 3400 3475 3550 1550 1625 1700 1775 1850 1925 2000 2075 2150 2225 2300 2375 2450 2525 2600 2675 2750 2825 2900 2975 3050 3125 3200 3275 3350 3425 3500 3575 –3. only the lox/LH2 propellant combination is considered.000 Total Mission Duration 1500 1575 1650 1725 1800 1875 1950 2025 2100 2175 2250 2325 2400 2475 2550 2625 2700 2775 2850 2925 3000 3075 3150 3225 3300 3375 3450 3525 3600 Spacecraft Mass (t) at SOI 350. The point at which the vehicle makes its rendezvous with the asteroid was varied from 500 to 3. which is defined as the number of days between the vehicle’s departure from Earth and the impending Earth-asteroid collision. In this manner. PBO. with an assumed vacuum Isp of 465 s. the parametric analysis was performed for this mission scenario. Also.800 Figure 83.The staged chemical tool uses the payload mass given by the mass driver component and the DV from the outbound trajectory code. also in increments of 25 days.000 200.100 days.500 days to 10 yr. 450. Staged chemical/mass driver vehicle mass at Earth departure. The total mission duration.300 –1.000 400. was varied from 1.000 150.300 –2. to size the vehicle that departs from Earth.000 250.000 100.000 300. 122 . Figures 83 and 84 show the resulting initial spacecraft mass at Earth departure.

5 1 1.900 days before the asteroid’s collision with the Earth.000 0 –800 –1. and arrives at the asteroid 500 days after launch.5 Asteroid at start –2 –2.20.000 2. Staged chemical/mass driver vehicle mass at Earth departure (expanded view).300 –1.800 –2.800 –3.000 6.5 0 0.000 10. This mission is launched 2.5 3 3.5 –2 –1. The minimum mass solution for this scenario is 6.000 18.5 Earth at finish Impact position Asteroid asc.850 t.300 –2.800 Asteroid Arrival Time (Days Before Earth Impact) Figure 84.000 4.5 2 2. node Transfer trajectory Start transfer Finish transfer Earth Earth perihelion Asteroid Asteroid perihelion Sun Figure 85.300 Minimum Mass Solution 1525 1600 1675 1750 1825 1900 1975 2050 2125 2200 2275 2350 2425 2500 2575 2650 2725 2800 2875 2950 3025 3100 3175 3250 3325 3400 3475 3550 1550 1625 1700 1775 1850 1925 2000 2075 2150 2225 2300 2375 2450 2525 2600 2675 2750 2825 2900 2975 3050 3125 3200 3275 3350 3425 3500 3575 –3. Optimal staged chemical/mass driver mission.5 0 –0.000 8.5 –3 –3.000 Total Mission Duration 1500 1575 1650 1725 1800 1875 1950 2025 2100 2175 2250 2325 2400 2475 2550 2625 2700 2775 2850 2925 3000 3075 3150 3225 3300 3375 3450 3525 3600 Spacecraft Mass (t) at SOI 16.5 –1 –1.000 12.5 2 1. The trajectory is shown in figure 85.5 3 2. –3 –2.5 1 0. 123 .5 –1 –0.000 14.

000 4.000 2. the overall vehicle mass remained <7. Staged chemical/mass driver vehicle mass versus chondrite asteroid diameter. the asteroid diameter was varied to determine its effect on mission mass. The results show that this scenario can be used to deflect chondrite asteroids up to 80 m in diameter before the staged chemical’s performance becomes inadequate. Mass driver deployed mass and total operating time required to deflect asteroids of 50 to 1. The resulting vehicle masses are given in figure 86. Because the mass driver system mass changed very little.000 Spacecraft Mass (t) at SOI 7. 100 95 Mass Driver Deployed Mass (t) 50 Deployed Mass Operating Time 45 40 35 30 25 20 15 10 5 0 Operating Time (days) 90 85 80 75 70 65 60 55 50 40 50 60 70 80 Asteroid Diameter (m) 90 100 110 Figure 87. 10.000 5.Once the optimal mission solution was found. these numbers may be used in future analysis with a different outbound propulsion system.000 1.000 6.000 m in diameter are given in figures 87 and 88. While the staged chemical propulsion system is not capable of delivering systems for the larger asteroids.000 9.000 3.000 0 0 20 80 40 60 Asteroid Diameter (m) 100 120 Figure 86.000 8.000 t. Mass driver deployed mass and total operating time versus chondrite asteroid diameter (50–100 m). 124 .

000 Deployed Mass Operating Time 45. stagedChemical3.000 m).500 2.2 Staged Chemical/Nuclear Deflection The cases considered in this scenario included both intercept and rendezvous with the asteroid.000 15.000 35.000 5. Starting with the asteroid diameter and type—obtained from the main script—and the required asteroid DV from PBI.000 1.000 30.000 25.000 0 0 100 200 300 400 500 600 700 800 900 1.250 1.000 10. The staged chemical tool then sizes the chemical rocket that is required to deliver the nuclear devices from Earth’s sphere of influence to the asteroid.250 Mass Driver Deployed Mass (mt) 50. Mass driver deployed mass and total operating time versus chondrite asteroid diameter (100–1.3. Main PBI km2m Explosion Ibm2kg PBO_v3 stagedChemical3 Figure 89. The overall layout is illustrated in figure 89.2. Diagram of the ModelCenter setup for the staged chemical/nuclear deflection option.500 1. The payload mass. is passed to the staged chemical tool. 8. 125 .3. explosion calculates the size and number of nuclear blasts necessary to deflect the asteroid.000 750 500 250 0 Figure 88.100 Asteroid Diameter (m) Operating Time (days) 2.000 40.1) is the replacement of the mass driver with the nuclear deflection tool called explosion. which is just the total mass of the nuclear devices.000 1. 8. The main difference between this model and that for the staged chemical/mass driver option (which was considered in sec.750 1.000 20.

This is accomplished by iterating over both the total mission time and the rendezvous time. Results of these runs for the intercept and rendezvous cases are illustrated in figures 90 and 91. given the large staged chemical system that was required for the mass driver option. the optimum rendezvous and total mission times are 910 days before impact and 1. the iteration process is likely to converge on nonoptimal solutions.620 kg for the rendezvous case (fig. the Optimization component of ModelCenter is used to find the minimum total system mass. but they do illustrate the benefits of using nuclear energy to deflect the asteroid. Here. 126 . Otherwise. These results are somewhat surprising.000 1. respectively. The trajectories for the optimum solutions are shown in figures 94 and 95. the minimum total system mass required to defeat a 100-m-diameter chondrite was found to be 847 kg for the intercept case (fig.400 7 6 5 4 3 2 1 0 500 1. respectively. the staged chemical system does not match the asteroid’s orbit at encounter. the optimum rendezvous and total mission times are 132 days before impact and 1.050 –1. Once the regions of interest are identified.509 days. Total system mass for the staged chemical/nuclear blast option versus total mission time for various rendezvous times. it is necessary to restrict the search to specific regions before the optimization tool can be used. Through a thorough set of analysis runs. 93).000 –1. For the rendezvous case.Various values of total mission time and rendezvous time are examined to determine those regions in which the global minimum total system mass resides.000 2.075 days. Intercept 10 9 8 Spacecraft Mass (t) at SOI Rendezvous Time (Days Before Impact) –100 –200 –500 –850 –900 –950 –1. With so many peaks and valleys in the total system mass graphs in figures 90 and 91. For the intercept case.500 Figure 90.500 Total Mission Time (Days) 2. 92) and 5.

8 0.3 1.480 1.9 1.1 0 1.520 1.509 847 kg Spacecraft Mass (t) at SOI –1.510 1.1 1 0.5 1. 127 .530 1. Total system mass for the staged chemical/nuclear blast option versus total mission time for various rendezvous times (zoomed).2 1. Minimum total system mass for the staged chemical/nuclear blast option.540 1. the staged chemical system matches the asteroid’s orbit at encounter.Rendezvous 20 18 16 Rendezvous Time (Days Before Impact) –100 –132 –150 –175 –200 –250 –300 –400 –500 –1. Here.500 Intercept Rendezvous Time (Days Before Impact) –100 –200 –500 –850 –900 –950 –1.470 1.6 0.500 Total Mission Time (Days) 2.000 1.3 0.550 Total Mission Time (Days) Figure 92.5 0. showing the optimum rendezvous and total mission times for intercept.6 1.7 0.2 0.4 1.500 Figure 91.7 1.400 –910 1.050 –910/1. 2 1.400 Spacecraft Mass (t) at SOI 14 12 10 8 6 4 2 0 500 1.4 0.000 –1.000 –1.9 0.000 2.8 1.490 1.

050 –132/1.5 –3 Sun Earth Earth perihelion Asteroid Asteroid perihelion Asteroid asc.5 5 1.075 5. –3 –2. showing the optimum rendezvous and total mission times for rendezvous.000 –1.5 1 1.070 1.5 7 6. Optimum intercept trajectory for the staged chemical/nuclear deflection option.5 2 2.5 –1 –0. Minimum total system mass for the staged chemical/nuclear blast option.5 –2 –1.5 –2 –2.100 Total Mission Time (Days) Figure 93.5 0 –0.5 0 0.060 1.5 8 7.Rendezvous 10 9.080 1.090 1.620 kg 1.5 9 Rendezvous Time (Days Before Impact) –100 –132 –150 –150 –200 –250 –300 –400 –500 –1.5 3 3 2.5 –1 –1.5 1 0.5 6 5. node Transfer trajectory Start transfer Finish transfer Asteroid at start Earth at finish Impact position Figure 94. 128 .400 Spacecraft Mass (t) at SOI 8.5 2 1.

With the rendezvous option. total system mass for the intercept case is much smaller and is less sensitive to changing asteroid mass.5 3 3 2.5 0 –0.000-t limit assumed for this study. 129 .–3 –2.5 –2 –2. the optimum times are nearly the same for both 100. Given the total system mass constraint of 1.5 1 0. the largest diameter M1999JT6 chondrite that this system can defeat has a diameter of 9.000 m for the rendezvous case. node Transfer trajectory Start transfer Finish transfer Asteroid at start Earth at finish Impact position Figure 95.5 2 1.000 t.5 –1 –1.5 –3 Sun Earth Earth perihelion Asteroid Asteroid perihelion Asteroid asc.5 0 0. With this information. The optimum rendezvous and total mission time vary little with changing asteroid mass.5 2 2.5 –1 –0.and 1. the total system mass increases rapidly and quickly exceeds the 1. However. it is relatively easy to determine the required size of the staged chemical rocket versus asteroid size.000 m for the intercept case and 1.000-m-diameter chondrites.5 –2 –1. The results are plotted in figures 96 and 97. Optimum rendezvous trajectory for the staged chemical/nuclear deflection option.5 1 1.

600 500 400 300 200 100 0 0 1.000 10.000 0 0 2.000 4. 25. 1.000 900 800 700 Total System Mass (t) Lack of data points is because the nuclear deflection tool does not interpolate between the tabular data used to estimate the size of nuclear blast required.000 8.000 5. 130 . Minimum total system mass for the staged chemical/nuclear blast option versus chondrite diameter for the smaller chondrites.000 20.000 6.000 Lack of data points is because the nuclear deflection tool does not interpolate between the tabular data used to estimate the size of nuclear blast required.000 Intercept Rendezvous 1.000 Chondrite Asteroid Diameter (m) Intercept Rendezvous 1.000 4.000-t limit Figure 97.000 Chondrite Asteroid Diameter (m) Figure 96.000 10.30.000 3.000 Total System Mass (t) 30. showing both intercept and rendezvous.000 2. Minimum total system mass for the staged chemical/nuclear blast option versus chondrite diameter for both intercept and rendezvous.000-t limit 15.

The projectile mass is varied until the momentum necessary to just achieve the desired DV is imparted to the asteroid. This feeds into PBO. PBI runs first to determine the asteroid’s position. The process flow for the analysis of this scenario in ModelCenter is given in figure 98. Hence. So. This time the number of stages is allowed to vary between one and five. which determines the impact velocity. All stages use the lox/LH2 propellant combination.3 Staged Chemical/Kinetic Deflection The staged chemical/kinetic deflection option consists of a chemical rocket that delivers a massive projectile to the asteroid.8. Main PBI PBI_dv PBO_v5 stagedChemical5b KineticDeflection DV_check ImpactorMassOptimizer Figure 98. PBI_dv then produces the DV required to deflect the asteroid. The two instances of PBI are required because the outbound trajectory code cannot run until the asteroid’s position at interception is known. the diagram contains two instances of the inbound trajectory code. this vehicle impacts the asteroid. First. which gives PBI_dv the deflection direction. Like the mass driver scenario. This projectile impacts the asteroid. Unlike the mass driver option. The resulting projectile mass is then input to the staged chemical tool as its payload mass. nudging it off of its collision course with the Earth. The process is similar to that of the mass driver option. there is an iteration loop to ensure that the deflection DV produced is equal to that required. The staged chemical tool then provides the overall vehicle mass required to complete the mission. the outbound trajectory. The reason for this apparent duplication is that the asteroid’s deflection direction is actually determined by the outbound trajectory of the projectile. has a significant influence on the subsequent deflection. delivering an Isp of 465 s. PBI and PBI_dv. but there are a few exceptions. 131 . Staged chemical/kinetic deflection model. it does not rendezvous with it.3.

200 days. this is the option that is most readily available to counter near-term threats with limited reaction times.000. it was decided to limit the trade space for this case to nearer term missions only.000. The justification for this limitation comes from the fact that a simple kinetic deflection system.. which would result in lower total mission masses than those presented here. With a limited amount of time to complete this study. See figure 99 for resulting vehicle masses for the missions considered.000 2. It can be activated at relatively short notice—certainly much sooner than the more complex deflection options.000 days before Earth impact would result in similar trends to those shown in figure 81.000 Spacecraft Mass (t) at SOI Total Mission Duration 150CF 200CF 250CF 300CF 350CF 400CF 450CF 500CF 525CF 550CF 575CF 600CF 625CF 650CF 675CF 700CF 725CF 750CF 775CF 800CF 825CF 850CF 875CF 900CF 925CF 950CF 975CF 1000CF 1025CF 1050CF 1075CF 1100CF 1125CF 1150CF 1175CF 1200CF 175CF 225CF 275CF 325CF 375CF 425CF 475CF 525 550 575 600 625 650 675 700 725 750 775 800 825 850 875 900 925 950 975 1000 1025 1050 1075 1100 1125 1150 1175 1200 5.000.000.The initial parametric study was performed for a 50-m-diameter asteroid of chondrite composition. is the simplest type of mitigation option that can be envisioned.000 1.g.100 days). placed on an impact trajectory by a staged chemical propulsion system.000 4.200 (a) 132 .000 or 3. diminishing in magnitude with each period.000. In short.000 Asteroid Arrival Time (Days Before Earth Impact) –1. e. the mass driver—and largely uses existing technology. One would expect that extending the trade-space to consider arriving at the asteroid 2. however.125 days before impact. and interception with the asteroid occurred between 100 and 1. It should be noted. 7.000 0 200 0 –200 –400 –600 –800 –1.000.000. that even in this limited trade space. the asteroid arrival times considered encompass one full period of the asteroid (ª1.000 3. Total mission time was varied from 150 to 1. The required deflection DV should oscillate.000 6.

missions that result in catastrophic fragmentation are denoted by a CF after the series label. (b) detailed view. and (c) detailed view—minimum mass solution.1.200 Asteroid Arrival Time (Days Before Earth Impact) 100 90 80 Total Mission Duration 150CF 200CF 250CF 300CF 350CF 400CF 450CF 500CF 525CF 550CF 575CF 600CF 625CF 650CF 675CF 700CF 725CF 750CF 775CF 800CF 825CF 850CF 875CF 900CF 925CF 950CF 975CF 1000CF 1025CF 1050CF 1075CF 1100CF 1125CF 1150CF 1175CF 1200CF 175CF 225CF 275CF 325CF 375CF 425CF 475CF 525 550 575 600 625 650 675 700 725 750 775 800 825 850 875 900 925 950 975 1000 1025 1050 1075 1100 1125 1150 1175 1200 Spacecraft Mass (t) at SOI 70 60 50 40 30 20 Minimum Mass Solution 10 0 (c) 200 0 –200 –400 –600 –800 –1. This was discussed in section 5. One additional complication associated with the kinetic deflection option is the possibility that the impact might cause the asteroid to fragment.000 –1. there is some threshold impact size above that which the asteroid will actually break—a process referred to as catastrophic fragmentation.000 –1. In figure 99(c). Small impacts will produce only craters.000 900 800 Total Mission Duration 150CF 200CF 250CF 300CF 350CF 400CF 450CF 500CF 525CF 550CF 575CF 600CF 625CF 650CF 675CF 700CF 725CF 750CF 775CF 800CF 825CF 850CF 875CF 900CF 925CF 950CF 975CF 1000CF 1025CF 1050CF 1075CF 1100CF 1125CF 1150CF 1175CF 1200CF 175CF 225CF 275CF 325CF 375CF 425CF 475CF 525 550 575 600 625 650 675 700 725 750 775 800 825 850 875 900 925 950 975 1000 1025 1050 1075 1100 1125 1150 1175 1200 Spacecraft Mass (t) at SOI 700 600 500 400 300 200 100 0 (b) 200 0 –200 –400 –600 –800 –1.200 Asteroid Arrival Time (Days Before Earth Impact) Figure 99. (a) Staged chemical/kinetic deflection vehicle mass at Earth departure.7. 133 . Larger impacts will produce larger craters. however.

–4 –3 –2 –1 0 1 2 3 4 6 Earth 5 Earth perihelion 4 Asteroid 3 Asteroid perihelion 2 Asteroid asc. 100). but might be addressed in future work. This option has not been considered under the present study. 134 . It requires four chemical stages to provide the necessary DV for the outbound trajectory (plotted in fig. determination was made that chondrite asteroids up to 400 m in diameter can be successfully deflected. however. This might allow a reduction in the projectile mass.853 kg. node 1 Transfer trajectory 0 Start transfer –1 Finish transfer –2 Asteroid at start –3 Earth at finish –4 Impact position –5 –6 Sun Figure 100.000 m in diameter. the resulting vehicle mass is very high for an asteroid of this size. This mission causes only cratering. As the system is currently conceived. The stage would then perform some small DV maneuver that would allow it to monitor the projectile’s impact with the asteroid and transmit useful data back to Earth. It corresponds to a 1.The minimum mass mission is 11. the asteroid size was again varied to determine its effect on the vehicle mass.025-day total mission duration— defined as the time from Earth launch to possible Earth-asteroid collision—and impacts the asteroid at 800 days before Earth impact. Optimal staged chemical/kinetic deflection mission. the final vehicle stage would jettison the projectile just before impact. it would serve to increase the energy and momentum available at impact. Such data might be needed to plan future impacts on either the same target asteroid or some other similar body in the future. Figure 101 shows the required spacecraft mass to deflect chondrite asteroids of up to 400 m in diameter. or 225 days after the vehicle departs Earth. Figure 102 shows the projectile masses required to deflect asteroids of 50–1. not complete fragmentation. The interceptor mass for this mission is 668 kg and the dry mass of the final lox/LH2 stage is 259 kg. Once the optimal mission was determined. thus making the entire system smaller. As a result of this investigation. If the final stage were to remain attached to the projectile.

Projectile mass versus chondrite asteroid diameter.000 0 0 200 400 600 800 Asteroid Diameter (m) 1.000 1.000 5.000 500 0 0 200 400 600 800 Asteroid Diameter (m) 1.500 3. 135 .000 Spacecraft Mass (t) at SOI 2. Staged chemical/kinetic deflection vehicle mass versus chondrite asteroid diameter.200 Figure 102.000 1.200 Figure 101.000 3.500 1.000 1. 6.000 Projectile Mass (t) 4.500 2.3.000 2.000 1.

Once again. which operates as both the outbound propulsion system and as the deflection device. This is the same approach that was used for the staged chemical/nuclear deflection option (see fig. For the rendezvous case. 90 for an example). Main FixMass PBI inboundNP kg2mT PBO_v3 PBMExtPulseMC Figure 103. Various values of total mission time and rendezvous time were examined to determine the regions in which to concentrate the search for the global minimum total system mass. it passes the asteroid size and type to inboundNP. The optimum trajectory is illustrated in figure 105.3. which is the tool that determines the number of nuclear pulses required to impart the required DV to the asteroid.8. The result of these runs is illustrated in figure 104.170 days. the main script is merely a central location in which to input parameters. the rendezvous and total mission times are 1. The overall layout is presented in figure 103. 136 .4 Nuclear Pulse Rendezvous with the asteroid or comet is required for the nuclear pulse option. which takes the mass from inboundNP. this DV is determined by PBI. As with the other models. Sizing of the nuclear pulse system for the outbound journey is done by PBMExtPulseMC. this mass is the payload that must be carried to the asteroid.200 and 2. respectively. Diagram of the ModelCenter setup for the nuclear pulse option.

Optimum rendezvous trajectory for the nuclear pulse option. Minimum total system mass for the nuclear pulse option.5 –1 –1.5 –3 Transfer trajectory Start transfer Finish transfer Asteroid at start Earth at finish Impact position 3 2.400 2. 137 .5 2 1.000 2.200 –1.500 –200 –500 –800 –1.5 2 2.100 90 80 Spacecraft Mass (t) at SOI Rendezvous Time (Days Before Impact) –100 70 60 50 40 30 20 10 0 2. node 1 0. –3 –2.000 –1. the optimum rendezvous and total mission time varies little with changing asteroid mass.400 Note: Several Other Rendezvous Times not Included for Clarity Min Found at –1.7 t Total Mission Time (Days) Figure 104. The results are plotted in figures 106 and 107.200/2.5 0 0.5 3 Sun Earth Earth perihelion Asteroid Asteroid perihelion Asteroid asc. it is relatively easy to determine the required size of the nuclear pulse system as a function of asteroid size. Therefore.5 –2 –2.100 –1.5 Figure 105. showing the optimum rendezvous and total mission times. As with the staged chemical/nuclear deflection option.5 –1 –0.5 –2 –1.300 2.300 –1.5 1 1.200 2.5 0 –0.100 2.170 Mass = 29.

Given the total system mass constraint of 1. Minimum total system mass for the nuclear pulse option versus chondrite diameter. 45 40 Total System Mass (t) 35 30 25 0 200 400 600 800 1.000 10.000 8. the largest diameter M1999JT6 chondrite that this system can defeat would have a diameter of 9.000 Total System Mass (t) 1.400 1.000 6.000 4.000 m.000-t Limit 800 600 400 200 0 0 2. 138 .000 t. 1.200 1.000 Chondrite Diameter (m) Figure 106.000 Chondrite Diameter (m) Figure 107. Minimum total system mass for the nuclear pulse option versus chondrite diameter for the smaller chondrites.

the minimum total system mass continued to decrease as the total mission time increased. 1 0.4 –0. For the solar collector.500 2.6 Difference in Total System Mass (t) Total Mission Time (days) 1. the analysis instead located the minimum total system mass for two specific total mission times: 3 and 10 yr. the RCS system.2 –0. It seems logical that there must be some combination of rendezvous time and total trip time that would result in the collector being just large enough for both the outbound and inbound journeys. Plot of the difference between required outbound and inbound solar collector sizes. dominates the required solar collector size. Negative values indicate that the inbound portion of the journey. but it does show that for some values of total trip time. rather than seeking a global minimum. since the collector is optimally designed for both stages of the mission. this does not yield an efficient solution. This analysis method helps to illustrate the benefit of very long mission times with this system. Negative values mean that the inbound requirement dominates. 139 . Therefore.2 0 –0. In such cases.000 0.8 0. Picking an arbitrary rendezvous and total trip time usually results in either the outbound journey or the inbound journey dominating the solar collector size requirement.8. Figure 108 shows the difference between inbound and outbound mass requirements for various total trip times.6 –0. there is no payload that the outbound system must deliver. however. the inbound and outbound systems are the same. other than some avionics components. which is the asteroid deflection portion. this plot does not include all of the total mission times that were examined.3. and some other minor components.8 –1 –500 –400 –300 –200 –100 0 Rendezvous Time (Days Before Impact) Figure 108. In fact. the total system for one of the solutions was always significantly less than the others.4 0.000 2. these times were not exactly 3 and 10 yr. the inbound and outbound solar collector sizes can match at more than one rendezvous time. but were allowed to fluctuate by up to 100 days or so about the nominal values. Also. unlike the other systems.000 1. In this case. For clarity.500 3. These points should yield the minimum total system mass.5 Solar Collector Operation of the solar collector system was analyzed in a somewhat different manner than were the other systems.

based on the rendezvous time. A plot of these locations for various total mission times. such as figure 109.000 –250 –300 –350 –400 500 –450 –500 0 1. Analyses far beyond the 10-yr limit for total mission time resulted in a continued decrease of the minimum total system mass. In those instances where this is not the case. The relatively simple ModelCenter model is presented in figure 110. All tools in the figure have been described in previous sections. the total system mass is within a few kilograms of the minimum value. The plot also shows that the local minimum total system mass continues decreasing as the total mission time increases. acceleration of the asteroid. and PBO_v3 determines the outbound trajectory requirement based on the outbound time. but rather to determine the minimum mass for two total mission times.000 0 1. PBI determines the asteroid deflection requirement. The solar collector tool gives both the inbound and outbound required system masses as output. the minimum required system mass. dimensions of the solar collector. the objective of this analysis was not to find the global minimum total system mass. As stated earlier. and the associated total system mass. which is the larger of the inbound and outbound requirements. As with the other models.The points at which the inbound and outbound size requirements are equal nearly always correspond to the points at which the total system mass is a minimum for a specific total mission time.000 Figure 109.000 4. and some additional data.000 2. which is the solar collector sizing tool. for 500-m-diameter chondrite. 140 . except for ssc. force on the asteroid. these two total mission times are around 10 and 3 yr—with some slight fluctuation to allow the ModelCenter optimizer to find the local minima. 0 –50 3. However.500 Rendezvous Time (Days Before Impact) Rendezvous 3. this is much like the other systems considered in this study. Combinations of total mission time and rendezvous time where inbound and outbound required solar sail sizes are equal.000 –100 –150 –200 System Mass 2.500 1.500 Total System Mass (t) 2.000 Total Mission Time (Days) 3. reveals the cyclic nature of the total system mass.

5 au and assumed to remain constant. This constant value.. with a total mission time of 3. and asteroid sizes.5 days before impact.Main PBI ssc PBO–v3 Figure 110. coupled with the inbound flight time.636 days. a fairly large performance margin would have to be built into the system. a much larger than optimal solar collector would be required. the minimum for a 1. For the staged chemical/nuclear pulse options. rendezvous times. The force imparted on the asteroid by the solar collector was determined at a heliocentric distance of 1. Since inbound and outbound times are determined by asteroid mass. Since PBI and PBO_v3 both determine impulsive DV requirements.711 days. Alternatively. Analysis of the solar collector option commenced near the end of the study. This is not the case with the solar collector option: the optimum points vary with asteroid size and the rendezvous time changes considerably. the optimum total mission and rendezvous times were very insensitive to asteroid size. the minimum total system mass to deflect a 100-m-diameter chondrite occurs at a rendezvous time of 112.e. Diagram of the ModelCenter setup for the solar collector option. 141 . For the rendezvous case. determined the inbound DV requirement. but it does highlight the trends in system performance for various total mission times. With limited time available for additional tool development. i. For example. the optimum trajectory for the 100-m-diameter chondrite is illustrated in figure 111. This allowed the determination of the required solar collector size. with a total mission time of 3. This analysis method does not accurately determine the required solar collector size. provided the DV requirement for asteroid deflection was available. the use of these values for analysis of the solar collector tool required the careful application of some conservative estimates. But. PBI and PBO_v3 were used to determine total DV requirements.000-m-diameter chondrite occurs at a rendezvous time of 824 days before impact. the size and type of an incoming asteroid would have to be determined quite accurately before the solar collector system was ever launched from Earth.

5 –2 –1.000 Asteroid Diameter (m) 1.000 2.000 10. 180.5 3 3 2. 142 .000 6.5 0 Sun Earth Earth perihelion Asteroid Asteroid perihelion Asteroid asc.5 –3 Figure 111.000 40.000 4. These results are plotted in figures 112 and 113.5 0 0.5 2 2.–3 –2.000 160.5 1 1.000 Total System Mass (kg) at Earth SOI 7. Optimum rendezvous trajectory for the solar collector option for a 100-m-diameter chondrite.000 4. node Transfer trajectory Start transfer Finish transfer Asteroid at start Earth at finish Impact position –0.000 Mass (≈3 yr) Mass (≈10 yr) Sail Size (≈10 yr) Sail Size (≈3 yr) 6.5 2 1.000 80.000 0 Figure 112. After extensive analysis.5 –2 –2.000 100.000 60.5 –1 –0.000 140.000 Solar Collector Diameter (m) 5.000 120.000 20. Minimum total system mass and size for the solar collector option versus chondrite diameter.5 –1 –1.000 3.5 1 0.000 0 0 2. the required total system mass versus asteroid size for a chondrite was determined for two cases: total mission time of ª3 yr and total mission time of ª10 yr.000 8.

400 1. rather than total system mass.750 2. Unfurling and controlling a collector of this size presents significant technical challenges.000 1.250 1.000 1. is the limiting factor for this option. the diameter of the solar collector is a remarkable 6 km.000-t limit imposed on the system mass.800 Mass (≈3 yr) Mass (≈10 yr) Sail Size (≈10 yr) Sail Size (≈3 yr) 1.2. 143 . Even for a 10-km-diameter chondrite.600 1. the total system mass is well within the 1.500 1. Solar collector size.200 1. Despite these problems.000 Asteroid Diameter (m) Figure 113. Minimum total system mass and size for the solar collector option versus chondrite diameter for the smaller chondrites. However.000 800 600 400 200 0 0 250 Solar Collector Diameter (m) 700 600 500 400 300 200 100 0 500 750 1. it is clear that the solar collector could still be effective in deflecting an incoming asteroid.000 900 800 Total System Mass (kg) at Earth SOI 1.

such collisions might take place perhaps only once or twice per century. Compared with other comparable threats. 9. conclusions and recommendations are given in sections 9.9. particularly those relating to acts of terrorism. It is interesting to note that the probability of dying due to a planetary body impact is about the same as that of dying due to an aircraft crash. say. a decade or a century. and a recovery time possibly extending over decades. 144 . Only when in full possession of the facts can the voting public make an informed decision about what steps should be taken. wide-scale destruction.2 Statistical Problem The lack of attention given to this threat is in part due to a statistical problem. 9.1 Public Awareness Despite the best efforts of Hollywood. must be considered. some appropriate parameters. Several possible combinations of mitigation techniques and transportation options have been analyzed in detail. proper account must also be taken of the likely consequences of such a collision. The chance of a significant-sized object striking the Earth is fairly low.1 through 9. However. To obtain proper assessment of the danger. While not advocating steps that could lead to hysteria and panic. This has led to the danger being downgraded when compared with other threats to public safety. the probability of an impact taking place cannot be considered in isolation.6. As table 29 shows. planetary body collision is still viewed as being a matter of science fiction rather than one of scientific fact. Consideration was also given to a range of transportation methods by which the mitigation hardware could be moved out to an approaching body for either rendezvous or impact. Even the impact of a relatively small body would probably be very severe with fatalities in the millions. the facts about this problem should be properly presented to the general public so as to raise public understanding of the threat and the ways in which it can be mitigated. the level of public awareness of this threat is still not high. CONCLUSIONS AND RECOMMENDATIONS A wide range of potential mitigation techniques by which threatening planetary bodies could be either deflected or fragmented have been modeled in this study. this approach presents an altogether more worrying perspective on the danger. Although there is much that still needs to be done. such as the expected number of fatalities over a period of. Table 29 shows the chances of death by a variety of causes for a typical resident of the United States.

A strong recommendation was made that funding for these efforts be increased. cost. as well as those of the DOD. research into new mitigation techniques is nonexistent. and other government agencies. Although funding is limited for NEO surveys and searches. which would represent an enlarged follow-on to this work. the Armed Forces.000 1 in 20. the study of mitigation techniques is—with the notable exception of this present effort—almost totally unfunded. most are proceeding with very limited funding. resident. as recounted. e.Table 29. it is not beyond our capabilities. actually scorned the detection efforts. sufficient high-quality instruments must be made available to conduct an all-latitude observation program with the aim of cataloging the entire NEO population. provided preparations are begun well in advance of an impact. Equally important.000 1 in 10. By comparison. In some countries.000 1 in 60.000 1 in 1.g.45 Cause of Death Motor vehicle accident Homicide Fire Firearms accident Electrocution Asteroid/comet impact Passenger aircraft crash Flood Tornado Venomous bite or sting Fireworks accident Food poisoning by botulism Drinking water with EPA limit of tricholoethylene Chance 1 in 100 1 in 300 1 in 800 1 in 2. A strong recommendation is made that a coherent study of mitigation techniques as well as their likely effectiveness.500 1 in 5.000 1 in 20.3 Funding of Future Work While a number of NEO search activities are currently underway.000. Causes of death and associated probabilities for a U.S.000 9. Despite the impression given by Hollywood.4 Development and Deployment of Mitigation Systems The technical work undertaken in this study shows clearly that.000. International collaboration and funding should also be actively sought. government agencies have declined to provide funding and.000 1 in 100. be undertaken in the very near future. 9. This study.000 1 in 3.000 1 in 30. although the mitigation challenge is formidable. etc. and deployment times. it is not practical to wait until a specific threat is identified 145 .. except in those cases where the technology is under study for some other application. In particular.000. the National Oceanic and Atmospheric Agency. the Federal Emergency Management Agency. at least it is nonzero. should involve and call upon all of NASA’s considerable resources.

after a deflection DV has been applied to the object.5 Accomplishments It was not the intent of this study to select a particular technical option for recommendation as a threat mitigation system. if not global. it was said that. it should be able to defeat those most likely to occur. the inbound tool takes a velocity vector at the point of impact and integrates the trajectory backward in time until the object is well outside the Earth’s sphere of influence. The data and tools available in the literature have been built on to create a threat assessment tool that calculates the percentage of the total threat that can be defeated using a given mitigation system. Instead. In the same way. It was also our intent to recommend future work. It then integrates forward. technology development. Most strongly recommended is that. These tools yielded first approximations for the performance and mass of each technical option. 9. Similarly. will take several years. Systems engineering. system deployment. A procedure for comparing all these technologies will be put into place in the future. a development program be initiated immediately. It was also our aim to categorize these options into different mission configurations and to propose a method for comparing the large number of possible combinations of mitigation options and mission configurations. At the outset of Project Apollo. following an appropriate study phase. our intent was to study the various options. failure to act will. using high thrust calculation methods. it might be said that. so as to determine the resulting miss distance from the Earth. None should be regarded as a finished product and all would benefit from further development and refinement. techniques. Several new tools were created during the course of this project. The program iterates until a specified closest approach to the Earth has been achieved. 146 . although more calculationally intensive. while the United States could not guarantee to come first in the race to the Moon. the outbound trajectory tool is designed to give a first approximation of the required DV. in the long run. however.before starting work on the mitigation system. while success in protecting the Earth against a cosmic impact cannot be guaranteed. guarantee a major catastrophe of regional. with a view to deploying an operational system as soon as possible. and in some cases. Both these tools would benefit from the use of more accurate. failure to act would guarantee that she would come last. in several cases. identification of mission categories for these technologies have been made and future analyses simplified by developing a procedure that deals with each category instead of attempting to deal with each individual technology combination separately. proportions. As an example. In so doing. Numerous outbound propulsion systems and threat mitigation options were considered and modeled using several tools that were created by combining some basic principles of physics with engineering data available in the open literature. It is already clear to us that a first-generation protection system will not be able to counter all possible threats. using improved and updated modeling techniques.

000/1.636 1. the relatively massive mass driver system coupled with the least efficient storedchemical system yielded unacceptably high initial masses. Table 30 summarizes the capability of each major system option. § The solar collector system is limited more by solar collector size than by total system mass. The mass driver could have been coupled with the solar sail/collector but would result in sail sides/diameters in the tens of kilometers. Table 30. ** Times are for 100-m-diameter chondrite. a preliminary assessment is possible. Rendezvous times are greater for larger asteroids. The mass driver was coupled with the staged chemical system to offer a non-nuclear threat mitigation option.7 0.076/65** 3. the combination of a nuclear blast system and a staged chemical outbound propulsion system also offered excellent performance.200 1.000 260/1.800 NA Rendezvous Rendezvous (≈3 yr) Rendezvous (≈10 yr) 2.6 9. Total System Mass at SOI (t) for Different Time Before Impact (days)/Outbound System Staged chemical/ mass driver Staged chemical/ nuclear deflection Staged chemical/ kinetic deflection Nuclear pulse Solar collector Maneuver Rendezvous Travel Time (days) 2.635/115** 29.000/1.075/132 1.9.000 Intercept Rendezvous Intercept 1.849 80/6.8 1.025/800 0.847 5. Thus. the mass driver is an attractive option for moving asteroids with the ultimate purpose of resource utilization.637 0. it was the staged chemical propulsion system that limited the system performance.000 NA 10.400 100 NA Asteroid Diameters (m) 1. this concept should be carried forward in further studies. although total missions times change little.6 Assessment of Mitigation Options Although it was not the purpose of this study to select mitigation options.000 § § * Maximum was constrained to a total system mass at Earth SOI of 1. See figure 112.000 NA Maximum Diameter of Asteroid*(m)/ Total System Mass at Earth SOI (t) 50/6.000/1. Summation of parametric results for mitigation concepts.62 73.240 167 34. perhaps with effort expended to reduce the required mass driver system mass. Once again.07 0.000 t. By comparison. However.550 41. However.8 8.27 568 NA 1.000 1.300 87.509/910 1.900/2. 147 .918 9.170/1.

The nuclear pulse option performs well because of its use of the same. This is by no means a recommendation but merely an observation based on the data at hand. is theoretically capable of deflecting large asteroidal bodies. perhaps at a Lagrange point. Each of these options may well find some application in the future. Nonetheless. Our threat assessment tool also requires further research into the consequences of an impact. may have nonuniform densities along their circumference. There are several other mitigation options that were not studied because of resource and time limitations. as well as other options suggested elsewhere. Many of the assumptions made during the development of our technology tools are in need of refinement. note that there are suggestions in the literature that cometary rings. However. two of these are of particular interest and merit some mention: (1) Laser ablation is used as either a remote station or as a rendezvous option. be carried forward into a higher fidelity analysis. 9. nuclear pulse offers the best prospect of providing an effective mitigation technique using existing and near-term technology. it has the advantage of using a single unified system for both outbound propulsion and deflection. but only at the expense of a very large sail area and the consequent operational problems. but our initial results indicate that the nuclear pulse option offers the best defensive capabilities in the near term. Our trajectory tools would benefit greatly from the ability to model continuous thrust propulsion systems. Of all the options considered during this study. Due to the level of fidelity and extensive assumptions that have been forced to be made in this limited study. (2) This second new option involves firing inert masses from a mass driver located in Earth orbit. This would combine our mass driver and kinetic deflection/fragmentation tools and would represent another remote station option. As a minor example of this. very effective technique for both outbound travel and deflection. there would be a synodic period on which the Earth would cross these higher density areas. it is recommend that all options discussed here. As with the nuclear pulse option.A kinetic deflection vehicle. such as the Leonid ring. The solar collector system showed itself to be capable. Since the Earth passes through such rings on a yearly basis. 148 . All of our tools would benefit from more detailed analysis procedures.7 Future Work A large amount of future work has been identified. deflection of a 100-m-diameter asteroid is possible. the interceptor vehicle mass required increases rapidly with asteroid size. This technique would allow deflection in a manner similar to that of the solar collector with a beam of high-energy coherent light being directed at the incoming asteroid or comet. yielding a higher probability of impact. Our threat assessment tool requires more research into the available data on the asteroid and comet population. carried out by a staged chemical system.

After completion of the more advanced tools above. It is well within humanity’s ability to effectively defend itself against this threat. 9. 149 .Finally. including the atmospheric reentry tool described earlier for fragmentation options. a method to combine the quantitative results from this analysis needs to be established and the qualitative issues outlined for each technology in the outbound propulsion and threat mitigation sections in order to compare architectures. Development of the necessary technologies would also offer considerable synergy with NASA’s other missions aimed at understanding the universe and exploring space.8 Summary Conclusion The threat posed by NEOs should be taken very seriously. the overall threat assessment flow chart could begin. The goal is to persuade those in positions of authority to continue the efforts presented here. The planetary defense mission is also one for which NASA is uniquely suited and could potentially offer the Agency a goal that both fires the public imagination and creates a sense of urgency comparable to that during the Apollo program in the 1960’s. as described in figure 79.

S. and Earth-approaching asteroid populations. and astrometric and physical studies. observation. and orbital calculation of NEOs. Its primary goal is to explore the various populations of small objects in the solar system. with particular attention to NEOs. It is also intended to promote and coordinate a ground network—Spaceguard system—backed up by a satellite network for discovery. observational. SpaceWatch is a pioneer in the use of CCDs and automation for asteroid and comet detection. and experimental levels of the physical and mineralogical characteristics of the minor bodies of the solar system. to the problem of detecting and cataloging NEOs that threaten the Earth. to promote and coordinate activities for the discovery. The principal instruments. 150 . Wide ranges of instruments are in use. A. located worldwide. located on Kitt Peak.8-m telescope. CCD systems scan the Centaur.9-m SpaceWatch telescope and the SpaceWatch 1.3 Lincoln Near-Earth Asteroid Research Lincoln Near-Earth Asteroid Research (LINEAR) is a Massachusetts Institute of Technology Lincoln Laboratory program funded by the U. Trojan. are the Steward Observatory 0.2 Spaceguard Spaceguard is an international association. Air Force and NASA. Note that Spaceguard is a coordinating body and that the technology available and effort expended vary widely between the various participating observatories. monitoring. and to study statistical data for asteroids and comets so as to understand the dynamical evolution of the solar system. Main Belt. SpaceWatch currently has the distinction of having detected the smallest known asteroid— 1993 KA2. The Spaceguard system is a collection of observatories engaged in NEO observations. It is intended to promote study activities at theoretical. Its goal is to demonstrate the application of technology. There are currently more than 70 observatories registered. SpaceWatch continues to detect some 20 to 30 new NEAs per year.APPENDIX A—CURRENT NEAR-EARTH OBJECT SEARCH PROGRAMS A. A. established in 1996. which approached to a distance of ª105.1 SpaceWatch The SpaceWatch program is run by The University of Arizona’s Lunar and Planetary Laboratory. established in 1980.000 km. originally developed for the surveillance of Earth-orbiting satellites. which is about 4 to 9 m in diameter—and has also observed the closest known approach of an asteroid to the Earth—1994 XM1. Trans-Neptunian.

Equipment consists of a pair of 1-m-diameter, ground-based, electro-optical deep-space surveillance (GEODSS) telescopes at Lincoln Laboratory’s Experimental Test Site on the White Sands Missile Range, Socorro, NM. The telescopes are equipped with Lincoln Laboratory-developed CCD electrooptical detectors and collected data are processed on site to generate observations. Survey results as of April 2002 are as follows: Number of observations to minor planet center Number of asteroid detections Number of new designations Number of confirmed NEOs Number of confirmed comets 7,416,832 1,127,759 157,920 951 82

A.4 Near-Earth Asteroid Tracking Near-Earth asteroid tracking (NEAT) observatory is an autonomous celestial observatory developed by JPL and funded by NASA to study asteroids and comets. It is based upon a specially designed CCD camera. The principal investigator is Dr. Eleanor F. Helin; co-investigators are Dr. Steven H. Pravdo and Dr. David Rabinowitz. NEAT is comprised of two autonomous observing systems at the Maui Space Surveillance Site (MSSS)—NEAT/MSSS and at the Palomar Observatory—NEAT/Palomar. At both sites, the NEAT cameras use 1.2-m (48-in) telescopes to find NEOs, NEAs, and comets. Nine new NEAs were discovered during July 2002: three Amors with one >1 km; five Apollos, including one PHA and one >1 km, and one Aten. A.5 Lowell Observatory Near-Earth Object Search The Lowell Observatory near-Earth object search (LONEOS) system can scan the entire sky every month, accessible from Flagstaff, AZ. It uses a 0.6-m Schmidt telescope and a CCD detector. It has been in operation since March 1998. The first new discovery was made on June 18, 1998, and is able to record objects to a magnitude limit near V=19.3, or ª100,000 times fainter than can be seen with the naked eye. As of August 2001, LONEOS had submitted more than 1 million asteroid observations to the International Astronomical Union Minor Planet Center. It is estimated that, after 10 yr of full-time operation, LONEOS could discover 500 of the 1 km or larger NEOs and perhaps twice as many smaller NEOs, thus substantially increasing our knowledge of these bodies. The asteroid discovery summary as of July 9, 2002, includes 10 Aten, 56 Apollo, 55 Amor, and 6 comets, for a total of 137 asteroids.

151

APPENDIX B—SOLAR ARRAY CALCULATIONS

The solar array forms an important part of the proposed mass driver system. The large distances anticipated between target asteroids and the Sun, coupled with the generally unfavorable incidence conditions, threaten to make the array one of the most massive system elements. Two distinct methods of sizing the array have been identified. B.1 Method I The electrical power (P) can be written as
P = FAε ,

(89)

where F = solar flux (W/m2) A = array area (m2) ε = efficiency. If σ = mass per unit area, then

A=
where M = array mass. Thus,

M , σ

(90)

σ=
and hence,

FMε , P

(91)

σ=

Fε .  P    M

(92)

For Earth orbit, F = 1,300 W/m2. For current generation arrays, ε = 0.15 is achievable, and can be bettered. Reference 46 (p. 333) gives (P/M) = 14 to 47 W/kg. This implies σ = 4 to 14.

152

B.2 Method II47 Achievable specific power = 130 W/kg (assumed achievable at Earth with ideal array orientation with respect to the Sun). Let RE = Earth’s orbital radius and RA = asteroid orbital radius. Hence, achievable specific power at asteroid = 130¥(RE/RA)2 W/kg. Again, it is assumed that the array is ideally oriented with respect to the Sun. Introducing an additional degradation factor (a) to take account of (1) nonideal orientation, (2) possible asteroid rotation, (3) dust obscuration of array, …, gives ÊR ˆ Specific power = 130a Á E ˜ Ë RA ¯
2

ÈW ˘ Í kg ˙ . Î ˚

(93)

153

APPENDIX C—MASS DRIVER

Appendix C contains details of a simple model for the key components of the mass driver system. C.1 Model of the Forces on a Bucket Coil Due to the Nearby Drive Coils As explained in section 5.7, although a bucket coil receives an accelerating force from the closest pair of drive coils, it experiences alternately retarding and attractive forces from each more distant pair (fig. 114). Of course, the more distant pairs of coils produce lower forces than do the nearer coils. For the force calculations in this analysis, only the effect of the first two pairs of coils will be considered. The nearest pair provides the major motive force; the next pair provides the major retarding force. By limiting consideration to these four coils, one essentially conducts a conservative analysis. This is because all subsequent drive coils—in theory, stretching out to infinity in both directions—can also be grouped into sets of four coils; the nearest and next-nearest sets are shown below.
Nearest Set of Four Coils— Included in Analysis

Direction of Motion

Next Nearest Set of Four Coils— Not Included in Analysis

Figure 114. Drive coils included in and omitted from the analysis. If one examines the coils which comprise the next-nearest set of four; i.e., the first coils to be excluded in the analysis, it is immediately apparent that the nearer pair exerts a motive force and the slightly more distant pair exerts a retarding force on the bucket coil. Hence, the net force from this set of four coils will be a motive one. The same argument can be applied to all subsequent; i.e., more distant, sets of four coils. Each produces a net motive force on the bucket coil, although declining in magnitude as distance increases. Hence, by their exclusion, the analysis neglects a portion of the overall motive force.

154

The analysis is conducted by considering the motion of a single bucket coil between two adjacent drive coils. ∂x F = N1N 2 I1I2 (94) where I1. The results can be multiplied to also include the effect of the remaining three-bucket coils.k E ˙ . Drive coils included in the analysis. Equation (94) can be derived from basic magnetic energy considerations for two interacting currentcarrying coils.2 = current in coil 1. Drive Coil n–1 Drive Coil Drive Coil n n+1 Bucket Coil Drive Coil n+2 x Direction of Motion S S S Figure 115. The general expression for M can be shown46 to be as follows: M ( x) = m 0 ( a1a 2 ) where 0. The force (F) between any two coils (designated as 1 and 2) is given by ∂M . Appendix D contains the derivation.2 N1.2 = number of turns in coil 1.5 ÈÊ 2 2 ˘ ˆ ÍÁ k .2 M = mutual inductance between one turn of coil 1 and one turn of coil 2.k˜ K . Figure 115 shows the bucket coil and surrounding four drive coils. Ë ¯ Î ˚ (95) 155 .

Note the minus signs in front of the n + 2d and n – 1st coils. ki   ki  (99) 156 .5 (96) and ϕ= π 2 E ( k ) = ∫ dϕ 1 − k 2 sin 2 ϕ ϕ =0 ( ) 0. The quantities K and E are. which provide retarding forces. n + 2). respectively. The mutual inductances are given by 0. elliptical integrals of the first and second kind and are given by: K (k ) = ∫ ϕ= π 2 ϕ = 0 1 − k 2 sin 2 ϕ ( dϕ ) 0. ∂x (98) ID = current in drive coil = current in bucket coil IB ND = number of turns per drive coil NB = number of turns per bucket coil Mi(x) = mutual inductance between one turn of the bucket coil and one turn of drive coil i (i = n – 1. … . which provide motive forces.c = speed of light a1.5  M i ( x ) = µ 0 ( aD a B )   2 2  − ki  K ( ki ) − E ( ki )  .2 k2 = 4×a1×a2/[x2+(a1+a2)2] x = distance between the two coils µ0 = permeability of free space. for the case under consideration here.2 = radius of coil 1. (97) Hence.5 . and the plus signs in front of the nth and n + 1st coils. the net force acting on the bucket coil due to the four drive coils is given by F = N D N B ID IB where ∂ (− M n + 2 + M n +1 + M n − M n −1) .

6 0. the two elliptical integrals. respectively). 4 3.5 3 2. K(k) and E(k). Both curves were fit to sixth-order polynomial equations.2 Distance From Bucket Coil to Drive Coil (m) Figure 116. is equal to k2. The resulting curves are shown in figure 116 (denoted by Poly(K(m)) and Poly(E(m)). (E(m)) Function 2 1. To facilitate calculational procedures.4 0.2 0. The curve fit equations are displayed in equations (100) and (101).8 1 1. have both been curve fitted. which in terms of the quantities given above. superimposed upon the raw data. (K(m)) Poly.5 K(m) E(m) Poly. Note that the independent variable in the graph is denoted by m.5 1 0.5 0 0 0. Elliptical function curve fits—K(m) and E(m) versus m. This was done using raw data48 accurate to the fourth decimal place.where aD = radius of drive coil aB = radius of bucket coil kn–12 = 4¥aD¥aB/[(S+x)2 + (aD + aB)2] kn2 = 4¥aD¥aB/[(x +(aD+aB)2] kn+12 = 4¥aD¥aB/[(S–x)2 + (aD+aB)2] kn+22 = 4¥aD¥aB/[({2¥S}–x)2 + (aD+aB)2]. 157 .

1. When it first enters the mass driver. the equations are as follows: 6 5 4 3 K k 2 = 86.2 Drive and Bucket Coil Currents Although the force equation is fairly straightforward.8625(k 2 ) 2 -0.N B M14 I0 .˜ .4009( k 2 ) + 1.9415(k 2 ) .1863 k 2 ( ) (101) C.7493 k 2 + 102.3540( k 2 ) + 1.Replacing m with k2. The guiding principle is that the net flux through a bucket coil remains constant.7180 k 2 .6024 6 5 ( ) ( ) (100) E k 2 = -1. where L = self-inductance of a bucket coil M1j = mutual inductance between bucket coil 1 and bucket coil j (where j = 2.9341 k 2 ( ) ( ) + 4. The drive current can be written as ID = ID 0 sin px . the current will vary according to the flux supplied by the drive coils.4) = initial current in a bucket coil (before any drive coils are energized).226.4. 4¯ Ë ro (103) (104) 158 . 4 3 + 2. S (102) where x is the distance between the nth drive coil and the bucket coil and ID0 is the maximum current in the drive coil.N B M12 I0 + N B M13I0 .3.0258( k 2 ) . Although this relatively simple functional dependence can be assumed for ID.0. it is assumed that (IB/ID) < 1. before any of the drive coils are energized. the calculations are complicated somewhat by the variations in both drive and bucket coil currents. In what follows. the bucket current (IB) is more complex. I0 The self-inductance of a bucket coil is given by:49 Ê 8a 7ˆ L = m0 aB Áln B .3862 k 2 + 223. As the bucket carries no power source. the total flux (F1) through the front bucket coil is given by F1 = N B LI0 .5701 .5435( k 2 ) .6237 k 2 ( ) ( ) ( ) 2 +22.

where ro = conductor wire radius. (106) 159 . Note that the signs of alternate mutual inductances are reversed to account for the current flow directions in alternate bucket coils.k ji ˜ K kij E kij ˙ . as shown in figure 117. The mutual inductances between bucket coils are given by ˘ ÈÊ 2 ˆ 2 Mij ( x) = m 0a B ÍÁ . The expression for total flux can be simplified to F1 = N B L1I0 . Coil Numbers 4 3 2 1 Alternating Current Flow Directions in Bucket Coils Figure 117. kij ˙ ¯ ÍË kij ˚ Î ( ) ( ) (105) where kii+12 = kii–12 = 4¥aB2/[S2+(2¥aB2)] kii+22 = kii–22 = 4¥aB2/[(2¥S2)+(2¥aB2)] kii+32 = kii–32 = 4¥aB2/[(3¥S2)+(2¥aB2)] and the K and E functions are the same as those given earlier. Bucket coil current directions.

will both have the same inductance as will coils 2 and 3. Inspection shows that the differences are not large and that. the middle coils. L1 = L4 and L2 = L3. (108) Drive Coil n –2 Drive Coil n –1 Drive Coil Drive Coil n n +1 Bucket Coil Drive Coil n +2 Drive Coil n+3 x Direction of Motion S S S S S Figure 118. The drive coils that are energized in the same sense as the bucket coil will add to the flux. given by L1 = L + M12 + M13 + M14 . n+1.where L1 is the total inductance of bucket coil 1.. the two end coils. the total flux through the coil is still F1. the average value is used. Note the signs before each of the mutual inductances. The more centrally located coils will have higher values than those on the ends. When the bucket undergoes acceleration (fig. typically. which only took into account the n–1. The n–2 drive coil is not included because it serves to increase the bucket coil current. In other words. Unlike the motive force equation.e. n. but this is now given in terms of the drive coil current and various mutual inductances between drive and bucket coils by F1 = N B L1IB + N D ID ( M n + 3 . 118). This is in keeping with the conservative intent throughout this analysis.M n + M n -1 ) . The lower current means that the motive force is reduced. Bucket and drive coil current directions during acceleration. the effect of the n+3 coil is also included. This is because it adds to the flux through the bucket coil and hence diminishes its current. but L1 π L2. i. L1 through L4 are not all identical. (107) Inspection of figure 117 shows immediately that the four coils will not all have the same total inductance. Those energized in the opposite sense will subtract from it. Coils 1 and 4. 160 . To avoid having to develop separate induction models for the end and central coils for the purpose of this model.M n + 2 + M n +1 . all four coils are within ª5 percent of the average induction. and n+2 drive coils.

A numerical integration. (112) where ID ( x ) = ID 0 sin πx S (113) and IB ( x ) = I0 − N D ID M n + 3 ( x ) − M n + 2 ( x ) + M n +1 ( x ) − M n ( x ) + M n −1 ( x ) . C.3 Analysis of Bucket Kinetic Energy and Acceleration The total mechanical work done on the bucket coil as it moves from the nth to the n+1st drive coil. is given by x=S ∆E = ∫ Fdx . As explained earlier. IB. x=0 (110) Although there are now expressions for ID.e.. N B L1 (114) 161 . (111) where δKE is the incremental increase in bucket kinetic energy between x and x + δx. is conducted instead: δKE = Fδx = N D N B ID IBδ (− M n + 2 + M n +1 + M n − M n −1 ) . at any instant.Eliminating Φ1 gives IB = I0 − N D ID M n + 3 − M n + 2 + M n +1 − M n + M n −1 N B L1 (109) as the bucket coil current. all four-bucket coils have current of the same magnitude. analytical integration of the force equation is impractical. i. using the following simple difference equation. from x=0 to x=S. and the various mutual inductances. This can be rewritten as δKE ( x → x + δx ) = N D N B ID ( x ) IB ( x ) {− [ M n + 2 ( x + δx ) − M n + 2 ( x )] +[ M n +1 ( x + δx ) − M n +1 ( x )] + [ M n ( x + δx ) − M n ( x )] − [ M n −1 ( x + δx ) − M n −1 ( x ) ] } . but with flow direction alternating between successive coils.

say at least 20. the average acceleration can be treated as being constant. when there are no drive coils behind the bucket. dt 2 S and thus 1 DKE / M B dv = . Note that although the bucket acceleration is not constant as it moves between two adjacent drive coils. as v ª 2 N DC DKE . the above expression should suffice. This means that DKE is independent of the bucket speed. then between the ith and i+1th drive coils.These three equations can be integrated in a stepwise fashion to determine the total kinetic energy increase that the bucket experiences due to a single one of its coils as it traverses from x=0 to x=S. MB (115) where NDC is the total number of drive coils. to maximize at approximately x=S/2. This can be seen as follows: 2 dv d v / 2 dv dx dv = =v = . 2 dt S (117) (118) which is a constant for all i. dv 1 v2 . For a design such as this. Hence. This is then multiplied by 4 to account for all the bucket coils. However. the force profile climbs from zero at x=0. and then subsequently declines. Note that the model is completely independent of the bucket speed. Similarly.v2 = i +1 i . thus giving the total increase in bucket kinetic energy. denoted by DKE. one can write the final bucket speed (v). it does not apply at the very end of the acceleration portion when there are no more accelerating coils ahead of the bucket. at the end of the accelerating portion of the mass driver. 162 . dx dx dt dt dx ( ) (116) where x is the distance along the mass driver traveled by the bucket. with equally spaced drive (and bucket) coils. this means that the same increase in kinetic energy is experienced between each pair of drive coils. Thus. The approximate equality symbol is used in the equation because the model employed here clearly does not apply at the very start of the mass driver. provided there is a sufficiently large number of drive coils. The average acceleration between any two adjacent drive coils is the same all the way along the coil gun if one neglects end effects. If one denotes the mean acceleration between two drive coils by ·dv/dtÒ.

Drive Coil Current Before Bucket Arrives Current = 0 Current at Maximum Positive Value for Transit of Bucket Coil No. instead.4 Drive Coils Drive coil current has been expressed as a function of the distance (x) traveled by the bucket between two drive coils: ID ( x ) = ID 0 sin px . If the effects of resistance are neglected. 163 . Ideal current versus time profile for a single drive coil. assume a simple oscillating inductance-capacitance-resistance (LCR) circuit. the current must be expressed as a function of time. of course. In practice. 4 Figure 119. 3 After Bucket Leaves Current=0 Time Current at Maximum Positive Value for Transit of Bucket Coil No.C. the ideal sinusoidal current versus time profile is as shown in figure 120. the ideal current versus time profile would be as depicted in figure 119. 2 Current at Maximum Positive Value for Transit of Bucket Coil No. a square profile will probably not be achievable and so. By following a square profile. S (119) In practice. 1 Current at Maximum Positive Value for Transit of Bucket Coil No. If one considers a single drive coil. the drive coil is always at the maximum value to attract or repel a bucket coil.

The capacitor is charged from a DC power source (the switch is shown in fig. 164 . Sinusoidal current versus time profile for a single drive coil. 4 Figure 120. as shown in figure 122. Oscillating drive circuit—charging capacitor. 121 in the charging position) and is then discharged to produce the required coil current.Drive Coil Current Before Bucket Arrives Current = 0 Positive Current for Transit of Bucket Coil No. this type of current profile can be produced relatively easily by incorporating the drive coil into the type of circuit shown in figure 121. Oscillating drive circuit—discharging capacitor. 2 Negative Current for Transit of Bucket Coil No. Switch Drive Coil Capacitor DC Power Figure 121. DC Power Figure 122. In theory. 1 Positive Current for Transit of Bucket Coil No. 3 After Bucket Leaves Current=0 Time Negative Current for Transit of Bucket Coil No.

to prevent this from happening. This frequency-specific aspect of the drive coil design has two implications: (1) A given drive coil interacts with each of the four bucket coils in succession. 165 . although adequate for this simple model.The LCR of the circuit are selected to give the required oscillation frequency. As a coil’s frequency is set by the time taken for the bucket to travel the standard intercoil distance (S). Its drive circuit should ideally be able to change its frequency slightly as each bucket coil passes. as shown in figure 123. This is because. which will be different for each drive coil—earlier coils having a lower frequency than later coils. 1 Positive Current for Transit of Bucket Coil No. the earlier coils will have a lower frequency than the later coils. 4 Figure 123. in principle. with multiple capacitors being discharged in a carefully timed sequence as the bucket coil passes. Real current versus time profile for a single drive coil. Resistance and other energy-loss mechanisms will result in a decaying current profile. Following the transit of one bucket. Drive coil circuit oscillation frequencies will vary along the length of the mass driver. the above drive coil circuit will need to be modified for practical use. the drive current’s sinusoidal profile for a simple circuit will be disrupted by the intrusion of bucket coil flux. 2 Negative Current for Transit of Bucket Coil No. This means that each coil’s circuit must be designed to produce the frequency (n) that is appropriate for the bucket’s speed when it passes the coil: v n= i . The possibility of modifying drive circuit design to facilitate this should be investigated. necessitating a slightly higher frequency. In practice. S (120) where vi is the mean speed of the bucket as it traverses the distance 2¥S. Drive Coil Current Before Bucket Arrives Current = 0 Positive Current for Transit of Bucket Coil No. unless (IB /ID) «1. 3 After Bucket Leaves Current=0 Time Negative Current for Transit of Bucket Coil No. Note that. This is because the bucket speed will increase slightly as successive coils pass a given drive coil. the circuit is switched back to the DC power source and the capacitor is charged again in preparation for the next bucket. the drive coil electrical circuit will probably need to be fairly complex. centered on the ith bucket.

(2) The total mass of the loaded bucket must not vary significantly between “shots” of the mass driver. This means that the mass of expellant added to the bucket must be accurately metered. Use of a bucket whose empty mass is significantly greater than the expellant mass it carries would clearly help reduce this sensitivity. An alternate approach, provided that some variation in drive coil frequency were possible, would be to weigh the bucket by some means after it has been loaded with expellant. This might be done by vibrating the bucket and measuring either its amplitude or response frequency. C.5 Braking Coils Analysis of the braking process is conducted using exactly the same equations as for the acceleration process. The sole difference is that all of the stationary coil currents are reversed. C.6 Bucket Design The bucket is assumed to be of cylindrical shape. As shown in figure 124, it consists of an inner structural layer surrounded by a dewer that contains liquid nitrogen (LN2) and the superconducting coils.
Dewer (With Insulation)

r h
Bucket Structure

LN2 Reservoir

Bucket Coil

Figure 124. Bucket conceptual design.

The bucket shape is determined by the ratio of the internal height to the internal radius,

h=

h ; r

(121)

thus, the internal volume (V), available to hold expellant, is given by V = hp r 3 . (122)

166

The radius is given in terms of the volume and h by

Ê V ˆ r=Á ˜ 3 . Ë hp ¯

1

(123)

The size of the LN2 reservoir is dictated by the total heat dissipated from the bucket coils during one acceleration-deceleration cycle. The LN2 will warm slightly during the acceleration-deceleration process due to the fact that bucket coil currents will vary due to inductive interaction with the stationary coils. If the total amount of heat generated is denoted by DQ and the allowable LN 2 temperature extremes are TL and TH, then, assuming that the thermal capacity of the coils may be neglected compared to that of the LN2, DQ = (TH - TL ) M LN 2 CLN 2 , where MLN is the mass of LN2 and CLN is specific heat.
2 2

(124)

Hence, the minimum mass of LN2 required is

M LN 2 =

DQ . (TH - TL ) CLN 2

(125)

The upper temperature limit is set by the critical temperature of the superconducting coils; i.e., the temperature above which their superconducting properties decline. The large number of so-called hightemperature superconductors now available, with critical temperatures at or above that of LN2 at 1 atm, suggest that the normal LN2 boiling point (77.4 K) as a good value for TH. To minimize the mass of the bucket, it is clearly desirable to minimize the mass of LN2 that must be carried. However, there are limits to the value of TL that are achievable. Some explanation of how the entire system works is appropriate here. Having completed its deceleration, the bucket LN2 reservoir will be near its maximum temperature (TH). The temperature will rise by an additional small amount as it makes its way via the return leg of the driver, back to the expellant loading hopper. This is simply because the ambient temperature on the asteroid surface is likely to be 50–100 K higher than that of the LN2 reservoir. Note that the bucket design must ensure that the reservoir temperature never exceeds TH. When the bucket has returned to the hopper, it will be placed into thermal contact with a cold plate at a sufficiently low temperature and for a sufficiently long time period to restore its LN2 reservoir to TL. For reference, the entire bucket return concept, including reservoir cooling and expellant loading, is shown in figure 125.

167

Bucket Rotates and Fully Engages With Frame Bucket Emerges From Final Braking Coils and Partially Engages With Handling Frame

Frame Moves Bucket on Return Path

Expellant Cold Plate

Cycle of Operation For Bucket

Hopper

Bucket Engages With Hopper and With Cold Plate

Frame Moves Bucket to Start of Mass Driver

Bucket Coils Engage With Acceleration Coils; Frame Releases and Returns to the Bucket Pickup Point

Figure 125. Bucket handling through a complete cycle. The flow of heat from the bucket LN2 reservoir to the cold plate can be written as

dQ = Kr 2 (T - TCP ) , dt

(126)

where K is a constant, r is the bucket external radius, which will probably determine the contact area with the cold plate, T is the instantaneous temperature of the bucket LN2, and TCP is the temperature of the cold plate. It is assumed that the cold plate is attached to a thermal reservoir and that its temperature may be taken as a constant. For a given DQ, the above equation simply shows that the time which needs to be spent in contact with the cold plate decreases with increasing contact area; i.e., with r2, and also decreases with decreasing cold plate temperature (TCP). This equation can be solved to give the time (t) that the bucket must spend in contact with the cold plate in order to have its reservoir temperature lowered from TH to TL: Ê T -T ˆ H t = - C ln Á L CP ˜ , Kr 2 Ë TH - TCP ¯ where HC is the total thermal capacity of bucket coils and LN2. (127)

168

The amount of heat dissipated in the bucket coils will, in some way, be proportional to the total time taken to accelerate and decelerate the bucket. Heat dissipation in the drive and braking coils is handled somewhat differently. The current in a drive coil is given by
Êp xˆ ID ( x ) = ID 0 sin Á ˜ . Ë S ¯

(128)

If R is the resistance of the coil, then the total heat dissipated as the drive circuit oscillates through one cycle is given by
t= 2p w

Heat =
hence,
2p t= w t =0

t =0

Ú

I 2 Rdt ;

(129)

Heat =
Therefore,

Ú

2 I D 0 R sin 2 (w t ) dt .

(130)

Heat = thus,

x =2 p x =0

Ú

2 I D0

R 2 sin xdx ; w

(131)

2 R Heat = 2 I D 0 w

x =p x =0

Ú

sin 2 xd x .

(132)

Hence,
2 Heat = 2 I D 0

Rp w2
R . w
(133)

2 = pID0

169

tuned to the frequency appropriate to each coil’s location along the length of the mass driver. This is a significant question given that. w (134) C. Thus.7 Interference Between Adjacent Drive Coils As outlined earlier. The spacing would almost certainly need to be less than the stationary coil diameter. Discharge of isolated drive coil. 126) will affect each other. Back emf = dF . This raises the question of how adjacent coils. As the capacitor begins to discharge. which must be at least four coil spacings long. 2 Total heat dissipated = 4 pID 0 R . each drive coil will probably oscillate through about four complete cycles. they would be very minor anyway for superconducting coils. The increasing flux induces a back emf. For simplicity. it is anticipated that the drive coils could each be part of a simple inductancecapacitance circuit. The discharging capacitor circuit is as shown in figure 126. which will have current flowing in opposite directions (fig. the increasing current induces an increasing magnetic flux (F) through the circuit (fig. Before considering the case of two interacting coils. the coils must be relatively close to each other. to avoid lengthening the bucket. acting against the direction of the increasing current: Magnetic Field +Q I –Q V1 V2 Figure 126. dt (135) 170 .Now in practice. 126). the effects of resistance are neglected. it is useful to consider first a single isolated coil.

C (136) where C denotes the capacitance of the capacitor. one can see that V2 = V1 - Q . V2 = V1 thus. the number of turns in the coil is assumed to have been taken into account in deriving the L value. dt (139) d 2F dt 2 =- I .From simple capacitor theory. From consideration of the back emf. C (140) The flux can be expressed in terms of the current and the circuit inductance (L) via F = LI . This gives d 2I dt 2 + 1 CI = 0 . dF . (141) To simplify the equations. L (142) which is the equation for simple harmonic motion and can be solved to give continuously oscillating solutions of the form I = I 0e i w t . dQ . dt (137) dF Q = . (143) 171 . dt C The capacitor charge and the discharge current are related via (138) I=so.

Ë LC ¯ Now consider the case of two adjacent coils with instantaneous AC flow directions. given by Ê 1 ˆ 0. (146) To simplify the equations. remains a constant. The total flux through coils 1 and 2 can be written as: F1 = L1I1 + MI 2 and (145) F2 = L2 I 2 + MI1 .5 w =Á ˜ . (144) Coil 1 Coil 2 Figure 127.2 = current in coils 1 and 2 M = mutual inductance of coils 1 and 2. Current flow directions of two adjacent drive coils.where I0 is the amplitude of the current oscillation and w its angular frequency. the number of turns in each coil is assumed to have been taken into account in the L and M values. where F1.2 = self-inductance of coils 1 and 2 I1. which. as the coils are both fixed. The interaction between these two coils can be described as follows. as shown in figure 127. 172 .2 = total flux through coils 1 and 2 L1.

ÍC Î ˚ ( ) (154) 173 .w 2 L˙ = Mw 2 . C (149) I1 = I10ei w t and I2 = I20ei w t . C (147) L1 Similarly. equation (147) can be written as d 2F1 dt from which F1 can be eliminated to give 2 I =. ÎC ˚ (153) These can be solved to give 2 È1 ˘2 .w 2 L ˙ = Mw 2 I1 ÎC ˚ (150) (151) (152) and È1 ˘ I2 Í . one can derive È1 ˘ I1 Í .As for the case of the single coil. C dt 2 dt (148) L2 Writing d 2 I2 dt 2 +M d 2 I1 dt 2 I + 2 =0 . for circuit 2.w 2 L˙ = Mw 2 I2 .1 . d 2 I1 d 2I I + M 22 + 1 = 0 .

the current will differ from that in an isolated coil. C(L . iw This means that.hence. C(L ± M ) (155) Substituting into one of the earlier equations relating the two currents gives I1 = ± I2 . for a given maximum capacitor charge.M ) (157) This means that the frequency is shifted upward when compared to an isolated coil that had w2 = 1 . iw where I10 is the maximum current. Hence. w2 = 1 . CL (158) In addition to the frequency shift. the minus sign applies and so w2 = 1 . (156) Hence. which is the situation here for adjacent drive coils. the maximum current (I10) is given by (161) I10 = iw Q10 . the maximum charge on the capacitor (Q10) is given by (160) I Q10 = 10 . This can be seen by taking the expression I1 = I10ei w t and integrating with respect to time to get an expression for the charge on the capacitor (Q1): (159) I Q1= 10 ei w t . when the two currents flow in opposite directions. (162) 174 .

M ˆ0. To avoid having this frequency shift take place while the bucket is very close. As soon as it has energized neighbors on both sides. the maximum current will be given by I10 = whereas for an interacting coil. If the interaction is modeled between a coil and those with counter-running current on either side. the factor will be increased further to 1 Ê 2 M ˆ 0. 175 .M )) 0.5 Ê1 .5 . (166) Note that. I10 = Q10 (CL) Q10 0. (164) Hence.5 Á1 ˜ Ë CL ¯ . it can be seen that the frequency of a coil will be somewhat lower when it is the endmost activated coil. its frequency will increase. the drive coils are activated as they are required.5 Á ˜ Ë CL ¯ CL (165) because of the interaction between the two coils. the discharge current is increased by a factor of 1 = (C(L .M )) 0.Hence. as the bucket proceeds along the mass driver. it may be necessary to activate each coil about a half period in advance and deactivate it about a half period after the ideal start and stop points.5 . Now from the above expression. for an isolated coil. (163) (C(L .

C. As before. Simple model of a drive coil circuit. d 2F dt 2 =- I C (167) for the stationary coil. (169) (170) 176 . Consider the situation where a bucket coil interacts with the stationary coil. the DC power system provides power to the stationary coil circuit. this quantity does not remain constant. it is not important to distinguish between the case of a drive coil and that of a braking coil. one can write F¢ = L¢I ¢ + MI . For present purposes. Note that. The mutual inductance between stationary and bucket coils is denoted by M. The current in the bucket coil circuit is denoted by I¢. if the bucket coil current flows in the same direction as that of the stationary coil. I¢ > 0. the DC power system accepts power from it. In the former case. (168) whereas before. L is the stationary coil self-inductance. Switch Stationary Coil Capacitor DC Power System Figure 128. in the latter case. I d2 + (LI + MI ¢ ) = 0 . Hence. As the bucket coil is in motion. C dt 2 Now for the bucket coil. In this case.8 Effect of Bucket Coil Motion on Stationary Coil Circuit Operation A stationary coil can be represented as an L-C circuit as shown in figure 128. the total flux is given by F = LI + MI ¢ .

F¢ . the equation can be rewritten as d 2I dt 2 f1 ( t ) + f 2 (t ) dI + f 3(t ) I = f 4 (t ) . M depends upon the distance between the two coils.thus. which is itself a function of time (t).MI ˆ I d2 Ê + 2 Á LI + M ˜=0 . C dt Ë L¢ ¯ which can be rewritten as F¢ ˆ I d 2 ÊÊ M2ˆ ˜I+ M ˜=0 .MI L¢ I¢ = (171) This enables one to eliminate I¢ from the main equation to give F¢ . + 2 ÁÁ L C dt ÁÁ L¢ ˜ L¢ ˜ ¯ ËË ¯ This can be expanded to give È M 2 ˘ È 1 d2 Ê M 2 ˘ d 2I dI d È M2 ˆ˘ d 2 Ê MF¢ ˆ ÍL ˙ ÍL ˙+ I Í + ÁL ˜˙ = +2 Á ˜ . L¢ ˚ ÍC dt 2 Á L¢ ˚ dt 2 dt dt Î L¢ ˜ ˙ Í ˙ Í ˙ Î dt 2 Ë L¢ ¯ Î Ë ¯˚ (172) (173) (174) Now L and L¢ are both constants. dt (175) where M2 L¢ f1 ( t ) = L - 2 2 d M f 2 (t ) = L¢ dt ( ) 177 . and F¢ is a constant. hence.

L¢ dt 2 Now. will not function properly. sinusoidal solutions with changing amplitude can be expressed as I = I 0e ( a +iw ) t .2 2 1 1 d M f 3(t ) = C L¢ dt 2 ( ) f 4 (t ) = - F¢ d 2 M . simply assume that satisfying this inequality is a challenge for the detailed design. it is necessary that the inequality 4 f1 ( t ) f 3 ( t ) > f 2 ( t ) 2 (181) be satisfied.4 f1 ( t ) f 3 ( t ) 2 l= 2 f1 ( t ) . dt (177) Solving this for solutions of the type I = Ae l t . . then the stationary coil current will not oscillate and the mass driver. where l is a constant. as designed here. (180) Now in order that the oscillating solutions be obtained. (178) (179) hence. If this is not met. 178 . (176) These solutions will only emerge from the complimentary equation corresponding to equation (169): f1 ( t ) d 2I dt 2 + f 2 (t ) dI + f 3(t ) I = 0 . one obtains f1 ( t ) l2 + f 2 ( t ) l + f 3 ( t ) = 0 .f 2 (t ) ± f 2 ( t ) . For present purposes.

If this condition is satisfied. First note that F = LI + M ¢ and F¢ = L¢I ¢ + MI .f 2 ( t ) .2 2 f1 ( t ) l= . then f (t ) 2 ± i 4 f1 ( t ) f 3 ( t ) . with (183) 2 d M f (t ) a =. it is useful to consider the sign of the denominator in equation (184). 2 f1 ( t ) which is of the form (182) l = a ± iw . Before proceeding further. = I ÁL Á L¢ ˜ L¢ Ë ¯ (187) M2 = LL¢ F- MF¢ L¢ . (185) (186) which means that Fhence.2 = L¢ dt . the sign of the above expression in a particular situation will determine whether the stationary coil current amplitude is increasing or decreasing. Ê 2 f 3(t ) M2ˆ ˜ 2 ÁL Á L¢ ˜ ¯ Ë ( 2) (184) Clearly. I (188) 179 . Ê M2ˆ MF¢ ˜ .

as the bucket is accelerating throughout. To summarize. Now. At first sight. and might lead one to conclude that the drive coil’s initial and final states are identical. • Bucket coil recedes from drive coil: In this case. the drive coil current amplitude increases as the bucket coil approaches and decreases as it recedes. C. d(M 2)/dt < 0. and the current amplitude decreases. L¢ (190) Thus far. 180 . and the current amplitude increases. This means that a > 0. Also. Closer consideration shows that this is not the case. this seems to imply a symmetry in the process. it can be said that F > F¢ and Hence. dM/dt > 0. L¢ (189) L- M2 >0 . dM/dt < 0. The magnitude of d(M 2)/dt is set by the bucket speed. the greater will be |d(M 2)/dt|.Now.9 Drive Coil The following scenarios are for the bucket coil as it approaches and recedes from the drive coil: • Bucket coil approaches drive coil: In this case. it has not been stated whether the bucket coil is approaching or receding from the stationary coil. hence. whether the stationary coil is a drive or a braking coil has not been specified. Consider two positions equidistant from and on either side of the drive coil. it follows that M <1 . This means that a < 0. d(M 2)/dt > 0. from the relative sizes of the drive and bucket coils as well as their relative currents. as shown in figure 129. its speed of recession will be greater than its speed of approach. The faster the bucket travels. hence.

d(M 2)/dt > 0. energy is supplied to the braking coil circuit by virtue of the bucket deceleration. • Bucket coil recedes from braking coil: In this case.10 Braking Coil The following scenarios are for the bucket coil as it approaches and recedes from the braking coil: • Bucket coil approaches braking coil: In this case. d(M 2)/dt is negative and of greater magnitude than it was at the first position. despite the fact that resistance is neglected. Thus. At the second (right-hand) bucket coil position. This means that. after the bucket coil has passed. the decrease in drive current amplitude at the second position exceeds the increase at the first position. once again. and the current amplitude decreases. dM/dt > 0. Note that. the initial and final bucket coil currents are the same and there is no change in energy. the drive coil current amplitude will be lower than it was initially. dM/dt < 0. At the first (left-hand) bucket coil position. so the amplitude increases while the bucket approaches and exceeds its decrease while the bucket recedes. and the current amplitude increases. although the bucket coil current rises and falls because of induction. neglecting resistance. d(M 2)/dt < 0. here the bucket is decelerating. hence. Hence. hence. simply by virtue of the bucket acceleration. C. energy has been extracted from the drive coil circuit. although these two positions are symmetric about the drive coil. The effect of bucket acceleration. This means that a > 0. d(M 2)/dt is positive. This means that a < 0. The difference is that. These results are identical to those for the drive coil. Hence.Bucket Coil Approaching at Speed V Bucket Coil Receding Drive Coil at Speed V+δV x x Direction of Motion Figure 129. the total magnetic flux through a bucket coil remains constant throughout the acceleration and deceleration processes. Hence. 181 .

In this case. The geometry is depicted in figure 130. hence.To summarize what is shown above. 2 where I and F are the current and flux through each circuit. The radial force on the entire coil (F) is best calculated by considering the change in energy associated with a slight increase in coil radius. This means that 1 U = NIF . respectively. each circuit is a turn of the coil. Self-induced magnetic field and resulting force. The total magnetic energy (U) of a system of N circuits is given by U= 1N Â Ii Fi . energy is lost from a drive coil circuit and supplied to a braking coil circuit simply due to the bucket acceleration and deceleration. even with a traditional zero-resistance model. (192) 182 .11 Self-Induced Coil Stresses A current-carrying coil loop will experience a radially outward force due to the interaction of its current with its self-generated magnetic field. which for an isolated circuit just produces an undamped sinusoidal oscillation. 2 i =1 (191) where Ii is the current through circuit i and Fi is the flux through circuit i. Current Force B Field B Field Coil Center Coil Loop With Clockwise-Directed Current and Self-Induced B Fields (Directed Into Plane of Figure) Radially Outward Force on Loop Element Figure 130. C. all are identical.

Now. the flux through one turn. 4˙ ˚ (197) Thus. 4˙ ˚ F= (198) The radial force per unit length of hoop per coil (f) is given by f = F . 4˙ ˚ (195) (196) hence. thus. The self-inductance of a single turn is given by equation (104): È Ê 8Rˆ L = m 0 R Íln Á ˜ Í Ë r0 ¯ Î 7˘ ˙ . 2 Therefore. (194) 1 dL F = I 2N 2 . 2 dR where R the radius of the coil. the total radial force on the entire coil is given by 1 2 2 È Ê 8Rˆ I N m 0 Íln Á ˜ 2 Í Ë r0 ¯ Î 3˘ ˙ . one circuit. È Ê 8Rˆ dL = m 0 Íln Á ˜ dR Í Ë r0 ¯ Î 3˘ ˙ . (193) where L is the self-inductance of a single turn. 2p RN (199) 183 .e. of the coil is due to the combined effects of the fields produced by each of the N turns. 1 U = LN 2 I 2 . i. F = NLI .. Hence.

thus. f= È Ê 8Rˆ 1 2 I Nm 0 Íln Á ˜ 4p R Í Ë r0 ¯ Î 3˘ ˙ . (201) so. 4˙ ˚ (200) The tension (T) within a loop is thus given by T=Rf . 4˙ ˚ T= (202) 184 . È Ê 8Rˆ 1 2 I Nm 0 Íln Á ˜ 4p Í Ë r0 ¯ Î 3˘ ˙ .

and resistance. (203) where ext Vk Vk Rk Ik Fk = voltage applied to circuit k from external sources = voltage across the capacitor of circuit k = resistance of circuit k = current in circuit k = total magnetic flux through circuit k. dt k (206) 185 . k k (205) where the sum over all values of k is denoted by  .ddtk = 0 . one can write F (Vkext + Vk ) . k Equation (203) can be used to eliminate Vkext from equation (205) to give W =  Ik k d Fk . The capacitance emf can be related to the current by Ik = -Ck where Ck is the total capacitance of circuit k. from both the circuit itself and all the other N–1 circuits. For circuit k. dt (204) The total available power in the entire system (W) is given by 2 ext W =  Vk Ik . dVk . Rk Ik .Rk Ik . capacitance. IkVk .APPENDIX D—COIL FORCE MODEL Consider an interacting set (N) of current-carrying circuits. and each subject to an externally imposed voltage. each with inductance.

dt (210) where the total magnetic energy. i (212) where Lki denotes the mutual inductance on circuit k due to circuit i. dt dt k (211) Next. Fk = Â Lki Ii . denoted by Um and Wmech . represents the mechanical work done in moving the circuits about. In view of the ultimate application of this model. the quantity W can be decomposed as follows: W = d (Ue + U m ) + W mech . 2k This gives (208) W = Â Ik k d Fk dUe + . this assumption may seem questionable. using equation (204) to give W = Â Ik k d Fk d 1 + Â CkVk2 . even though the circuits may move relative to one another. note that the total magnetic flux through a circuit is comprised of contributions due to its own inductance as well as that of each of the remaining circuits. dt dt (209) Now. dt dt k 2 (207) where it is assumed that none of the capacitances vary with time. however. it is only the fixed stationary circuits that have large capacitors built into their circuits. in general. Combining equations (209) and (210) gives d Fk dU m + W mech = Â Ik . 186 .which can be rewritten. The bucket coils will have relatively low capacitances. Noting that the total electrostatic energy (Ue) is given by 1 Ue = Â CkVk2 . Hence.

r=s W mech = Â Fr r =1 dar . The expression is analogous to that for electrostatic energy: 1 U m = Â Â Ik Lki Ii . where r=1. (214) dI dU m = Â Â Ik Lki i . then it is clear that: Wmech = 0 and dLki =0 dt for all values of k and i. equation (211) can be written as d ( Lki Ii ) dU m + W mech = Â Â Ik .Hence. where the circuits are permitted to engage in relative motion. if one considers a restricted situation in which all the circuits are held securely and cannot engage in any relative motion. Let the instantaneous location of all the circuits be defined by a set of parameters ar. dt dt k i from which it follows that (215) dU m d 1 = Â Â Ik Lki Ii . dt dt k i 2 (216) This permits one to write the following expression for the magnetic energy. consider the more general situation. In this case. 2k i (217) Now. the work done in a movement (dar) can be written as Work done = Fr dar . dt (219) 187 . s. …. Then. dt dt k i (213) Now. (218) hence.

thus.  Lki Ii k . it is also known that W mech =   Ik k i d ( Lki Ii ) dU m . d ( Lki Ii ) 1 dI 1 W mech =   Ik . dar dt dt 2 k i r (225) From this. Lki denotes the mutual inductance on circuit k due to circuit i. dt 2k i (222) (223) Now.But from equation (213). dt dt 2 k i (221) hence. dt dt (220) which can be rewritten using equation (217) as W mech =   Ik k i d ( Lki Ii ) d 1   Lki Ik Ii . one can write dLki dL dar =  ki . it follows that dL 1 Fr =   Ik Ii ki . Its variation with time is due solely to the relative motion of the circuits. dt 2k i dt 2k i This can be further simplified to dL 1 W mech =   Ik Ii ki . dar 2k i (226) 188 . dt r dar dt (224) which can be used together with equations (219) and (223) to give  Fr r dL dar dar 1 =    Ik Ii ki .

dar (228) For two coaxial coils. respectively. Also. dx (229) where M is the mutual induction between a single turn of each coil. so the two self-inductance terms (L11 and L22) are constant. this can be rewritten as Fx = N1N 2 I1I2 dM . equation (226) is applied to the specific case of two circuits that are in relative motion: dL dL 1 Ê 2 dL11 2 dL ˆ + I1I2 12 + I2 I1 21 + I2 22 ˜ . having N1 and N2 turns. Hence. and with motion along only the x axis. Á I1 dar ¯ dar dar 2 Ë dar Fr = (227) Now.Now. 189 . Fr = I1I2 dL12 . it can be shown from basic energy considerations that Lki = Lik for all i and k. there is no distortion of the coils. with two circular coils in relative motion. in the case of interest here.

R.harvard. and Haggerty.R.). 827–857.ciesin. Hills.” T. 791–826.” Hazards Due to Comets and Asteroids. Cited March 2003. Rampino. 1994. Morrison. pp. S. 8. The Cyrus Company. Cited December 12.barringercrater. B. 779–789. Zahnle. 1994.REFERENCES 1.). pp. 1994. pp.. Gehrels (ed.: “Extraterrestrial Impacts and Mass Extinctions of Life.: “INTROS: Integrated Rocket Sizing Program. New York. AZ. French. 12. to be published Spring 2004. 59–91.M. The University of Arizona Press.R. and Straw. 2003.. 11.: “Environmental Perturbations Caused by Asteroid Impacts..V.: Traces of Catastrophe. A. and Slovic.. 721– 778. AZ. 4. “The Barringer Meteorite Crater. 1994.. I.” Chapter 3 in Space Launch and Transportation Systems.V. 79–214. AL. Gehrels (ed. and Covey.: Impact Jupiter: The Crash of Comet Shoemaker–Levy 9. Tucson. Popov. Gehrels (ed. November 2003.. and Nemchinov.gif. 1998.” NASA/TP—2003–212691. J.” T. 3. B.: “Tsunami Generated By Small Asteroid Impacts. A. http://cfa-www.. V. Lunar and Planetary Institute. and Teterev.. 6. Tucson.V. 14–15. Alexander. 2003.: “The Impact Hazard.. Toon. The University of Arizona Press.” T.” Marshall Space Flight Center.org Cited March 2003. I. 190 .” Harvard-Smithsonian Minor Planet Center. Tucson.” http://www. Tucson. Chapman. 5. C. Columbia University.. Adams. The University of Arizona Press. E. The University of Arizona Press.: “Launch Vehicle Conceptual Design.G.: “Conceptual Design of In-Space Vehicles for Human Exploration of the Outer Planets.). “The Inner Solar System. Marshall Space Flight Center. R. K. D.M. Turco. Levy. AL.. pp. pp. 7.com. J. 10.A. P. Plenum Press. http://www. Nemchinov. Adushkin. R. et al. 3. 13.B. M. pp. Chapman.. Tucson. 2.P. 1994. Adams. O.. R. Center for International Earth Science Information Network.B.edu/ iau/Animations/Inner. C. MSFC. 9. Hazards Due to Comets and Asteroids..V.P. The University of Arizona Press. 1995. D. Lynn.” Hazards Due to Comets and Asteroids. Hazards Due to Comets and Asteroids. Hazards Due to Comets and Asteroids.M.: “Consequences of Impacts of Cosmic Bodies on the Surface of the Earth. pp.

1089–1110. J. E. T..” Hazards Due to Comets and Asteroids. Soc.A. 1986. 2000.” American Institute of Physics.. Vol. 22.” AIAA–2000–3610. February 1965. pp. G. 2000. pp.: Project ORION: The True Story of the Atomic Spaceship.shtml Cited April 14.: Solar Sailing. McInnes. pp. A.. “Dan’s History. “External Pulsed Plasma Propulsion (EPPP) Analysis Maturation. Soc. P. J. 33–109. Chichester. Physics for Scientists and Engineers. Reynolds.” http://danshistory. 23.C. J. Solem.” J. Martin. 24. Ahrens. pp. J. 32.J. Tucson.” J.” http://www. Cotter. and Irvine. Praxis Publishing Ltd...: “Rotating Cable Pusher for Pulsed-Propulsion Space Vehicle. Schmidt. A.: “The MagOrion—A Propulsion System for Human Exploration of the Outer Planets.shtml Cited April 14. Nance.W. 897–924.: “Application of Nuclear Propulsion to NEO Interceptors.P.. 2002.W.. and Schmidt. 20. 1979. 17. Bonometti.: “Nuclear Pulse Propulsion. T. 27. British Interpl.: “Deflection and Fragmentation of Near-Earth Asteroids. NASA TN D–6984. P. 21. 28. Dyson.danshistory.” Hazards Due to Comets and Asteroids. May–June 2002. G. J.com/lgb. 2003.com/nuke. 26. and Morton. 15. Bonometti.. C. 1972. Tucson. LA–4666–MS UC–33. 283–310. Gustafson. 2003. 29. 1993. 225–239. Propulsion and Power. 25. Philadelphia. Inc. G.R. J.14. pp.” IEEE Transactions on Nuclear Science. 16. Vol. Bonometti. New York..A.. The University of Arizona Press. Venetoklis. 30. Saunders College Publishing.R. and Powell. D. 1533–1537. Propulsion Systems and Energy Conversion TID–4500. J.: “Nuclear Pulse Propulsion: A Historical Review of an Advanced Propulsion Concept. 1971. T. and Bond. pp. 1236–1241. 18. “Dan’s History. P. 21–26. 1994.: Space Sailing. 497–504..J.. 1992. Maise. G.: “External Pulsed Plasma Propulsion and Its Potential for the Near Future. pp. 1994. 191 . pp. Morton. C.J. Vol. British Interpl.: “Medusa: Nuclear Explosive Propulsion for Interplanetary Travel. pp. Andrews. 1999. 46.” NASA Technical Note. and Harris.: “Project Orion and Future Prospects for Nuclear Pulse Propulsion.C.. J.: “Effective Specific Impulse of External Nuclear Pulse Propulsion Systems. A. Taylor and Francis. 19. Serway.A.” J.” Los Alamos Scientific Laboratory.” American Institute of Physics.: and Andrews. New York. The University of Arizona Press.A.R. 2000. 18. UK.. Wright. Owl Books.

1994.: “Non-Nuclear Strategies for Deflecting Comets and Asteroids. Vol. January/February 2001.. pp. and Rafikov. Mueller.R. pp. Ahrens. Academic Press. 35. 39.com/e-bomb..: “Impacts on the Earth by Asteroids and Comets: Assessing the Hazard. 32. T. pp. pp.. 41. Manley. Tucson.O.. 43.R. 2749–2784. S.. J. Tedeschi.” in Proc. D.W. Chesley.: Comet and Asteroid Impact Hazards on a Populated Earth. B. 1994. Brown. 159. p.: “The Flux of Small Near-Earth Objects Colliding with Earth.” Annual Rev.: Fundamentals of Astrodynamics. Chodas. I. Jeffers.K.. P. 42.. J. 327. AZ. O’Neill. Harris.M. Bate. Vol. 44.: “SHIELD—A Comprehensive Earth Protection System: A Phase I Report to the NASA Institute for Advanced Concepts. p. p. 192 . 327. Tabachnik... 122. Y.howstuffworks.” http://science.. 6 January 1994. E. 1999. Tucson.: “Solar System Objects Observed in the Sloan Digital Sky Survey Commissioning Data. CA. Chapman. 11. S. 423–432. et al. 367. and White. et al. 36. 313–323. Astron.” Hazards Due to Comets and Asteroids...” Nature.. 2002. Lawrence Livermore National Laboratory. November 2001. Nelson. Ivezic. R.E.. and Asher.V. T. and Morrison. R. M. C.htm Cited December 2002.I. ReVelle. New York. Vol. Soc.P. 34. Bailey. May 22– 26.. J. W. Planetary Defense Workshop.. Nemchinov. A. P. Lewis. 142.. 21 November 2002. R. 33. 1971.31. R.D.A. 45. 2001. and Harris.J. Not. Vol. D. Inc. and Zetzer.” Mon.” Icarus. Spalding.” Report No.E. 4. 1111– 1132.” Progress in Astronautics and Aeronautics. 37. May 1983.E. 1999. A.: “Deflection and Fragmentation of Near-Earth Asteroids.E. Livermore. pp. The University of Arizona Press. and O’Leary.J.: “Mitigation of the NEO Impact Hazard Using Kinetic Energy. 40. p. Melosh. Dover Publications.J.S. R. H.” FSA Public Interest Report. 1977.. Z.: “How Stuff Works.: “Near-Earth Object Velocity Distributions and Consequences for the Chicxulub Impactor. Vol. 57.: “Low-Yield Earth Penetrating Nuclear Weapons. May 28..” Hazards Due to Comets and Asteroids.: “Space-Based Manufacturing From Nonterrestrial Materials. 33–40. Gold.: “Asteroid and Comet Bombardment of the Earth.. Vol. S. S.: “Quantifying the Risk Posed by Potential Earth Impacts. R. 38. 461–494. D. 897–927. 1995.. pp. Shoemaker. The University of Arizona Press.J. published by the American Institute of Aeronautics and Astronautics. Milani..” Astron. G. Earth and Planetary Sci. D.V. SDO–10974.” Nature.

Abramowitz. New York.: Handbook of Mathematical Functions.gov/osstech/solar. NASA Glenn Research Center Web-page: http://spacescience. 47. 1999. CA. Reeves. Dover Publications Inc.htm Cited December 2002. 1965.: Electromagnetic Fields and Interactions. I. M. Wertz and Wiley J. New York.I. NY.46. pp. NY.” Chapter 10 in Space Mission Analysis and Design. 1982. El Segundo.: “Spacecraft Design and Sizing. 608–609. R.nasa. edited by James R.A. 193– 197. pp. Microcosm Press.. 49. E.. 48.. Becker. Larson. Dover Publications Inc. and Stegun. 193 .

REPORT DATE July 2004 3. and to the Office of Management and Budget. VA 22202-4302. It is hoped that this study will raise the level of attention about this very real threat and also demonstrate that successful defense is both possible and practicable. 14. Suite 1204. AUTHORS R. postdeflection path are given. 0704-0188 Public reporting burden for this collection of information is estimated to average 1 hour per response.B. Various ongoing projects intended to survey and catalog the NEO population are also reviewed. J.. trajectory modeling tools Unclassified 18. ABSTRACT (Maximum 200 words) Several recent near-miss encounters with asteroids and comets have focused attention on the threat of a catastrophic impact with the Earth. Statham. DISTRIBUTION/AVAILABILITY STATEMENT Unclassified-Unlimited Subject Category 18 Availability: NASA CASI 301–621–0390 13. AL 35812 9.REPORT DOCUMENTATION PAGE Form Approved OMB No. Marshall Space Flight Center Marshall Space Flight Center. AGENCY USE ONLY (Leave Blank) 4. Send comments regarding this burden estimate or any other aspect of this collection of information. White* 8. Space Transportation Directorate *ERC. Adams. SPONSORING/MONITORING AGENCY REPO NUMBER National Aeronautics and Space Administration Washington. M. Polsgrove. comets. Details of analytical tools. and a tool that was created to model both the undeflected inbound path of an NEO as well as the modified. Details are given of a Marshall Space Flight Center-led study intended to develop and assess various candidate systems for protection of the Earth against NEOs. Huntsville. 1215 Jefferson Davis Highway. G. Directorate for Information Operation and Reports. Alexander. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) M–1100 10. SECURITY CLASSIFICATION OF THIS PAGE 15. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) George C. T. SUPPLEMENTARY NOTES NASA/TP—2004–213089 Prepared by the Advanced Concepts Department. LIMITATION OF ABSTRACT 17. DC 20503 1. R. AL 35816 12b. J. bombardment of Earth. Arlington. TITLE AND SUBTITLE 2. DC 20546 11. trajectory tools. Suite 1622. DISTRIBUTION CODE 12a. provided appropriate steps are taken. to Washington Headquarters Services. REPORT TYPE AND DATES COVERED Technical Publication 5.* and S. Kalkstein. Paperwork Reduction Project (0704-0188). SECURITY CLASSIFICATION OF REPORT Unclassified 19. near-Earth object. SECURITY CLASSIFICATION OF ABSTRACT Unclassified Unlimited Standard Form 298 (Rev. including the time for reviewing instructions. PERFORMING ORGANIZATION REPORT NUMBER 7. including suggestions for reducing this burden. NUMBER OF PAGES 222 16. and completing and reviewing the collection of information. A representative selection of these possible options was modeled and evaluated. 2-89) Prescribed by ANSI Std. gathering and maintaining the data needed. Hopkins. FUNDING NUMBERS Survey of Technologies Relevant to Defense From Near-Earth Objects 6. Executive Plasma. R. This Technical Publication reviews the historical impact record and current understanding of the number and location of near-Earth objects (NEOs) to address their impact probability. searching existing data sources. Inc. Chapman. PRICE CODE 20. Bonemetti. Washington. 239-18 298-102 NSN 7540-01-280-5500 . S. Fincher. SUBJECT TERMS asteroids. 555 Sparkman Drive.