You are on page 1of 65

UFC 3-130-06

16 January 2004

UNIFIED FACILITIES CRITERIA (UFC)

CALCULATION METHODS FOR


DETERMINATION OF DEPTH OF
FREEZE AND THAW IN SOIL: ARTIC
AND SUBARCTIC CONSTRUCTION

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED


UFC 3-130-06
16 January 2004

UNIFIED FACILITIES CRITERIA (UFC)

CALCULATION METHODS FOR DETERMINATION OF DEPTH OF FREEZE AND


THAW IN SOIL: ARTIC AND SUBARCTIC CONSTRUCTION

Any copyrighted material included in this UFC is identified at its point of use.
Use of the copyrighted material apart from this UFC must have the permission of the
copyright holder.

U.S. ARMY CORPS OF ENGINEERS (Preparing Activity)

NAVAL FACILITIES ENGINEERING COMMAND

AIR FORCE CIVIL ENGINEER SUPPORT AGENCY

Record of Changes (changes are indicated by \1\ ... /1/)

Change No. Date Location

This UFC supersedes TM 5-852-6, dated 25 January 1988. The format of this UFC does not
conform to UFC 1-300-01; however, the format will be adjusted to conform at the next revision.
The body of this UFC is the previous TM 5-852-6, dated 25 January 1988.
UFC 3-130-06
16 January 2004
FOREWORD
\1\
The Unified Facilities Criteria (UFC) system is prescribed by MIL-STD 3007 and provides
planning, design, construction, sustainment, restoration, and modernization criteria, and applies
to the Military Departments, the Defense Agencies, and the DoD Field Activities in accordance
with USD(AT&L) Memorandum dated 29 May 2002. UFC will be used for all DoD projects and
work for other customers where appropriate. All construction outside of the United States is
also governed by Status of forces Agreements (SOFA), Host Nation Funded Construction
Agreements (HNFA), and in some instances, Bilateral Infrastructure Agreements (BIA.)
Therefore, the acquisition team must ensure compliance with the more stringent of the UFC, the
SOFA, the HNFA, and the BIA, as applicable.

UFC are living documents and will be periodically reviewed, updated, and made available to
users as part of the Services responsibility for providing technical criteria for military
construction. Headquarters, U.S. Army Corps of Engineers (HQUSACE), Naval Facilities
Engineering Command (NAVFAC), and Air Force Civil Engineer Support Agency (AFCESA) are
responsible for administration of the UFC system. Defense agencies should contact the
preparing service for document interpretation and improvements. Technical content of UFC is
the responsibility of the cognizant DoD working group. Recommended changes with supporting
rationale should be sent to the respective service proponent office by the following electronic
form: Criteria Change Request (CCR). The form is also accessible from the Internet sites listed
below.

UFC are effective upon issuance and are distributed only in electronic media from the following
source:

Whole Building Design Guide web site http://dod.wbdg.org/.

Hard copies of UFC printed from electronic media should be checked against the current
electronic version prior to use to ensure that they are current.

AUTHORIZED BY:

______________________________________ ______________________________________
DONALD L. BASHAM, P.E. DR. JAMES W WRIGHT, P.E.
Chief, Engineering and Construction Chief Engineer
U.S. Army Corps of Engineers Naval Facilities Engineering Command

______________________________________ ______________________________________
KATHLEEN I. FERGUSON, P.E. Dr. GET W. MOY, P.E.
The Deputy Civil Engineer Director, Installations Requirements and
DCS/Installations & Logistics Management
Department of the Air Force Office of the Deputy Under Secretary of Defense
(Installations and Environment)
ARMY TM 5-852-6
AIR FORCE AFR 88-19, Volume 6

TECHNICAL MANUAL

ARCTIC AND SUBARCTIC


CONSTRUCTION
CALCULATION METHODS
FOR DETERMINATION OF
DEPTHS OF FREEZE AND
THAW IN SOILS

DEPARTMENTS OF THE ARMY AND THE AIR FORCE


JANUARY 1988
REPRODUCTION AUTHORIZATION/ RESTRICTIONS
This manual has been prepared by or for the Government and is public property
and not subject to copyright .
Reprints or republications ofthis manual should include a credit line substantially
as follows : "Joint Departments of the Army and Air Force USA, Technical Manual
TM 5-852-6/A FR 88-19, Volume 6, Calculation Methods for Determination of Depths
of Freeze and Thaw in soils, date."
TECHNICAL MANUAL HEADQUARTERS
NO. 5-852-6 DEPARTMENTS OF THE ARMY
AIR FORCE REGULATION AND THE AIR FORCE
AFR 88-19, VOLUME 6 WASHINGTON, DC, 25 January 1988

ARCTIC AND SUBARCTIC CONSTRUCTION-CALCULATION METHODS FOR


DETERMINATION OF DEPTHS OF FREEZE AND THAW IN SOILS

Paragraph Page
CHAPTER 1. GENERAL
Purpose and scope . . . . . . . . . . . . . . . . . .... . . ..... .... ..... . 1-1 1-1
References and symbols . . . . . . . . . . ... .. . ..... . .... .... . . 1-2 1-1
Background . . . . . . . . . . . . . . . . . . . . . . . .... . . ..... .... . .... . 1-3 1-1
CHAPTER 2. DEFINITIONS AND THERMAL PROPERTIES

Definitions 2-1 2-1


Thermal properties of soils
and other construction
materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... .... . . 2-2 2-2
Fundamental considerations . . . . . . . . . . . . . . . . . . .. . .... .. 2-3 2-3
Freezing and thawing
indexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . .... .. 2-4 2-5
CHAPTER 3. ONE-DIMENSIONAL LINEAR AND
PERIODIC MEAT FLOW

Thermal regime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . 3-1 3-1


Modified Berggren equation . . . . . . . . . . . . . . . . . . . . . . . . .. . . 3-2 3-1
Homogeneous soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . 3-3 3-3
Multilayer soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . 3-4 3-3
Effect of snow and
vegetative cover . . . . . . . . . . . . ..... ...:. ...... . . ... . ...... 3-5 3-6
Surface temperature
variations . . . . . . . . . . . . . . . . . . ..... ..... ...... . . ... .... ... 3-6 3-6
Converting indexes into
equivalent sine waves of
temperature . . . . . . . . . . . . . . . ... .... . . .... . .............. 3-7 3-8
Penetration of freeze or
thaw beneath buildings . . . . .. . .... . ..... . ...... ... . .... 3-8 3-11
Use of thermal insulating
materials . . . . . . . . . . . . .... .. . .... . .... . . ......... ..... 3-9 3-18
CHAPTER 4. TWO-DIMENSIONAL RADIAL HEAT FLOW

General . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . .. ... . . . .. ... . . . ... . . . ... . . ... 4-1 4-1


Pile installation in
permafrost . . . . . . . . . . . . . . . . . . . . . . . .. . . ... . . . . ... . . . . . ... . . ... . . . ... 4-2 4-2
Utility distribution systems
in frozen ground . . . . . . . . . . . . . . . . . . . .... . . . . ... . . . . . ... . . ... . . . ... 4-3 4-7
Discussion of multi-
dimensional heat flow . . . . . . . . . .. . .... . . . .. . . . . .. ... . . ... . . . ... 4-4 4-13

APPENDIX A. LIST OF SYMBOLS A-1


APPENDIX B. THERMAL MODELS FOR COMPUTING
FREEZE AND THAW DEPTHS B-1
APPENDIX C. REFERENCES C-1
BIBLIOGRAPHY BIBLIO-1

"This manual supercedes TM 5-852-6/AFM 88-19, Chap . 6, dated January 1966 .


LIST OF FIGURES

Figure 2-1 . Average thermal conductivity for sands and gravels, frozen. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3
2-2. Average thermal conductivity for sands and gravels, unfrozen. . . . . . . . . . . . . . . . . . . . . . . . . 2.4
2-3 . Average thermal conductivity for silt and clay soils, frozen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4
2-4. Average thermal conductivity for silt and clay soils, unfrozen. . . . . . . . . . . . . . . . . . . . . . . . . . 2.4
2-5. Average thermal conductivity for peat, frozen. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5
2-6. Average thermal conductivity for peat, unfrozen. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5
2-7. Average volumetric heat capacity for soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.8
2-8.7dolumetriclatentheatforsoils. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  .  2.9
2-9. Average monthly temperatures versus time at Fairbanks, Alaska- . . . . . . . . . . . . . . . . . . . . . 2-10
2-10Relationship between wind speed and n-factor during thawing season. . . . . . . . . . . . . . . . 2-11
2-11.Relationship between mean freezing index and maximum freezing index for
10 years of record, 1953-1962 (arctic and subarctic regions). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-12
2-12Relationship between mean thawing index and maximum thawing index for
10 years of record, 1953-1962 (arctic and subarctic regions). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-13
3-1 . .1 coefficient in modified Berggren formula. . . . . . . . . . . . . ... . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-2
3-2. Relationship between (x/2;att-) and erf (x/287_1). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-7
3-3. Sinusoidal temperature pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ., . . . 3.9
3-4. Indexes and equivalent sinusoidal temperature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-10
3-5. Average monthly temperatures for 1949-1950 and equivalent sine wave,
Fairbanks, Alaska. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-11
3-6. Long-term mean monthly temperatures and equivalent sine wave,
Fairbanks, Alaska. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-12
3-7. Schematic of ducted foundation. . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  , . . . . . . 3-14
3-8. Properties of dry air at atmospheric pressure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-16
4-1. Illustration for example in paragraph 4-la. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2
4-2. Temperature around a cylinder having received a step change in temperature. . . . . 4-3
4-3. General solution of slurry freeze-back . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-4
4-4. Specific solution of slurry freeze-back. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-5
4-5. Freezeup of stationary water in an uninsulated pipe. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-9
4-6. Temperature drop of flowing water in a pipeline. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-12
LIST OF TABLES

Table 2-1 . Specific heat values of various materials. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........ . . .... . . .. 2.2
2-2. Thermal properties of construction materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....... . . . ... . . . .. 2.6
2-3. Calculation of cumulative degree-days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ..... . . .... . . .. 2.9
2-4. n-factors for freeze and thaw. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....... . .. .... . .... 2 .11
3-1. Multilayer solution of modified Berggren equation. . . . . . . . . . . . . . . . . . . . . ...... . . . ..... . .... 3.5
3-2. Thaw penetration beneath a slab-on-grade building constructed
on permafrost. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .. . ..... . .... 3.13
3-3. Insulated pavement design, no frost penetration. . . . . . . . . . . . . . . . . . . . . . . . . ..... .. . . . ... . .. . 3.20
3-4. Insulated pavement design, frost penetration. . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .. . ..... . ... 3.20
B-1 . Thermal, fluid and electric analogs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ... .. . . . .... . . .. B-2
*TM 5-852-6/AFR 88-19, Volume 6

CHAPTER 1
GENERAL

1-1 . Purpose and scope. underlying theory, see appendix C and


This manual contains criteria and the bibliography .
methods for calculating the depths of b. Heat transfer in soils involving
freeze and thaw insoils,with considera- phase change of pore water is an ex-
tion of the effects of other adjacent tremely complex process and many
materials, for the design of military problems defy rigorous mathematical
facilities in seasonal frost, arctic and treatment. The methods presented here
subaretic regions. The contents are are simplified procedures developed
applicable to both Army and Air Force for the solution of engineering design
construction in arctic, subaretic and problems.
seasonal frost areas. The data pre- c. Several assumptions have been
sented in this manual relate to arctic made in developing practical methods
and subaretic facility design presented ofcalculating depths of freeze or thaw
in the other manuals of the Arctic and in soils . It is assumed that each layer of
Subarctic Construction series . material is homogeneous and isotropic,
and that the average thermal proper-
ties of frozen and unfrozen soils are
1-2. References and symbols. applicable. Unless specific data are
Appendix C lists the references for available, it is also assumed that all soil
this manual ; appendix A contains a list water is converted to ice, or all ice is
of symbols . converted to water, ata temperature of
32F. This latter assumption is sub-
1-3. Background. stantially correct for coarse-grained
a. The depths to which soils may soils but only partially true for fine-
freeze and thaw is very important in grained soils.
the design of pavements, structures d. The services of the U.S. Army Cold
and utilities in areas of seasonal frost Regions Research and Engineering
and permafrost. Methods ofcalculating Laboratory (USACRREL), Hanover, New
such depths, based on heat-transfer Hampshire, are available to assist in
principles, are presented here . For the development of solutions for heat-
derivation of basic equations, and the flow problems in soils.
*TM 5-852-6/AFR 88-19, Volume 6

CHAPTER 2

DEFINITIONS AND THERMAL PROPERTIES

2-1 . Definitions . e. thermalresistance, R. The recipro-


Definitions of certain specialized terms cal of the time rate of heat flow through
_ applicab le to arctic and subarctic a unit area ofa soil layer of given thick
regio ns are contained in TM 5-8_5_2/ ness d per unit temperature difference :
AFR _S-1G,_Volume 1. Following are d
additional terms used specifically in (eq 2-5)
heat-transfer calculations . K
a. Thermal conductivity, K. The f. Thermal diffusivity, a. An indica-
quantity of heat flow in a unit time tor ofhoweasily a material willundergo
through a unit area of a substance temperature change:
caused by a unit thermal gradient. K
b. Specific heat, c. The quantity of (eq 2-6)
heat absorbed (or given up) by a unit c
weight ofa substance when its tempera g. Thermal ratio, a-
ture is increased (or decreased) by 1
degree Fahrenheit (F) divided by the a (eq 2-7)
quantity of heat absorbed (or given up)
by a unit weight of water when its where
temperature is increased (or decreased) vo absolutevalue ofthe difference
by 1F. ' between the mean annual tem-
c. Volumetric heat capacity, C. The perature below the ground
quantity of heat required to change surface and 32F.
the temperature of a unit volume by vs one of two possible meanings,
1F. depending on the problem
For unfrozen soils, being studied:
C
W (1) vs = nF/t (or nl/t)
u =yd(c+1.0 100
(eq 2-1)
where
Forfrozen soils, n=conversion factor from air
W
C f = yd (c + 0.5 index to surface index
100 ) (eq 2-2) F= air freezing index
Average values for most soils, I =air thawing index
W
t = length of freezing (or thawing)
c avg. = Yd (c + 0.75
100 ) (eq 2-3) season .
where c = specific heat ofthe soil solids (2) vs absolute value of the dif-
(0.17 for most soils) ference between the mean
Yd = dry unit weight of soil annual groundsurface temp-
w = water content of soil in per- erature and 32F.
cent of dry weight . In the first case, vs is useful for
d. Volumetric latent heat offusion, computing the seasonal depth
L. The quantity of heat required to of freeze or thaw . In the second
melt the ice (or freeze the water) in a case, it is useful in computing
unitvolume of soil without achange in multiyear freeze or thaw
temperature-in British thermal units depths that may develop as a
(Btu) per cubic foot (ft3): result of some long-term
W change in the heat balance at
L=144 yd (eq 2-4) the ground surface.
100 h. Fusion parameter, /j"
*TM 5-852-6/AFR 88-19, Volume 6
C
2-2. Thermal properties of soils and other
N -_
"s (eq 2-8) construction materials.
L
a. The basic thermal properties of
whereve has the two possible meanings soils and other construction materials
noted above. used to calculate depths of freeze and
i. "Lambda" coefficient, JI " A factor thaw are specific heat, thermal con-
allowing for heat capacity and initial ductivity and volumetric latent heat of
temperature ofthe ground (see fig. 3-1). fusion. Other terms used in heat-flow
Aadj). calculations are derived from these
(eq 2-9)
dataand the elements ofweight, length,
j. Thermal regime. The temperature temperature and time.
pattern existing in a soil body in rela- b. the specific heat of most dry soils
tion to seasonal variations. near the freezingpoint maybe assumed
k. British thermal unit, Btu. The to be constantat the value of0.17 Btu/lb
quantity of heat required to raise the F. Specific heats of various materials
temperature of1 pound (lb) ofwater 1F are given in table 2-1; see theASFIRAE
at about 40F. Guide and Data Book of the American

Table 2-1. Specific heat values of various materials .*


(U.S. Army Corps of Engineers).

Temperature Specific heat


Material (OF) (Btullb OF)

Aluminum -27 .4 0.20


Asbestos fibers -- 0.25
Concrete (avg. stone) -- 0.20
Concrete (dams) - 0.22
Copper 44.6 0.20
Corkboard 91 0.43
-19 0.29
Cork, granulated -- 0.42
Fiberglas board 111 0.24
-22 0.19
Foamglas -20 0.16
Glass block, expanded 112 0.18
Glass sheets -- 0.20
Glass wool -- 0.16
Ice 32 0.48
Iron (alpha) 44.6 0.11
Masonry __ 0.22
Mineral wool __ 0.22
Perlite, expanded __ 0.22
Polystyrene, cellular foam __ 0.27
Polyurethane foam -- 0 .25
Sawdust -- 0.60
Snow -- 0 .50
Steel __ 0.12
Straw -- 0.35
Water -- 1 .00
Woods (avg .) Y 68 0.33
Woods fiberboard 148 0 .34

Specific heat values shown to nearest 0.01 . Average values listed where temperature is not shown .

2-2
*TM 5-852-6/AFR 88-19, Volume 6
Society of Heating and Air Condition- dry unitweight of1201b/ft 3 and awater
ing Engineers for the specific heat content of 15 percent. Its volumetric
values of common materials . latent heat of fusion, L, is (144 X 120 X
c. The thermal conductivity of soils 0.15 =1 2592 Btu/ft3 and its average volu-
is dependent upon anumber offactors: metric heat capacity, C, is (120[0.17 +
density ; moisture content; particle 0.75 X 0.15] =) 33.9 Btu/ft3 OF. The quan-
shape; temperature ; solid, liquid and tity of heat required to change the
vapor constituents; and the state ofthe phase of pore water in 1 ft3of this soil at
pore water, whether frozen or unfrozen. 32F is the same as that required to
Average values, expressed in Btu/ft cause a temperature change of
hour F, for frozen andunfrozen granu- (2592/33 .9 =) 76 .4'F when a phase
lar soils, silts and clays should be read change is not involved.
from figures 2-1 through 2-4. The 2-3 . Fundamental considerations.
charts for sands and gravels are applic-
able when the silt and clay content a. Theoretical basis. The freezing
together make up less than 20% of the or thawing of soils is the result of re-
soil solids. The charts for silt and clay moving or adding heat to an existing
are applicable when that fraction is at soil mass . The movement of heat is
' least 50%. For intermediate silt-clay always inthe direction of lower temper-
fractions, it is recommended that the ature. The time rate of change of heat
simple average of the values for the content depends on the temperature
two sets of charts be used. In all cases, differential in the direction of heatflow
the error in the thermal conductivity and on the thermal properties of the
estimates may be 25%, and even soil.
higher when the percentage of quartz b. Physical factors and data re-
grains in the soil is exceptionally high quired. Calculation of the depth of
or low. Figures 2-5 and 2-6 presentesti- freeze or thaw is based on knowledge
mates of the average thermal conduc- of the physical and thermal properties
tivity of frozen and unfrozen peat. An of the soil in the profile, the existing
excellent source of data for dry con- thermal regime, and the nature and
struction materials is the ASHRAH duration of boundary conditions
Guide and Data Book. Thermal con-
ductivity values for a number of com-
mon construction materials are listed 150
in table 2-2. 0I I
d. The latent heat of fusion is the 14

amount of heat required to cause a 13


phase change in soil moisture. This
amount of heat does not change the ~I2.0
temperature of the system when \\
freezing or thawing takes place. The 1 .8
gravimetric latent heat of fusion of 0
water is assumed to be 144 Btu/lb . The 90 1
amount ofheat energy required to con- 8
T 100%/ Saturation

vert 1 ft of water to -ice is (144 X 62.4 =) e0


9000 Btu/ft3 and to change 1 ft3 of ice to 0 0.2 o. s o.s \
water is (144 X 0.917 X 62.4 =) 8240
Btu/ft3. (Note: The density ofwater 62.4 6

lb/ft3 and that of pure ice Is 57.2 lb .ft3). 50


e. Figures 2-7 and 2-8 may be used 0 10 2030 40 50 60 70

to determine the average volumetric Moisture Content (%)


Thermal conductivity K is expressed in Btu per hour per square foot per
heat capacity and volumetric latent unit thermal gradient in F per foot .
heat of fusion, respectively, of moist Dashed line represents extrapolation .
soils.
f. The following example illustrates (U .S . Army Corps of Engineers)
the significance of latentheatof fusion
relative to the volumetric heat capacity Figure 2-1. Average thermal conductivity for sands and
for a moist soil. Assume a soil having a gravels, frozen.

2-3
*TM 5-852-6/AFR 88-19, Volume 6

nK"li
~.eGil~y .yrnHl
130

120

;'
E~
BEEN, rrrr
INNOM
/1I usit POW
140
W 100
v
a
90
130
La
u
W 3 80
70

i
0 60
3 50
H
2 WIV 40
i 300
3- 100
10 20 30 40 50 60 70
0
' 1- Moisture Content (/,)

w
90 d 90 120 100% saturated
LIM 50 N/A N/A 1 .20
40 1.00 N/A 1 .23
so 30 0 .98 1 .26
0 10 20 30 40 20 0 .61 1 .26
pure ice 1 .26
MOISTURE CONTENT, Percent Thermal
conductivitY K is expressed'iq :$tu perhour per square foot per
(U .S . Army Corps of Engineers) unit thermal gradient in F per foot.
Dashed line represents extrapolation .
Figure 2-2. Average thermal conductivity for sands andgravels, unfrozen .

(U .S . Army Corps of Engineers)


Figure 2-3. Average thermal conductivity for silt and clay
soils, frozen.

iav

roenwr it"r. ..,nmn

20111 I"
.r H. hr .MI lMrwal
4 rs~
.~
i

w 110 I" 
NE

I 00 1111,101k0m --
C7
3 90

z
so 111W 0

70 01,,
_ ' ~~-

aa ""
0 10 t0 30 40

MOISTURE CONTENT, Percent


(U .S . Army Corps of Engineers)
Figure 2-4. Average thermalconductivity forsilt and clay soils, unfrozen .

2-4
'TM 5-852-6/AFR 88-19, Volume 6

_0
200 400 600 800 1000 1200 1400 1600 1800 2000
Water Content (%dry weight)

(U .S . Army Corps of Engineers)


Figure 2-5. Average thermal conductivity for peat, frozen .

600 800 1000 1200 1400


Water Content (%dry weight)
(U.S. Army Corps of Engineers)

Figure 2-6. Average thermal conductivity for peat, unfrozen.

causingachange in the thermal regime . sional heat flow may be assumed, there-
Data pertinent to the soil profile in- by simplifying the problem and its
clude grain-size distribution, classifica- solution.
tion, density, water or ice content, and
temperature of each soil stratum . 2-4. Freezing and thawing indexes.
Knowledge of the thermal properties a. Physical concept and quantita-
of all the materials in the heat flow tive measurement. The penetration of
path is also required . Measured or as- freezing or thawing temperatures into
sumed temperatures within the soil soil partly depends on the magnitude
mass determine the initial conditions . and duration of the temperature dif-
Ifsurface temperatures can be assumed ferential at the air-ground interface.
to be spatially uniform and the thermal The magnitude of the temperature dif-
influence ofany buried structures can ferential is expressed as the number of
be considered negligible, one-dimen- degrees that the temperature in the. air
2-5
*TM 5-852-6/AFR 88-19, Volume 6

Table 2-2. Thermal properties of construction materials


(U.S . Army Corps of Engineers) .

k K
conductivity conductivity
Type of Unit weight (Btulft2 -hr" F (Btulft "
material Description (Iblit3) per in .) hr-OF)

Asphalt Mix with 6% by


paving . weight cut-
mixture back asphalt 138 10.3 0 .86
Concrete With sand and
gravel or stone
aggregate (oven-
dried) 140 9 .0 0.75
With sand and
gravel or stone
aggregate (not
dried) 140 12.0 1 .00
With lightweight
aggregates, in- 120 5 .2 0.43
cluding expanded
shale, clay or 100 3 .6 0 .30
slate ; expanded 80 2 .5 0 .21
stags ; cinders ; 60 1 .7 0 .14
pumice; perlite ; 40 1 .15 0 .096
vermiculite ; also 30 0 .90 0 .075
cellular con- 20 0 .70 0 .058
cretes .
Wood Maple, oak and
similar hard-
woods 45 1 .10 0 .092
Fir, pine and
similar soft-
woods 32 0 .80 0 .067
Building Asbestos-
boards cement board 120 4 .0 0 .33
Plywood 34 0 .80 0 .067
Wood fiberboard,
laminated or
homogeneous 26,33 0.42,0 .55 0 .035,0.046
Wood fiber-
hardboard type 65 1 .40 0 .12
Blanket Mineral wool,
and batt fibrous form,
insula- processed from 1 .5-4.0 0 .27 0 .022
tion rock, slag, or
glass
Wood fiber 3.2-3.6 0 .25 0 .021
Board Cellular glass 9.0 1' 0.39 0 .032
and Corkboard (with-
slab out added binder) 6.5-8.0 10.27 0 .022
insula- Glass fiber 9.5-11 .0 0 .25 0 .021
tion Wood or cane fiber-
interior finish 15.0 0 .35 0 .029
(plank, tile, lath)
Expanded polystyrene 1 .6 0.29 0 .024
Expanded ureaformal-
dehyde 1 .0 0.25 --
Expanded perlite 9 .5 " 11 .5 0 .34 --
Polyurethane foam 1 .5-3.0 0.17 --
Mineral wool with
resin binder 15.0 0 .28 0 .023
Mineral wool with
2-6 asphalt binder 15.0 0 .31 0.026
'TM 5-852-6/AFR 88-19, Volume 6 '
Table 2-2. Thermal properties of construction materials, Continued
k K
conductivity' conductivity
Type of Unit weight (Btulft2-hr"F (Btulft-
material Description (Iblft3) per in.) hr-F)

Loose Cork, granulated 5-12 0.25-0.36 --


fill Expanded perlite 3-4 0.28 --
Insula- Mineral wool
tion (glass, slag, or rock) 2-5 0.30 0.025
Sawdust or shavings 8-15 0.45 0.037
Straw 7-8 0.32
Vermiculite (ex-
panded) 7.0-8.2 0.48 0.040
Wood fiber: redwood,
hemlock, or fir 2.0-3.5 0.30 0.025
Miscel- Aluminum 168 1416 118
laneous Copper 549 2640 220
Ductile Iron 468 360 30
Glass 164 5.5 0 .46
Ice 57 ~15.4 1 .28
Snow, new, loose 5.3 0.6 0 .05
Snow on ground 18.7 1.56 0.13
Snow, drifted and
compacted 31 .2 4.8 0.40
Steel 487 310 25 .8
Water, average 62.4 4.2 0.35

Values fork are fordrybuilding materials at a mean temperature of 75F except as noted; wet conditions will adversely affect values of
many of these materials.
t Mean temperature of 60 F.
Mean temperature of 320 F.

or at the ground surface is above (posi- freezing degree-days, is usually


tive) or below (negative) 32F, the as- omitted.
sumed freezing point of water. The (2)Calculation fromaverage month-
duration is expressed in days. ly temperatures . The freezing or
b. Air freezing index (F) and air thawing index may be calculated by
thawing index(1). The air freezing and determining the area between the 32F
air thawing indexes, as defined in TM line and the curve of average monthly
5-862-1/AFR 88-19, Volume 1, may be temperature and time, taken over the
determined by the following methods. appropriate season . The area may be
(1) Summation of degree-days of determined by planimeter or a simple
freeze and thaw from average daily approximation rule (Simpson's rule,
temperatures. If T1 is the maximum midordinate rule, etc.). The areas are
daily air temperature and T2 is the expressed in units of degree-days, re-
minimum daily air temperature. the sulting in a summation of degree-days
average daily air temperature T, may or a freezing or thawing index. For an
be taken as 1/2 (TI + T2), and the number example refer to figure 2-9, a plot of
of degree-days for the day is (T -_32) . the monthly average temperatures at
The summation of the degree-days for" Fairbanks, Alaska, from September
a freezing or thawing season gives the 1949 to October 1950 . Determination of
air freezing or air thawing. Index. areas by planimeter gave a freezing
Table 2-3 illustrates the method used index of -5240 degree-days and a
to obtain the summation ofdegree-days thawing index of +3420 degree-days.
for a1-week period, assuming that-456 The use ofSimpson's rule gave a freez-
degree-days had been accumulated ing index of -5390and a thawing index
since freezing began. An average daily of +3460 degree-days . Either pair of
temperature based onhourly tempera- indices is adequate for computations.
tures would be slightly more accurate, c. Surfacefreezing and surface-
but such precision is not usually war- thawing indexes. For determining the
ranted. The negative sign, indicating heat flow within the soil, it is neces-
2-7
*TM 5-852-6/AFR 88-19, Volume 6

W
90

130 -40

Btu/cu ft F
120-" SO
30

40

-20
100 -
30 RC .C .
28 A.C .

90 --

!- 10
20

16 -5

7'o -L- -

Avre.' Specific hoot of save solids assumed to be O./T ON //&. *F

(U .S . Army Corps of Engineers)

Figure 2-7. Average volumetric heat capacity for soils.

sary to determine or estimate the temp- acteristics, and subsurface thermal


erature condition at the ground sur- properties. Heat balance algorithms
face . Since air temperatures are gen- that approximate many of these inter-
erally available and surface tempera- relationships exist but they are often
tures are not, a correlation between unwieldly to use and the inputs are
these temperatures helps establish the often difficult to characterize . It is
thermal boundary condition at the recommended that the ratio ofsurface
ground surface . The combined effects index to air index, designated as the
of radiative, convective and conductive "n-factor," be used to represent a
heat exchange at the air-groundinter- monthly or seasonal correlation . Re-
face often must be considered in deter- liable determination ofthe n-factor for
mining surface temperature . a specific location requires concurrent
observations of air and surface tem-
d. Correlation of air and surface peratures throughoutanumberofcom-
indexes. No simple correlation exists plete freezing and thawing seasons
between air and surface indexes. The plus anticipation of future changes to
difference between air and surface conditions existing during the period
temperatures at any specific time is of measurement. Such determination
influenced by latitude, cloud cover, is often not feasible, so n-factors must
time of year, time of day, atmospheric generally be estimated conservatively
humidity and stability, wind speed, from n-factors tabulated for other, pref-
snow cover and ground surface char- erably similar, sites.
2-8
"TM 5-852-6/AFR 88-19, Volume 6

r
W
Ib/cu ft
L
140 -I
0 Btu/cu ft -t- 40

130-

KEY
120-

too-

too-

90-

Go-

-PO -"A.-

(U .S . Army Corps of Engineers)

Figure 2-8. Volumetric latent heat for soils.

Table 2-3. Calculation of cumulative degree-days


(U.S. Army Corps of Engineers).

Temperature ( F) Degree-days Cumulative


Day Maximum Minimum Average per day degree-days

1 29 1 15 -17 -473 *
2 9 -11 - 1 -33 -506
3 10 - 8 1 -31 -537
4 15 - 1 7 -25 -562
5 30 16 23 - 9 -571
6 38 30 34 2 -569
7 30 18 24 -8 -577

Prior accumulation of -456 degree-days assumed.

2-9
*TM 5-852-6/AFR 88-19, Volume 6

60

50

W 40
w

S
I-
Q 32
30
W
0-
M
W
E-
20
JS
I-
Z
O 10

W
Q
S 0
W
Q

-20

(U.S . Army Corps of Engineers)


Figure 2-9. Average monthly temperatures versus time at Fairbanks, Alaska.

(1) Freezing conditions. The n- basis ofobservations and studies made


factor is very significant in analytical to date, the n-factors given for average
ground studies. It generally increases conditions in table 2-4 should be used
with wind speed. Snow cover reflects a to convert the air freezing index to the
large part of incoming solar radiation surface freezing index in the absence
resulting in higher freezing indexes of specific measurements at the site of
at the snow surface, but its insulating planned construction .
effect can greatly reduce the freezing (2) Thawing conditions. The n-
indexat the ground surface. The effects factor for thawing conditions is af-
of turf or an organic ground cover on fected by the same factors as those for
the heat flow processes at the air- freezing conditions . It is the ratio of
ground interface are extremely vari- surface degree-days of thaw (degrees
able and difficult to evaluate. On the above 32F) to air degree-days of thaw.
2-10
"TM 5-852-6/AFR 88-19, Volume 6

Incoming shortwave radiation may in- in figure 2-10 and are based on field
troduce heat into the surface to an studies conducted in Alaska and
extent that the surface becomes a Greenland. The n-factors given for
source ofheatconducted notonly down- average conditions in table 2-4 should
ward but upward into the air. In such a be used to convert air thawingindexes
case the n-factor may become signifi- to surface thawing indexes in the ab-
cantly larger than 1.0 . The effect of sence of specific measurements at the
latitude is not particularly significant planned construction sites.
in arctic and subarctic areas, but con- e. Design indexes. For design ofper-
sideration should be given to the effect manent pavements, the design freezing
of wind speed. Recommended curves (or thawing) index should be the aver
for n-factors versus wind speed for age air freezing (or thawing) index of
portland-cement-concrete and bitu- the three coldest winters (or warmest
minous-concrete pavements are shown summers) in the latest 30 years of

Table 2-4. n-factors for freeze and thaw (ratio of surface index to air index)
(U.S. Army Corps of Engineers).

For freezing For thawing


Type of surface conditions conditions

Snow surface 1 .0 _-
Portland-cement concrete 0.75 1 .5
Bituminous pavement 0.7 1 .6 to 2t
Bare soil 0.7 1 .4 to 2t
Shaded surface 0.9 1 .0
Turf 0.5 0.8
Tree-covered 0.3** 0.4

Surface exposed directly to sun or air without any overlying dust, soil, snow or ice, except as noted otherwise, and with no building
heat involved .
t Use lowest value except in extremely high latitudes or at high elevations where a major portion of summer heating is from solar
radiation.
Data from Fairbanks, Alaska, for single season with snow cover permitted to accumulate naturally .

3.0
I
X

2.6
I
S CANCRETE
a~rc 2.0 ~~ ,
N

C
`PoWTCAA9 CEMENT CONCRETE

1.0
0 2 4 6 6 10 12 14 16
AVERAGE WIND SPEED, mph, DURING THAWING SEASON

(U .S . Army Corps of Engineers)


Figure 2-10. Relationship between wind speed and n- factor during thawing
season .
2-1 1
'TM 5-852-6/AFR 88-19, Volume 6

record . If 30 years of record are not struction need not be changed more
available, the air freezing (or thawing) often than once in 5 years unless the
index for the coldest winter (or more recent temperature records in-
warmest summer) in the latest 10-year dicate a significant change. The distri-
period may be used. For design of bution ofdesign freezingand thawing
foundations for average permanent index values in North America is
structures, the design freezing (or presented in TM 5-862-1/AFR 88-19,
thawing) index should be computed Volume 1. Therelatively linear relation-
for the coldest (or warmest) winter in ship between recorded maximum in-
30 years of record or should be esti- dexes and mean freezing and thawing
mated to correspond with this fre- indexes shown in figures 2-11 and 2-12
quency ifthe number of years of record may be used in conjunction with dis-
is limited . Periods of record used tribution of mean and freezing and
should be the latest available. To avoid thawing indexes in TM 5-862-1/AFR
the necessity for adopting a new and 88-19, Volume 1 to determine the design
only slightly different freezing (or index values for arctic and subarctic
thawing) index each year, the design regions.
index at a site with continuing con-

I Cg000
I (BARROW!o
W
h
(rHIXE) o
9000

/6(Fr YUKON/
. e0001-
x (NORrNWAY) o
(BETTLES)
W
O
Z

Z 7000F-
N
W
W / o(KOrZESUE)
(FAIRBANKS!
W
(6ULKANA) o
6000
(B/6 DELTA! o,
Q

(AmAloo

X 5000
Q o(BEMEL )

NOTE .' rhulr, bnenlond dolo from /952 through 1961 .


400,0 I I I I I
3000 4 000 5000 6000 7000 8000 9000

MEAN AIR-FREEZING INDEX, Degree-days F


(U .S . Army Corps of Engineers)

Figure 2-11 . Relationship between mean freezing index and maximum freezing index for 10
years of record, 1953-1962 (arctic and subarctic regions) .
*TM 5-852-6/AFR 88-19, Volume 6

4000
(BW OELW T
(FAIRBAAWS)

IL 3500
(FT. YUIfQY)
r
(ANIAK) 0
Ir (MOPrHWAY1
(BETNEL) a' VG7/LKANA)
3000 0 wrnESI

z 2500

Z
0(KOTIEBUE)

1500

X ( THUL E) 0
Q

1000 -IBARROWI 0/

NOTE .' Mule, areonlend dolo from 1952 /Amuph /of/ .

500 Imo_- L. j I 1 I 1 I I I t
0 500 1000 1500 2000 2500 3000 3500

MEAN AIR-THAWING INDEX, D"rwdoysF


(U.S. Army Corps of Engineers)
Figure 2-12. Relationship between mean thawing index and maximum thawing index for 10 years of
record, 1953-1962 (arctic and subarctic regions) .
'TM 5-852-6/AFR 88-19, Volume 6

CHAPTER 3

ONE-DIMENSIONAL LINEAR AND PERIODIC HEAT FLOW

3-1 . Thermal regime . = coefficient that considers the


The seasonal depths of frost and thaw effect of temperature changes
penetration in soils depends upon the in the soil mass (dimensionless).
thermal properties of the soil mass, the The A coefficient is a function of the
surface temperature (upper boundary freezing (or thawing) index, the mean
condition) and the thermal regime of annual temperature of the site, and the
the soil at the start of the freezing or thermal properties of the soil. Freeze
thawing season . Many methods are and thaw of low-moisture-content soils
available to estimate frost and thaw in the lower latitudes is greatly in-
penetration depths and surface tem- fluenced by this coefficient . Itisdeter-
peratures . Some of these are sum- mined by two factors : the thermal ratio
marized in appendix B. This chapter a and the fusion parameter /j. These
concentrates on some techniques that have been defined in paragraph 2-1 .
require only relatively simple hand Figure 3-1 shows A as a function of a
calculations . For the computational and N .
methods discussed below, the initial b. A complete development of this
ground temperature isassumed to uni- equation and a discussion of the neces-
formly equal the mean annual air tem- sary assumptions and simplifications
perature of the particular site under made during its development are not
consideration . The upper boundary presented here. A few of the more im-
condition is represented by the sur- portant assumptions and some of the
face freezing (or thawing) index. equation limitations are discussed
below. The assumptions and limita-
3-2. Modified Berggren equation. tions apply regardless of whether the
a. The depth to which 32F tem- equation is used to determine the depth
peratures will penetrate into the soil of freeze or the depth of thaw.
mass is based upon the "modified" (1)Assumptions. The mathematical
Berggren equation, expressed as: model assumes one-dimensional heat
48KnF flow with the entire soil mass at its
mean annual temperature (MAT) prior
L to the start of the freezing season . It
or (eq 3-1) assumes that when the freezing season
48 K nl starts, the surface temperature
changes suddenly (as a step function)
L from the mean annual temperature to
where a temperature v3 degrees below freez-
X= depth of freeze or thaw (ft) ing and that it remains at this new
K= thermal conductivity of soil temperature throughout the entire
(Btu/ft hr F) freezing season. Latent heat affects the
model by acting as a heat sink at the
L = volumetric latent heat of fusion
(Btu/ft3) moving frost line, and the model as-
sumes that the soil freezesata tempera-
n=conversion factor from air ture of 32F .
index to surface index (dimen- (2) Limitations. The modified
sionless) Berggren equation is able to determine
F= air freezing index (F-days) frost penetration in areas where the
I = air thawing index (F-days) ground below a depth of several feet
*TM 5-852-6/AFR 88-19, Volume 6

IIIMIdMI M%I/
ImpAPEFAMUFAF
I19,aPAPIMAPOAF

N MINE 10110. 115111111211 Ap

1111 ~II~D~p "/~~~~~

111111111111111!1 1 0
;
,
lI/I%/I%I MEN
.MMwMII0;WAMIIW,I "
I%IL//w%/

MAI

OPFA
'A NFAO-IE 'A POPPAP
FOR JIPPAFA
IF-

I's 04 AP04
0

n~~r~i~r/y
a~inztiRai~azra~i~w~an
W
4-1

Alai
N 0

a
WONFA 0
U

NII~~ ,IPI 'A ll


~~I
4ua!o1j;aOO
Figure 3-1. A coefficient in the modified Berggren formula.

3-2
*TM 5-852-6/AFR 88-19, Volume 6
remains permanently thawed, or to -Surface freezing index (nF) _
determine thaw penetration in areas 2600 degree-days.
where the ground below a depth .of -Length of freezing season (t)
several feet remains permanently = 160 days.
frozen. These two conditions are simi- -Soil properties: yd =1001b/ft3, w
lar in that the temperature gradients = 16%.
are of the same shape, although re- The soil thermal properties are as
versed with respect to the 32F line. No follows:
simpleanalytical method exists to deter-
mine the depth of thaw in seasonal -Volumetric latent heat of fusion,
frost areas or the depth of freeze in L = 144(100)(0.16) = 2160 Btu/ft3. (eq. 3-2).
permafrost areas, and such problems , -Average volumetric heat
should be referred to HQDA (DAEN- capacity,
ECE-G) or HQ AFESC . Numerical tech- Cavg =100[0.17++(0.75X0.l5)]
niques and computer programs are = 28 .3 Btu/ft3 -F . (eq 3-3)
available to solve more complex prob- -Average thermal conductivity,
lems. Appendix B discusses some ther- Kf = 0.80 Btu/ft hr F (fig. 2-3)
mal computer models for computing Ku = 0.72 Btu/ft hr F (fig. 2-4)
freeze and thaw depths. The modified
Berggren equation cannot be used suc- Kav = 1/2 (K u + Kf) --' 0.76 Btu/ft
cessfully to calculate penetration over hr f'.
parts of the season. The modified The ,l coefficient is as follows:
Berggren equation does not account -Average surface temperature
for any moisture movement that may differential,
occur within the soil. This limitation vs = nF/t - 2600/160
would tend to result in overestimated - 16.6F (16.6F below 32F). (eq 3-4)
frost penetration (if frost heave is sig- -Initial temperature differential,
nificant) or underestimated thaw
vo - MAT -32 - 37.2-32.0 =
penetration. 6.2F (6 .2 above 32F). (eq 3-6)
(3) Applicability. The modified
Berggren equation is most often applic- -Thermal ratio,
able in either of two ways: to calculate a = vo/vs = 6.2/16.6 - 0.33. (eq 3-6)
the multi-year depth of thaw in perma- -Fusion parameter,
frost areas or to calculate the depth of p - vs (C/L) - 16.6(28.2/2160) - 0.20. (eq 3-7)
seasonal frost penetration in seasonal -Lambda coefficient,
frost areas. It is also sometimes used to
calculate seasonal thaw penetration .1 - 0.89 (fig. 3-1) . (eq3-8)

J
(active layer thickness) in permafrost Estimated depth of frost penetration,
areas . 48(0.76)(2600)
48 K nF
X =~ = 0.89 5.8 ft
L 2160 (eq 3-8
3-3. Homogeneous soils.
The depth of freeze or thaw in one 3-4. Multilayer soils.
layer ofhomogeneous soilmay be deter- A multilayer solution to the modified
mined by means of the modified Berggren equation is used for non-
Berggren equation. A thin bituminous homogeneous soils by determining
concrete pavement will not affect the that portion of the surface freezing (or
homogeneity of this layer in calcula- thawing) index required to penetrate
tions, but a portland-cement-concrete each layer. The sum of the thicknesses
pavement greater than 6 inches thick of all the frozen (or thawed) layers is
should be treated as a multilayered the depth of freeze (or thaw) . The
system . In this example for homoge- partial freezing (or thawing) index re-
neous soils, determine the depth of quired to penetrate the top layer is
. frost penetration into a homogeneous given by
sandy silt for the following conditions: Lid, Ry
-Mean annual temperature F1 (or 11) - 2 ( ) (eq 3-10)
(MAT) = 37.2F. 24a 1 2

3-3
'TM 5-852-6/AFR 88-19, Volume 6
where Since a wind speed of 7-1/2 mph results
dl = thickness of first layer (ft) in an n-factor of 2.0 (fig. 2-10), a surface
R1 = dl/K1 = thermal resistance of thawing index nI of 1560 degree-days
first layer. is used in the computations . The vs, vo
and a values are determined in the same
The partial freezing (or thawing) index way as those for the homogeneous case:
required to penetrate the second layer vs- 1560/105 = 14.8F (eq 3-15)
is
L2d2 vo= 12.0 - 32.0 = 20.0F (eq 3-16)
R2 a s 20.0/14 .8 - 1.35.
F2 (or I2) = . (111+ -) . (eq 3-11) (eq3-17)
24x2 2 The thermal properties C, K and L of
the respective layers are obtained from
The partial index required to penetrate figures 2-1 through 2-8.
the nth layer is: b. Table 3-1 facilitates solution of
the multilayer problem, and in the fol-
Lnd n Rn lowing discussion, layer 3 is used to
Fn (or I n) - (MR + - ) (eq 3-12) Illustrate quantitative values. Columns
24.12 2 9, 10, 12 and 13 are self-explanatory.
Column 11, T:, represents the average
where IR is the total thermal resistance value of L for a layer and is equal to
above the nth layer and equals ELd/ld (2581/5.0 = 517). Column 14, Zs,
represents the average value of C and
R1 + R2 + R3 ... + R n.1 " (eq 3-13)
is obtained from XCd/ld (145/5.0 = 29).
The summation of the partial indexes, Thus T: and -C represent weighted
values to a depth of thaw penetration
F 1 +F2 + F3 .. .+Fn(orI 1 +I2+13 ... given by Ed, which is the sum of all
+ In) (eq 3-14) layer thicknesses to that depth .
The fusion parameter p for each layer
Is equal to the surface freezing index is determined from
thawing index) .
a. In this example, determine the v 8 (C /L) = 14.8 (29/517) = 0.83 . (eq 3-18)
depth of thaw penetration beneath a The A coefficient is equal to 0.508 from
bituminous concrete pavement for the figure 3-1. Column 18, Rn, is the ratio
following conditions: d/K and for layer 3 equals (3.0/2.0) or
-Mean annual temperature (MAT) 1.5. Column 19, FR, represents the sum
= 12F. of the Rn values above the layer under
consideration . Column 20, IR + (Rn/2),
-Air thawing index (I) equals the sum of the R n values above
= 780 degree-days. the layer plus one-half the Rn value of
-Average wind speed in summer the layer being considered . Fog ,ayer 3
7.5 miles per hour (mph). this is [1.32 + (1.50/2)] = 2.07. Column 21,
-Length of thaw season (t) nI, represents the number of degree-
= 105 days. days required to thaw the layer being
-Soil boring log-. considered and is determined from

Depth Dry unit weight Water content


Layer (ft) Material' (Iblft3) (%)

1 0.0-0.4 Asphaltic concrete 138 _-


2 0 .4-2.0 G W-G P 156 2.1
3 . 2.0-5.0 GW-G P 151 2.8
4 5.0-6.0 SM 130 6.5
5 6.0-8.0 SM-SC 122 4.6
6 8.0-9.0 SM 116 5.2

~In accordance with Unified Soil Classification System .

3-4
TM 5-852-6/AFR 88-19, Volume 6

Table 3-1, Multilayer solution of modified Berggren equation (U .S . Army Corps of Engineers) .

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22)

R w
Layer Yd w d Ed C K L Ld ELd L Cd Md C It .~ ~2 Rn BR D2 +2 nl PnI

c
1 138 -- 0,4 0,4 28 0,86 0 0 0 -- 12 12 -- -- -- -- 0,46 0 0,23 -- -- m
2 156 2,1 1,6 2,0 29 1,85 470 751 751 376 46 58 29 1,15 0,455 0,207 0,86 0,46 0,89 134 134
3 151 2,8 3,0 5,0 29 2,00 610 1830 2581 517 87 145 29 0,83 0,508 0,258 1,50 1,32 2,07 612 746 0
4 130 6,5 1,0 6,0 28 1 .65 1220 1220 3801 633 28 173 29 0,68 0,537 0,288 0,60 2,82 3,12 551 1297
5a 122 4,6 1,0 7,0 25 0,64 808 808 4609 658 25 198 28 0,63 0,552 0,305 1,56 3 .42 4,20 465 1762
5b 122 4,6 0,6 6,6 25 0,64 808 485 4286 650 15 188 28 0,64 0,550 0,303 0,94 3,42 3 " 89 260 1557 "
3
0
a
a = 1,35 vs = 14,8F ni = 1560 degree-days

__ (470) (1,6)
12 (24) (0,207) (0 " 89) = 134

(610)(3,0)
I = (2,07) = 612
3 (24) (0 .258) o
_ (1220)(1,0
(3'12) = 551
14 (24) (0,288)

(808)(1,0)
1 5a (24) (0,305) (4 " 20 ) = 465
70
_ (808)(0,6)
1 5b 24) (0,303) (3 " 89) = 26

Total thaw penetration = 6,6 feet


O
C
*TM 5-852-6/AFR 88-19, Volume 6
Ld equation, with snow cover thickness
n1 - (7-R + (eq 3-19) estimated seasonally . The tabulation
24,2 below presents average thermal proper-
For layer 3, ties of snow applicable for calculation
(610)(3.0) in the noted regions if a better data
n13 = (2.07) (eq 3-20) base is not available. In the absence of
24(0 .608)2 site-specific data, figures 2-5 and 2-6
= 612 degree-days should be used to estimate the thermal
conductivities of vegetative surface
The summation of the number of cover.
degree-days required to thaw layers 1
through 4 is 1297, leaving (1560 -1297 =) 3-6. Surface temperature variations .
263 degree-days to thaw a portion of The temperatures at the air-ground
layer 5. A trial-and-error method is used interface are subject to daily and sea-
to determine the thickness of the sonal fluctuations. Precipitation, insola-
thawed part of layer 5. First, it is as- tion, air temperature and turbulence
sumed that 1.0 feet of layer 5 is thawed contribute to these variations in sur-
(designated as layer 5a). Calculations face temperature . To facilitate mathe-
indicate 465 degree-days are needed matical calculations, two assumptions
to thaw 1.0 foot of layer 5 or (465 - 263 =) are commonly made regarding the tem-
202 degree-days more than available. peratures at the upper boundary : 1) a
A new layer, 5b, is then selected by the sudden step change occurs in surface
following proportion temperature or 2) the surface tempera-
ture change is sinusoidal. The sinu-
(263/466)1 .0 = 0.57 ft (try 0.6 ft). (eq 3-21) soidal variation of temperature over a
year closely approximates actual condi-
This new thickness results in 260 tions; however, it is amenable to hand
degree-days required to thaw layer 5b calculations only if latent heat effects
or 3 degree-days less than available. are negligible . Solutions and examples
Further trial-and-error isunwarranted for both conditions are given below .
and the total estimated thaw penetra- a. Sudden step change. This involves
tion would be 6.6 feet. A similar tech- a sudden change in the surface tem-
nique is used to estimate frost penetra- perature of a mass that was initially at
tion in a multilayer soil profile. a constant, uniform temperature . The
3-5. Effect of snow and vegetative cover. sudden step change was used to estab-
Thermal properties of snowand vegeta- lish the boundary conditions for heat
tive covers are extremely variable in flow in the modified Berggren equation
both time and space. Both materials given in paragraph 3-2. If the influence
tend to act as insulators and retard of latent heat is not involved, or is as-
heat transfer at the air-ground inter- sumed negligible, the following equa-
face.In freeze-thawcomputations, snow tion may be used:
and vegetative surface materials are
treated as separate layers in the multi- T(X t) = Ts + (To - Ts) erf ( ) (eq 3-22)
layer solution ofthe modified Berggren 2 -fat

Unit K
weight (Btulft) C L
Region (Iblft 3 ) hr F) (Btulft 3 F) (Btulft 3 )

Interior Alaska 16 0.11 8 2300

Canadian Archipelago,
N . Alaskan coast,
and temperate regions 20 0.18 10 2880

Northern Greenland 22 0.20 11 3170

3-6
*TM 5-852-6/AFR 88-19, Volume 6
where a = thermal diffusivity of the
mass (ft2/day = K/C)
T (X,t) = temperature at depth x, at
time (F) t = time after application of
sudden change in surface
TS = suddenly applied constant temperature (days).
surface temperature (F) Figure 3-2 gives the relationship be-
To =initial uniform tempera- tween (x/2 at) and erf (x/2 1rat). The
ture of the mass (F) expression for the error function is
erf =mathematical expression, shown in appendix A and figure 3-2.
- termed the error function, In thisexample ofasudden step change,
which is frequently used a highly frost-susceptible subgrade is
in heat flow computations covered with a 2-foot thick, non-frost-
(dimensionless) susceptible gravel pad. Both soils are
x = depth below surface (ft) at an initial temperature of 20F . If the

I 0
N

h64'
1

t0
O

..

v
N
O M
b0
G
W
W
O

A.
H
O
U

Figure 3-2. Relationship between (x/2 Y-a-t) and erf (x/2 at).

3- 7
*TM 5-852-6/AFR 88-19, Volume 6
surface ofthe gravel is suddenly heated the concrete . The average annual tem-
to and maintained at 70F fora number perature is [60 + (-40)1/2 = 10F and the
of days, estimate the temperature at surface amplitude is (60 - 10) = 50F.
the gravel-subgrade interface after one The amplitude at an 8-foot depth equals
day. The gravel material is very dry
and latent heat may be ignored . The
thermal conductivity of the gravel is
50 exp [-8
.V (1.0)(365)
= 50 e'0.742

1.0 Btu/ft hr F and the volumetric heat 24F . (eq 3-24)


capacity is 25 Btu/ft3. The thermal dif- The maximum temperature at 8 feet is
fusivity of the gravel is (K/C = 1.0/25) (10 + 24) = 34F. The time lag tx between
== 0.04 ft2 /hr., or 0.96 ft2/day, and the maximum temperature at the sur-
x/2 at = (2.0/2 1r0.~-X 1) = 1.02. From face and 8 feet is
figure 3-2, erf (x/2 -a6t) is equal to 0:85,
and the interface temperature T is tx
8 ~ 365
=43 days
[70 + (20 - 70)0.851 = 27.5F. 2 n(1.0) (about 6 weeks) . (eq 3-25)
b. Sinusoidal change. Asurfutce tem- (Note : latent heat would increase the
perature variation that is nearly sinu- time lag and decrease the amplitude.)
soidal repeats itself periodically for a c. Nonuniform layers. The method
surface exposed to the atmosphere. For of equivalent thickness is used to find
most problems in this manual, the sinu- the temperature at a point below a
soidal variation ofconcernoccurs over number of layers of different thermal
an annual cycle. If latent heat is not properties. This technique assumes a
involved or is assumed negligible, the negligible effect of latent heat, and in-
following equation may be used: volves reduction of each layer to an
equivalent material thickness by set-
Ax = A o exp (-x ) (eq 3-23) ting the ratio of the thicknesses equal
aP to the ratio of square roots of the ther-
where mal diffusivities . For example, deter-
A= amplitude oftemperature wave mine the equivalent gravel thickness
at depth x (F). for the three layers shown, assuming
Ao amplitude of the surface tem- all materials are unfrozen .
perature wave above or below The following table shows that 4.75
the average annual tempera- feet of the nonuniform materials can
ture (F) be considered equivalent to 5.4 feet of
x = depth below surface (ft) gravel for heat-flow purposes . This
equivalent thickness and the thermal
a = thermal diffusivity of the mass diffusivity of the gravel are used to
(ft2/day) calculate temperatures at the base of
P= period of sine wave (365 days) . the gravel layer by either the step-
The sinusoidal temperature pattern is change or sinusoidal method.
assumed to exist at all levels to a depth
where there is no temperature change. 3-7. Converting indexes into equivalent sine
The temperature waves lag behind the wave of temperature.
surface wave, and the amplitude of the Some problems may require the use of
sinusoidal waves decreases with depth the sinusoidal temperature variation
below the surface. The phase lag is technique, given only the freezing or
determined by t x = (x/2) ( -1r 6U
3 a) . thawing indexes and the average
Typical temperature-time curves for a annual temperature . These indexes
surface and at a depth x are shown in may be converted into a sine curve of
figure 3-3. temperature to give the same index
In the following example of a sinu- values and the sam emean temperature .
soidal temperature change, the surface For example, convert the monthly aver-
temperature ofan 8-foot-thick concrete age temperature data for Fairbanks,
slab varies from 60 to -40F during the Alaska, shown in figure 2-9 into an
year. Determine the maximum tem- equivalent sine wave. The relationship
perature at the base of the slab as- between the sinusoidal amplitude,
suming a diffusivity of 1.0 ft2/day for freezing index, thawing index and aver-
3-8
"TM 5-852-6/AFR 88-19, Volume 6

FREEZING INDEX Ar DEPTH (X)

TIME
(U.S . Army Corps of Engineers)

Figure 3-3. Sinusoidal temperature pattern .

Volumetric
Dry Unit Thermal heat Thermal
weight, Water conductivity, capacity, diffusivity,
yd content, K C a = K/C
Material (Ib/ft) w (%) (Btulft hr F) (Btulft3 F) (ft2/hr)
Concrete -- -- 1 .0 33.0 0.033
Sand 120 2 0.8* 23 t 0.035
Gravel 135 4 1 .5* 28 tx 0.054
From figure 2-2 .
tC = y d (0 .17 + w/100) .

Thermal
Thickness diffusivity Equivalent gravel
Material (ft) (ft2lhr) thickness (ft)
Concrete 1 .75 0.033 1 .3 2.3 (1 .3 X 1 .75)
Sand 0.50 0.035 1 .2 0.6
Gravel 2.50 0.054 1 .00 2.50
Total thickness 4.75 5.4
The subscript g refers to the gravel layer and the subscript m refers to the other material layer .

3-9
*TM 5-852-6/AFR 88-19, Volume 6

age annual temperature is shown in of 5240 degree-days, the sinusoidal


figure 3-4 . amplitude is found to be 37.0F. The
The average temperature at Fairbanks equation of the sine wave is
for October 1949 to September 1950 T = 27.0 - 37.0 sin 2vft (eq 3-26)
was 27F . By use of the freezing index = 27.0 - 37.0 sin 0 .0172 t (radians)

,1,' -1VI1N3a3ddl0 3an1Va3dW31 '1VIIINI 4 ?


J1
M A O
O ,$ N N 2 o O

w
T

4
N
P
O
x
W
0
Z
0
EWA VAF,FA :~ ~ ~~~NIMIN&e Z
INIMA iurua~ L11 ~RAM RAW\TAXU
a
2ManiMriiM "t~~n llWIW1 11R 0
MANFAM iiINTERIM""~~~~ ~NaI Z
a
I WHEN N

O Of N O r ON f?
b 10
10 M f W~ M N

d, '38f11Va3dW31 -lVnNNV 3JV83AV 'l

Figure 3-4 . Indexes and equivalent sinusoidal temperature.

3- 10
*TM 5-852-6/AFR 88-19, Volume 6

where 3-8 . Penetration of freeze or thaw beneath


f = frequency, 1/365 cycles per day buildings.
t = time from origin in days. (Origin The penetration of freeze or thaw
of curve is located at a. point beneath buildings depends largely on
where T intersects the average the presence or absence ofan airspace
annual temperature on its way between the building floor and the
downward toward the yearly ground as discussed below and in TM
minimum .) 5-852-4/AFM 88-19, Chapter 4.
If the-thawing index of 3240 degree- a. Building floorplaced on ground.
days had been used,the sinusoidal tem- When the floor of a heated building is
perature amplitude would be 35.50 F. placed directly on frozen ground, the
The actual temperature curve for Fair- depth of thaw is determined by the
banks, Alaska, and the equivalent sine same method as that used to solve a
wave computed from the freezing index multilayer problem when the surface
are plotted in figure 3-5. This illustra- is exposed to the atmosphere, except
tion makes use ofair indexes buta sine that the thawing index is replaced by
wave could be determined for surface the product of the time and the dif-
indexes by multiplying the air indexes ferential between the building floor
with appropriate n-factors. Note that temperature and 32'F. For example,
the mean annual ground surface tem- estimate the depth of thaw after 1 year
perature may be substantially different for a building floor consisting of 8
(frequently higher) from the mean an- Inches of concrete, 4 inches of insula-
nual air temperature because the tion and 6 inches of concrete, placed
freezingn-factor is generally not equal directly on a5-foot-thick sand pad over-
to the thawing n-factor . If the long- lying permanently frozen silt for the
term mean monthly temperature had following conditions :
been used instead of the average -Mean annual temperature = 20F.
monthly temperatures for the 1949-1950 -Building floor temperature = 650F.
period, the correlation between the -Sand pad: Yd= 133 lb/ft 3, w = 5%.
actual temperature curve and the equi- -Frozen silt: y d = 75 Ib/ft3, w = 45%.
valentsine curvewould practically coin-
cide, as shown in figure 3-6. -Concrete: K = 1.0 Btu/ft hr
'F, C = 30 Btu/ft3 oF.
-Insulation: K = 0.033 Btu/ft hr'F,
C = 1.5 Btu/ft 3 'F.
60 Sine Wave
The resistances of the three floor
of Equivalent Indexes
50 layers are in series, and the floor re-
sistance B.f isthe sum ofthe three layer
.d 40 THAWING resistances:
INDEX
Rf . d + 8
+
4
+ a
30 k (12)(1 .0) (12)(0.03:3) (12)(1.0)
27
- -- TAvg Annual = 11.2 ft 2 ft (eq 3-27)
r 20 /~ Temp
a FREEZING
INDEX / The average volumetric heat capacity
fd 10 of the floor system is
a0
" 0
Cr (30)(8) + (1,6)(4) + (30)(F3)
8+4+g
-10 Avg. = 23.7 Btu/ft 3,F (eq 3-28)
Temperature
-2 0 The solution to this problem, shown in
L_ IL table 3-2, predicts a total thaw depth of
QD~lM 138~ o
- . 1949 1950 7.8 feet. This solution did not consider
edge effects, i.e.,a long narrow building
(U .S . Army Corps of Engineers) will have lesser depth of thaw than a
Figure 3-5. Average monthly temperatures for
square building with the same floor
1949-1950 and equivalent sine wave, Fairbanks, because ofthe difference in lateral heat
Alaska . flow.
3-1 1
*TM 5-852-6/AFR 88-19, Volume 6

TT--T_ - _i

SINE WAVE EOUI VAL EN T


60 (AMPL/TUOE 37 F)

50

IJL
0
THAWING INDEX
W 4 3400 DEGREE-DAYS
MEAN TEMPERATURE
Q 32
W 30
a MEAN ANNUAL TEMPERATURE
26 .2
W
H

J 20 FREEZING
5300 DEGREE-DAYS
INDEX

H
Z
O
IO
Z
W
0

NOTE. Areas between curves and 32F line


represent the indexes.
-10 i*

SEPT I OCT I NOV -I DEC JAN I FEB I MAR I APR I MAY JUNE__ 1 JULY 1AUG A SEPT OCT

(U .S . Army Corps of Engineers)


Figure 3-6 . Long-term mean monthly temperatures and equivalent sine wave, Fairbanks, Alaska.

b. Airspace below building (2) There is no simple mathemati-


(1) An unskirted airspacebetween cal expression for analyzing the heat
the heated floor of building and the flow in a ventilated floor system that
ground will help prevent degradation has ducts or pipes installed within the
ofunderlying permafrost.The airspace floor or at some depthbeneath the floor,
insulates the building floor from the with air circulation induced by stack
ground and acts as a convective pas- effect. The depth to which freezing
sage for flow of cold air that dissipates temperatures will penetrate is com-
heat from the floor system and the puted with the modifiedBerggren equa-
ground .The depth ofthaw is calculated tion, except that the air freezing index
by means of the modified Berggren at the outlet governs. This index is
equation for either a homogeneous or influenced by a number of design vari-
multilayered soil system, as applicable. ables, i.e., average daily air tempera-
An n-factor of 1.0 is recommended to tures, inside building temperatures,
determine the surface thawing index floor and duct or pipe system design,
beneath the shaded areaof an elevated temperature and velocity of air in the
building . system, and stack height. Cold air
3-12
TM 5-852-6/AFR 88-19, Volume 6
m
w
N
Table 3-2 . Thaw penetration beneath a slab-on-grade building constructed on permafrost (U .S . Army Corps of Engineers),

m
a
R 3
~2 R
Layer d w d Ed C K L Ld ELd L Cd Xd C u n )R ~R +2 nl Fn I ""
w
J
Floor -- -- 1,5 1,5 24 -- 0 0 0 0 36 -- - -- -- 11,20 0 0 -- -- m
Sand 133 5 .0 5,0 6 .5 28 1,54 960 4800 4800 738 140 176 27 1 .21 0 .68 0 .463 3,25 11,20 12,82 5540 5540 m
Silt a 72 45 .0 1,5 8,0 37 0 .90 4650 6970 11770 1470 55 231 29 0,65 0,77 0,593 1,67 14,45 15 .29 7480 13020 w
Silt b 72 45,0 1,3 7,8 37 0,90 4650 6050 10850 1390 48 224 29 0,69 0,765 0 .586 1,44 14,45 15,17 6520 12060
a
0
v0 = 32 - 20 = 12F vs = 65-32 - 33F a = 12/33 = 0 .36
dn.

Surface thawing index (nl) = 33 x 365 = 12050 degree-days a


4800
ni(Sand) - (12 .82) = 5540 degree-days
24(0,463)

6970
nI(Silt a) = (15 .29) = 7480 degree-days
24(0 .593) ma
0
6050
ni(Silt b) - (15,17) - 6520 degree-days m N
24(0 .586)
3m a
Total thaw penetration = 7,8 feet oN
70

w
w
*TM 5-852-6/AFR 88-19, Volume 6

passing through the ducts acquires Outlet mean annual tempera-


-
heat from the duct walls and experi- ture = 32F 0, w = 2.51.
ences a temperature rise as it moves (conservative assumption).
through the duct, and the air freezing - Minimum site freezing index
index is reduced at the outlet. Field = 4000 degree-days.
observations indicate that the inlet air - Freezing season = 215 days.
freezing index closely approximates - Thawing season = 150 days
the site airfreezing index. The freezing (period duringwhich ducts are
index at the outlet must be sufficient closed).
to counteract the thawing index and - Building floor temperature
ensure freeze-back of foundation soils. = 60F.
(3) As an example, determine the - Thermal conductivity of con-
required thickness of a gravel pad be- crete, K. = 1.0 Btu/ft hr F.
neath the floor section shown in figure - Thermal conductivity ofinsula-
3-7 to contain all thaw penetration. Also tion, Ki = 0.033 Btu/ft hr F.
determine the required stack height to (a) The required thickness is
ensure freeze-back of the pad on the determined by the following equation,
outlet side of the ducts. The conditions derived from the modified Berggren
for this example follow: equation :
- Duct length, l = 220 ft. 48*21 f

- Gravel pad : yd = 125 lb/ft 3 ,


w = 2.51.
X = KRf [
V 1+
KL(R f)2
. 11 (eq 3-29)

Figure 3-7 . Schematic of ducted foundation .

3-14
`TM 5-852-61AFR 88-19, Volume 6
where pad surface temperature at the outlet
K = average thermal conductivity of end equals the ratio
gravel Required freezing index
=1/2(0 .7 + 1.0) = 0 .85 Btu/ft hr F Length of freezing season
RT = thermal resistance of floor 1420
system
18 4 12 215
= 6.6F below 32F or 25 .4F.
12(1 .0) 12(0.033) 12(1.0)
12.5 ft2 hr F/Btu The inletair during the freezing season
(eq 3-30
has an average temperature of
(In the computations the dead airspace Air freezing index 4000
is assumed equivalent to the thermal
resistance of concrete of the same Length of freezing season 215
thickness.) = 18.6F below 32F or 13.4F.
=factor in modified Berggren
Therefore, the average permissible tem-
equation = 0.97 (conservative
assumption) perature rise TR along the duct is
(25.4 - 13.4) = 12.0F.
I f = thawing index at floor surface (c) The heat flowing from the
= (60 - 32)(150) = 4200 degree- floor surface to the ductair during the
days winter is equal to the temperature dif
L=latent heat of gravel = ference between the floor and duct air
144(125)(0 .025 = 450 Btu/ ft3 divided by the thermal resistance be-
then (48)(0 .97) 2'(4200) * tween them. The thermal resistance R
X - (0.85(12 .5) [ _ 1 + .11 is calculated as follows:
(0.85)(450)(12 .5)2 Xe Xi 1 14
R
- 1.1.0 ft. (eq 3-31) Ke Ki hre (12)(1 .0)
(b) Thus the total amount of heat + 4. +1
to be removed from the gravel pad by (12)(0 .033) 1.0 = 12.3 hr ft2 OF/Btu (eq 3-33)
cold-air ventilation during the freezing where
season with ducts open is equal to the XC = thickness of concrete (ft)
latent and sensible heat contained in Xi = thickness of insulation (ft)
the thawed pad. The heat content per hre =surface transfer coefficient
square foot of pad is determined as between duct wall and duct
follows: air
- Latent heat, (X)(L) = (11.0)(450) (For practical design, hrC = 1.0 Btu/ft2
= 4950 Btu/ft2 hr F and represents the combined
- Sensible heat (10 percent of effect of convection and radiation . At
latent heat, based upon ex- much higher air velocities, this value
perience) = 495 will be slightly larger; however, using
- Total heat content: a value of 1.0 will lead to conservative
5445 Btu/ft2. designs). The average heat flow be-
The ducts will be open during the tween the floor and inlet duct air is
freezing season (215 days),and the aver- [(60 - 13 .4)/12 .3] = 3.8 Btu/ft2 hr, and be- -
age rate of heat flow from the gravel twe.en the floor and outlet duct air is
during this season is equal to 5445/215 [(60 - 25.4)/12 .3] = 2.8 Btu/ft2 hr. Thus
X 24 = 1.0 Btu/ft2 hr. The average thaw- . the average rate of heat flow from the
ing index at the surface of the pad is gravel pad to the duct air is 1.0 Btu/ft2
hr. The total heat flow 0 to the duct air
L X2 (450)(11 .0) from the floor and gravel pad is
= 1420 degree-days.
4812 K = 48(0.97) 2(0.85) (eq 3-32) (3 .3 + 1.0) = 4.3 Btu/ft2 hr. The heat flow
This thawing index must be com- to the duct air must equal the heat
removed by the duct air-
pensated for by an equal freezing index
at the duct outlet on the surface of the Heat added - heat removed
pad to assure freeze-back. The average O fm . 60V Ad p e pTR. (eq 3-34)

3-15
*TM 5-852-6/AFR 88-19, Volume 6

Thus the average duct air velocity Substitution of appropriate values


required to extract this quantity of heat gives a required air velocity
(4.3 Btu/ft2 hr) is determined by the (4.3)(220)(2.68)
equation : V=
(60)(1.68)(0.083)(0 .24)(12 .0)
V = ft/minute (eq3-36) = 111 ft/minute. (eq 3-36)
60 A d p C pTR (d) The required air flow is ob
where tained by a stack or chimney effect,
total heat flow to duct air (4.3 which is related to the stack height.
Btu/ft2 hr) The stack height is determined by the
.Q = length of duct (220 ft) equation
m = duct spacing (2.66 ft) li d = h V + h r (eq 3 .37)
Ad = cross-sectional area of duct where
(1.68 ft2) p e H(TC . To)
= density of air (0.083 lb/ft3 hd 5.2(T e + 460) Inches of water
[figure 3-10] ) (natural draft head)
cp = specific heat of air at constant p - density of air at average duct
pressure (0.24 Btu/lb F) temperature (lb/ft3)
TR =temperature rise in duct air e = efficiency of stack system
(12F). This factor provides for

Figure 3-8 . Properties of dry air at atmospheric pressure.

3-16
'TM 5-852-6/AFR 88-19, Volume 6

friction losses within the factor is small, as noted by examining


chimney the-equation below used to calculate
H = stack height (ft) the friction factor f. Reynolds number
Te = temperature of air in stack is obtained from the equation
(F) V(a' + 0 .25 De)
To = temperature of air NR (eq 3-40)
surrounding stack (F)
V 2 where
_( (111 X 80)(1.0 + 0.25 X 1.221
)2 inches of water NR - -17,700
4000 (velocity head) 0 .49
V = velocity of duct air (ft/minute) V - average duct velocity (ft/hr)
a - shortest dimension (ft)
Is V - kinematic viscosity
hf = f- by inches of water
De (friction head) (ft2/hr at 19.4F [fig. 3.81).
The friction factor f' is obtained by
= friction factor solving the equation
(dimensionless)
= equivalent duct length (ft) f= 0.0066[1 + (20,000 X -+ -
le De NR
De = equivalent duct diameter (ft). 0.0 01 106
-0.005511+(20,000x -+-)1/31
The technique used to calculate the 1.22 17.700
friction head is -0.0285. (eq 3-41)
4(cross-sectional area Therefore, the friction head is
of duct in ft2)
Dc =
4(1.58)
- hf - f X De X by (eq 3-42)
perimeter of duct in ft
2 ( 18+20 + 12) 423
12 2 - 0.0285 X
1.22 by - 9.8hv.
= 1.22 ft. X
(eq 3-38)
The equivalent length of the duct is The draft head required to provide the
equal to the actual length 1 s plus an desired velocity head and to overcome
allowance l b for bends and entry and the friction head is furnished by the
exit. Each right-angle bend has the chimney or stack effect. The draft head
effect of adding approximately 65 hd is obtained as follows:
diameters to the length of the duct, and h d - by + hf - by + 9.8 by (eq 3-43)
entry and exit effects add about 10 - 10.8 by
diameters for each entry or exit. In this v
example the total allowance 1b for these a 10.8 (-)2
effects is [2(65 + 10) =] 150 diameters, 4000
which is added to the length of the 111
straight duct . The estimated length of a 10.8 (-)2 = 8.31X10.3 inches of wat
straight duct I s is
The stack height required to produce
5 ft (assumed inlet open length) this draft head is
220 ft (length of duct beneath 5.2 h d(Tc + 460)
floor) H
15 ft (assumed stack height) pE(Tc - To)
(5.2)(8.31x10-3)(25.4 + 460)
240 ft
fe = ks + fb
(eq 3-39) (0.083)(0.80)(25.4 - 13.4)
/ e m 240 + (150 X 1.22) - 423 ft. - 26 ft (eq 3-44)
The friction factor f' s a function of where
Reynolds number NR and the ratio P = 0.08315/ft3
e/De. Areasonable absolute roughness To - 25.4F
factor e of the concrete duct surface is
0.001 feet, based on field observations .
To =13.4F
Suggested values of e for other types e - 8095 (found to be a reasonable
of surfaces are given in the ASHRAB designvalue based on observa-
Data and Guide Book . The effect of tions over an entire season)
minor variations in e on the friction hd = 8.31X103 inches of water.
347
'TM 5-852-6/AFR 88-19, Volume 6
(e) If the stack is too high for the TM 5-852-4/AFM 88-19, Chap. 4) . Limited
structure, a greater thickness of insula- field tests indicate that the heat-flow
tion could be used. In this example, the resistance of a Portland-cement-
effectof increasing the insulation thick- concrete pavement overlying a high-
ness by one-half would result in lower- quality insulating layer is more compli-
ing the stack height by five-eighths. cated than simple addition of resist-
(f) This firstapproximated stack ances, but until sufficient data are
height is next incorporated in the cal- obtained for validation, treatment of
culation of the length of straight duct resistances in series is recommended.
.25, and the newly obtained f e is used to a. Example. A pavement consists of
recalculate the friction head hf. By trial- 14 inches of Portland-cement concrete
and-error, the final calculated stack placed on a 6-foot gravel base course .
height is found to be 26.5 ft. Frost penetrated 3.2 feet into the under-
(g) The stack height is an im- lyingsiltsubgrade. Determine the thick-
portant variable because an increase ness of insulation required to prevent
in stack height will increase the duct frost penetration into the subgrade for
airflow. Circulation of air through the the following conditions .
ducts results from 1) a density dif- - Mean annual temperature =
ference between the air inside the duct 35.3F .
and that outside the building, 2) a pres- - Air freezing index =
sure reduction at the outlet end attrib- 3670 degree-days
utable to the stack effect, 3) a positive - Freezing season = 170 days
pressure head at the inlet end when - Concrete: K = 1.0 Btu/ft hr F,
wind blows directly into the intake C = 0.30 Btu/ft3 F
stack opening, and 4) a negative pres- - Insulation: K = 0.024 Btu/ft hr
sure head at the outlet when wind F, C = 0.28 Btu/ft3 F
passes over the exhauststack opening. - Gravel base : yd = 130 lb/ft 3,
Draft caused by wind is highly erratic w=4%
and unpredictable and should not be - Silt subgrade : yd= 100lb/ft3,
considered in design; however, the w = 10%.
vents should be cowled to take ad- Surface freezingindex = 0.75 X 3670 =
vantage of any available velocity head 2752 degree-days. From the known
provided by the wind. If sufficient air data
cannot be drawn through ducts by 2752
natural draft, mechanical blowers vs = 16.2'F (eq 3-45)
could be specified or consideration 170
given to alternatingairflow in the ducts. vo 35.3 - 32 .0 = 3.3'F (eq 3-46)

3-9. Use of thermal insulating materials. 3.3


a = =0.20 (eq 3-47)
An insulating layer may be used in 16.2
conjunction with a non-frost-suscep- b. Trial 1. Use a2-inch layer of insula-
tible material to reduce the thickness tion and a 6-inch concrete leveling
of fill required to keep freezing or course.
thawing temperatures from pene-
-Pavement section : 14 inches of concrete
tratinginto anunderlying frost-suscep-
tible soil. As in the example of para- 2 Inches of Insulation
graph 3-8a, the thermal resistance of 6 inches of concrete
leveling course
the pavementand insulation layers are
added to obtain total resistance, and
the latent heat effect of a combined 22 inches total
pavement and insulation layer is as- d 14 2 6
R =-= + +
sumed negligible . (TM 5-818-2/AFM 88- K 12x1.0 12x0.024 12x1.0
6, Chap. 4 discusses in detail the design 8.60 hr ft2, F/Btu (eq 3.48)
of insulated pavements.) If the insu- (14X30) + (2X0.28) + (6X30)
lating material will absorb water, its Cp
insulating effectiveness will be re- 22
duced considerably (as discussed in 27.3 Btu/ft3 - F (eq 3-49)

3-18
'TM 5-852-6/AFR 88-19, Volume 6 '

The calculation appears in table 3-3 d


-=-+
14 1.5 6
and indicates that this pavement sec- RP =
K 12X1 .0 12x0 .024
+
12 x 1.0
tion has an excess of (2752 - 3033 =) 481 6.86 hr ft2 F/Btu (eq 3-50)
degree-days to prevent frost - pene-
tration into the silt subgrade . Cp =
(14X30) + (1 .5X0 .28) + (6X30)
c. Trial 2. Use a 1.5-inch layer of 21 .5
insulation and a6-inch concrete level- = 27.9 Btu/ft3 F (eq 3-51)
ing course. The calculation (see table 3-4) indicates
-Pavement section: 14 inches of concrete that this pavement section will not
1.5 inches of insulation prevent frost penetration into the silt
6 inches of concrete subgrade as (2752 - 2553 =) 199 degree-
leveling course days remain for subgrade penetration.
The 2-inch thickness of insulation is
21.5 inches total therefore required .
'TM 5-852-6/AFR 88-19, Volume 6

Table 3-3. Insulated pavement design, no frost penetration. Table 3-4. Insulated pavement design, frost penetration .

3-20
*TM 5-852-6/AFR 88-19, Volume 6

CHAPTER 4

TWO-DIMENSIONAL RADIAL HEAT FLOW

4-1 . General
= . r3 radius to outer edge of insula-
Radial flow of heat is considered in tion
thermal =problems involving the design r4 radius to outer edge of gravel
of pile foundations in permafrost (TM K1-2
5-852-4/AFM 88-19, Chapter 4), the con- thermalconductivity ofconcrete
struction of utility supply lines for the K2-3 = thermal conductivity of insula-
transport of waterand sewage in perma- tion
frost areas and seasonal frost areas K3-4 =thermal conductivity of gravel.
(TM 5-852-5/AFR 88-19, Volume 5), and 1 r2 1 r3
the design ofartificially frozen ground R-0.367( - log + - log (eq 4-1)
'1 -2 r K2-3 r2
for retaining structures during con- 1 r4
struction. A number of the basic con- + - log -
cepts and techniques used to calculate K3-4 r3
radial heat flow from cylindrical sur- 1 5.5 1 5.83 1 9.83
faces are discussed below. =0 .367 ( - log - + - log _+_109 -+)
1.0 5.0 0.033 5.5 1.5 5.83
a. ThermalResistance. In analyzing
heat flow for areas with cylindrical -0.367 (0.041 + 0.767 + 0.151) -0.352 ft.2 hr F/Btu.
cross sections, the effective thickness
for radial flow from a unit length of the If the temperature in the conduit were
cylinder is 45F and the temperature at the outer,
1 r2 r2 face of the gravel were 35F, the heat
In -,I or0.367log r flow per linear foot of conduit would
27r 1 1
equal
where
rI = inside wall radius (ft) (45 - 35) - 28.4 Btu/hr. (eq 4-2)
r2= outside wall radius (ft) . 0.352
The thermal resistance R is equal to b. Temperature field surrounding
the effective thickness divided by the a cylinder. The sudden or step change
conductivity of the material between in surface temperature discussed for
the two radii. As an example, a concrete semi-infinite slabs in paragraph 3-6a
conduit with a wall thickness of 6 has application to heat-flow problems
inches and an inside diameter of 10 feet associated with pile foudations in
is surrounded by 4 inches of cellular permafrost. A mathematical solutiop
glass insulation and 4 feetofdry gravel. is available for the problem where the .
Calculate the thermal resistance be- surface temperature of a cylinder is
tween the inside concrete wall and the suddenly changed from the uniform
outer edge of the gravel material . The temperature of the surrounding
following thermal conductivities are medium, as long as there is no phase
given: change. Figure 4-2 is used to determine
- Concrete, K =1 .00 Btu/ft hr F. the temperature T at a distance r from
- Insulation, K = 0.033 Btu/ft hr the center of acylinder of radius rh at a
F. time t after the surface temperature of
- Gravel, K = 1.5 Btu/ft hr F. the cylinder is changed from To to T8.
The temperatureTorepresents the uni-
Let (see fig. 4-1 for values of rI - r4) form temperature of the medium prior
=
rI radius to inner wall of conduit to the sudden change in surface
=
r2 radius to outer wall of concrete temperature.
'TM 5-852-6/AFR 88-19, Volume 6

Figure 4-1 . Illustration for example in paragraph 4-1a .

4-2 . Pile installation in permafrost . tions, artificial refrigeration may be


At many arctic and subarctic sites, pile required to ensure freeze-back within
foundations are commonly placed in a reasonable time . The volumetric
preaugered holes, and the annular specific heat of the slurry and the effect
space between the oversized hole and of vertical heat flow are assumed to
pile is backfilled with a slurry of soil have a negligible effect in computing
and water. The tangential adfreeze required freeze-back time . With proper
strength of the pile is principally a pile spacing, the slurry temperature
function of the bond developed be- reaches thatof the surrounding perma-
tween the pile and the frozen slurry . frost in time. Surrounding tempera-
Dissipation of the sensible and latent tures, natural freeze-back time, proper
heat of the slurry into permafrost is a pile spacing, and heatremoval by refrig-
major design factor because construc- eration, as well as heat transfer by ther-
tion scheduling depends upon the time mal piles, are discussed below.
required for slurry freeze-back . Pile a. Surrounding temperatures. The
spacing is important as each pile adds increase in permafrost temperatures
heat, i.e., piles spaced too closely may during slurry freeze-back is deter
increase permafrost temperatures, mined using the technique described
with a reduction of pile adfreeze in paragraph 4-1b. . For example, a
strength and an increase in freeze- preaugered hole for a pile installation
back time. This is particularly true in is 16 inches in diameter and the slurry
relatively warm permafrost (above is placed at a temperature of 32F. The
30F) . The following discussion as- surrounding permafrost has a thermal
sumes first that the slurry will freeze diffusivity of 0.06 ft2/hr and an initial
back naturally because of heattransfer temperature of 28F . Calculate the
between the surrounding permafrost ground temperature at a distance of 3
and the slurry, second that the time feet from the center of the pile after an
required for freeze-back at a particular elapsed time of 48 hours . Given
depth is predominantly influenced by rl = 0.667 ft t = 48 hr
the permafrost temperature at that r = 3.00 ft To = 28F
depth, and third that the permafrost
does not thaw. Under certain condi- a = 0.06 ft2/hr TS = 32F
4-2
"TM 5-852-6/AFR 88-19, Volume 6

Figure 4-2. Temperature around a cylinder having received a step change in temperature.

therefore, From figure 4-2,


0.667
r1 T - To
~ - = 0.39 (eq 4 .3) = -..- (eq 4-6)
V (0.06)(48) TS' TO

T = 0 .18 (T S - To)+T o = 0 .18 (32-28)+28


r 3 .00
- _ - = 4 .60 . (eq 4-4)
rl 0.667 = 28.7F . (eq 4-6)

4-3
'TM 5-852- 6/AFR 88-19, Volume 6

This technique is also used to predict a semi-infinite medium, with asuddenly


the increase in permafrost temperature applied constant temperature source
during slurry freeze-back. Since it as- (32F) that dissipates radially into
sumes a constant surface temperature frozen ground ofknown initial tempera-
for a cylinder, it is applicable only to ture . The general solution for the
the time offreeze-back . After the slurry natural freeze-back problem, based
has frozen, the permafrost temperature upon the latent heat content of the
decreases and the model is not valid. slurry, is shown in figure 4-3.
b. Natural freeze-back time. This (1) For example, calculate the time
heat transfer problem assumes a slur- required to freeze back a 12 .5-inch
ried pile to be a finite heat source inside diameter timber pile placed in an 18-

Figure 4-3. General solution of slurry freeze-back.

4-4
'TM 5-852-6/AFR 88-19, Volume 6

inch hole preaugered in permafrost be increased by 50 percent to permit


and backfilled with a slurry for the an element of safety in the design .
following conditions : (Note: the sensible heat introduced by
- Permafrost: Silty sand the pile and slurry was negligible in
Initial temperature - 27F comparison to the latent heat intro-
Dry unit weight = 941b/ft3 duced by the slurry and was not con-
Water content - 259 . sidered in calculations .)
- Slurry backfill: Silt, water (2) Permafrost temperature varia-
Placement tempera- tions with depth, as discussed in TM
ture - 33.5F 5-852-4/AFM 88-19, Chap. 4, should be
Dry unit weight - 72 ib/ft3 considered in calculating freeze-back
Water content - 45%. time . Figure 4-4 illustrates the effect
In figure 2-3, the thermal conductivity of permafrost temperature on freeze-
of the permafrost is determined to be back for the above example. Since heat
1.1 Btu/ft hrF. The volumetric heatcapa- input is governed principally by the
city is calculated to be latentheat of slurry backfill, which is a
function of slurry volume, moisture
25
[94 (0.17 + 0.5 100 )] - 27.7 Btu/ft3 17 (eq 4-7) content and dry unit weight, a family
of curves relating volumetric latent
and the thermal diffusivity to be heat of slurry, permafrost tempera-
tures and freeze-back time may be de-
1.1 veloped for a specific site to account
- 0.0397 ft2/hr. (eq 4-8)
27.7
The volumetric latentheatof theslurry
is
(144 X 72 X 0.45) - 4970 Btu/ft3 17 (eq 4-9)
and the latent heat per linear foot of
backfill is
n
[
4 (1 .52 .1 .042)] 4970 - 4280 Btu. (eq 4.10)

When figure 4-3 is entered with avalue


of

1 .55 (eq 4.11)


Cr2At (27.7)(0.752)(5)
then,
at
- .12.4. (eq 4-12)
r22
The time required to freeze back the
slurry backfill is
12 .4 X 0.752
- 179 hours or about
0.0397 7.3 days. (eq 4-13)
At this time the slurry temperature is
32F. Subsequent to freeze-back, the
temperature of the slurry will continue
to decrease and will approach the
permafrost temperature . Ninety per-
cent of the temperature difference ; will
disappear in about twice the time re-
quired for freeze-back to 32F. In this
example, after a period of [7.3 + (2
X 7.3) =122 days, the slurrytemperature
would be approximately [32 - 0.90 (32 -
27) =1 27.5F. The time (22 days) should Figure 4-4. Specific solution of slurry freeze-back

4-5
*TM 5-852-6/AFR 88-19, Volume 6

for varying pile shapes and preaugered permafrost temperature rise for the
hole diameters. To minimize the heat volumetric latent heat of the slurry
introduced by the slurry, the water backfill introduced into the drill hole .
content should be the minimum re- A family of curves may be developed to
quired for complete saturation. This account for variation of slurry volumet-
can be bestaccomplished by backfilling ric latent heat.
with the highest dry unit weight mate- (2) In this example the slurry back-
rial that can be processed and placed, fill was placed ata temperature slightly
i.e., a well-graded concrete sand with a above freezing (33 .5F) and, theoreti
6-inch slump. cally, the sensible heat of the slurry
c. Pile Spacing. The effect of pile should be considered . The volumetric
spacing on the overall rise of perma- capacity of the unfrozen slurry was
frost temperature resulting from instal [72(0.17 + 1.0 X 0 .45) _] 44.6 Btu/ft3 F,
lation of piles in preaugered holes is and with a temperature difference of
found by equating the latent heat of (33.5 - 32 =) 1 .5F, this represents a
slurry backfill with the allowable sen- sensible heat of (1.5 X 44.6 =) 67 Btu/ft 3 .
sible heat (temperature) rise of the sur- A comparison of this quantity with the
rounding permafrost . For example, volumetric latent heat of the slurry
calculate the minimum allowable pile (4670 Btu/ft 3) shows that its heat may
spacing in the preceding example so be considered negligible, as long as it
that the permafrost temperature will near the freezing point.
not rise above 31F.The following equa- d. Artificialfreeze-back time. Ifper-
tion, equating the latent heat of the mafrost is temperatures are marginal,
slurry to the change of sensible heat in it may be necessary to refrigerate the
a permafrost prism of side S, is used to pile to accelerate slurry freeze-back
determine the pile spacing: time and to have refrigeration available
Q if permafrost temperatures rise after
s = (nr2) + C AT (eq 4-14) construction. The following example
shows calculations required to deter-
where mine the amount of heat to be extracted
S = grid pile spacing (ft) from the ground . The average volume
r2 = radius of augered hole (ft) of slurry backfill for a group of piles is
Q = latent heat of slurry per lineal foot 31 cubic feet each. The slurry is placed
(Btu/ft) at a n average temperature of 48F and
C = volumetric heat capacity ofperma- must be frozen to 23F. A silt-water
frost (Btu/ft3 F) slurry of SO lb/ft 3 dry weight and 40
percent water content is used as back-
AT= temperature rise of permafrost fill material, and an available refrigera-
(F). tion unit is capable of removing225,000
Substitution of appropriate values from Btu/hr. Calculate the length of time
the above exampleand a maximum allow- required to freeze back a cluster of 20
able permafrost temperature rise AT piles .
of 4F give a minimum pile spacing S of -Volumetric latent heat of back-
4280 fill:
(3.14)(0.75) 2 + = 6.4 ft. (eq 4-15) L = (1-14 X 80 X 0 .40) = 4600 Btu/ft3. (eq 4-16)
(27.7)(4)
This spacing may not keep local tem- -Volumetric heat capacity of
perature from rising to more than 31F; frozen backfill :
Cf = 80 (0 .17 + (0 .5X0 .4)) = 29 .6 Btu/ft 3 F .
however, itwill keep the entire mass of (eq 4-17)
permafrost from reaching that tempera- -Volumetric heat capacity of un-
ture. frozen backfill :
(1) Numerical analysis ofa number Cu = 80 [0.17 +(1 .0X0 .4)) = 46 .6 Btu/ft 3 F .
of pile installations indicates that pile (eq 4-18)
spacing should be at least five diam -Heatrequired to lower the slurry
eters of the drill hole size. A plot, similar temperature to the freezing point-
46 .6 X 31 (48 - 32) = 22,618 Btu/pile . (eq 4-19)
to that shown in figure 4-4, may be -Heat required to freeze slurry:
prepared to relate pile spacing and 31 X 4600 = 142,600 Btu/pile . (eq 4-20)

4-6
*TM 5-852-6/AFR 88-19, Volume 6

-Heat required to lower the niques are presented in TM 5-852-


slurry temperature from the freezing 5/AFR 88-19, Volume 5.
point to 23F; b. Freezing of standing water in a
29.6 X 31 (32 - 23) = 8258 Btu/pile . (eq 4-21) buried pipe. Problems with freezeup
-Total heat to be removed from of stationary water must take into
the slurry: account the initial time required to
20 (22,600 + 142,700 + 8200) = 3,470,000 Btu.
(eq 4-22) lower the water temperature to the
-Time required for artificial freezing point and the amount of time
freeze-back, excluding allowances for required to form an annulus of ice in
system losses: the pipe. In most instances the danger
3,470,000/225,000 - 15.5 hours. (eq 4-23) point is reached when water begins to
e. Meat transfer by thermal piles. freeze .
Artificial freeze-back may be accom- (1) The time required to lower the
plished also by use of two types ofself- temperature of nonflowing water in
refrigerated heat exchangers : a single- an insulated pipe to the freezing point
phase liquid-convection heat transfer is given by the equation
device and a two-phase boiling-liquid 31 .2 ri TN,-Ts
and vapor convection heat transfer
rp 32-T
t= In )In- s (eq 4-24)
(r'
device . TM 5-852-4/AFM 88-19, Chapter K,

4 presents heat transfer rates for the where


two-phase system . There are few heat t = time (hr)
transfer field data available for the Ki = thermal conductivity of in-
single-phase system. sulation (Btu/ft hr F)
rp = radius of pipe (ft)
4-3. Utility distribution systems in frozen
ground. r i = radius to outer edge ofinsula-
General considerations for the design tion (ft)
of utility systems in cold regions are T N, = initial water temperature (F)
given in TM 5-852-5/AFR 88-19, Volume T s = temperature of surrounding
5. soil (F).
a. Burying water pipes in frozen For example, a 12-inch diameter iron
ground. Water pipes that are buried in pipe containing water at42F is buried
frozen ground may be kept from freez in 28F soil. Determine the time re-
ing by any one of the following quired to lower the water temperature
methods : 1) placing the waterline in an to 32F if the pipe is insulated with 3
insulated utilidor, which is a contin- Inches of cellular glass (K i = 0.033
uous closed conduit with all lines, such Btu/ft hrF) .
as water, sewage and steamlines, in- 31.2 - _ 0.75 42-28
stalled away from direct contact with t- - (0.502 In - ) in - - 120 hours (5 days)
0.033 0 .50 32-28
frozen ground, 2) providing a suf- eq 4-25

ficient flow velocity such that the water (2) Once the water temperature
temperature at the terminus ofthe pipe- has been lowered to the freezing point,
line does not reach the freezing point ice begins to form in an annular ring
or 3) heating the water at the intake or inside the pipe. The following assump-
at intermediate stations along the line . tions are made to solve this problem:
A layer of insulation around a pipeline -The water is initially at 32F .
will retard, but not prevent, freezing of -The heatreleased bythe freezing
standing water in a pipe. The thermal of water does not affect the sur-
analysis of a pipeline buried in frozen rounding ground temperatures .
ground is complicated by the changing -The volumetric heat capacity of
thermal properties, ice content, sea- the ice may be ignored .
sonal and diurnal changes of tempera- -The thermal resistance of the
ture and the intermittent water flow. pipe wall is negligible.
Some calculation techniques applic- The solution predicts the time required
able to the problem of buried utilities to form an annulus of ice around the
are presented below for standing and inner wall of the pipe. Knowledge of
for flowing water. Additional tech- pipe radius, insulation thickness and
4-7
'TM 5-852- 6/AFR 88-19, Volume 6

and thermal properties, thermal con- inches and the time required to com-
ductivity of ice, latent heat of fusion of pletely freeze the water. Assume the
water, and surrounding ground tem- rate of flow does not influence freezing.
peratures are necessary to solve this The time required to reduce the bore
problem. The temperatuze of ground to 6 inches will be
surrounding the pipe and the time 9000 (0.5)2 1.33 0.75
during which the ground remains t= r
.50 + 0..~)
below freezing is difficult to estimate. 2(1.33)(32-38) [( 0.033 1n 0
The relationship between time and the 0.252 0.25 2 0.50
1_
radius of ice formed inside an insulated 6.502)- ( 0.50 ) in0.25
pipe is given by the expression - 2640 hours. (eq 4-30)

T [( rp )In P-].
KL r2 r ri 2 2 r
t= -1n-+1/2)(1--
r2 )-( The time required to completely freeze
2K0 i p p the water in the pipe will be
(eq 4-26)
where 9000 (0.5)2 1 .33 0.75
t = time (hr) t
2(1.33)(32-38) [( 0.033 1I~ 0.50 +
0.5)
L = latent heat of water (9000 O O 0 .50
Btu/ft3) (1-0
.502)-( 0.50 )21 ol
rp = radius of pipe (ft) - 3550 hours. (eq 4-31)
K = thermal conductivity of ice
(1.33 Btu/ft hr F) The calculation is simplified since the
term"r" for the inner radius of the pipe
AT = temperature difference be- goes to zero. For an uninsulated pipe,
tween water and surround- the calculations assume the simplified
ingsoil (F, assume water tem- form of
perature is 32F)
K i = . thermal conductivity of in- (1690)(0 .5)2
sulation (Btu/ft hr F) (32 28) ~Y~
r i =radius to outer edgeof insula-
tion (ft) where r/r p = O and y = 1 .0 (fig. 4-5) .
r =inner radius of ice annulus Therefore,
(ft). t = 106 hours.
If the pipe is not protected by insula- This example illustrates the effective-
tion, the equation is ness of insulation in retarding the
freezeup of water in pipes, but as stated
L r2 r2 r r
t [1/2(1- ) -( )21n P 1. (eq 4-27) above, the assumptions used to develop
= 2 KIT 2
rp rp these equations are conservative and
the actual length offreezing time would
This expression for an insulated pipe be greater .
may be simplified by rearrangement e. Thawing offrozen soil around a
and substitution of numerical values suddenly warmed pipe. In the pre-
for the latent heat and the thermal con- ceding example it was assumed that
ductivity of ice. This yields water initially at 32F was placed in
r frozen ground and the time relation-
t = 1690 QT (eq 4-28) ship for freezing of the water in the
jyj
pipe was determined . If the water was
where maintained above freezing, the frozen
r2 r2
Y = [1- (1-1n A- (eq 4-29) soil surrounding the ppe would thaw .
x2
p p To formulate a mathematical expres-
The relationship between y and r/rp is sion relating the time with the radius
given in figure 4-5. of thaw, it is assumed that: 1) the volu-
metric heat capacity of the soil is neg-
Following is an example. A 12-inch iron ligible, 2) both the surrounding soil
pipe, insulated with 3 inches of cellular and pipe are initially at 32F, and 3) the
glass (Ki = 0.033 Btu/ft hr F), is placed pipe temperature is suddenly raised to
in 28F soil. Calculate the time required a temperature above 32F. The formula
to reduce the bore of the pipe to 6 for an insulated pipe is
4-8
'TM 5-852-6/AFR 88-19, Volume 6

Figure 4-5. Freezeup of stationary water in an uninsulated pipe .

L r2 KU ri KU = thermal conductivity of un-


2 KU pT V
Ki In
.0 .5) frozen soil (Btu/ft hr F)
rP
r2 r QT = temperature difference be-
[I -
r2 )+'n
1 (e9 4-33) tween pipe and surrounding
i soil (F, assume soil at 32F)
where Ki = thermal conductivity of in-
t = time (hr) sulation (Btu/ft hr F)
L = latent heat of soil (Btu/ft3) ri =radius to outer edge of insula-
r = radius to outer edge of thaw- tion (ft)
ing soil (ft) rP = radius of pipe (ft).
4-9
*TM 5-852-6/AFR 88-19, Volume 6

If the pipe is not protected by insula- preceding equations, the time t can be
tion, the expression is multiplied by the temperature differen-
r L r2 2 tial AT to give either a freezing or
t [-0 .5 (1 -
r2 ) + In . 1. (eq 4-34) thawing index, and the radius of ice
2I{u 4T rp
formation or thawed ground radius
For example, a 7- by 7-foot concrete can be determined by trial-and-error .
utilidor has 9-inch concrete walls with It was shown that a temperature dif-
an outer covering of 6 inches ofinsula- ferential of (32-28) = 4F lasting for 3150
tion (K i = 0.033 Btu/ft hr F). The frozen hours would result in complete freeze-
soil around the utilidor is a sandy up of the water in thepipe; this is equiva-
gravel with a dry density of 115 lb/fO lent to a freezing index of [(4 X 3150)/24
and a water content of 7.8 percent at a =1 524 degree-days . If the freezing
temperature of 32F. Neglect the ther- index at the depth of pipe burial were
mal resistance of the concrete, and less than 524, the standing water in the
determine the time required to thaw 1 pipe would not completely freeze in
foot of soil when the temperature of that time. Thus, the freezing index at a
the utilidor walls is suddenly raised to particular depth can be used to fore-
50F. cast freezeup of stationary water in
Ku =1.2 Btu/ft hr F pipes located in the annual frost zone.
L =1290 Btu/ft3 (2) Even insulated water lines
K i = 0.033 Btu/ft hr F. located in frozen ground usually re-
For calculation, square sections may quire an inlet water temperature sig
be treated as cylinders of the same nificantly above freezing. Whether the
perimeter. water lines are insulated or not, this
may thaw some of the surrounding
Dimension-square Equivalent radius frozen ground. This thawed annulus
Symbol ON (ft) will retard water freezeup in the pipe
if and when flow conditions change.
The situation is practice is generally
r 9.5+24/12=11 .5 7.33
8 .5 + 12/12 = 9.5 complicated by the intermittentcharac-
r; 6 .04 ter of water demand. In some northern
rp 7.0 + 18/12 = 8.5 5.42 communities the problems of irregular
water demand are solved by con-
(1290)(7.33) 2 1.2 6 .04 structing the water lines in a contin-
t
2(1.2)(50-32)
U 0 .033 In
5.42
.0.5) uous loop with provisions for periodic
6 .042 7 .33
flow reversals . Water temperatures
should be closely monitored and water
) + In
7.332 64 usage patterns considered in esti-
= 2060 hours . (eq 4-35) mating water freezeup .
d. Practical considerations. e. Freezing and flowing water in
(1) The above-mentioned formulas buried pipes. Problems involving freez-
indicate the relationship between time, ing of flowing water in buried pipes
radius of freeze or thaw, and tempera require knowledge of the distance the
ture difference between the water in water will flow before the temperature
the pipe and the ground. In sufficient of the water lowers to the freezing
time, standing water in the pipe will point. By providing enough above-
freeze or frozen ground will thaw, freezing water, the loss of heat to the
depending on temperature differen- surrounding frozen soil can be
tials. For practical problems the as- balanced to provide an outlet tempera-
sumed constant temperature differen- ture slightly above freezing. The prob-
tials will not exist for a long time but lems of freezing of flowing water in
will vary with season and even with insulated and bare pipe are illustrated
the hour at shallow depths. The freez- below.
ing and thawing index concept con- (1) Insulated pipe. It is assumed
siders the intensity of temperature dif- that 1) the temperature of the frozen
ferential from freezing (32F) and the ground surrounding the pipe is con
duration of this differential . In the stant for the period of flow over the
" 4-10
*TM 5-852-6/AFR 88-19, Volume 6
entire length of pipe, 2) the effect of eventually become reasonably stable
friction heat developed by water flow with time. An expression relating these
is negligible, 3) the thermal resistance variables is
ands heat capacity of the pipe wall are
T1, T S h 1
negligible, and 4) the temperature dis-
tribution of the water in the pipe is
T2 -
TS
= exp
2rp
X
v X 5 (eq 4-38)
.6X104
uniform at each cross-sectional area. where
The velocity required to prevent freeze- Tl = inlet water temperature (F)
up of flowing water in a pipe is given Ts = frozen soil temperature (F)
by -
s Ki T2 =outletwater temperature (F)
v s = length of pipeline (ft)
ri TI-Ts
112,000 r2 (In - )(In T ) (eq 4-36) 2rp diameter of pipe (ft)
p T2 - S
h =heat transfer coefficient
where (Btu/ft 2 hr F)
V = velocity of flow (ft/s) V = velocity of flow (ft/s).
s = length of pipeline (ft) A nomogram of this equation is shown
= thermal conductivity of in- in figure 4-6.
sulation (Btu/ft hr F)
(b) Limited field experiments in
rn = radius of pipe (ft) clay and sandy clay soils suggest
r i = radius to outer edge ofinsula- values of h for metal pipelines subject
tion (ft) to normal use (conditions or intermit-
Ti = inlet water temperature (F) tent flow) of 6.0 for the initial period of
T2 =outletwater temperature (F) operation an 2 .0 thereafter. These
Ts = temperature of surrounding values are not applicable for pipes
frozen soil (F). smaller than 4-inches in diameter . The
For example, an 11,000-ft long, 6-Inch- h value is dependent upon the thermal
diameter pipe is buried in 10Fr soil. properties of the surrounding soil, the
The pipe is covered with a2-inch layer diameter of the pipe, the type of pipe
of insulation (Ki = 0.03 Btu/ft hr F) and material and the temperature gradient
the inlet water temperature is 39F. in the ground around the pipe's radius.
Calculate the velocity of flow required The value for h given above provides a
to keep the water from freezing. reasonable basis for design ofpipelines
(11,000)(0 .03) in which the total quantity of water
v= consumed per day is at least eight times
0.417 39-10
112,000 (0 .25) 2(ln 0,25 )(In the volume ofpipes in the entire system.
32-10 ) The time of operation required for the
= 0.33 ft/s (20 ft/min). (eq 4-37) temperature distribution in the water
To provide for temporary reductions to stabilize is approximately .
in flow and in recognition of the un- S
certainties concerning the manner of
ice formation within the pipe, it is recom-
t o = 0.005
v (eq 4-39)

mended that the velocity of flow be where


doubled in design. to = time (hr)
(2) Uninsulated pipe. s = length of pipeline (ft)
(a) When flowing water is first V = velocity of flow (ft/s).
introduced into a bare pipe buried in
frozen ground, the heatloss from water (c)For example, water atan inlet
is greater than it is after the system temperature of 40F flows at 2 ft/s in a
has been in operation for a period of 12-inch iron pipeline, 2.2 miles long.
time. The initial heat loss is greater The ground temperature surrounding
because the pipe wall and the soil im- the pipe is 25F . Estimate the outlet
mediately adjacent to the pipe are water temperature during the initial
colder than they are after water has period of flow (h = 6.0) and after sur-
flowed over a time. Soil temperatures rounding temperatures have stabilized
surrounding the pipe increase and (h = 2 .0) .
4-1 1
*TM 5-852-6/AFR 88-19, Volume 6
-Initial period (see fig. 4-6): -Stabilized period:
s 2 .2 X 5280 s
:- =
(eq 1 = 1 .16X .104 4-40) = 1.16X104
2 P 2r
A
V = 2 ft/s. V ='.3 ft/s
h- 2 .0 Btu/ft2 hr F
h = 6 .0 Btu/ft2 hr F
thus thus
T 1 - TS, 40-25 T l - TS .40-25
= T - (eq 4-41) 1.22 =
T2 -TS _11" 2 25 T 2' T S = (eq 4-42)
T2- 25
T 2 - 33 .5F . T 2 - 37 .3F .

Figure 4-6. Temperature drop of flowing water in a pipeline.

4- 12
*TM 5-852-6/AFR 88-19, Volume 6
To operate on the safe side, the outlet lysis and to promote adequate design.
temperature T 2 should remain at or The assumptions involved and tech-
above 35F. The calculated initial water nique limitations have been em-
temperature of 33.5F would be con- phasized .
sidered unsafe and either the flow velo- b. Heat flow beneath heated struc-
city should be increased to approxi- tures is multidimensional because of
mately 3 ft/s or the inletwater tempera- the finite boundaries of such struc
ture should be raised to about 43F. tures.The ground surface temperature
These- precautions would be needed adjacent to the south side of the build-
only for an initial period, i.e., ing is generally higher than that on
s 11 .600 the north side, and the ground surface
to = 0 .005
V
= 0 .005
2 temperature on the west side is gen-
= approximately 29 hours. (eq 4-43)
erally higher than on the east side; in
addition to this influence, the three-
f. Design considerations. The cal-
culation techniques presented above dimensional temperature distribution
beneath the building will be affected
indicate the principal factors to be con-
sidered in design of water distribution by the plan dimensions of the floor .
Three-dimensional solutions are avail-
lines placed in frozen ground . The use able to the problem of heat flow in
ofthese techniques together with recog- homogeneous materials beneath the
nition of the complexities of actual in- surface of a heated finite area sur-
pipe ice formation and sound engi- rounded by an infinite area subject to
neering judgment provides a basis for adissimilar surface temperature condi-
design ofpipelines in areas of seasonal tion; however, the solutions consider
frostand permafrost. Changing the sur- only the effects of temperature change
face cover over an installed pipeline and notthe effects of phase transforma-
will affect the distribution of tempera- tions. Such solutions tend to be rather
tures with depth and may result in complex and unwieldy, and their
depressing the temperatures adjacent neglect of latent heat generally results
to the pipe. This is particularly true ifa in an over-estimation of the depth of
natural vegetative cover is stripped and freeze or thaw . The magnitude of this
replaced by a snow-free pavement .The over-estimation is dependent on the
influence of new construction above quantity of moisture in the frozen or
an existing pipeline may require a thawed soil.
change in operating procedures for c. The example given in paragraph
the system, such as an increase in the 3-8a for calculating the depth of thaw
velocity or flow or additional heating beneath a heated slab-on-grade build
at the inlet. ing considered only one-dimensional
4-4. Discussion of multidimensional heat vertical heat flow and excluded lateral
flow. heat flow from the soil beneath the,
a. The relatively simple analytical building to the surrounding soil mass
techniques discussed in this manual in the winter . The amount of lateral
are not always sufficient for consider heat flow would depend on the building
ing the concurrent thermal effects of dimensions and the wintertime soil tem-
multidimensional temperature change perature gradient . Slab-on-grade,
and soil water phase transformation . heated structures usually prevent
The one-dimensional and radial heat- frost penetration under the center of
flow computation techniques pres- the building and result in a thaw bulb
ented in this manual were based on in the foundation soil that may cause
field observations and the use ofreason- permafrost degradation with time.This
able simplifying assumptions. The tech- type of construction is discussed in
niques are intended to facilitate ana- TM 5-852-4/AFM 88-19, Chapter 4.

4- 1 3
'TM 5-852-6/AFR 88-19, Volume 6

APPENDIX A

Sumbot Description Units


a Thermal diffusivity ft2/day or ft2/hr
a' Shortest dimension ft
A Sinusoidal temperature amplitude F
Ad Duct cross-sectional area ft2
c Specific heat Btu/lb F
Specific heat of air at constant pressure Btu/lb F
C Volumetric heat capacity Btu/ft3 F
Cf Volumetric heat capacity in frozen
condition Btu/ft3 F
Cu Volumetric heat capacity in
unfrozen condition Btu/ft3 F
d Thickness of soil layer ft
De Equivalent duct diameter ft
e Roughness factor ft
erf Error function; dimensionless
erf z = (2/Y77_r)a(ze-Y2dp,
where ert = 1 and. erf (-z)
= -erf z
exp(x) ex dimensionless
f Frequency of sine wave cycles/day
F Air freezing index degree-days
f. Friction factor dimensionless
h Heat transfer coefficient Btu/ft2 hr F
hre Surface conductance for combined
radiation and convection Btu/ft2hr F
hd Draft head inches of water
hf Friction head inches of water
hV Velocity head inches of water
H Stack height ft
I Air thawing index degree-days
if Floor thawing Index,, degree-days
k Coefficient of thermal conductivity Btu/ft2 hr F per in.
K Thermal conductivity Btu/ft hr F
Kf Thermal conductivity in frozen
condition Btu/ft hr F
Ku Thermal conductivity in unfrozen
condition Btu/ft hr F
L Volumetric latent heat of fusion Btu/ft3
m Duct spacing ft
In Natural logarithm dimensionless
f. Duct length ft
'TM 5-852-6/AFR 88-19, Volume 6

symbol Description Units

1b Allowance for bends,, etc ., in duct ft


le Equivalent duct length ft
Is Length of straight duct ft
MAT Mean Annual temperature F
surface index
n "n"-factor = air index dimensionless
nF Surface freezing index degree-days
nI Surface thawing index degree-days
NR Reynolds number dimensionless
P Period of sine wave 366 days
Q Latent heat per linear foot of slurry baekfill Btu/ft
r Radius ft
R Thermal resistance ft2 hr F/Btu
S Pile spacing ft
s Length of pipeline ft
t Time hours or days
T Temperature F
To Initial temperature F
TR Temperature rise in duct air F
TS Surface temperature F
Ts Soil temperature F
Tw Water temperature F
V Velocity of flow ft/s
Vo Initial temperature of the soil with respect
to 32F F
vs Average surface temperature with respect
to 32F F
w Water content percent dry weight
X Depth of freeze or thaw ft
a Thermal ratio - vo/v$ dimensionless
Yd Dry unit weight lb/ft3
s Efficiency percent
Lambda coefficient dimensionless
u Fusion parameter dimensionless
v Kinematic viscosity ft2/hr
Pi 3.14
P Density of air lb/ft3
Heat flow to duct Btu/ft2 per hr
*TM 5-852-6/APR 88-19, Volume 6

APPENDIX B

THERMAL MODELS FOR COMPUTING FREEZE AND THAW DEPTHSI

B-1 . Analytical Solutions. where


a. The Oxford University Press X i = ice thickness (ft)
publication, Conduction of Heat in K i = thermal conductivity of ice
Solids,2 presents solutions to many (Btu/ft hr F)
one-, two- and three-dimensional heat
flow problems. Homogeneous isotropic F =freezing index (degree-days)
materials are used in most solutions, Li =volumetric latent heat of
but some solutions are presented for fusion of ice (Btu/ft3).
layered systems. U.S.G.S. Bulletin 1083- The Stefan equation has been modified
A3 developed a one-dimensional tech- by many individuals and agencies, and
nique for predicting the damping of a many similar equations have been
periodic surface temperature at dif- developed . Some of the equations use
ferent depths in a two- and three- functions or initial conditions slightly
layered soil system. Unidirectional different from those used in the
heat flux was considered. U.S.G.S. Bul- original Stefan model. The most widely
leti n 1052-B4 developed a method for used equation to estimate seasonal
estimating the three-dimensional ther- freeze and thaw depths is the modified
mal regime in ahomogeneous isotropic Berggren equation. Application of this
soil beneath a heated structure . None equation, given below, has been very
of these techniques considers phase widespread in North America .
change of the soil moisture . Neglect of USACRREL Special Report 122 5
the effects of latent heat of fusion of developed a computer program for cal-
the soil moisture normally does not culating freeze and thaw depths in
cause substantial error in predication layered systems using this equation:
of frost depths in soil of low water X = 48K n /L or X = 11 48K n I/L
content. Differences between actual (eq B-2)
and computed thaw depths increase where
rapidly with increasing water content X = depth of freeze or thaw (ft)
because of the increased volumetric K = thermal conductivity of soil
heat capacity and greater latent heat (Btu/ft hr F)
of the wetter soil_. L = volumetric latent heatof fusion
b. Several empriical and semi-empir- (Btu/ft3)
ical equations have been developed. that n = conversion factor from air in-
consider latent heat of fusion of the dex to surface index (dimen-
soil. the Stefan equation was originally sionless)
developed to calculate the thickness of F - air freezing index (degree-
ice on a calm body of water (isothermal
at the freezing temperature, 32F) days)
expressed as: I = air thawing index (degree-
days)
(eq Xi = 48KjF/Li B-1) = coefficient that considers the
effect of temperature changes
within the soil mass. It is a
The documents mentioned in this appendix function of the freezing (or
are sources for additional information and
thawing) index, the mean
are found in the bibliography.
2 Carslaw and Jaeger,1959. annual temperature and the
3 Lachenbruch,1959 . thermal properties of the soils.
4 Lachenbruch,1957 . 5 Aitken and Berg, 1968.
*TM 5-852-6/AFR 88-19, Volume 6

c. An equation very similar to the B-3. Numerical techniques.


Stefan equation is used in the USSR to a. Because of the widespread avail-
calculate the "standard" depth of freez ability of electronic digital computers,
ing for foundation designs. Many other their application to numerical solu
closed-form analytical techniques are tions ofthe continuity equation is com-
also used in the USSR. monplace . Numerical procedures are
B-2 . Graphical and analog methods. approximations to the partial differen-
a. Graphical methods have been tial equation ; however, they are much
used to estimate depths of freeze and more accurate and versatile in solving
thaw. The flow net technique can be complex transient heat flow problems
used to estimate steady-state tempera- than are the analytical techniques.
ture conditions. National Research Many computer programs allow flex-
Council of Canada, Technical Paper. ible definition of boundary and initial
No . 163,g presents a graphical means to conditions for both one- and two-
determine temperature in the ground dimensional problems. General back-
under and around natural and engi- ground information on the finite dif-
neering structures lying directly on ference methods available to solve heat
the ground surface . flow problems are discussed in Inter-
b. Analog techniques are also used national Textbook Company's Heat
to estimate freeze and thaw depths . Transfer Calculations by Finite
Table B-1 shows thermal, fluid and elec Differences. 7 Since rectangular ele-
trical analogies. Electric analog com- ments are normally used, complex
puters are available, cost relatively geometries are difficult to simulate ac-
little and are reasonably simple to use. curately unless small element sizes are
The primary disadvantage of these used.
machines are that reprogramming is; b. Use ofthe finite element technique
normally necessary for each problem is also widespread (see The Finite
and complex geometries are difficult Element Method in Engineering
to simulate adequately . Hydraulic Sciences or The Finite Element Method
analogs are also available. The primary in Structural and Continuum
disadvantages of these computers are Mechanics, 9 both from McGraw-Hill) .
their complex tubing systems, space Elements ofvarious shapes can be used
requirements, and the necessity to with this technique; but the triangular
shape is commonly used for two-
thoroughly clean and reconstruct them.
for each problem. At any instant of dimensional problems. Complex bound-
ary geometries can be more closely
time, however, hydraulic analogs
graphically show the temperature
distribution . 7 Dusinberre,1961.
Zienkiewicz,1967 .
Brown, 1663. 9 Zienkiewicz,1971 .

Table B-1 . Thermal, fluid and electric analogs (U.S. Army . Corps of Engineers) .

Medium

Item Thermal Fluid Electric

A - Variables (1) Heat u Volume S Charge density _p


(2) Heat flux 4 Flow Q Current density -'j
(3) Temperature T Head H Voltage e

B - Principles :
au as ap
Continuity (1) +V-a=o +V . d=o +V .i=o
at at at
Conductivity (2) q = -kVT d= -kVH j= -UVe
Capacitance (3) du=CdT dS = AdH pdV = Cde

B-2
`TM 5-852-6/AFR 88-19, Volume 6
simulated using finite element pro- putational capabilities, and acquisition
cedures. For multidimensional heat costs and restraints. USACRREL has
flow problems, the finite element pro- completed and documented a model
cedure is frequently more efficient, i .e., that may be useful in solving many
it requires less computer time than the analytical problems related to construc-
finite difference technique. _ tion in the Arctic and Subarctic; other
c. Many flexible computer programs models are under development. Con-
exist thatsimulate heatconduction and tact HQ (DAEN-ECE-G) or HQ AFESC
phase change in soils. Each has its for assistance in selecting an appro-
own particular data requirements,com- priate model.
'TM 5-352-6/AFR 88-19, Volume 6

APPENDIX C

REFERENCES

Government Publications
Departments of the Army and the Air Force.
TM 5-818-2/AFM 88-6, Chapter 4 Pavement Design for Seasonal Frost
Conditions
TM 5-852-1/AFR 88-19, Volume 1 Arctic and Subarctic Construction,
General Provisions
TM 5-852-4/AFM 88-19, Chapter 4 Arctic and Subarctic Construction,
Building Foundations
TM 5-852-5/AFR 88-19, Volume 5 Arctic and Subarctic Construction,
Utilities
Nongovernment Publications
American Society of Heating, Refrigeration, andAir-Conditioning Engineers
(ASHRAE)
345 East 47th Street, New York,
N.Y. 10017
ASHRAE Guide and Data Book (1963)
*TM 5-852-6/AFR 88-19, Volume.6

BIBLIOGRAPHY

Aitken, G.W. and R.L. Berg (1968) Digital Heating, Refrigerating and Air-
solution of modified Berggren equa. Conditioning Engineers .
tion'to calculate depths of freeze and Barthelemy, J.L. (1980) Performance of
thaw in multilayered systems. Han- natural convection heat exchange
over, N.H.: U.S. Army Cold Regions system for subgrade cooling ofperma-
Research and Engineering Labora- frost . CRREL Special Report 80-40.
tory, Special Report 122. Hanover, N .H.: U.S. Army Cold Re-
Albert, M .R. (1983) Computer models gions Research and Engineering
for two-dimensional transient heat Laboratory .
conduction . CRREL Report 83-12. Baumeister, T. (ed .) (1968) Mechanical
Hanover, N.H.: U.S. Army Cold Re- Engineers Handbook, 6th Ed. New
gions Research and Engineering York: McGraw-Hill.
Laboratory. Berg, R.L. (1976) Thermoinsulating
Aldrich, H.P. and H .M. Paynter (1963) mediawithin embankments on peren-
Frostinvestigations, fiscal year 1963, nially frozen soil. CRREL Special
first interim report, analytical studies Report 76-3. Hanover, N.H.: U.S. Army
of freezing and thawing of soils . Cold Regions Research and Engi-
ACFELTechnicalReport42 .Hanover, neering Laboratory .
N .H.: U .S. Army Cold Regions Re- Brown, W.G. (1963) Graphical determina-
search and Engineering Laboratory . tion of temperature under heated or
Alter, A.J. (1969) Water supply in cold cooled areas on the ground surface.
regions . CRREL Monograph III-C5a . Ottawa: Division ofBuildingResearch,
Hanover, N.H.: U .S . Army Cold Re- National Research Council ofCanada,
gions Research and Engineering Technical Paper No. 163.
Laboratory .
Alter, A.J. (1969) Sewerage and sewage Carey, K.L. (1982) The freezing and
disposal disposal in cold regions. blocking of water pipes. Cold Regions
Technical Digest, 82-1. Hanover, N.H .-.
CRREL Monograph III-C6b. U.S. Army U.S. Army Cold Regions Research and
Cold Regions Research and Engi- Engineering Laboratory .
neering Laboratory.
Arctic Construction and Frost Effects Carslaw, H.S . and J.C. Jaeger (1969)
Laboratory (1966) Design and opera- Conduction of Heat in Solids, 2nd
tion ofan hydraulic analog computer edition. London: Oxford University
for studies of freezing and thawing Press, Inc.
of soils. ACFEL Technical Report 62. Corruccini, R.J . and J.J. Gnlewek (1960)
Hanover, N.H.: U.S. Army Cold Re- Specific heats and enthalpies of tech-
gions Research and Engineering nical solids at low temperatures .
Laboratory . National Bureau of Standards Mono-
Arctic Construction and Frost Effects graph 21. Washington, D.C.
Laboratory (1968) Frost investiga- Crory, F.E. (1963) Pile foundations in
tions, 1962-1963. Cold room studies, permafrost. In Permafrost: Pro-
third interim report of investigations. ceedings, International Conference,
ACFEL Technical Report 43/1. Han- Purdue University. NAS-NRCPublica-
over, N.H.: U.S. Army Cold Regions tion No. 1287. Washington, D.C., pp.
Research and Engineering Labora- 467-472 .
tory. Dusinberre, G.M. (1961) Heat Transfer
ASHRAE (1963) Air Conditioning Re- Calculations by Finite Differences.
frigeration DataBook, Design Volume. Pennsylvania: International Text
New York: American Society of Book Company.
Biblio- 1
*TM 5-852-6/AFR 88-19, Volume 6
Farouki, O.T. (1981) Thermal properties ments, Cold Regions Specialty Con-
of soils. CRREL Monograph 81-1. Han- ference, Anchorage, Alaska . New
over, N .H .: U.S. Army Cold Regions York: ASCE.
Research and Engineering Labora- Kersten, M.S. (1949) Laboratory re-
tory. search for the determination of the
Gilpin, R.R. (1977) The effects of den- thermal properties of soils. ACFEL
dritic ice formation in water pipes. Technical Report 23. Hanover, N.H.:
International Journal of Heat and U.S. Army Cold Regions Research and
Mass Transfer, vol. 20, pp. 693. Engineering Laboratory.
Gilpin, RA. (1977) A study ofpipe freez- Lachenbruch, A.H. (1957) Three-dimen-
ing mechanisms. In Utilities Delivery sional heat conduction in Permafrost
in Arctic Regions. Ottawa : Environ- beneath heated buildings . U.S. Geo-
mental Protection Service, Environ- logical Survey Bulletin 1062-B.
"ment Canada, Report No. EPS 3-WP- Lachenbruch, A.H. (1959) Periodic heat
77-1, pp. 207-220 . flowin stratified medium with applica-
Gilpin, R.R. (1978) A study of factors tion topermafrost problems. U.S. Geo-
affecting the ice nucleation tempera- logical Survey Bulletin 1083-A.
ture in a domestic water supply. The Luikov, A.V . (1966)Heat and Mass
CanadianJournal of Chemical Bngi- Transferin Capillary-PorousBodies.
neering, vol. 56, pp.466-471 . (Translated from Russian by P.W.B.
Gilpin, R.R. (1979) The morphology of Harrison.) London: Pergamon Press,
ice structure in a pipe at or near Ltd.
transition Reynolds numbers. Heat Lunardini, V.J. (1981) Heat Transfer in
Transfer. San Diego: American In- Cold Climates . New York: Van
stitute of Chemical Engineers Sym- Nostrand Reinhold Co.
posium Series 189, vol . 75, pp. 89-94. Mallory, S.F. (1969)Thermallnsulation.
Gilpin, R.R. (1981) Ice formation in a New York: Van Nostrand Reinhold
pipe containing flows in the transition Co.
and turbulent regimes . ASMB M.I.T. (1957) Frost penetration in multi-
Journal ofHeat Transfer, vo1.103, pp. layer soil profiles . ACFEL Technical
363-368. Report 67. Hanover, N.H.: U.S. Army
Ingersoll, L.R., O.J. Zobel and A.C. Inger- Cold Regions Research and Engi-
soll (1954) HeatConduction: With .Engi- neering Laboratory .
neering, Geological and Other Ap- McAdams (1954) Heat Transmission,
plications. Madison : University of 3rd Ed. New York: McGraw-Hill .
Wisconsin Press. Porkhaev, G.V. and V .F. Zhukov (1972)
Jakob, M. and G.A. Hawkins (1957) Ble- Standard depth of freezing of soils
ments of Heat Transfer and Insula- for foundation engineering purposes.
tion, 3rd Ed. New York: John Wiley Osnovaniya, Fundamenty i Mek-
and Sons, Inc. hanikaGruntov, no.6 (Translated from
Johnson, P .R. (1971) Empirical heat Russian by Plenum Publishing Corp.,
transfer rates ofsmall, Longand Balch Consultants Bureau). New York.
thermal piles and thermal convection Quinn, W.F. and E.F. Lobacz (1962) Frost
loops. Fairbanks: Institute of Arctic penetration beneath concrete slabs
Environmental Engineering, Univer.. maintained free ofsnow and ice, with
sity of Alaska, Report 7102. and without insulation. Highway
Jennings, B.H. and S .R. Lewis (1968) Air Research Board Bulletin 331,
Conditioning and Refrlgeration,. 4th. pp. 98-115.
Ed. Pennsylvania: International Text- Sanger, F.J . (1968) Ground freezing in
book Co. construction . CRREL Miscellaneous
Kaplar, C.W. (1978) Effects of moisture Paper 404 . Hanover, N.H.: U.S. Army
and freeze-thaw on rigid thermal in- Cold Regions Research and Engi-
sulations : A laboratory investigation. neering Laboratory .
In Proceedings of the Conference on Sanger, F.J. and F.H. Sayles (1978)
Applied Techniques forColdBnviron- Thermal and rheological computa-
Biblio-2
'TM 5-852-6/AFR 88-19, Volume 6
tions for artificially frozen ground book No. 14. London: Constable and
construction. CRREL Miscellaneous Company, Ltd.
Paper 1216 . Hanover, N.H.: U.S. Army
Cold Regions Research and Engi. Wilkes, G.B. (1960) HeatInsulation New
York: John Wiley and Sons, Inc.
neering Laboratory.
Scott, R.F. (1964) Heat exchange at the Zienkiewicz, O.C. (1967) The Finite Ele-
ground surface. CRREL Monograph ment Method in Structural and Con-
II-A1. Hanover, N.H.: U .S. Army Cold tinuum Mechanics. London: MeGraw.
Regions Research and Engineering Hill Publishing Co., Ltd.
Laboratory . Zienkiewicz, O.C. (1971) The Finite Ele-
The Engineering Equipment Users mentMethod inEngineering Science.
Association (1966) Thermal insulation London: McGraw-Hill Publishing Co.,
of industrial buildings . EEVA Hand- Ltd.

U.S. GJVEXA4Nf PRINTING OMQ{ 1988,!06-982/80019

Biblio- 3
The proponent agency of this publication is the Office of the Chief
of Engineers, United States Army. Users are invited to send
comments and suggested improvements on DA Form 2028
(Recommended Changes to Publications and Blank Forms) direct to
HODA(CEEC-EG), WASH, DC 20314-1000 .

By Order of the Secretaries of the Army and the Air Force:

CARL E. VUONO'
General, United States Army
Official: Chiefof Staff
R.L. DILWORTH
Brigadier General, United States Army
The Adjutant General

LARRY D. WELCH, General, USAF


Official: Chief of Staff
WILLIAM O. NATIONS, Colonel, USAF
Director of Information Management
and Administration

DISTRIBUTION :
To be distributed in accordance with DA Form 12-34B, Require-
ments for Arctic and Subarctic Construction .
PIN: 024920 000