This action might not be possible to undo. Are you sure you want to continue?
HOW TO CALCULATE MASS PROPERTIES
(An Engineer's Practical Guide)
by
Richard Boynton, President
Space Electronics, Inc.
Berlin, CT 06037
and
Kurt Wiener, Assistant Chief Engineer
Space Electronics, Inc.
Berlin, CT 06037
Abstract There are numerous text books on dynamics which devote a few pages to the
calculation of mass properties. However, these text books quickly jump from a very brief
description of these quantities to some general mathematical formulas without giving
adequate examples or explaining in enough detail how to use these formulas. The
purpose of this paper is to provide a detailed procedure for the calculation of mass
properties for an engineer who is inexperienced in these calculations. Hopefully this
paper will also provide a convenient reference for those who are already familiar with
this subject.
This paper contains a number of specific examples with emphasis on units of
measurement. The examples used are rockets and reentry vehicles. The paper then
describes the techniques for combining the mass properties of subassemblies to yield the
composite mass properties of the total vehicle. Errors due to misalignment of the stages
of a rocket are evaluated numerically. Methods for calculating mass property corrections
are also explained.
Page 2
Calculation of Mass Properties using Traditional Methods
Choosing the Reference Axes
The first step in calculating mass properties of an object is to assign the location of the
reference axes. The center of gravity and the product of inertia of an object can have any
numerical value or polarity, depending on the choice of axes that are used as a reference
for the calculation. Stating that a CG coordinate is "0.050 inches" means nothing unless
the position of the reference axis is also precisely defined. Any reference axes may be
chosen. For example, the center of gravity of a cylinder may be 4.050 inches from one
end, 0.050 inches from its midpoint, and 3.950 inches from the other end. Furthermore,
each end of the cylinder may not be perpendicular to the central axis, so that a means of
determining the "end" of the cylinder would have to be further defined.
Three mutually perpendicular reference axes are
required to define the location of the center of
gravity of an object. These axes are usually
selected to coincide with edges of the object,
accurately located details, or the geometric center
of the object.
It is not sufficient to state that an axis is the
centerline of the object. You must also specify
which surfaces on the object define this centerline.
Moment of inertia is a rotational quantity and
requires only one axis for its reference. Although
this can theoretically be any axis in the vicinity of
the object, this axis usually is the geometric center,
the rotational center (if the object revolves on
bearings), or a principal axis (axis passing through
the center of gravity which is chosen so the
products of inertia are zero).
Product of inertia requires three mutually perpendicular
reference axes. One of these axes may be a rotational axis
or a geometric centerline.
For maximum accuracy, it is important to use reference
axes that can be located with a high degree of precision. If
the object is an aerospace item, then we recommend that
this object be designed with two reference datum rings per
section, which can be used to define the reference axes.
These rings can be precision attachment points that are
used to interface the object with another section of a
spacecraft or rocket, or they can be rings that were
provided solely for the purpose of alignment and/or measurement
of mass properties. The accuracy of calculation (and the
subsequent accuracy of measurement of an actual piece of hardware) is only as good as
Figure 2  Datum Rings
Figure 1 Center of gravity (and
product of inertia) are defined
relative to orthogonal axes
Page 3
the accuracy of the means of locating the reference axes. We
have found that the single largest source of error in mass
properties calculations is the uncertainty of the reference. The
dimensional data provided to the mass properties engineer must
be sufficiently accurate to permit mass properties tolerances to be
met.
For example, if you are asked to make precise calculations of
mass properties of a projectile, you should establish the error due
to reference misalignment as the first step in your calculations. If
you are required to calculate CG within an accuracy of 0.001
inch and the reference datum is not round within 0.003 inch, then you
cannot accomplish your task. There is no sense in making a detailed
analysis of the components of an object when the reference error prevents
accurate calculations. Furthermore, it will be impossible to accurately measure such a
part after it is fabricated and verify the accuracy of your calculations. The location and
accuracy of the reference axes must be of the highest precision.
If your task is to calculate the mass properties of a vehicle that is assembled in sections,
then serious thought should be given to the accuracy of alignment of the sections when
they are assembled. Often this can be the biggest single factor in limiting the degree of
balance (if the vehicle was balanced in sections because the total vehicle is too big for the
balancing machine). Alignment error is amplified for long rockets . . . a 0.001 inch lean
introduced by alignment error on a 12 inch diameter can result in a 0.007 inch CG error
on a 15 foot long rocket section. This is discussed in detail in the sections of this paper
that present the math for combining the mass properties of subassemblies.
The accuracy required for various types of calculations is summarized in later sections of
this paper.
Choosing the Location of the Axes
The axes in Figure 3 do not make a good reference because a small error in squareness
of the bottom of the cylinder causes the object to lean away from the vertical axis. The
axes below (Figure 4) make a better choice.
Figure 3
The first step in calculating mass properties is to establish the location of the X,
Y, and Z axes. The accuracy of the calculations (and later on the accuracy of
the measurements to verify the calculations) will depend entirely on the
wisdom used in choosing the axes. Theoretically, these axes can be at any
location relative to the object being considered, provided the axes are mutually
perpendicular. However, in real life, unless the axes are chosen to be at a
location that can be accurately measured and identified, the calculations are
meaningless.
Figure 4
Page 4
Reference axes must be located at physical points on
the object that can be accurately measured. Although
the center line of a ring may exist in midair, it can be
accurately measured and is therefore a good reference
location as can the center of a close tolerance hole
which could be identified as the zero degree reference
to identify the X axis (Fig. 4).
An axis should always pass through a surface that is
rigidly associated with the bulk of the object. In Figure
5 it would be better to locate the origin at the end of the
object rather than the fitting that is loosely dimensioned
relative to the end.
Calculating CG Location
General Discussion
The center of gravity of an object is:
! also called the "center of mass" of the object.
! the point where the object would balance if placed on a knife edge
! the single point where the static balance moments about three mutually
perpendicular axes are all zero.
! the centroid of the volume of the object, if the object is homogeneous.
! the point where all the mass of the object could be considered to be concentrated
when performing static calculations.
! the point about which the object rotates in free space
! the point through which the force of gravity can be considered to act
! the point at which an external force must be applied to produce pure translation of
an object in space
Center of gravity location is expressed in units of length along each of the three axes (X,
Y, and Z). These are the three components of the vector distance from the origin of the
coordinate system to the CG location. Center of gravity of composite masses is
calculated from moments taken about the origin. The fundamental dimensions of
moment are typically Force times Distance. Alternatively, Mass moment may be used
with any units of Mass times Distance. For homogeneous elements, volume moments
may also be used. Care must be taken to be sure that moments for all elements are
expressed in compatible units.
When combining mass elements, a useful technique is to use "offset moments" about
each of the three orthogonal axes. The X offset moment of one element (such as M
X1
=
+3W
1
) can be easily added to the X offset moments of other elements of mass, the sum
divided by the total weight, and the result will be the X component of the CG location of
the composite mass. Likewise, the Y and Z offset moments (M
Y1
=5W
1
, and M
Z1
=+7W
1
)
can be combined with similar Y and Z offset moments of other elements to determine the
Y and Z components of the CG location. Unfortunately, the term "X offset moment" is
Figure 5
Page 5
frequently described as "moment along X". This does not make mathematical sense, but
like the term "pound mass" most engineers will understand the meaning.
Component distances for center of gravity location may be either positive or negative,
and in fact their polarity depends on the choice of reference axis location.
The center of gravity of a homogeneous shape is calculated by determining
the centroid of its volume. In real life, most objects are not homogeneous,
so that the center of gravity must be computed by summing the offset
moments along each of the three axes. These processes are described in
detail in the following sections.
The center of gravity of an object can be located in "midair". For example,
the center of gravity of a piece of pipe is on the centerline half way along its
length, even though there is no metal in the center of the pipe (Figure 6).
The composite CG of an object can be computed if the CG of each
component is known. Examples follow.
CG along a single axis
Consider the round metal rod with two cylindrical weights shown (Figure 7).
Note that the elements do not have to be the same diameter to be symmetrical
along the length. In fact the elements could overlap (such as sliding one pipe
inside another). From symmetry, the CG of the object is on its centerline (since
the CG of a homogeneous mass is at its centroid of volume). The CG location
along the length can be determined by summing moments about the reference
axis at the bottom of the figure (x = 0).
Assume that the element weights are; W
a
=12.250 lb, W
b
=4.613 lb, W
c
=2.553 lb.
A
Figure 7 – CG
along a single axis
Figure 6
M W z M lbs x inch lb in
M W z M lbs x inch lb in
M W z M lbs x inch lb in
Total Weight lbs
Total Moment lb in
CG
Total Moment
Total Weight
lb in
lb
inch from A
a a a a
b b b b
c c c c
z
· · · −
· · · −
· · · −
·
· −
· ·
−
·
12 250 6319 77 408
4 613 2 445 11279
2553 8666 22124
19 416
110811
110811
19 416
5707
. . .
. . .
. . .
.
.
.
.
.
Page 6
CG of Unsymmetrical Three Dimensional Body
The center of gravity of an unsymmetrical body may be
calculated in the same manner as the single axis example
above. Each axis may be considered separately (Figure 8).
Consider the cylinder with attached rectangles. The CG of
each component is known by symmetry, computation, or
measurement. A convenient frame of reference is assigned,
in this case such that the CGs of each component fall on the
axes, and offset moments are summed along each axis.
Dimensions shown are to the CG of each component from
the origin.
M M M M lb in
CG lb in lb in
x a b c
x
· + + · + + · −
· − ·
04 175 0 0 07
0 70 48 0146
. ( . ) .
. / . .
CG of a Complex Shape Similar to a Standard Shape
Consider the hollow cone shown below. From symmetry, the CG lies along the center
line. The CG distance along the length could be calculated using calculus. However, the
CG of a solid cone is given in the SAWE handbook. Using the observation that a hollow
cone can be created by removing a small solid cone from a larger one, we can calculate
the CG by subtracting the moment due to the smaller cone from the larger one. Volume
moments are taken around the center of the base to find the centroid of the hollow cone.
When the cone is combined with other elements to find the overall CG, its actual weight
and calculated centroid location are combined with those of the other elements.
V
r H
V
r H
1
1
2
1
2
2
2
2
2
2
3
4 20
3
3351
3
36 18
3
244 3 · · · · · ·
π π π π ( )( )
.
( . )( )
.
V V V in
CG
a V a V
V
CG in
NET
NET
· − ·
·
−
·
−
·
1 2
3
1 1 2 2
908
5 3351 45 244 3
908
635
.
( )( . ) ( . )( . )
.
.
Figure 8 – CG of an
unsymmetrical body
M M M M lb in
CG lb in lb in
M M M M lb in
CG lb in lb in
y a b c
y
z a b c
z
· + + · + + − · − −
· − − · −
· + + · + + · −
· − ·
0 0 18 125 2 25
2 25 48 0469
04 4 25 2 6 25 18 05 91
91 48 1896
. ( . ) .
. / . .
. ( . ) . ( . ) . ( . ) .
. / . .
Figure 9 – Hollow Cone
X
Page 7
CG of a Complex Unusual Shape
If you encounter a shape which is not in the handbook and which cannot be created from
known shapes, then it will be necessary to use calculus to calculate its CG. The basic
concept of the calculation is the same as the previous examples, except the moments that
are summed are moments involving a small differential slice of the object rather than
moments of discrete objects. The trick to simplifying this process is to chose the right
differential shape, so that triple integration can be avoided. Your differential element
should not be a small cube unless there is no symmetry of any kind. Generally you can
use a rectangular bar that covers the full length of the part, or a thin disc or annular ring
whose diameter is a function of location.
To illustrate the CG calculation using calculus, we will use the same hollow cone
discussed in the previous section.
Do· ′′ 8
X
o
· ′′ 20
V dV dV T
x
x
x
x
· +
−
−
−
−
∫ ∫
1
0
2
2
2
20
M xdV xdV T
x
x
x
x
· +
−
−
−
−
∫ ∫
1
0
2
2
2
20
dV
D
dx D
D
X
X
o
o
1
1
2
1
4
· ·
π
( ) ( ) dV D D dx D
D
X
x
o
o
2 1
2
2
2
2
4
2 · − · −
π
dV
D
X
x dx
1
0
2
0
2
2
4
·
π
( )
[ ]
dV
D
X
x x x dx
2
0
2
0
2
2 2
4
4 4 · − − +
π
( ) M
D
X
x dx x x dx
D
X
x x x
in
T
· + −
]
]
]
·
]
]
]
]
+ −
]
]
]
]
]
]
]
]
·
∫ ∫
π π
0
2
0
2
3 2
2
20
0
2
0
2
0
2
0
2
4
2
20
3 2
4
4
4 4
4 4
4
3
4
2
1240 0 .
Location of Centroid from tip
M
V
in
in
OR inch fromBase
T
T
· · ·
− ·
12400
908
1365
20 1365 635
4
3
.
.
.
. .
( )
3
2
20
2
3
2
0
2
0
2
0
20
2
2
0
2
2
0
2
0
8 . 90 4
2
4
3
4
4 4
4
in x
x x
X
D
dx x dx x
X
D
V
T
·
]
]
]
]
]
]
− +
]
]
]
·
]
]
]
− + ·
∫ ∫
π π
Page 8
Rectangular to Polar Conversion
When first calculated, the CG data is in rectangular form. Often it is useful to convert
this data into polar coordinates. Most computers and scientific calculators will do this
automatically. However, if one is not available, the following method can be used (two
axis):
Magnitude
2 2
Y X M + ·
Angle
A = arcTAN (Y/X) if X = (+) and Y = (+) (1st quadrant)
A = 180
o
 arcTAN (Y/X) if X = () and Y = (+) (2nd quadrant)
A = 180
o
+ arcTAN (Y/X) if X = () and Y = () (3rd quadrant)
A = 360
o
 arcTAN (Y/X) if X = (+) and Y = () (4th quadrant)
Polar to Rectangular Conversion
After the data has been converted to polar form, sometimes you may then need to convert
it back to rectangular form, using a different set of axes. This might occur if you wanted
to adjust the CG offset of a reentry vehicle, so that it was balanced about its centerline.
In this case, the available locations for correction weights would usually not fall on the
reference axes.
Step 1 Add appropriate offsets to the X and Y rectangular components.
Step 2 Convert to polar form (magnitude and angle).
Step 3 Add the appropriate offset angle to rotate the vector to the new XY coordinate
system.
Step 4 Convert the new vector to rectangular coordinates using the following formula:
where X
1
and Y
1
are the new axes and A
1
is the angle between the
unbalance moment vector and the X
1
axis.
Correcting Static Unbalance
The satellite shown has a calculated static unbalance (CG offset
moment) of X= 4.65 lbinch and Y = +12.32 lbinch. It is necessary
to add ballast weights to the vehicle so that this unbalance is reduced
to zero. If weights could be added at 0
o
and 270
o
, then we would need
4.65 lbin at 0
o
to compensate for the 4.65 lbin X axis unbalance and
12.32 lbin at 270
o
to compensate for the +12.32 lbin Y axis unbalance.
However, these locations are not available. This is the general situation in
most aerospace balancing. In this example, the only locations where these
( ) ( ) X M A Y M A
1 1 1 1
· · cos sin
Figure 7 – Static Unbalance
Page 9
weights can be added are at 33
o
and 255
o
. The radius of the correction weight at 33
o
is
8.25 inches and at 255
o
is 7.60 inches. What weights should be added to each location to
compensate for the unbalance?
The following example outlines the method used to determine the new weights at the
allowed locations:
We will first calculate the resultant polar magnitude and angle, then calculate rectangular
coordinates of correction moments. We will then divide the correction moments by their
radii to obtain correction weights.
General Correction Equations
where: C = Correction Moments
A
c
= Allowable Correction Angles
M = Static unbalance Moment
A = Angle of Unbalance Moment
Note that these calculations involving static unbalance are concerned with weight rather
than mass. In our example above, the figure did not show the height at which the weights
were to be added. In general, the weights should be added at a height that is as close as
possible to the CG height of the vehicle, so that the addition of these weights will not
produce a large product of inertia unbalance.
Calculating Correction Weights
M M M lb in
x y
· + · −
2 2
1317 .
A TAN y x ·
−1
( / )
A TAN
o
· − ·
−1
12 32 4 65 1107 ( . / . ) .
To find Correction Moments C
1
, and C
2
at 33
o
and 255
o
locations:
C C Cos C Cos
C C Sin C Sin
x
y
· + ·
· + · −
1 2
1 2
33 255 4 65
33 255 12 32
.
.
[ ] ( ) Sum C A M A
c
sin sin · − 180 [ ] ( ) Sum C A M A
c
cos cos · − 180
. . .
. . .
84 26 4 65
54 97 12 32
1 2
1 2
C C
C C
− ·
− · −
Page 10
Multiply both sides of eq. (2) by: (0.84/0.54) and add to eq. (1)
0 125 2381
19 05
4 65 26 84
1143
1 2
2
1 2
1
C C
C lb in
C C
C lb in
+ ·
· −
· +
· −
. .
.
( . . )/.
.
Weights to be added:
Note that total weight is more than twice as large as would be required
if ballast weights could have been added at the angle of the unbalance
at 8” rad.
Combining CG Data from Subassemblies
Let us consider the case of a three stage rocket. The center of gravity
of the individual stages had been originally calculated to be on the
centerline. After construction, the CG of each section has been
measured and found to be:
X Y Z W
Stage 1 +0.004" 0.012" 27.436" 167 lb.
Stage 2 0.007" +0.012" +32.771" 96 lb.
Stage 3 0.004" +0.012" +12.115" 43 lb.
TOTAL 306 lb.
Two views of the rocket are shown. The top view shows the X and Y
axes; side view shows X and Z axes.
The X and Y coordinates are measured from the centerline of the
section; the Z coordinates of stages 1 and 2 are measured from their
intersection, while the Z coordinate of the third stage is measured from
the intersection of stages 2 and 3. In order to calculate the Z center of gravity location,
we will first have to translate the stage 3 coordinate to the same reference as
stages 1 and 2. Since the length of stage 2 is 51.125 inches, the Z coordinate
becomes 12.115 + 51.125 = 63.240 inches.
If the three stages were perfectly aligned at assembly, then the combined CG
of the total rocket could be calculated by summing X, Y, and Z offset moments about the
origin:
W
lb in
in
lb
W
lb in
in
lb
1
2
1143
825
139
19 05
7 6
251
·
−
·
·
−
·
.
.
.
.
.
.
Stage 3
Stage 2
Stage 1
Figure 8 – CG of Three
Stage Rocket
Page 11
X Y Z
Stage 1 +0.668 lbin. 2.004 lbin. 4581.812 lbin.
Stage 2 0.672 lbin. +1.152 lbin. +3146.016 lbin.
Stage 3 0.172 lbin. +0.516 lbin. +2719.320 lbin.
Total Moment 0.176 lbin. 0.336 lbin. +1283.524 lbin.
Moment/306 lb 0.00058 in. 0.00110 in. +4.1945 in.
If the three stages are assembled with an alignment error, then:
1. Select one of the stages to be the reference. For this example we will choose stage 2.
The CG coordinates for this stage will therefore remain unchanged.
2. Recalculate the CG coordinates for stages 1 and 3 to reflect the alignment error. If the
stages did not assemble tightly along their length, so that there is a 0.006 inch gap
between stages 2 and 3, then the 63.240 Z dimension becomes 63.246. If the X axis is
shifted sideways on the first stage by +0.003 inch, then the 0.004 inch X dimension for
the first stage becomes 0.001 inch, etc. If the stages are tilted relative to each other, then
the offset due to the tilt must be determined at the CG height of the stage. For example,
if stage 3 is tilted so that the error on the Y axis at a Z dimension of 24.5 inch is +0.020,
then the Z axis correction for stage three is: .020 x 12.115/24.500 = 0.00989 inch.
The new value of Y for stage three is therefore: Y = 0.012 + 0.00989 = 0.0219 inch
3. After the revised table for the CG coordinates for stages 1 and 3 are complete, then the
calculation proceeds in a manner identical to the example with perfect alignment.
CALCULATING MOMENT OF INERTIA
General Comments
Moment of inertia ("MOI") is similar to inertia, except it applies to rotation rather than
linear motion. Inertia is the tendency of an object to remain at rest or to continue moving
in a straight line at the same velocity. Inertia can be thought of as another word for mass.
Moment of inertia is, therefore, rotational mass. Unlike inertia, MOI also depends on the
distribution of mass in an object. The greater the distance the mass is from the center of
rotation, the greater the moment of inertia.
A formula analogous to Newton's second law of motion can be written for rotation:
F = Ma (F = force; M = mass; a = linear acceleration)
T = IA (T = torque; I = moment of inertia; A = rotational acceleration)
Page 12
Choosing the Reference Axis Location
Three reference axes were necessary to define center of gravity. Only one axis is
necessary to define moment of inertia. Although any axis can be chosen as a reference, it
is generally desirable to choose the axis of rotation of the object. If the object is mounted
on bearings, then this axis is defined by the centerline of the bearings. If the object flies
in space, then this axis is a "principal axis" (axis passing through the center of gravity and
oriented such that the product of inertia about this axis is zero (see discussion of product
of inertia). If the reference axis will be used to calculate moment of inertia of a complex
shape, choose an axis of symmetry to simplify the calculation. This axis can later on be
translated to another axis if desired, using the rules outlined in the section entitled
"Parallel Axis Theorem".
Polarity of Moment of Inertia
Values for center of gravity can be either positive or negative, and in fact their polarity
depends on the choice of reference axis location. Values for moment of inertia can only
be positive, just as mass can only be positive.
Units of Moment of Inertia
In the United States, the word "pound" is often misused to describe both mass and
weight. If the unit of weight is the pound, then the unit of mass cannot also be a pound,
since this would violate Newton's second law. However, for reasons which have been
lost in antiquity, in the USA an object weighing 1 pound is often referred to as having a
mass of 1 pound. This leads to units of moment of inertia such as lbin
2
, where the "lb"
refers to the weight of the object rather than its mass. Correct units of moment of inertia
(or product of inertia) are: MASS x DISTANCE
2
When lbin
2
or lbft
2
are used to define MOI or POI, the quantity MUST be divided by
the appropriate value of "g" to be dimensionally correct in engineering calculations.
Again, dimensional analysis will confirm if correct units are being used.
The following table shows some of the units in use today for moment of inertia and
product of inertia:
UNIT COMMENTS
lbin
2
lb = weight; must be divided by g = 386.088 in/sec
2
lbinsec
2
lbinsec
2
= distance
2
x weight/g; weight/g = mass; dimensionally correct
slugft
2
slug = mass; dimensionally correct
kgm
2
Kg = mass; dimensionally correct
The most common units used in the U.S. are lbin
2
, even though this is dimensionally
incorrect.
Page 13
RULE 1. If moment of inertia or product of inertia are
expressed in the following units, then their values can be
used in engineering calculations as they are:
Slugft
2
, lbinsec
2
, kgm
2
, lbftsec
2
, ozinsec
2
RULE 2. If moment of inertia or product of inertia are
expressed in the following units, then their values must be
divided by the appropriate value of "g" to make them
dimensionally correct.
lbft
2
, lbin
2
, ozin
2
Value of g : 32.17405 ft/sec
2
or 386.088 in/sec
2
Do not use local value of g to convert to mass!
Calculating the Moment of Inertia
MOI, sometimes called the second moment, for a point mass
around any axis is: I = Mr
2
where I = MOI (slugft
2
or other masslength
2
units)
M = mass of element (Slugs or other mass unit)
r = distance from the point mass to the reference axis
Radius of Gyration
The moment of inertia of any object about an axis through its CG can be
expressed by the formula: I = Mk
2
where I = moment of inertia
M = mass (slug) or other correct unit of mass
k = length (radius of gyration) (ft) or any other unit of length
The distance (k) is called the Radius of Gyration. The method of calculating
radius of gyration is outlined in the following sections.
Consider first the body consisting of two point masses each with a mass of
M/2 separated by a distance of 2r. The reference axis is through a point
equidistant from the two masses. The masses each have a MOI of Mr
2
/2. Their
combined MOI is therefore Mr
2
. The second example shows a thin walled tube of
radius r. By symmetry, the CG lies on the centerline of the tube. Again, all the mass is
located at a distance r from the reference axis so its MOI = Mr
2
. In these examples, the
radius of gyration is k = r. This leads to the definition:
"The radius of gyration of an object, with respect to an axis through the CG, is the
distance from the axis at which all of the mass of an object could be concentrated without
changing its moment of inertia. Radius of gyration is always measured from CG."
Figure 9
Figure 10
Page 14
Parallel Axis Theorem
If in the example above we wanted to determine the MOI of the
object about the axis X
a
rather than the axis X, through the CG,
then the value can be determined using the parallel axis theorem:
I
a
= I + d
2
M, Since I = k
2
M, then I
a
= M (d
2
+ k
2
)
where k is the radius of gyration.
This parallel axis theorem is used very frequently when calculating
the MOI of a rocket or other aerospace item. The MOI of each
component in the rocket is first measured or calculated around an axis through its CG,
and the parallel axis theorem is then used to determine the MOI of the total vehicle with
these components mounted in their proper location. The offset "d" is the distance from
the CG of the component to the centerline of the rocket.
Useful Approximations
Since the moment of inertia of an object displaced from its
reference axis is proportional to (d
2
+ k
2
), we can make two
observations that will simplify the job of calculating MOI:
RULE 1. If the radius of gyration of an object is less than 1% of
its offset distance "d", then the MOI of the object around its CG
can be ignored when calculating total MOI, and the value becomes
d
2
M. For example if a gyro with a mass of 0.1 slug is located near
the outer surface of a rocket and the offset to the CG of the gyro is
3 feet while the radius of gyration of the gyro is only 0.02 ft, then
the MOI about the center line of the rocket due to the gyro is d
2
M
= 0.9 slugft
2
. The error using this approximation is less than 0.01%.
RULE 2. If the radius of gyration of an object is more than 100 times its offset distance
"d", then the offset of the object can be ignored when calculating total MOI, and the
value becomes k
2
M. For example if a rocket motor with a mass of 100 pounds is located
near the center line of the rocket and the offset to the CG of the rocket motor is 0.100
inches, while the radius of gyration of the rocket motor is 12 inches, then the MOI about
the center line of the rocket due to the rocket motor is k
2
M = 14400 lbin
2
(or more
properly 37.3 lbinsec
2
). Again the error of approximation is less than 0.01%
Rule 2 can also be applied to alignment errors when calculating or measuring MOI. If
the offset or misalignment is less than 1% of the radius of gyration, then the alignment
error is insignificant.
K
MK
Figure 11
Figure 12
Page 15
Combining Moment of Inertia of Two Objects
If the object contains more than one mass, then the moment of inertia
is the sum of the individual moments of inertia taken about the same
axis. The radius of gyration is:
,
`
.

·
total
total
M
I
k
The moment of inertia of the two examples (fig. 13) is the same.
Note that it makes no difference what angle the masses have relative
to each other. Radius is the only factor affecting their moment of
inertia.
These examples illustrate that moment of inertia depends only on the
radius of the masses within an object. However, if the object were
flying in space, since the CG, radius of gyration, and principle axis
would be different for the two examples, their flight characteristics
would differ.
Basic Formula Using Differential Elements of Mass
The basic technique for calculating moment of inertia of an object is
to consider each element of mass and its radius, apply the formula I =
Mr
2
to each, and then add up all the moments of inertia of the
elements.
If this were done as described, then the computation would be of the
form:
I
1
= M
1
(r
1
)
2
where r
1
is the radius of M
1
, etc.
I
2
= M
2
(r
2
)
2
I
n
= M
n
(r
n
)
2
etc.

I = Total of above
If the object is a homogeneous solid, then this process can be accomplished by choosing a
suitable differential element and integrating over the limits of the radius:
Figure 13
Each weight = 1 lb.
Radius from X axis=2 in.
I
x
= 1 x 2
2
+1 x 2
2
= 8 lbin
2
The MOI about the X axis is
the same for both examples
I r dM ·
∫
2
Page 16
Combining axial MOI values
If the axial moment of inertia of two cylindrical rocket sections about their mutual
centerline are 10 slugft
2
and 20 slugft
2
respectively, then the total moment of inertia of
both sections when assembled is 30 slugft
2
. Moment of inertia values are simply added
to obtain the total. Before adding the values, make certain that they are both calculated
about axes which are coincident when assembled and that the units for each are consistent
and correct. Alignment is relatively unimportant. The moment of inertia error due to
misalignment is proportional to the ratio of the square of the misalignment offset to the
square of the radius of gyration of the object. For example, if a rocket has a radius of
gyration of 15 inches, and it is laterally misaligned by 0.002 inch, then the resulting error
is only 0.000,002% (2 millionths of 1 percent)!
Combining Transverse MOI Values
The combination of MOI around transverse axes is a more complex procedure.
1. The MOI of each component around an axis through its own CG parallel to the desired
axis must be determined by computation or measurement
2. The location of the composite CG must be calculated
3. The MOI of each component around the composite CG must be calculated using the
parallel axis theorem
4. The MOIs are added to find the total MOI around the desired axis through the
composite CG.
Standard Shapes
The moment of inertia of a variety of standard shapes has been published in most of the
textbooks and handbooks that cover dynamics such as the SAWE Handbook,
Machinery's Handbook etc.
Moment of Inertia of Objects Similar to Standard Shapes
Since total moment of inertia can be calculated by simply summing the values of the
component parts, it is possible to derive the MOI of many shapes by modifying the values
for standard shapes. This often eliminates the need for calculus and greatly speeds up the
calculation of MOI. For example to determine MOI for a hollow cone, note that MOI of
the inner conical space can be subtracted from the outer as a technique for simplifying
calculations even though there is no such thing as negative moment of inertia.
Composite MOI Example
An example finding composite MOI around the Z (longitudinal) axis is shown using a
reentry vehicle consisting of a hollow cone and other components. The parallel axis
theorem is used to calculate I
z
for off center components.
Page 17
CALCULATE MOI AROUND Z AXIS
200 lb(15)
2
in
2
ft
2
sec
2
I
A
= = 9.6 lbftsec
2
= 9.6 slugft
2
144 in
2
32ft
450 lbin
2
sec
2
ft
I
B
= =0.097 lbftsec
2
=0 .1 Slugft
2
386.088 in 12 in
Ic = 25 Slugft
2
2
2
15 . 0
12
sec 8 . 1
ft Slug
in
ft in lb
I
o
D
− ·
− − −
·
Iz TOTAL = 9.6 + .1 + 25 + 2.03 + 1.92 = 38.65 SlugFt
2
Effects of Misalignment
When misalignment results in tilt errors, or if other effects cause the CG as well as MOI
to change, more complex analysis must be performed. This will be discussed further
under "Product of Inertia".
I
WK
g
W lb K in
I lb in
I Slug Ft
I lb in
I lb in
A
B
C
D
E
· · ·
· −
· −
· − −
· − −
2
2
2
2
2
200 15
450
25
18
35
. sec
. sec
2
2
03 . 2
12
12
32
60
15 . ft slug I
z
D
− ·
,
`
.

,
`
.

+ ·
2
2
92 . 1
12
10
32
75
29 .
) (
ft Slug I
z
E
− ·
,
`
.

,
`
.

+ ·
2
2
29 . 0
12
sec 5 . 3
ft Slug
in
ft in lb
I
o
E
− ·
− − −
·
Page 18
Calculating Product of Inertia
General Comments
Consider the homogeneous balanced cylinder to which two equal
weights have been attached 180
o
apart, and spaced equidistant
along the length from the CG of the cylinder. The addition of
these weights will not alter the CG of the cylinder, and the cylinder
remains statically balanced. However, if we spin this cylinder
about the vertical Z axis, then centrifugal force acts through the
two weights and produces a couple. If the cylinder is mounted on
bearings, then this couple causes a sinusoidal force to be exerted
against the bearings as the cylinder rotates. If the cylinder is
spinning in space, then the axis of rotation of the cylinder shifts to
align itself to a condition where the centrifugal forces are equalized (i.e.,
it shifts toward the unbalance weights slightly). The mass distribution
which results in a couple moment when the object is spinning is called
"product of inertia."
Basically, product of inertia ("POI") is a measure of dynamic unbalance. POI is
expressed in the same units as moment of inertia, but it can have either a positive or
negative polarity. Product of inertia is generally not covered in undergraduate dynamics
courses, and consequently many engineers are unfamiliar with this concept.
Choosing the Reference Axis Location
Like center of gravity, three mutually perpendicular reference axes are necessary to
define products of inertia (only one axis is necessary to define moment of inertia).
Although any axis can be chosen as a reference, it is generally desirable to select the axis
of rotation of the object as one axis. If the object is mounted on bearings, then this axis is
defined by the centerline of the bearings. If the object flies in space, then this axis is
often defined by the location of thrusters. On reentry vehicles, the axis may be
coincident with the path of flight resulting from axis of symmetry of the outer surface of
the vehicle. If the reference axis will be used to calculate product of inertia of a complex
shape, choose an axis of symmetry to simplify the calculation. This axis can later on be
translated to another axis if desired, using the rules outlined in the section entitled "POI
Parallel Axis Theorem."
Polarity of Product of Inertia
Values for product of inertia can be either positive or negative, and in fact their polarity
depends on the choice of reference axis location. In this respect, POI is similar to CG.
Values for moment of inertia can only be positive, just as mass can only be positive.
Generally, the product of inertia of one component is offset by a negative product of
inertia due to another component, so that the composite product of inertia of a composite
object will be much smaller than the product of inertia of many of its elements.
Figure 14– CGz=0, but
weights cause couple
moment
Page 19
Units of Product of Inertia
Product of inertia is expressed in the units of mass times distance
squared. For CG calculations we used the weight of the object;
product of inertia calculations use the mass. Unfortunately, the word
"pound" can mean either weight or mass, so the engineer must use
caution when applying product of inertia values to engineering
equations. (See section entitled “Units of Moment of Inertia”).
Sometimes engineers will encounter bogus units for product of
inertia such as "ozinch." Although such units are incorrect, they
have meaning in context with the machine used to measure dynamic
unbalance. This machine may have a readout for a specific
correction plane height which gives the moment required at this
plane height to reduce the product of inertia to zero. This data can
be converted to valid units by multiplying the "ozinch" moment by
the height between the correction plane and the test object CG (and
then converting the weight in ounces to a unit of mass). This is
explained in more detail in the section on correction of dynamic
unbalance.
Principal Axis
On any object there will be three mutually perpendicular axes
intersecting at the CG for which the products of inertia will be zero.
For a perfect cylinder, these axes correspond to the centerline of the
cylinder plus two mutually perpendicular axes through the CG at any
orientation (since the cylinder has perfect symmetry). These axes
are called the "principal axes." The moment of inertia of the object is at a
maximum about one principal axis and at a minimum about another
principal axis. A spin stabilized vehicle will rotate about a principal axis
(usually the axis of minimum moment of inertia).
Calculating Product of Inertia
A perfectly balanced cylinder rotates on a set of bearings. A small weight
whose POI is zero is mounted on this cylinder. The product of inertia due
to this weight is:
P
zx
= M Z X = 0.01 x 2 x 1 = 0.02 slugft
2
where M = mass of weight = 0.01 slug
X = radius of CG of weight = 1 foot
Z = height between CG of cylinder & CG of weight = 2 feet
Figure 16  calculation of P
zx
& P
zy
Figure 15  X, Y and Z are
principal axes
Page 20
This calculated POI is in the XZ plane of the cylinder. If a similar weight were then
added on the Y axis at a location above the CG of the cylinder, the value for P
zx
would
not change, since the X coordinate of this weight would be zero. The second POI
component, P
zy
, could be calculated as shown below.
The product of inertia due to this weight is: P
zy
= MZY = 0.01 x 2 x (1) = 0.02 slugft
2
where M = mass of weight = 0.01 slug
Y = radius of CG of weight = 1 foot
Z = height between CG of cylinder & CG of weight = 2 feet
Note that the value for P
zy
is negative.
Rectangular to Polar Conversion
The previous examples were for a special case where the unbalance was located directly
on either the X or Y axis. When this occurs, then the mathematics is simplified, because
the unbalance can be analyzed as a twodimensional problem on a plane rather than a
threedimensional problem. A realistic object generally contains a couple unbalance that
does not fall directly on any axis. However, this unbalance can be converted into
rectangular components that fall directly on the axes, so that simplified calculations are
possible
The real object under test is shown in A. The unbalance equivalent of this object can be
simulated by a single weight in each of the two planes as shown in B (angular difference
between upper and lower planes can be any angle). As an aid to analysis, each single
weight in the previous example can be replaced by two weights located on the X and Y
axes as shown in C (i.e., polar to rectangular conversion). Each plane can now be
analyzed separately.
At the conclusion of all calculations, the resulting P
zx
and P
zy
can then be converted
back into polar coordinates if desired. The procedure for these rectangulartopolar
transformations is described in the section of this paper dealing with center of gravity.
Figure 17 – Any POI can be simulated by two weights in upper plane and two
weights in lower plane
Page 21
The product of inertia of the two components in the previous example can be resolved
into a single resultant P
zr
in the ZR plane which passes through the equivalent unbalance
mass and the Z axis.
2 2
zy zx zr
P P P + ·
Angle between resultant and X axis = arcTAN (P
zy
/ P
zx
)
Difference between CG Offset and Product of Inertia
The figures illustrate the difference between static unbalance (CG
offset) and couple unbalance (product of inertia). In the example
shown in figure 18, a 5 lb weight is added in the plane of the CG,
creating a static unbalance but no product of inertia.
P
zx
= 0 lbin
2
CG
x
= 25 lbin
CG
z
= 0 lbin
In the example shown in figure 19, a weight is added outside the
plane of the CG, creating both static and dynamic unbalance
(product of inertia not zero). (This is sometimes called "quasistatic
unbalance," since a single correction weight can be used to correct
for this unbalance.)
P
zx
= +75 lbin
2
CG
x
= +25 lbin
CG
z
= +15 lbin
In the example in figure 20, a second weight is added at 180
o
at the
bottom of the cylinder in the above example, creating a static balance
but not a dynamic balance.
P
zx
= 5 lb x 3 in x 5 in + 5 lb x (3) x (5) = 150 lbin
2
CG
z
= +15 lbin  15 lbin = 0
CG
x
= +25 lbin  25 lbin = 0
Figure 18 CG offset
Figure 20 – Static &
dynamic unbalance
Figure 19 – Quasistatic unbalance
Page 22
In figure 21, we move this lower weight up to the plane
of the first, creating both static and dynamic balance
about z.
P
zx
= +75 lbin
2
 75 lbin
2
= 0
CG
z
= +15 + 15 = 30 lbin
CG
x
= +25  25 = 0 lbin
The previous discussion has assumed that the small
unbalance weights were perfectly symmetrical, and
therefore the product of inertia of the weight itself could be
ignored. In real life, the "weights" consist of various
components of a rocket or spacecraft, and their product of
inertia is usually not zero. Even if the component products of inertia are small, they
cannot be ignored since the product of inertia of the total vehicle is usually very small,
and even the smallest couple unbalance can tilt the principal axis of the vehicle.
POI Parallel Axis Theorem
When determining the product of inertia of a vehicle, it will be
necessary to first calculate or measure the product of inertia of the
component parts of the vehicle, and then translate these values to
the effective POI about the axes of the vehicle.
To translate the product of inertia of an object relative to the X', Y',
Z' axes to the X, Y, Z axes:
P P M z x
P P M z y
zx z x
zy z y
· +
· +
′ ′
′ ′
where M = mass of object
x, y, z = CG offsets from coordinate origin along the X, Y, Z axes
It is much more difficult to use this theorem than the equivalent one for moment of
inertia, because there are two formulas required, and because each term has a polarity
associated with it. The following example illustrates this type of translation:
Example for the illustration shown, let z = 4, x = +5, and y = 6 inches. The product of
inertia of the object about the X', Y', Z' axes is P
z'x'
= 2 lbinsec
2
, P
z'y'
= 0 lbinsec
2
.
The weight of the object is 4 lbs. Calculate the effective product of inertia of the object
relative to the X, Y, Z axes:
Figure 22– POI can be
translated to another set of
parallel axes
Figure 21 – Static & dynamic balance
Page 23
Calculate mass:
Comparison Between MOI and POI
There are some similarities and some differences between this axis translation formula
and the formula to translate moment of inertia to a different parallel axis:
1. Both formulas are dimensionally similar:
(mass) (length)
2
= (mass) (length)
2
+ (mass) (length)
2
However, the polarity of the values of (mass) (length)
2
can only be positive for moment
of inertia, whereas they can be either positive or negative for product of inertia.
2. In the case of moment of inertia, it is possible to ignore the MOI of the object about its
CG if the translation term is large. This is not so for product of inertia!
If the POI of the object about its own CG is not zero, it cannot be
ignored, even if the translation term value, Mxy, is large, since small
values of product of inertia can be very significant if one large term is
subtracted from another, leaving a small difference.
3. In the case of moment of inertia, the value for the MOI about the CG
of an object always has a value greater than zero. It is possible for the
product of inertia of an object to be zero, so that the translation formula
becomes P
zx
= M z x.
4
386088
001036
2
( . )
. sec
·
− lb
in
P P Mz y
zy z y
· +
′ ′
P lb in
zy
· + − − · − − 0 001036 4 6 024864
2
( . )( )( ) . sec
P P Mz x
zx z x
· +
′ ′
P lb in
zx
· − + − + · − − − 2 001036 4 5 2 2072
2
( . )( )( ) . sec
Figure 23 P
zx
is zero for all of
the examples above
Page 24
Axes and Planes of Symmetry
The product of inertia of a homogenous body with respect to any pair of perpendicular
axes is EQUAL TO ZERO if the plane determined by either of the axes and the third
coordinate axis is a plane of symmetry of the body. This rule is hard to visualize when
put into words. The examples on the right illustrate some symmetrical shapes that have a
zero P
zx
.
Determining Product of Inertia of a Volume
The basic concept of determining the product of inertia of a homogeneous volume is
identical to the previous method involving discrete objects, except the objects are now
differential elements of a solid. The formula becomes:
P M yxdV
yx
·
∫
where M = total mass of object; dV = differential volume
As in the case for moment of inertia and center of gravity, the
solution to the problem can be simplified by choosing the right
differential element. For example, the elliptical wing tip shown
can be analyzed using a small square element dX by dY. This
leads to a double integral. If a rectangular slice parallel to the X
axis is chosen instead, then the POI of the element is zero, and the
product of inertia of dA is:
The CG of the rectangular slice is at x/2
From the equation of an ellipse:
x
a b y a
b
2
2 2 2 2
2
·
−
Therefore:
( . )
.
0 05
05
2
+
·
·
∫
x y dA
Since dA x dy
then P x y dy
zy
y
x
x
Figure 24 – Calculation of POI
P
a b y a
b
y dy
xy
·
−
∫
05
2 2 2 2
2
.
Page 25
Combining the POI of Two Bodies
If two sections of a rocket are combined, what is the resulting
product of inertia? If the sections are aligned perfectly, so
that the Z reference axis for the lower section is exactly
coincident with the Z reference axis for the upper section,
then the following method can be used:
1. Transform the product of inertia of the lower section into
values for P
zx
and P
zy
(polar to rectangular conversion). If
the X and Y axes for the upper section do not correspond to
those chosen for the lower section, rotate the data by
converting temporarily in polar form and then back into the
new rectangular axes, so that upper and lower axes are at the
same angular location. Analysis will be done by planes and
transformed into polar coordinates if desired after all
calculations have been made.
2. Sum the values for P
zx
upper and P
zx
lower. Do the same
for P
zy
. Note: observe the polarity of the data! This will
yield new product of inertia values for the composite vehicle.
(The values for the total can be larger or smaller than the
individual values.)
Effect of Lateral Misalignment
What happens if the two sections are not aligned, so that the axes are parallel to
each other, but the axis of one is not coincident with the axis of the other? Here
is the method we recommend for analyzing this type of problem:
1. The location of the upper section must be defined in terms of the lower
reference axis. This requires the measurement of offset of the upper section
when both sections are assembled. This can be accomplished by placing the
total rocket on a rotary table and dial indicating both lower and upper section; or
if this is not possible, then measurements can be made of the individual sections
and their interface ring concentricity and the offset calculated (not as accurate a
method).
2. Calculate the X, Y, and Z coordinates of the CG of the total rocket. The new
reference axis for the total rocket will pass through the new CG and be parallel
to the reference axes of the two sections. The centerline of the two sections will
be offset from this axis by some small but not insignificant amount.
Often, the magnitude of POI resulting from misalignment will be greater
than the POI of the individual sections, so that it is not satisfactory to
ignore this effect. The X or Y axis distance will be small, but the mass
will be very large since it is the entire mass of a section.
Figure 25 – Combining POI
Figure 26 – Lateral misalignment
Page 26
3. P
zx
relative to the new combined axis due to the P
zx
of the upper section may be
calculated by using the parallel axis translation formula:
where P
zx
= POI relative to combined CG
P
z'x'
= POI relative to upper section
M = mass of upper section x
z = distance between composite CG and upper section CG
x = offset between new combined reference and upper reference
In the view shown, both z and x are positive, so that the POI due to the upper offset is
positive.
4. Repeat the calculation for the lower section. In the view shown, both z and x are
negative, so that the POI due to the lower section offset is also positive.
5. Sum the values for P
zx
upper and P
zx
lower to yield P
zx
total.
6. Repeat Steps 3, 4, and 5 for P
zy
.
7. If desired, convert P
zy
and P
zx
into P
zr
, the resultant polar representation.
Effect of Tilt on MOI and POI
If a perfectly balanced cylinder is tilted by an angle "a," then P
zx
, I
z
, and
I
x
will change. As the cylinder leans to an angle of 90
o
, I
z
becomes I
x
,
and I
x
becomes I
z
. For a lean angle of 0
o
, P
zx
is zero. For a lean angle
of 45
o
, P
zx
is a maximum determined by the values of I
z
and I
x
. At a
lean angle of 90
o
, P
zx
is again zero. This unique relationship between
moment of inertia and product of inertia is discussed in the SAWE paper
No. 1473 entitled "Determining Product of Inertia Using a Torsion
Pendulum." This paper outlines a method of measuring POI of objects
using a moment of inertia instrument.
When the MOI or POI of an object is determined about its center line,
and the object is then installed in a vehicle in such a way that there is a
lean angle between the center line of the object and the reference of the
vehicle, then it is very useful to be able to convert the calculated values for the object into
mass properties relative to the new reference without having to recalculate the object
itself. These "lean angle formulas" are given below.
P P Mz x
zx z x
· +
′ ′
Figure 27– Axis Tilt
Page 27
MOI Inclined Axis Formulas
For the balanced cylinder shown, the moment of inertia about an axis Z' displaced from
the center line of the cylinder by an angle "a":
( ) I I I I I a
z z x z x ′
· + + − 05 05 2 . ( ) . cos( )
Note that for this example, P
zx
was zero. This formula is only valid for the orientation of
the axes shown. There are some interesting observations to be made regarding the
change in MOI:
1. The MOI at an angle of 45
o
is the average of I
z
and I
x
.
2. The sensitivity to tilt angle is a function of the difference between I
x
and I
z
. If there is
very little difference, then tilt angle can be ignored. If the object is tall and slender, then
tilt angle is very critical.
In addition to being useful in the calculation of MOI, this formula also can be used to
determine fixturing accuracy required when measuring MOI. When measuring tall,
slender rockets, the axial MOI error will be large unless the rocket is fixtured very
carefully. Transverse MOI (I
y
and I
x
) can be measured on a vee block fixture without the
need to adjust the rocket's position, since the sensitivity to lean is very small in this case.
MOI Incline Axis with POI
The previous analysis assumed that the Z and X axes were principal axes and the POI
was zero. If this is not the case, then the formula becomes:
( ) ( ) I I I I I a P a
z z x z x zx ′
· + + − + 05 05 2 2 . . cos( ) sin( )
This formula reflects the fact that the principal axis is no longer through the centerline of
the cylinder, so that the maximum and minimum MOI are no longer axes X and Z.
The formulas presented also assume that there is no tilt in the Y direction, so that the
problem can be analyzed from a twodimensional standpoint. If this is not the case, then
the coordinate system must be transformed so that it is. Furthermore, note that the origin
of the two axes is at the CG of the object. Equations can be written for the more general
case. However, it is easier to manipulate the axes than to solve the general equations.
Page 28
POI Inclined Axis Formulas
Since POI and MOI are related, it might be assumed that a similar formula can be written
for the POI of an object when tilted. In this case, we are starting with a value of P
zx,
which is zero and returning to a value of zero at an angle of 90
o
. The formula is:
P I I a P a
z x x z zx ′ ′
· − − 05 2 2 . ( ) sin( ) cos( )
If the X and Z axes are principal axes, then the formula becomes:
P I I a
z x x z ′ ′
· − 05 2 . ( ) sin( )
These formulas are only valid for the orientation of the axes shown (fig. 27) and for the
direction and definition of positive tilt angle (CCW from the Z axis).
Mohr's Circle
A graphical representation of the relationship between MOI and POI was originated in
the 19
th
century by a German engineer, Otto Mohr. A copy of this aid is reproduced from
the SAWE handbook and is shown on the next page. With the advent of the personal
computer, graphical solutions to engineering problems are no longer necessary; but
Mohr's circle still is useful in visualizing the effect of tilt.
Mohr's Circle for Moments of Inertia
Given (1) The moment of inertia values I
X,
I
Y
for an object about its center of gravity,
where the center of gravity lies at the origin of a set of mutually perpendicular
axes XY.
(2) The corresponding value for the product of inertia, P
XY
Mohr's circle is then constructed using the layout geometry shown below. The following
information may then be obtained.
(1) The location of the principal axes about which the moments of inertia
are maximum and minimum and the products of inertia are zero.
(2) The corresponding maximum and minimum values of moments of
inertia.
(3) The moments and products of inertia for any other set of mutually
perpendicular axes AB whose origin lies at the center of gravity of the
given object and rotated C degrees from the original axes XY (reference,
the figure to the right).
(4) The maximum values for the products of inertia about axes located 45
o
from the
principal axes.
Figure 28
Page 29
Layout Geometry
( )
The Radius of the circle is R
I I
P
x y
xy
: ·
−

.
`
,
+
2
2
2
Figure 29 – Mohr’s Circle
Page 30
Effect of Angular Misalignment (Tilt)
If the upper section is tilted relative to the lower section, then two
factors tend to increase the effective POI of this section: the tilt results
in a CG offset similar to the case described previously, and the tilt also
alters the POI of the upper section itself. The method for calculating
the total POI is as follows:
1. Using the center line of the lower section as a reference, calculate
the Y axis offset of the CG of the upper section from the
formula:
Y = H sin a
Where: H is the CG height of upper section
a is the tilt angle in the zy plane
Using a similar concept, calculate the X offset of the upper section.
2. Calculate the X, Y, and Z coordinates of the CG of the total rocket. The new
reference axis for the total rocket will pass through the new CG and be parallel to the
reference axes of the lower section.
3. Recalculate the P
zx
of the upper section by applying the axis tilt formula. Add this
POI to the P
zx
of the upper section relative to its center line (observe signs; value for P
zx
may be either larger or smaller than value without considering tilt).
4. P
zx
relative to the new combined axis due to the P
zx
of the upper section may be
calculated by using the parallel axis translation formula:
P P Mz x
zx z x
· +
' '
where P
zx
= POI relative to combined CG reference axes
P
z'x'
= POI relative to upper section after effect of tilt has been added
M = mass of upper section
z = distance between composite CG and upper section CG
x = offset between new combined reference and upper reference
In the view shown, both z and x are positive, so that the POI due to the upper offset is
positive.
5. Repeat the calculation in Step 4 for the lower section. Since this section is not tilted,
the P
z'x'
is the value through the center line. In the view shown, both z and x are
negative, so that the POI due to the lower section offset is also positive.
X
Figure 30 – Angular Misalignment
Page 31
6. Sum the values for P
zx
upper and P
zx
lower to yield P
zx
total.
7. Repeat Steps 3, 4, 5 and 6 for P
zy
.
8. If desired, convert P
zx
and P
zy
into P
zr
, the resultant polar representation.
Angle of Inclination of Reentry Vehicle
An aerospace vehicle will often have an axis defined by the minimum air
resistance of the vehicle. This axis corresponds to the axis of symmetry of
the outer surface of the vehicle. Since the vehicle is not homogeneous, the
product of inertia about this axis may not be zero, resulting in a principal
axis at an angle to the axis of symmetry of the vehicle. The angle is known
as the "angle of inclination" of the vehicle. Generally, it is desirable to
make this angle as small as possible so the vehicle will "fly straight".
Sometimes, however, this angle is deliberately adjusted to a specific value,
so a reentry vehicle will have a "coning" motion upon entering the
atmosphere, and the resulting drag will slow the reentry.
Angle of inclination may be calculated using the following formula:
Example: Given P
zx
= 0.002 lbinsec
2
I
xx
= 8.95 lbin
2
I
zz
= 2.40 lbin
2
What is the angle of inclination in the XZ plane? First, the units of moment of inertia
must be converted to lbinsec
2
to be consistent with the units for product of inertia (or
the units for product of inertia could be converted into lbin
2
):
I lb in lb in
xx
· − · − − 895 0023155
2 2
. . sec
I lb in lb in
zz
· − · − − 2 40 0006216
2 2
. . sec
A arcTAN
P
I I
zx
zz xx
·
−
05
2
.
(
)
Figure 31 – Angle of
Inclination
Page 32
Then, using the formula:
A arcTAN
P
I I
A arcTan rees
zx
zz xx
·
−
]
]
]
]
·
−
]
]
]
· −
05
2
05
0004
0006216 0023155
674
.
(
.
.
( . .
. deg
)
The previous example was for a single plane. This would also be the angle of inclination
for the object if the product of inertia of the ZY plane were zero. If there were a product
of inertia for the ZY plane, then the combined solution for the object would be:
A arcTAN
P
I I
zr
zz rr
·
−
]
]
]
05
2
.
( )
where P
zr
is the resultant of P
zx
and P
zy
.
I
rr
is the moment of inertia about axis "r" (resultant)
For a typical projectile or reentry vehicle, I
yy
= I
xx
, so that I
xx
can be used in place of
I
rr
. For a vehicle with wings, this is not the case, and I
rr
must be calculated from the
values of I
yy
and I
xx
. (See section on moment of inertia.)
Controlling Tilt Angle
Since the amount of axis tilt which results from a given product of inertia is a function of
the difference between I
zz
and I
xx
, the effect of a product unbalance can be adjusted by
altering the difference in moment of inertia. This leads to two conclusions:
1. If you want to stabilize the vehicle and have it resistant to the effects of product
unbalance, make the difference in moment of inertia as large as possible. This is
accomplished by designing the vehicle to be long and slender, and by placing the heavy
items near the ends of the vehicle.
2. If you want to steer the vehicle with as little correction force as possible, then make
the moment of inertia difference as small as can be tolerated. The minimum moment of
inertia difference will be limited by the skill with which you can correct product of inertia
unbalance (a function of the sensitivity of the balancing machine used, and the stability of
the components in the vehicle).
Page 33
Unbalance of Rotating Objects
Unbalance Forces Due to Offset CG of Rotating Object
If a rotating object is mounted on bearings, then CG offset from the axis of rotation will
produce a sinusoidal force on the bearings. This unbalance force will cause premature
wear of the bearings, noise, additional friction, and may lead to errors if the rotating
object is part of a guidance system.
The force exerted by a CG offset is a function of rotation speed and the magnitude of
unbalance moment:
F Mr w ·
2
where F = unbalance force in pounds
M = mass in slugs = W/g
r = CG offset in feet
w = angular velocity in radians per second
If we examine the force exerted on a bearing system along a single radial axis, it varies
sinusoidally:
F Mrw a ·
2
sin
where a = angle of rotation relative to the axis
Since a rotating system usually has two bearings, the force on each bearing would be a
proportion of the total force (if the CG were equidistant from both bearings, then half the
force would be applied to each bearing).
This analysis assumes that the product of inertia is zero. Such would be the case if the
object were a thin flywheel, or if the object were dynamically balanced and a single
weight was then added in a plane perpendicular to the axis of rotation and passing
through the center of gravity of the object. If the product of inertia is not zero, then the
forces on the bearings will be different from this example. The following section
describes the forces due to product of inertia unbalance ("couple unbalance").
Unbalance Forces Due to POI of Rotating Object
For a CG offset, the forces on the bearings do not depend on the spacing between the
bearings but do depend on the axial location of the CG of the rotor. If the rotor CG is
located between the bearings, then the unbalance force is in the same direction for both
upper and lower bearings; but the magnitude of the force is proportional to the CG
location relative to the bearings and is, in general, different for each bearing. In contrast,
a product of inertia unbalance will result in equal and opposite forces on the bearings and
the force is not a function of the CG location. For a given POI, the magnitude of the
bearing force will increase as the spacing between the bearings is made smaller.
Page 34
One method of calculating the force due to product unbalance is to first determine the
equivalent mass at the bearings that would result in the magnitude of POI. The force on
the bearings can then be determined using the formulas given above for CG offset.
If a shaft with a P
zx
of 100 lbin
2
is supported on bearings which are 10 inches apart,
then the equivalent product of mass times distance at each bearing is 10 lbin. If the
rotation speed is 300 RPM, then:
300
300
60
5 5 2 31416 31416 RPM rev x x rad · · · · / sec . . / sec
10
10
32174 12
00259 lb in
x
slug ft − · · −
( . )
.
F Mrw slug ft x lbs force peak · · − ·
2 2
00259 31416 2556 . ( . ) . ( )
These relationships form the basis for a spin balance machine. An analysis of centrifugal
forces acting against the bearings of a spin balance machine results in measured values
for center of gravity offset and product of inertia.
Mass vs Weight (and English vs Metric)
In 1999, the Mars Climate Orbiter crashed as a result of a confusion over the system of
units. The software program which controlled the thrusters was supplied with thrust data
in poundseconds but interpreted it as if it were newtonseconds, result in an
underestimation of the thruster impulse by a factor of 4.45. This is only one of many
thousands of errors that have occurred as a result of the confusion between Metric and
English units.
Many of these errors are the result of a misunderstanding regarding the difference
between mass and weight. If you place an object on a scale in Europe, you will read its
mass (generally expressed in kg). However, if you place an object on a scale in the USA,
you will read a value equal to the force exerted by the acceleration of gravity (generally
expressed in lbf). Since you are really trying to use the scale to measure mass, when you
weigh yourself on an American bathroom scale, it should read 6.22 slug rather than 200
lbf.
Traditionally, a dimensionally inconsistent correction factor is used to convert from one
set of units to the other. The expression 1 kg = 2.205 lb is not valid. It is like comparing
apples to oranges. Mass does not equal force. This traditional conversion factor is based
on the value of standard gravity, which is 9.80665 m/sec
2
.
Page 35
Mass is related to weight through Newton's second law:
where W = the weight of the object (gravity force)
M = the mass of the object
g = the acceleration of gravity
MASS is the QUANTITY OF MATTER in an object (its inertia), while WEIGHT is the
FORCE that presses the object down on a scale due to the acceleration of gravity. The
mass of an object is a fixed quantity; its weight varies as a function of the acceleration of
gravity. The mass properties of an object are related to mass, not weight. Mass
properties do not change as a space vehicle leaves the attraction of the earth and enters
outer space.
If different names are used for weight and mass, then the problem of distinguishing
between the two is minimized. The Metric SI system uses the word "Newton" for weight
and the word "Kilogram" for mass. The Newton is defined as the force required to
accelerate a 1 Kilogram mass by 1 meter per second
2
. The aerospace industry has created
a unit of mass called the "Slug." A one pound force is required to accelerate a one Slug
mass at one ft/sec
2
. If an object weighs 32.17405 lbf on earth, then its mass is one Slug.
Unfortunately, not all systems of units adequately differentiate between mass and weight.
In the USA, the word "pound" is commonly used for both mass and weight, resulting in
endless confusion and errors in calculating mass properties and dynamic response.
Officially “pound” refers to mass (see, for example, NIST documents). However, the
common usage of the word pound is the value you read on a scale, which is actually lbf.
If the term "pound" is used to describe a mass whose measured weight is one pound
(force), this quantity MUST be divided by the acceleration of gravity in appropriate units
to convert it to proper mass dimensions if it is to be used in mass properties calculations.
Similarly, in metric countries the terms Kilogram and Gram are often, incorrectly, used
to describe force as well as mass. To avoid confusion and uncertainty, an analysis of
fundamental dimensions will confirm if correct units of measurement are being used and
if conversion factors are being applied correctly to achieve desired results.
The various metric systems are fundamentally MASS, LENGTH, TIME systems with
force being a defined or derived term. The U.S. systems are fundamentally FORCE,
LENGTH, TIME systems with mass being defined or derived. Table One shows the
three most commonly used systems of measurement. Time in seconds is used throughout
DIMENSIONALLY CORRECT MEASURING SYSTEMS
MASS LENGTH WEIGHT g
SI (Metric) Kg Meter Newton 9.80665 M/sec
2
U.S. (inch) Weight in lbf
386.0886 in.
Inch Pound (lbf) 386.0886 in/sec
2
U.S. (foot) Slug Foot Pound 32.17405 ft/sec
2
W Mg ·
Page 36
The U.S. inch system has no common name for the mass whose weight equals one pound,
although this is sometimes called a “pound mass”. One pound mass is equal to one pound
force divided by 386.088 inches/sec2. Applying W = Mg shows that the system is
dimensionally consistent.
1
1
38609
38609
1
2
2
lbmass x g
lb
in
x
in
lbweight ·
−
·
sec
.
.
sec
The acceleration of gravity used to convert weight to mass is a fixed number which has
been established as an international standard.
Conversion Factors
Standard acceleration of gravity
g = 32.17405 ft/sec
2
or 9.80665 m/sec
2
Dimensionally inconsistent conversion factors based on standard
acceleration of gravity
1 oz (ounce) = 28.349 52 gram
1 oz tr (troy ounce) = 31.103 48 gram
1 lbf (pound force) = 0.453 592 37 kg
1 kg = 2.204 622 6 lbf
1 kg = 9.806 65 Newtons
1 sh tn (short ton, US) = 907. 184 7 kg
1 ton (long ton, UK) = 1016. 047 kg
Force conversion factors
1 dyne = 10
5
N
1 lbf (poundforce) = 4.448 22 N
1 kp (kilopond) = 9.806 65 N
Mass conversion factors
1 kg = 0.068 521 76 lb mass
1 t (tonne, metric) = 68.521 76 lb mass
1 lb mass = 14.593 904 kg
1 t (tonne, metric) = 1000 kg
Space Electronics, Inc.
81 Fuller Way, Berlin, CT 06037
8608290001 fax 8608290005
www.spaceelectronics.com
info@spaceelectronics.com
Calculation of Mass Properties using Traditional Methods
Choosing the Reference Axes The first step in calculating mass properties of an object is to assign the location of the reference axes. The center of gravity and the product of inertia of an object can have any numerical value or polarity, depending on the choice of axes that are used as a reference for the calculation. Stating that a CG coordinate is "0.050 inches" means nothing unless the position of the reference axis is also precisely defined. Any reference axes may be chosen. For example, the center of gravity of a cylinder may be 4.050 inches from one end, 0.050 inches from its midpoint, and 3.950 inches from the other end. Furthermore, each end of the cylinder may not be perpendicular to the central axis, so that a means of determining the "end" of the cylinder would have to be further defined. Three mutually perpendicular reference axes are required to define the location of the center of gravity of an object. These axes are usually selected to coincide with edges of the object, accurately located details, or the geometric center of the object. It is not sufficient to state that an axis is the centerline of the object. You must also specify which surfaces on the object define this centerline. Moment of inertia is a rotational quantity and requires only one axis for its reference. Although this can theoretically be any axis in the vicinity of the object, this axis usually is the geometric center, the rotational center (if the object revolves on bearings), or a principal axis (axis passing through the center of gravity which is chosen so the products of inertia are zero).
Figure 1 Center of gravity (and product of inertia) are defined relative to orthogonal axes
Product of inertia requires three mutually perpendicular reference axes. One of these axes may be a rotational axis or a geometric centerline. For maximum accuracy, it is important to use reference axes that can be located with a high degree of precision. If the object is an aerospace item, then we recommend that this object be designed with two reference datum rings per section, which can be used to define the reference axes. These rings can be precision attachment points that are used to interface the object with another section of a spacecraft or rocket, or they can be rings that were Figure 2  Datum Rings provided solely for the purpose of alignment and/or measurement of mass properties. The accuracy of calculation (and the subsequent accuracy of measurement of an actual piece of hardware) is only as good as Page 2
the accuracy of the means of locating the reference axes. We have found that the single largest source of error in mass properties calculations is the uncertainty of the reference. The dimensional data provided to the mass properties engineer must be sufficiently accurate to permit mass properties tolerances to be met. For example, if you are asked to make precise calculations of mass properties of a projectile, you should establish the error due to reference misalignment as the first step in your calculations. If you are required to calculate CG within an accuracy of 0.001 inch and the reference datum is not round within 0.003 inch, then you cannot accomplish your task. There is no sense in making a detailed Figure 3 analysis of the components of an object when the reference error prevents accurate calculations. Furthermore, it will be impossible to accurately measure such a part after it is fabricated and verify the accuracy of your calculations. The location and accuracy of the reference axes must be of the highest precision. If your task is to calculate the mass properties of a vehicle that is assembled in sections, then serious thought should be given to the accuracy of alignment of the sections when they are assembled. Often this can be the biggest single factor in limiting the degree of balance (if the vehicle was balanced in sections because the total vehicle is too big for the balancing machine). Alignment error is amplified for long rockets . . . a 0.001 inch lean introduced by alignment error on a 12 inch diameter can result in a 0.007 inch CG error on a 15 foot long rocket section. This is discussed in detail in the sections of this paper that present the math for combining the mass properties of subassemblies. The accuracy required for various types of calculations is summarized in later sections of this paper.
The first step in calculating mass properties is to establish the location of the X, Y, and Z axes. The accuracy of the calculations (and later on the accuracy of the measurements to verify the calculations) will depend entirely on the wisdom used in choosing the axes. Theoretically, these axes can be at any location relative to the object being considered, provided the axes are mutually perpendicular. However, in real life, unless the axes are chosen to be at a location that can be accurately measured and identified, the calculations are meaningless.
Choosing the Location of the Axes The axes in Figure 3 do not make a good reference because a small error in squareness of the bottom of the cylinder causes the object to lean away from the vertical axis. The axes below (Figure 4) make a better choice.
Figure 4
Page 3
These are the three components of the vector distance from the origin of the coordinate system to the CG location. volume moments may also be used. Alternatively. ! the point where all the mass of the object could be considered to be concentrated when performing static calculations. and Z). The X offset moment of one element (such as MX1 = +3W1) can be easily added to the X offset moments of other elements of mass. and MZ1=+7W1) can be combined with similar Y and Z offset moments of other elements to determine the Y and Z components of the CG location. the sum divided by the total weight. ! the point where the object would balance if placed on a knife edge ! the single point where the static balance moments about three mutually perpendicular axes are all zero. ! the centroid of the volume of the object. The fundamental dimensions of moment are typically Force times Distance. Likewise. it can be accurately measured and is therefore a good reference location as can the center of a close tolerance hole which could be identified as the zero degree reference to identify the X axis (Fig. Unfortunately. and the result will be the X component of the CG location of the composite mass. Center of gravity of composite masses is calculated from moments taken about the origin. 4). the Y and Z offset moments (MY1=5W1. When combining mass elements.Reference axes must be located at physical points on the object that can be accurately measured. a useful technique is to use "offset moments" about each of the three orthogonal axes. ! the point about which the object rotates in free space ! the point through which the force of gravity can be considered to act ! the point at which an external force must be applied to produce pure translation of an object in space Center of gravity location is expressed in units of length along each of the three axes (X. Figure 5 Calculating CG Location General Discussion The center of gravity of an object is: ! also called the "center of mass" of the object. Mass moment may be used with any units of Mass times Distance. the term "X offset moment" is Page 4 . Care must be taken to be sure that moments for all elements are expressed in compatible units. Y. if the object is homogeneous. An axis should always pass through a surface that is rigidly associated with the bulk of the object. For homogeneous elements. Although the center line of a ring may exist in midair. In Figure 5 it would be better to locate the origin at the end of the object rather than the fitting that is loosely dimensioned relative to the end.
the center of gravity of a piece of pipe is on the centerline half way along its length. even though there is no metal in the center of the pipe (Figure 6). so that the center of gravity must be computed by summing the offset moments along each of the three axes. the CG of the object is on its centerline (since the CG of a homogeneous mass is at its centroid of volume). Assume that the element weights are. but like the term "pound mass" most engineers will understand the meaning.707 inch from A CG z = Total Weight 19. The center of gravity of an object can be located in "midair".445inch = 11279 lb − in . and in fact their polarity depends on the choice of reference axis location.613 lb.319 inch = 77.frequently described as "moment along X". The composite CG of an object can be computed if the CG of each component is known. most objects are not homogeneous.553 lb. Component distances for center of gravity location may be either positive or negative. The CG location along the length can be determined by summing moments about the reference axis at the bottom of the figure (x = 0).811 lb − in Total Moment 110. Examples follow. The center of gravity of a homogeneous shape is calculated by determining the centroid of its volume.124 lb − in Total Weight = 19.416 lbs Total Moment = 110. CG along a single axis Consider the round metal rod with two cylindrical weights shown (Figure 7). These processes are described in detail in the following sections.666 inch = 22. Note that the elements do not have to be the same diameter to be symmetrical along the length. From symmetry.250 lbs x 6. In fact the elements could overlap (such as sliding one pipe inside another).416 lb Figure 7 – CG along a single axis Page 5 .250 lb. This does not make mathematical sense. For example. Figure 6 M a = Wa za M a = 12. In real life. Wa=12.811 lb − in = = 5. M c = Wc zc M c = 2.613 lbs x 2. Wc=2.408 lb − in A M b = Wb zb M b = 4. Wb=4.553 lbs x 8.
CG y = − 2. 70 lb − in / 4. computation.5)(244.6(2.4(4.25lb − in / 4. The CG of each component is known by symmetry.8 in 3 CG = CG = a1V1 − a 2V2 V NET (5)(3351) − (4.25) + 2. When the cone is combined with other elements to find the overall CG. or measurement.CG of Unsymmetrical Three Dimensional Body The center of gravity of an unsymmetrical body may be calculated in the same manner as the single axis example above. the CG of a solid cone is given in the SAWE handbook. A convenient frame of reference is assigned. M x = M a + M b + M c = 0. in this case such that the CGs of each component fall on the axes. Volume moments are taken around the center of the base to find the centroid of the hollow cone.3) . . Using the observation that a hollow cone can be created by removing a small solid cone from a larger one.8lb = − 0. Dimensions shown are to the CG of each component from the origin. From symmetry. The CG distance along the length could be calculated using calculus.5) + 18(0. However. CG x = 0.1lb − in / 4. its actual weight and calculated centroid location are combined with those of the other elements.35 in 90. CG z = 9. 3 3 V2 = π r2 2 H2 π (3.8 lb = 0146 in . X Figure 8 – CG of an unsymmetrical body CG of a Complex Shape Similar to a Standard Shape Consider the hollow cone shown below.8 lb = 1896 in .4(175) + 0 + 0 = 0.1 lb − in . the CG lies along the center line. V1 = π r12 H1 π (4 2 )(20) = = 3351 . M y = M a + M b + M c = 0 + 0 + 18( − 125) = − 2.469in M z = M a + M b + M c = 0. and offset moments are summed along each axis.3 3 3 V NET = V1 − V2 = 90.25lb − in . Each axis may be considered separately (Figure 8). we can calculate the CG by subtracting the moment due to the smaller cone from the larger one.7 lb − in . Consider the cylinder with attached rectangles.62 )(18) = = 244.8 Figure 9 – Hollow Cone Page 6 . = 6.5) = 9.
0 in 4 Location of Centroid from tip = = = 13.8 in 3 OR 20 − 13. The basic concept of the calculation is the same as the previous examples.65 = 6.CG of a Complex Unusual Shape If you encounter a shape which is not in the handbook and which cannot be created from known shapes. so that triple integration can be avoided. we will use the same hollow cone discussed in the previous section. then it will be necessary to use calculus to calculate its CG.65 VT 90.8in 3 VT = x dx + ∫ (4 x − 4 )dx 2 ∫ 4X 2 0 3 2 2 2 4X 0 0 0 MT = π D0 2 2 3 π D0 2 20 x dx + ∫ 4 x 2 − 4 x dx = ∫ 2 4 X 02 4 X 02 0 ( ) 2 x 4 20 4 x 3 4 x 2 4 + − = 1240. The trick to simplifying this process is to chose the right differential shape. Do = 8 ′′ VT = X o = 20 ′′ x −2 ∫x−0 dV 1 + ∫ x −20 x −2 dV 2 MT = ∫ 2 π D1 dV 1 = dx 4 x −2 x −0 xdV 1 + ∫ x −20 x −2 xdV 2 Do ( x − 2) Xo D1 = Do X Xo dV2 = π 2 2 D1 − D2 dx 4 ( ) D2 = dV1 = 2 π D0 2 4X0 x 2 dx 2 π D0 2 2 4 4 dx dV2 = 2 x − x − x+ 4 X0 [ ( )] 2 2 3 2 2 20 20 πD0 2 2 = πD0 x + 4 x − 4 x = 90. except the moments that are summed are moments involving a small differential slice of the object rather than moments of discrete objects.35 inch from Base Page 7 . Your differential element should not be a small cube unless there is no symmetry of any kind. To illustrate the CG calculation using calculus.0 in 2 0 4 2 3 M T 1240. Generally you can use a rectangular bar that covers the full length of the part. or a thin disc or annular ring whose diameter is a function of location.
32 lbinch. If weights could be added at 0 and 270 .arcTAN (Y/X) o A = 180 + arcTAN (Y/X) o A = 360 . This is the general situation in most aerospace balancing. In this case. However. sometimes you may then need to convert it back to rectangular form.65 lbin at 0 to compensate for the 4.65 lbin X axis unbalance and o 12. the available locations for correction weights would usually not fall on the reference axes. This might occur if you wanted to adjust the CG offset of a reentry vehicle.Rectangular to Polar Conversion When first calculated. Step 3 Add the appropriate offset angle to rotate the vector to the new XY coordinate system. It is necessary to add ballast weights to the vehicle so that this unbalance is reduced o o to zero. Step 4 Convert the new vector to rectangular coordinates using the following formula: X 1 = M cos( A1 ) Y1 = M sin ( A1 ) where X1 and Y1 are the new axes and A1 is the angle between the unbalance moment vector and the X1 axis. these locations are not available. then we would need o 4. Correcting Static Unbalance The satellite shown has a calculated static unbalance (CG offset moment) of X= 4.32 lbin at 270 to compensate for the +12. if one is not available. Step 1 Add appropriate offsets to the X and Y rectangular components. the following method can be used (two axis): Magnitude M = Angle A = arcTAN (Y/X) o A = 180 . so that it was balanced about its centerline. In this example. the only locations where these Figure 7 – Static Unbalance Page 8 .arcTAN (Y/X) if X = (+) and Y = (+) if X = () and Y = (+) if X = () and Y = () if X = (+) and Y = () (1st quadrant) (2nd quadrant) (3rd quadrant) (4th quadrant) X 2 +Y2 Polar to Rectangular Conversion After the data has been converted to polar form. Often it is useful to convert this data into polar coordinates. Step 2 Convert to polar form (magnitude and angle). However.65 lbinch and Y = +12. Most computers and scientific calculators will do this automatically.32 lbin Y axis unbalance. the CG data is in rectangular form. using a different set of axes.
97C2 = − 12. then calculate rectangular coordinates of correction moments. A = TAN −1 ( y / x ) A = TAN −1 (12.26C2 = 4. General Correction Equations Sum C sin Ac = M sin (180 − A) where: C = Correction Moments Ac = Allowable Correction Angles M = Static unbalance Moment A = Angle of Unbalance Moment Note that these calculations involving static unbalance are concerned with weight rather than mass.60 inches. o o o [ ] Sum C cos Ac = M cos (180 − A) [ ] Calculating Correction Weights M= M x 2 + M y 2 = 1317 lb − in .65) = 110. the figure did not show the height at which the weights were to be added.32 Page 9 . and C2 at 33o and 255o locations: Cx = C1 Cos33 + C2 Cos 255 = 4.25 inches and at 255 is 7.32 .weights can be added are at 33 and 255 . The radius of the correction weight at 33 is o 8. the weights should be added at a height that is as close as possible to the CG height of the vehicle.65 . We will then divide the correction moments by their radii to obtain correction weights.7 o To find Correction Moments C1. In general. In our example above.54C1 − .84C1 − .65 C y = C1 Sin33 + C2 Sin 255 = − 12. so that the addition of these weights will not produce a large product of inertia unbalance. What weights should be added to each location to compensate for the unbalance? The following example outlines the method used to determine the new weights at the allowed locations: We will first calculate the resultant polar magnitude and angle.32 / − 4.
6 in Stage 3 W2 = Note that total weight is more than twice as large as would be required if ballast weights could have been added at the angle of the unbalance at 8” rad.240 inches.84/0. Two views of the rocket are shown. 8.125 = 63. and Z offset moments about the origin: Page 10 . Y. . then the combined CG of the total rocket could be calculated by summing X. = 139 lb .05 lb − in C1 = (4.84 C1 = 1143 lb − in .26C2 )/.115" 43 lb.65+ .436" 167 lb. the Z coordinates of stages 1 and 2 are measured from their intersection.004" 0. we will first have to translate the stage 3 coordinate to the same reference as stages 1 and 2. Since the length of stage 2 is 51.771" 96 lb.51lb 7. TOTAL 306 lb. Combining CG Data from Subassemblies Let us consider the case of a three stage rocket. Stage 2 +32.012" +0. the CG of each section has been measured and found to be: Z W Stage 1 27. The center of gravity of the individual stages had been originally calculated to be on the centerline.115 + 51. After construction. (2) by: (0.25in 19.012" +0.54) and add to eq. Weights to be added: W1 = 1143 lb − in . X +0. In order to calculate the Z center of gravity location. the Z coordinate Figure 8 – CG of Three Stage Rocket becomes 12.004" Y 0. side view shows X and Z axes.Multiply both sides of eq. If the three stages were perfectly aligned at assembly. The top view shows the X and Y axes. while the Z coordinate of the third stage is measured from the intersection of stages 2 and 3.012" Stage 1 Stage 2 The X and Y coordinates are measured from the centerline of the section. C2 = 19.05 lb − in = 2. (1) 0C1 + 125C2 = 2381 . Stage 3 +12.007" 0.125 inches.
320 lbin. The greater the distance the mass is from the center of rotation.672 lbin. For example.5 inch is +0. +4.006 inch gap between stages 2 and 3. Total Moment 0.668 lbin. A formula analogous to Newton's second law of motion can be written for rotation: F = Ma T = IA (F = force.812 lbin. 0. CALCULATING MOMENT OF INERTIA General Comments Moment of inertia ("MOI") is similar to inertia. +0.012 + 0. +2719. After the revised table for the CG coordinates for stages 1 and 3 are complete. if stage 3 is tilted so that the error on the Y axis at a Z dimension of 24. The new value of Y for stage three is therefore: Y = 0. I = moment of inertia.003 inch. For this example we will choose stage 2.1945 in.524 lbin. The CG coordinates for this stage will therefore remain unchanged.336 lbin. then: 1. Y Z 2. Inertia is the tendency of an object to remain at rest or to continue moving in a straight line at the same velocity. If the X axis is shifted sideways on the first stage by +0. etc. Select one of the stages to be the reference. If the three stages are assembled with an alignment error. Stage 2 0. then the calculation proceeds in a manner identical to the example with perfect alignment. +3146. then the 63. If the stages are tilted relative to each other. therefore. 4581.00110 in.246. MOI also depends on the distribution of mass in an object. 0. 2.004 lbin.516 lbin.00989 inch.00058 in. so that there is a 0. then the 0.X Stage 1 +0. a = linear acceleration) (T = torque. Inertia can be thought of as another word for mass. then the Z axis correction for stage three is: . +1.0219 inch 3.500 = 0. the greater the moment of inertia.00989 = 0. M = mass. Unlike inertia.004 inch X dimension for the first stage becomes 0. then the offset due to the tilt must be determined at the CG height of the stage. rotational mass. If the stages did not assemble tightly along their length. Moment of inertia is. +1283. Stage 3 0.020. Recalculate the CG coordinates for stages 1 and 3 to reflect the alignment error.176 lbin.152 lbin. except it applies to rotation rather than linear motion.001 inch.240 Z dimension becomes 63.172 lbin.016 lbin.115/24. A = rotational acceleration) Page 11 . Moment/306 lb 0.020 x 12.
The following table shows some of the units in use today for moment of inertia and product of inertia: UNIT 2 lbin 2 lbinsec 2 slugft 2 kgm COMMENTS 2 lb = weight. Correct units of moment of inertia 2 (or product of inertia) are: MASS x DISTANCE When lbin or lbft are used to define MOI or POI. are lbin . Units of Moment of Inertia In the United States. using the rules outlined in the section entitled "Parallel Axis Theorem". Although any axis can be chosen as a reference. choose an axis of symmetry to simplify the calculation. Page 12 . weight/g = mass. must be divided by g = 386. where the "lb" refers to the weight of the object rather than its mass. dimensionally correct 2 2 2 The most common units used in the U.S. If the unit of weight is the pound. Values for moment of inertia can only be positive. then this axis is defined by the centerline of the bearings. Polarity of Moment of Inertia Values for center of gravity can be either positive or negative. dimensionally correct Kg = mass. dimensional analysis will confirm if correct units are being used. the quantity MUST be divided by the appropriate value of "g" to be dimensionally correct in engineering calculations. If the reference axis will be used to calculate moment of inertia of a complex shape. This leads to units of moment of inertia such as lbin . for reasons which have been lost in antiquity. Only one axis is necessary to define moment of inertia. If the object is mounted on bearings. then this axis is a "principal axis" (axis passing through the center of gravity and oriented such that the product of inertia about this axis is zero (see discussion of product of inertia). even though this is dimensionally incorrect. since this would violate Newton's second law. in the USA an object weighing 1 pound is often referred to as having a 2 mass of 1 pound. If the object flies in space. it is generally desirable to choose the axis of rotation of the object. This axis can later on be translated to another axis if desired. the word "pound" is often misused to describe both mass and weight. dimensionally correct slug = mass. Again. and in fact their polarity depends on the choice of reference axis location.Choosing the Reference Axis Location Three reference axes were necessary to define center of gravity. then the unit of mass cannot also be a pound.088 in/sec lbinsec2 = distance2 x weight/g. just as mass can only be positive. However.
The second example shows a thin walled tube of radius r.088 in/sec2 Do not use local value of g to convert to mass! Calculating the Moment of Inertia MOI. The reference axis is through a point 2 equidistant from the two masses. In these examples. ozin 2 2 2 Value of g : 32. If moment of inertia or product of inertia are expressed in the following units. the CG lies on the centerline of the tube. lbftsec2. is the distance from the axis at which all of the mass of an object could be concentrated without changing its moment of inertia. The masses each have a MOI of Mr /2. Consider first the body consisting of two point masses each with a mass of M/2 separated by a distance of 2r. lbinsec . the radius of gyration is k = r. ozinsec 2 2 2 2 RULE 2. By symmetry. Their 2 combined MOI is therefore Mr . for a point mass 2 around any axis is: I = Mr where I = MOI (slugft or other masslength2 units) M = mass of element (Slugs or other mass unit) r = distance from the point mass to the reference axis Radius of Gyration The moment of inertia of any object about an axis through its CG can be 2 expressed by the formula: I = Mk where I = moment of inertia M = mass (slug) or other correct unit of mass k = length (radius of gyration) (ft) or any other unit of length The distance (k) is called the Radius of Gyration. kgm . with respect to an axis through the CG. This leads to the definition: "The radius of gyration of an object. Again. then their values must be divided by the appropriate value of "g" to make them dimensionally correct. lbft . The method of calculating radius of gyration is outlined in the following sections. lbin . Radius of gyration is always measured from CG.RULE 1.17405 ft/sec2 or 386. then their values can be used in engineering calculations as they are: Slugft ." 2 Figure 9 Figure 10 Page 13 . If moment of inertia or product of inertia are expressed in the following units. sometimes called the second moment. all the mass is 2 located at a distance r from the reference axis so its MOI = Mr .
then the MOI of the object around its CG can be ignored when calculating total MOI.1 slug is located near the outer surface of a rocket and the offset to the CG of the gyro is 3 feet while the radius of gyration of the gyro is only 0. then Ia = M (d + k ) where k is the radius of gyration. Figure 12 RULE 2. and the value becomes 2 d M. Useful Approximations Since the moment of inertia of an object displaced from its 2 2 reference axis is proportional to (d + k ). while the radius of gyration of the rocket motor is 12 inches. we can make two observations that will simplify the job of calculating MOI: RULE 1.Parallel Axis Theorem If in the example above we wanted to determine the MOI of the object about the axis Xa rather than the axis X.01%. Since I = k M.3 lbinsec ). If the radius of gyration of an object is more than 100 times its offset distance "d". and the 2 value becomes k M. For example if a gyro with a mass of 0. then the alignment error is insignificant. For example if a rocket motor with a mass of 100 pounds is located near the center line of the rocket and the offset to the CG of the rocket motor is 0. The MOI of each component in the rocket is first measured or calculated around an axis through its CG. then the offset of the object can be ignored when calculating total MOI.02 ft. and the parallel axis theorem is then used to determine the MOI of the total vehicle with these components mounted in their proper location.9 slugft . through the CG. then the value can be determined using the parallel axis theorem: Ia = I + d M. then 2 the MOI about the center line of the rocket due to the gyro is d M 2 = 0. Page 14 .01% Rule 2 can also be applied to alignment errors when calculating or measuring MOI.100 inches. The error using this approximation is less than 0. If the offset or misalignment is less than 1% of the radius of gyration. 2 2 2 2 K Figure 11 MK This parallel axis theorem is used very frequently when calculating the MOI of a rocket or other aerospace item. Again the error of approximation is less than 0. If the radius of gyration of an object is less than 1% of its offset distance "d". then the MOI about 2 2 the center line of the rocket due to the rocket motor is k M = 14400 lbin (or more 2 properly 37. The offset "d" is the distance from the CG of the component to the centerline of the rocket.
13) is the same. then the moment of inertia is the sum of the individual moments of inertia taken about the same axis. if the object were flying in space. then the computation would be of the form: I1 = M1 (r1) I2 = M2 (r2) 2 2 2 Each weight = 1 lb. The radius of gyration is: I k = total M total The moment of inertia of the two examples (fig. Ix = 1 x 22 +1 x 22 = 8 lbin2 The MOI about the X axis is the same for both examples In = Mn (rn) etc. where r1 is the radius of M1. their flight characteristics would differ. radius of gyration. and principle axis would be different for the two examples. Radius is the only factor affecting their moment of inertia. then this process can be accomplished by choosing a suitable differential element and integrating over the limits of the radius: I = ∫ r 2 dM Page 15 . These examples illustrate that moment of inertia depends only on the radius of the masses within an object. apply the formula I = 2 Mr to each. since the CG. Figure 13 If this were done as described. Radius from X axis=2 in. Note that it makes no difference what angle the masses have relative to each other.Combining Moment of Inertia of Two Objects If the object contains more than one mass. I = Total of above If the object is a homogeneous solid. However. etc. Basic Formula Using Differential Elements of Mass The basic technique for calculating moment of inertia of an object is to consider each element of mass and its radius. and then add up all the moments of inertia of the elements.
Moment of inertia values are simply added to obtain the total. Alignment is relatively unimportant. Standard Shapes The moment of inertia of a variety of standard shapes has been published in most of the textbooks and handbooks that cover dynamics such as the SAWE Handbook. For example. Composite MOI Example An example finding composite MOI around the Z (longitudinal) axis is shown using a reentry vehicle consisting of a hollow cone and other components. Moment of Inertia of Objects Similar to Standard Shapes Since total moment of inertia can be calculated by simply summing the values of the component parts. The moment of inertia error due to misalignment is proportional to the ratio of the square of the misalignment offset to the square of the radius of gyration of the object. 1.Combining axial MOI values If the axial moment of inertia of two cylindrical rocket sections about their mutual 2 2 centerline are 10 slugft and 20 slugft respectively. then the total moment of inertia of 2 both sections when assembled is 30 slugft . Page 16 .002 inch. The MOI of each component around the composite CG must be calculated using the parallel axis theorem 4. it is possible to derive the MOI of many shapes by modifying the values for standard shapes. note that MOI of the inner conical space can be subtracted from the outer as a technique for simplifying calculations even though there is no such thing as negative moment of inertia. make certain that they are both calculated about axes which are coincident when assembled and that the units for each are consistent and correct. The MOI of each component around an axis through its own CG parallel to the desired axis must be determined by computation or measurement 2. Machinery's Handbook etc. and it is laterally misaligned by 0. The location of the composite CG must be calculated 3. For example to determine MOI for a hollow cone. The MOIs are added to find the total MOI around the desired axis through the composite CG. This often eliminates the need for calculus and greatly speeds up the calculation of MOI. then the resulting error is only 0.002% (2 millionths of 1 percent)! Combining Transverse MOI Values The combination of MOI around transverse axes is a more complex procedure. Before adding the values.000. The parallel axis theorem is used to calculate Iz for off center components. if a rocket has a radius of gyration of 15 inches.
6 + . Page 17 .92 Slug − ft 2 32 12 2 I Eo I E( z ) Iz TOTAL = 9. This will be discussed further under "Product of Inertia".1 + 25 + 2.IA WK 2 = g W = 200 lb K = 15 in I B = 450 lb − in 2 I C = 25 Slug − Ft 2 I D = 18 lb − in − sec 2 .15 + = 2. I E = 35 lb − in − sec 2 . more complex analysis must be performed.088 in 12 in Ic = 25 Slugft2 1.15Slug − ft 2 12in 3.1 Slugft2 I Do I Dz 60 12 = .29 + = 1.6 lbftsec2 = 9. CALCULATE MOI AROUND Z AXIS 200 lb(15)2 in2 ft2 sec2 IA = = 9.03slug − ft 2 32 12 75 10 = .03 + 1.6 slugft2 144 in2 32ft 450 lbin2 sec2 ft IB = 386.8lb − in − sec 2 − ft = = 0.92 = 38.29 Slug − ft 2 12 in 2 =0.5lb − in − sec 2 − ft = = 0.65 SlugFt2 Effects of Misalignment When misalignment results in tilt errors. or if other effects cause the CG as well as MOI to change.097 lbftsec2 =0 .
choose an axis of symmetry to simplify the calculation. but weights cause couple moment Basically. then centrifugal force acts through the two weights and produces a couple. but it can have either a positive or negative polarity. The mass distribution which results in a couple moment when the object is spinning is called "product of inertia. On reentry vehicles. POI is expressed in the same units as moment of inertia.. so that the composite product of inertia of a composite object will be much smaller than the product of inertia of many of its elements. if we spin this cylinder about the vertical Z axis. Although any axis can be chosen as a reference. three mutually perpendicular reference axes are necessary to define products of inertia (only one axis is necessary to define moment of inertia). and consequently many engineers are unfamiliar with this concept. Page 18 . using the rules outlined in the section entitled "POI Parallel Axis Theorem. the product of inertia of one component is offset by a negative product of inertia due to another component. then this couple causes a sinusoidal force to be exerted against the bearings as the cylinder rotates. If the object flies in space. If the reference axis will be used to calculate product of inertia of a complex shape. and the cylinder remains statically balanced. If the cylinder is mounted on bearings." Figure 14– CGz=0. the axis may be coincident with the path of flight resulting from axis of symmetry of the outer surface of the vehicle. This axis can later on be translated to another axis if desired. Product of inertia is generally not covered in undergraduate dynamics courses.Calculating Product of Inertia General Comments Consider the homogeneous balanced cylinder to which two equal o weights have been attached 180 apart. product of inertia ("POI") is a measure of dynamic unbalance. then the axis of rotation of the cylinder shifts to align itself to a condition where the centrifugal forces are equalized (i.e. The addition of these weights will not alter the CG of the cylinder. In this respect. If the object is mounted on bearings. If the cylinder is spinning in space." Polarity of Product of Inertia Values for product of inertia can be either positive or negative. and in fact their polarity depends on the choice of reference axis location. Choosing the Reference Axis Location Like center of gravity. then this axis is defined by the centerline of the bearings. POI is similar to CG. then this axis is often defined by the location of thrusters. just as mass can only be positive. However. it shifts toward the unbalance weights slightly). it is generally desirable to select the axis of rotation of the object as one axis. Generally. Values for moment of inertia can only be positive. and spaced equidistant along the length from the CG of the cylinder.
(See section entitled “Units of Moment of Inertia”). A spin stabilized vehicle will rotate about a principal axis (usually the axis of minimum moment of inertia). These axes are called the "principal axes. Principal Axis On any object there will be three mutually perpendicular axes intersecting at the CG for which the products of inertia will be zero. This data can be converted to valid units by multiplying the "ozinch" moment by the height between the correction plane and the test object CG (and then converting the weight in ounces to a unit of mass).02 slugft 2 where M = mass of weight = 0. these axes correspond to the centerline of the cylinder plus two mutually perpendicular axes through the CG at any Figure 15 ." Although such units are incorrect.calculation of Pzx & Pzy Page 19 . Y and Z are principal axes orientation (since the cylinder has perfect symmetry). the word "pound" can mean either weight or mass. product of inertia calculations use the mass. For a perfect cylinder. The product of inertia due to this weight is: Pzx = M Z X = 0. so the engineer must use caution when applying product of inertia values to engineering equations.X. This machine may have a readout for a specific correction plane height which gives the moment required at this plane height to reduce the product of inertia to zero. they have meaning in context with the machine used to measure dynamic unbalance.01 x 2 x 1 = 0. For CG calculations we used the weight of the object. Calculating Product of Inertia A perfectly balanced cylinder rotates on a set of bearings. A small weight whose POI is zero is mounted on this cylinder. This is explained in more detail in the section on correction of dynamic unbalance.01 slug X = radius of CG of weight = 1 foot Z = height between CG of cylinder & CG of weight = 2 feet Figure 16 ." The moment of inertia of the object is at a maximum about one principal axis and at a minimum about another principal axis. Unfortunately.Units of Product of Inertia Product of inertia is expressed in the units of mass times distance squared. Sometimes engineers will encounter bogus units for product of inertia such as "ozinch.
This calculated POI is in the XZ plane of the cylinder. this unbalance can be converted into rectangular components that fall directly on the axes. polar to rectangular conversion). However. the resulting Pzx and Pzy can then be converted back into polar coordinates if desired. A realistic object generally contains a couple unbalance that does not fall directly on any axis. When this occurs. The product of inertia due to this weight is: Pzy = MZY = 0. so that simplified calculations are possible 2 Figure 17 – Any POI can be simulated by two weights in upper plane and two weights in lower plane The real object under test is shown in A. could be calculated as shown below.01 slug Y = radius of CG of weight = 1 foot Z = height between CG of cylinder & CG of weight = 2 feet Note that the value for Pzy is negative. At the conclusion of all calculations. Rectangular to Polar Conversion The previous examples were for a special case where the unbalance was located directly on either the X or Y axis. Pzy.e. since the X coordinate of this weight would be zero. If a similar weight were then added on the Y axis at a location above the CG of the cylinder. The second POI component. because the unbalance can be analyzed as a twodimensional problem on a plane rather than a threedimensional problem. As an aid to analysis. The unbalance equivalent of this object can be simulated by a single weight in each of the two planes as shown in B (angular difference between upper and lower planes can be any angle).01 x 2 x (1) = 0. Each plane can now be analyzed separately. the value for Pzx would not change.02 slugft where M = mass of weight = 0. each single weight in the previous example can be replaced by two weights located on the X and Y axes as shown in C (i. The procedure for these rectangulartopolar transformations is described in the section of this paper dealing with center of gravity. then the mathematics is simplified.. Page 20 .
15 lbin = 0 CGx = +25 lbin . Pzx = 5 lb x 3 in x 5 in + 5 lb x (3) x (5) = CGz = +15 lbin . In the example shown in figure 18. creating a static balance but not a dynamic balance.The product of inertia of the two components in the previous example can be resolved into a single resultant Pzr in the ZR plane which passes through the equivalent unbalance mass and the Z axis.25 lbin = 0 150 lbin 2 Figure 20 – Static & dynamic unbalance Page 21 . Pzr = Pzx + Pzy 2 2 Angle between resultant and X axis = arcTAN (Pzy / Pzx) Difference between CG Offset and Product of Inertia The figures illustrate the difference between static unbalance (CG offset) and couple unbalance (product of inertia). a second weight is added at 180o at the bottom of the cylinder in the above example. Pzx = 0 lbin 2 Figure 18. a 5 lb weight is added in the plane of the CG.) Pzx = +75 lbin 2 CGx = +25 lbin CGz = +15 lbin Figure 19 – Quasistatic unbalance In the example in figure 20.CG offset CGx = 25 lbin CGz = 0 lbin In the example shown in figure 19. creating both static and dynamic unbalance (product of inertia not zero). creating a static unbalance but no product of inertia." since a single correction weight can be used to correct for this unbalance. (This is sometimes called "quasistatic unbalance. a weight is added outside the plane of the CG.
Y. Y. Z axes: Pzx = Pz ′x ′ + M z x Pzy = Pz ′y ′ + M z y where M = mass of object x. Calculate the effective product of inertia of the object relative to the X. Z axes Figure 22– POI can be translated to another set of parallel axes It is much more difficult to use this theorem than the equivalent one for moment of inertia. Y'. creating both static and dynamic balance about z. they cannot be ignored since the product of inertia of the total vehicle is usually very small. Y. Z' axes is Pz'x' = 2 lbinsec . it will be necessary to first calculate or measure the product of inertia of the component parts of the vehicle. Pz'y' = 0 lbinsec . z = CG offsets from coordinate origin along the X. Even if the component products of inertia are small. 2 2 POI Parallel Axis Theorem When determining the product of inertia of a vehicle. In real life. the "weights" consist of various components of a rocket or spacecraft. To translate the product of inertia of an object relative to the X'. x = +5. Z' axes to the X. and then translate these values to the effective POI about the axes of the vehicle.25 = 0 lbin The previous discussion has assumed that the small unbalance weights were perfectly symmetrical. Pzx = +75 lbin .75 lbin = 0 CGz = +15 + 15 = 30 lbin CGx = +25 . Y'. and y = 6 inches. and Figure 21 – Static & dynamic balance therefore the product of inertia of the weight itself could be ignored. Z axes: 2 2 Page 22 . y. and their product of inertia is usually not zero. we move this lower weight up to the plane of the first. The weight of the object is 4 lbs. because there are two formulas required. let z = 4. and even the smallest couple unbalance can tilt the principal axis of the vehicle.In figure 21. The product of inertia of the object about the X'. The following example illustrates this type of translation: Example for the illustration shown. and because each term has a polarity associated with it.
2. This is not so for product of inertia! If the POI of the object about its own CG is not zero.24864 lb − in − sec 2 Pzx = Pz ′x ′ + M z x Pzx = − 2 + (0. Mxy. it cannot be ignored. In the case of moment of inertia. whereas they can be either positive or negative for product of inertia.088) in Pzy = Pz ′y ′ + M z y Pzy = 0 + (0. so that the translation formula becomes Pzx = M z x. since small values of product of inertia can be very significant if one large term is subtracted from another.01036)( − 4)( + 5) = − 2. leaving a small difference.01036lb − sec 2 = (386.01036)( − 4)( − 6) = 0. the polarity of the values of (mass) (length) can only be positive for moment of inertia. it is possible to ignore the MOI of the object about its CG if the translation term is large. Both formulas are dimensionally similar: (mass) (length) = (mass) (length) + (mass) (length) 2 2 2 2 However. is large. Figure 23. 3.Calculate mass: 4 0. even if the translation term value.Pzx is zero for all of the examples above Page 23 .2072 lb − in − sec 2 Comparison Between MOI and POI There are some similarities and some differences between this axis translation formula and the formula to translate moment of inertia to a different parallel axis: 1. the value for the MOI about the CG of an object always has a value greater than zero. In the case of moment of inertia. It is possible for the product of inertia of an object to be zero.
5∫ b2 Page 24 . dV = differential volume As in the case for moment of inertia and center of gravity. The examples on the right illustrate some symmetrical shapes that have a zero Pzx. then the POI of the element is zero. This rule is hard to visualize when put into words. except the objects are now differential elements of a solid. The formula becomes: Pyx = M ∫ yxdV where M = total mass of object. If a rectangular slice parallel to the X axis is chosen instead. For example. the solution to the problem can be simplified by choosing the right differential element. the elliptical wing tip shown can be analyzed using a small square element dX by dY. This leads to a double integral.Axes and Planes of Symmetry The product of inertia of a homogenous body with respect to any pair of perpendicular axes is EQUAL TO ZERO if the plane determined by either of the axes and the third coordinate axis is a plane of symmetry of the body. and the product of inertia of dA is: The CG of the rectangular slice is at x/2 (0 + 0.5 x y dA) Since dA = x dy then Pzy = 0.5∫ x 2 y dy x y x Figure 24 – Calculation of POI From the equation of an ellipse: a 2b 2 − y 2 a 2 x = b2 2 Therefore: a 2b 2 − y 2 a 2 y dy Pxy = 0. Determining Product of Inertia of a Volume The basic concept of determining the product of inertia of a homogeneous volume is identical to the previous method involving discrete objects.
If the X and Y axes for the upper section do not correspond to those chosen for the lower section. (The values for the total can be larger or smaller than the individual values. so that it is not satisfactory to ignore this effect. The X or Y axis distance will be small. what is the resulting product of inertia? If the sections are aligned perfectly. The location of the upper section must be defined in terms of the lower reference axis. but the axis of one is not coincident with the axis of the other? Here is the method we recommend for analyzing this type of problem: 1. Do the same for Pzy. then the following method can be used: 1. rotate the data by converting temporarily in polar form and then back into the new rectangular axes. Transform the product of inertia of the lower section into values for Pzx and Pzy (polar to rectangular conversion). 2. This requires the measurement of offset of the upper section when both sections are assembled. the magnitude of POI resulting from misalignment will be greater Figure 26 – Lateral than the POI of the individual sections. Often. Note: observe the polarity of the data! This will yield new product of inertia values for the composite vehicle. This can be accomplished by placing the total rocket on a rotary table and dial indicating both lower and upper section. The centerline of the two sections will be offset from this axis by some small but not insignificant amount. then measurements can be made of the individual sections and their interface ring concentricity and the offset calculated (not as accurate a method). misalignment Page 25 . Calculate the X. The new reference axis for the total rocket will pass through the new CG and be parallel to the reference axes of the two sections.Combining the POI of Two Bodies If two sections of a rocket are combined. and Z coordinates of the CG of the total rocket. but the mass will be very large since it is the entire mass of a section. so that upper and lower axes are at the same angular location. so that the Z reference axis for the lower section is exactly coincident with the Z reference axis for the upper section.) Figure 25 – Combining POI Effect of Lateral Misalignment What happens if the two sections are not aligned. 2. Sum the values for Pzx upper and Pzx lower. so that the axes are parallel to each other. Analysis will be done by planes and transformed into polar coordinates if desired after all calculations have been made. Y. or if this is not possible.
and Ix becomes Iz. Iz. the resultant polar representation. and the object is then installed in a vehicle in such a way that there is a Figure 27– Axis Tilt lean angle between the center line of the object and the reference of the vehicle. Repeat Steps 3. In the view shown." then Pzx. and 5 for Pzy . Pzx is again zero. 7. 6. then it is very useful to be able to convert the calculated values for the object into mass properties relative to the new reference without having to recalculate the object itself." This paper outlines a method of measuring POI of objects using a moment of inertia instrument. 4. Pzx relative to the new combined axis due to the Pzx of the upper section may be calculated by using the parallel axis translation formula: Pzx = Pz ′x ′ + M z x where Pzx = POI relative to combined CG Pz'x' = POI relative to upper section M = mass of upper section x z = distance between composite CG and upper section CG x = offset between new combined reference and upper reference In the view shown. As the cylinder leans to an angle of 90 . At a lean angle of 90 . both z and x are negative. Effect of Tilt on MOI and POI If a perfectly balanced cylinder is tilted by an angle "a. Iz becomes Ix. 5. This unique relationship between moment of inertia and product of inertia is discussed in the SAWE paper No. If desired. When the MOI or POI of an object is determined about its center line.3. Sum the values for Pzx upper and Pzx lower to yield Pzx total. Repeat the calculation for the lower section. Pzx is a maximum determined by the values of Iz and Ix. 1473 entitled "Determining Product of Inertia Using a Torsion Pendulum. For a lean angle of 0o. Pzx is zero. both z and x are positive. and Ix will change. 4. convert Pzy and Pzx into Pzr . For a lean angle of 45o. so that the POI due to the lower section offset is also positive. so that the POI due to the upper offset is positive. o o Page 26 . These "lean angle formulas" are given below.
Pzx was zero.5( I z − I x ) cos(2a ) + Pzx sin( 2a ) This formula reflects the fact that the principal axis is no longer through the centerline of the cylinder. The MOI at an angle of 45 is the average of Iz and Ix. so that the maximum and minimum MOI are no longer axes X and Z. the moment of inertia about an axis Z' displaced from the center line of the cylinder by an angle "a": I z′ = 0. If this is not the case. When measuring tall. slender rockets. this formula also can be used to determine fixturing accuracy required when measuring MOI. The formulas presented also assume that there is no tilt in the Y direction. then the coordinate system must be transformed so that it is. This formula is only valid for the orientation of the axes shown.5( I z − I x ) cos(2a ) Note that for this example. then tilt angle can be ignored.5( I z + I x ) + 0. then tilt angle is very critical. Equations can be written for the more general case. If the object is tall and slender. o Page 27 . There are some interesting observations to be made regarding the change in MOI: 1. the axial MOI error will be large unless the rocket is fixtured very carefully. The sensitivity to tilt angle is a function of the difference between Ix and Iz.5( I z + I x ) + 0. If there is very little difference. MOI Incline Axis with POI The previous analysis assumed that the Z and X axes were principal axes and the POI was zero. so that the problem can be analyzed from a twodimensional standpoint. Transverse MOI (Iy and Ix) can be measured on a vee block fixture without the need to adjust the rocket's position. it is easier to manipulate the axes than to solve the general equations. However. 2. note that the origin of the two axes is at the CG of the object. Furthermore.MOI Inclined Axis Formulas For the balanced cylinder shown. then the formula becomes: I z′ = 0. since the sensitivity to lean is very small in this case. If this is not the case. In addition to being useful in the calculation of MOI.
POI Inclined Axis Formulas Since POI and MOI are related. we are starting with a value of Pzx. A copy of this aid is reproduced from the SAWE handbook and is shown on the next page. (2) The corresponding value for the product of inertia. o Figure 28 Page 28 . 27) and for the direction and definition of positive tilt angle (CCW from the Z axis). In this case. where the center of gravity lies at the origin of a set of mutually perpendicular axes XY. (1) The location of the principal axes about which the moments of inertia are maximum and minimum and the products of inertia are zero. Otto Mohr. (3) The moments and products of inertia for any other set of mutually perpendicular axes AB whose origin lies at the center of gravity of the given object and rotated C degrees from the original axes XY (reference. (4) The maximum values for the products of inertia about axes located 45 from the principal axes. o which is zero and returning to a value of zero at an angle of 90 . With the advent of the personal computer. then the formula becomes: Pz ′x ′ = 0. PXY Mohr's circle is then constructed using the layout geometry shown below. Mohr's Circle for Moments of Inertia Given (1) The moment of inertia values IX. graphical solutions to engineering problems are no longer necessary. The formula is: Pz′x ′ = 0. but Mohr's circle still is useful in visualizing the effect of tilt. the figure to the right). (2) The corresponding maximum and minimum values of moments of inertia. The following information may then be obtained. it might be assumed that a similar formula can be written for the POI of an object when tilted. Mohr's Circle A graphical representation of the relationship between MOI and POI was originated in th the 19 century by a German engineer.5 ( I x − I z ) sin (2a ) − Pzx cos(2a ) If the X and Z axes are principal axes. IY for an object about its center of gravity.5 ( I x − I z ) sin (2a) These formulas are only valid for the orientation of the axes shown (fig.
Layout Geometry The Radius of the circle is: R = 2 Ix − Iy + Pxy 2 ( ) 2 Figure 29 – Mohr’s Circle Page 29 .
both z and x are positive. so that the POI due to the upper offset is positive. Page 30 . 4. The new reference axis for the total rocket will pass through the new CG and be parallel to the reference axes of the lower section. 5. so that the POI due to the lower section offset is also positive. value for Pzx may be either larger or smaller than value without considering tilt). 3.Effect of Angular Misalignment (Tilt) If the upper section is tilted relative to the lower section. Repeat the calculation in Step 4 for the lower section. and Z coordinates of the CG of the total rocket. calculate the X offset of the upper section. Pzx relative to the new combined axis due to the Pzx of the upper section may be calculated by using the parallel axis translation formula: Pzx = Pz ' x ' + M z x where Pzx = POI relative to combined CG reference axes Pz'x' = POI relative to upper section after effect of tilt has been added M = mass of upper section z = distance between composite CG and upper section CG x = offset between new combined reference and upper reference In the view shown. Y. Calculate the X. Add this POI to the Pzx of the upper section relative to its center line (observe signs. Since this section is not tilted. Using the center line of the lower section as a reference. the Pz'x' is the value through the center line. 2. then two factors tend to increase the effective POI of this section: the tilt results in a CG offset similar to the case described previously. Recalculate the Pzx of the upper section by applying the axis tilt formula. The method for calculating the total POI is as follows: X 1. and the tilt also alters the POI of the upper section itself. In the view shown. both z and x are negative. calculate the Y axis offset of the CG of the upper section from the formula: Figure 30 – Angular Misalignment Y = H sin a Where: H is the CG height of upper section a is the tilt angle in the zy plane Using a similar concept.
it is desirable to make this angle as small as possible so the vehicle will "fly straight". the units of moment of inertia 2 must be converted to lbinsec to be consistent with the units for product of inertia (or 2 the units for product of inertia could be converted into lbin ): I xx = 8. Sometimes. the product of inertia about this axis may not be zero. If desired. convert Pzx and Pzy into Pzr .40lb − in 2 = 0. however. Since the vehicle is not homogeneous. 5 and 6 for Pzy .6. Sum the values for Pzx upper and Pzx lower to yield Pzx total. Generally.95 lbin Izz = 2. 8. 4. so a reentry vehicle will have a "coning" motion upon entering the atmosphere. 7. The angle is known as the "angle of inclination" of the vehicle. Angle of inclination may be calculated using the following formula: 2 Pzx A = 0.002 lbinsec Ixx = 8. Angle of Inclination of Reentry Vehicle An aerospace vehicle will often have an axis defined by the minimum air resistance of the vehicle. this angle is deliberately adjusted to a specific value.95 lb − in 2 = 0.40 lbin 2 2 2 Figure 31 – Angle of Inclination What is the angle of inclination in the XZ plane? First.5arcTAN ( I zz − I xx ) Example: Given Pzx = 0. the resultant polar representation.006216 lb − in − sec 2 Page 31 . This axis corresponds to the axis of symmetry of the outer surface of the vehicle. resulting in a principal axis at an angle to the axis of symmetry of the vehicle. and the resulting drag will slow the reentry.023155 lb − in − sec 2 I zz = 2. Repeat Steps 3.
023155 The previous example was for a single plane. make the difference in moment of inertia as large as possible.5 arcTan = − 6. this is not the case.Then. This leads to two conclusions: 1. The minimum moment of inertia difference will be limited by the skill with which you can correct product of inertia unbalance (a function of the sensitivity of the balancing machine used.5 arcTAN ( I zz − I rr ) where Pzr is the resultant of Pzx and Pzy. using the formula: 2Pzx A = 0. This is accomplished by designing the vehicle to be long and slender.74 deg rees (0. (See section on moment of inertia. and by placing the heavy items near the ends of the vehicle. and the stability of the components in the vehicle). Irr is the moment of inertia about axis "r" (resultant) For a typical projectile or reentry vehicle. If you want to stabilize the vehicle and have it resistant to the effects of product unbalance.) Controlling Tilt Angle Since the amount of axis tilt which results from a given product of inertia is a function of the difference between Izz and Ixx. then make the moment of inertia difference as small as can be tolerated. Iyy = Ixx. then the combined solution for the object would be: 2 Pzr A = 0. If you want to steer the vehicle with as little correction force as possible. so that Ixx can be used in place of Irr. and Irr must be calculated from the values of Iyy and Ixx.004 A = 0.006216 − 0. For a vehicle with wings.5 arcTAN ( I zz − I xx ) 0. This would also be the angle of inclination for the object if the product of inertia of the ZY plane were zero. If there were a product of inertia for the ZY plane. the effect of a product unbalance can be adjusted by altering the difference in moment of inertia. 2. Page 32 .
then half the force would be applied to each bearing). the force on each bearing would be a proportion of the total force (if the CG were equidistant from both bearings. then CG offset from the axis of rotation will produce a sinusoidal force on the bearings. Unbalance Forces Due to POI of Rotating Object For a CG offset. If the product of inertia is not zero. then the unbalance force is in the same direction for both upper and lower bearings. The force exerted by a CG offset is a function of rotation speed and the magnitude of unbalance moment: F = M r w2 where F = unbalance force in pounds M = mass in slugs = W/g r = CG offset in feet w = angular velocity in radians per second If we examine the force exerted on a bearing system along a single radial axis. and may lead to errors if the rotating object is part of a guidance system. In contrast. This unbalance force will cause premature wear of the bearings. Such would be the case if the object were a thin flywheel. a product of inertia unbalance will result in equal and opposite forces on the bearings and the force is not a function of the CG location. but the magnitude of the force is proportional to the CG location relative to the bearings and is. For a given POI.Unbalance of Rotating Objects Unbalance Forces Due to Offset CG of Rotating Object If a rotating object is mounted on bearings. Page 33 . the forces on the bearings do not depend on the spacing between the bearings but do depend on the axial location of the CG of the rotor. If the rotor CG is located between the bearings. the magnitude of the bearing force will increase as the spacing between the bearings is made smaller. The following section describes the forces due to product of inertia unbalance ("couple unbalance"). or if the object were dynamically balanced and a single weight was then added in a plane perpendicular to the axis of rotation and passing through the center of gravity of the object. in general. it varies sinusoidally: F = Mrw 2 sin a where a = angle of rotation relative to the axis Since a rotating system usually has two bearings. additional friction. then the forces on the bearings will be different from this example. noise. This analysis assumes that the product of inertia is zero. different for each bearing.
If you place an object on a scale in Europe.0259 slug − ft x (31416) 2 = 2556 lbs force ( peak ) .One method of calculating the force due to product unbalance is to first determine the equivalent mass at the bearings that would result in the magnitude of POI. If the rotation speed is 300 RPM. Since you are really trying to use the scale to measure mass. if you place an object on a scale in the USA.0259 slug − ft (32. 60 10 10 lb − in = = 0. Mass does not equal force. Page 34 . Many of these errors are the result of a misunderstanding regarding the difference between mass and weight. However. which is 9. If a shaft with a Pzx of 100 lbin is supported on bearings which are 10 inches apart. An analysis of centrifugal forces acting against the bearings of a spin balance machine results in measured values for center of gravity offset and product of inertia. then the equivalent product of mass times distance at each bearing is 10 lbin. it should read 6. This is only one of many thousands of errors that have occurred as a result of the confusion between Metric and English units. The force on the bearings can then be determined using the formulas given above for CG offset. result in an underestimation of the thruster impulse by a factor of 4.174 x12) 300 RPM = F = Mrw 2 = 0. you will read its mass (generally expressed in kg). . . The software program which controlled the thrusters was supplied with thrust data in poundseconds but interpreted it as if it were newtonseconds. a dimensionally inconsistent correction factor is used to convert from one set of units to the other. you will read a value equal to the force exerted by the acceleration of gravity (generally expressed in lbf). 2 Mass vs Weight (and English vs Metric) In 1999. when you weigh yourself on an American bathroom scale. Traditionally. It is like comparing apples to oranges.22 slug rather than 200 lbf.205 lb is not valid.45. This traditional conversion factor is based on the value of standard gravity. These relationships form the basis for a spin balance machine. the Mars Climate Orbiter crashed as a result of a confusion over the system of units. The expression 1 kg = 2. then: 300 = 5 rev / sec = 5 x 2 x 31416 = 31416 rad / sec .80665 m/sec2.
(foot) Kg Weight in lbf 386.80665 M/sec2 386. systems are fundamentally FORCE. The various metric systems are fundamentally MASS. used to describe force as well as mass.0886 in. Slug LENGTH Meter Inch Foot WEIGHT Newton Pound (lbf) Pound g 9.S. the word "pound" is commonly used for both mass and weight.S. TIME systems with mass being defined or derived. The mass of an object is a fixed quantity." A one pound force is required to accelerate a one Slug 2 mass at one ft/sec . an analysis of fundamental dimensions will confirm if correct units of measurement are being used and if conversion factors are being applied correctly to achieve desired results. not weight. Similarly. LENGTH. while WEIGHT is the FORCE that presses the object down on a scale due to the acceleration of gravity. If different names are used for weight and mass. In the USA. The Metric SI system uses the word "Newton" for weight and the word "Kilogram" for mass. Mass properties do not change as a space vehicle leaves the attraction of the earth and enters outer space. Officially “pound” refers to mass (see. NIST documents). The aerospace industry has created a unit of mass called the "Slug. incorrectly. the common usage of the word pound is the value you read on a scale. then its mass is one Slug.S. then the problem of distinguishing between the two is minimized. However. If an object weighs 32. Unfortunately. The U.Mass is related to weight through Newton's second law: where W = the weight of the object (gravity force) M = the mass of the object g = the acceleration of gravity W = Mg MASS is the QUANTITY OF MATTER in an object (its inertia). TIME systems with force being a defined or derived term. The Newton is defined as the force required to 2 accelerate a 1 Kilogram mass by 1 meter per second . Table One shows the three most commonly used systems of measurement. its weight varies as a function of the acceleration of gravity. (inch) U. If the term "pound" is used to describe a mass whose measured weight is one pound (force). The mass properties of an object are related to mass. LENGTH.17405 ft/sec2 Page 35 . resulting in endless confusion and errors in calculating mass properties and dynamic response. which is actually lbf. this quantity MUST be divided by the acceleration of gravity in appropriate units to convert it to proper mass dimensions if it is to be used in mass properties calculations. in metric countries the terms Kilogram and Gram are often. To avoid confusion and uncertainty.0886 in/sec2 32. for example.17405 lbf on earth. not all systems of units adequately differentiate between mass and weight. Time in seconds is used throughout DIMENSIONALLY CORRECT MEASURING SYSTEMS MASS SI (Metric) U.
Inc.S. Applying W = Mg shows that the system is dimensionally consistent.09 in 1lb mass x g = x = 1lb weight 386.068 521 76 lb mass 1 t (tonne.09 in sec 2 The acceleration of gravity used to convert weight to mass is a fixed number which has been established as an international standard.204 622 6 lbf 1 kg = 9.The U.com Page 36 . One pound mass is equal to one pound force divided by 386.453 592 37 kg 1 kg = 2.spaceelectronics. 1lb − sec 2 386. US) = 907.806 65 Newtons 1 sh tn (short ton. UK) = 1016.521 76 lb mass 1 lb mass = 14. inch system has no common name for the mass whose weight equals one pound.349 52 gram 1 oz tr (troy ounce) = 31.806 65 N Mass conversion factors 1 kg = 0. metric) = 1000 kg Space Electronics. CT 06037 8608290001 fax 8608290005 www.448 22 N 1 kp (kilopond) = 9.088 inches/sec2.80665 m/sec2 Dimensionally inconsistent conversion factors based on standard acceleration of gravity 1 oz (ounce) = 28. metric) = 68. although this is sometimes called a “pound mass”.17405 ft/sec2 or 9.103 48 gram 1 lbf (pound force) = 0. Conversion Factors Standard acceleration of gravity g = 32. 184 7 kg 1 ton (long ton. Berlin.com info@spaceelectronics. 047 kg Force conversion factors 1 dyne = 105 N 1 lbf (poundforce) = 4. 81 Fuller Way.593 904 kg 1 t (tonne.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.