Skip to main content
 Previous Next
  • Zoom In (+)
  • Zoom Out (-)
  • Rotate CW (r)
  • Rotate CCW (R)
  • Overview (h)
Civil Engineering: Encyclopedia Arctica 2a: Permafrost-Engineering
Stefansson, Vilhjalmur, 1879-1962

Civil Engineering

Cold Weather Operations

EA-I. (Palmer W. Roberts)

COLD WEATHER OPERATIONS

Increased activity in the arctic region has focused attention on the
engineering problems encountered. In considering cold weather operations, it
would be hard to find a field of human knowledge in which the past reveals so many
practical guideposts for the present and upon which plans can be made for the
future.
Operations in the Arctic and Subarctic do not vary greatly from those
in the North Temperate zone. There are, however, certain variations in field
practices, construction methods, and engineering design necessary to meet the
requirements imposed by climate, terrain, and other local factors. In approach–
ing such problems, the peculiarities of the area must be accepted and the usual
design, materials, and construction techniques modified accordingly. The effi–
cient and effective use of all resources and conditions in the Arctic must
greatly influence the application of all engineering principles. It is necessary
to work with nature and not against her in the Arctic.
In the space assigned the subject of engineering in the Encyclopedia
Arctica , it would not be possible to cover the many subjects in the fields of
engineering for each country or area where special problems may be encountered.
It has, therefore, been necessary for the authors to treat many of the subjects
only in specific areas. In the main, the articles are concerned with operations
in the northern regions of the North American continent. However, the princi–
ples and techniques discussed may serve as a guide for other areas, after

EA-I. Roberts: Cold Weather Operations

analyzing local conditions.
In planning for operations in the Arctic, logistics, or providing the
means for supporting and accomplishing the task, may often prove to be the
limiting factor. The absence of rail and road nets, the difficulty of move–
ment overland both in winter and summer, the lack of ice-free ports in the
winter and the short period in the summer when vessels may operate, and the
limitations imposed on flying by poor visibility near bodies of water during
certain periods of the year, are a few of the many factors relating to the
area which must be considered. The maintenance and operation of equipment
under the variety of conditions imposed by extreme weather and terrain, and
the providing of at least the minimum of bodily comfort in the manner of
clothing and shelter for personnel, as well as proper food and adequate train–
ing, are problems that must be solved for each operation.
A thorough knowledge of the physical character of the Arctic during
all seasons of the year is also essential to the success of any operation
(see “Work Feasibility”). Surface and subsurface conditions vary greatly
with the season in the Arctic. Vast quantities of ice form in the winter
and much of it remains in the seas during the summer. The presence of ice as
a rock, or as a rock constituent of frozen ground, presents a major subsurface
problem for study (see “Permafrost”).
Palmer W. Roberts

Work Feasibility

EA-I. Roberts; Work Feasibility

LIST OF FIGURES
Page
Fig. 1. Work feasibility chart 5-a

EA-I. (Palmer W. Roberts)

WORK FEASIBILITY
In the northern regions and seasonal variations in climate and terrain
are extreme, yet the cold does not establish a barrier for human activity.
Climatic conditions do become a major factor when dealing with problems in
design and construction engineering. Climate has been said to be the annual
or long-term result of weather, and weather the result of nature’s effort to
balance the heat equation of the hemisphere. To provide a general background
of weather as it relates to engineering in the northern regions, major climate
factors - temperature, wind, precipitation, visibility, and light - are briefly
discussed herein.
Temperature . In the northern areas great temperature differences prevail.
The coldest temperature ever recorded in the northern regions was at Verkhoiansk,
Siberia, where −93.5°F. was reported. The lowest temperature in North America
ever recorded was −81°F. at Snag. Yukon Territory. Fort Yukon, just north of
the Arctic Circle in Alaska, has a U.S. Weather Bureau record of 100°F. in the
shade and considerably higher extremes have been reported at several places in
the North American and European Subarctic. Thus, it is apparent that the extreme
temperature range (Fahrenheit) for the Arctic can be placed at approximately
195 degrees; in Alaska the spread is approximately 185 degrees.
Wind . The frequency and velocity of high winds in the Arctic is are not as

EA-I. Roberts: Work Feasibility

great as in the other zones. Stronger winds, and more frequent strong winds,
are experienced in the winter along the coast than are found in the interior
at that time; in summer the differences are not so great. The frequency of
strong winds in the North American Arctic resembles that of Siberia. At
Point Barrow, Alaska, the maximum velocity so far recorded (up to 1950) for
the winter months is something over 100 m.p.h. but for the summer only 41 m.p.h.,
with the average annual mean at 12 m.p.h.
Precipitation is light in the Arctic. The annual total of melted snow
and rain may not exceed 10 inches per year in most of the Alaska and Canadian
Arctic, and is probably even less in the Soviet Arctic. It is difficult to
obtain accurate measurements of snow, due to the light and fluffy nature of
the crystals and their tendency to migrate under less than average wind velocity.
Snow may fall in any month. In the winter the precipitation is mainly in the
form of snow and is generally very light, except that fairly heavy snowfall occurs
in certain foothill and mountain areas. At Point Barrow, where accurate
aerological measurements are maintained throughout the year, the maximum snowfall
occurs in October, eight inches having been recorded during this month. Very
light rains are experienced throughout the Arctic in the summer, with the
maximum occurring during July and August. Thunder is rare.
Visibility . Fog and poor visibility provide difficulties for arctic
operations and ofttimes limit e a viation and ship operations along the coast.
Fogs are usually most dense at the borders of the ice. The land senses of
the Arctic are warmer than the sea in the summer and colder in the winter.
Thus winds passing from either belt across the other are liable to produce fog.
The most severe fog conditions are experienced in spring and in the fall
when a visibility of a mile or less occurs more than 20% of the time. Cloudless

EA-I. Roberts: Work Feasibility

days (0 to 0.2 coverage) are not common along the arctic coasts, but do occur
during December and January, the clearest months.
Light is a rather peculiar problem is polar regions; it is received from
the direct and refracted rays of the sun, the moon, the stars, the planets,
and from the aurora borealis. The sun remains above the horizon, because of
refraction, for something over six months at the Pole, nearly four and one-half
months at 80° N., and for over eighty days at Point Barrow, 71°23′ N. It is
never pitch dark at night in the North and there is more effective light in
these areas than in the tropics or temperate zones. This is because of
several factors, namely:
2. Cold, clear, dust-free air transmits light more readily.
4. During the period of no direct sunlight, the snow and ice cover
reflects and magnifies all light that it receives from the various light sources.
6. When the ground is not covered with snow, there is perpetual direct
and reflected sunlight.
8. The starts, moon, and aurora succeed in delivering a high percentage
of their produced light to the earth during the winter.
Maximum arctic darkness occurs in winter when the sky is densely overcast,
when there is no moon or twilight, and when there is no aurora behind the
clouds. However, even under such maximum darkness conditions, a man clad in
dark clothing can be seen a distance of about one hundred yards against a
snow or ice background. In winter the aurora borealis display is frequent;
in the summer it can be seldom be seen because of continuous light. They Auroras are of
varied description and undergo rapid changes in form. Some of these forms may
be described as arcs, rays, streaks or lances, bands, curtains or draperies,

EA-I. Roberts: Work Feasibility

coronas, diffused light, crowns, and cloud-shaped. They c v ary greatly in
color (see “The Aurora Borealis”).
The effect of the extreme variations in climate on engineering operations
can best be realized by comparing the physical character of the arctic surface
and subsurface during the extreme periods of winter and summer. Permanently
frozen ground, called permafrost, is a widespread phenomenon, common through–
out North America and northern Asia; about one-fifth of the land area of the
world is underlain by permafrost. Temperature within permafrost varies during
the year generally from 16° f t o 30°F. In the upper layer of the ground, which
is not permanently frozen, the effects of seasonal thawing are usually more
severe than in areas not underlain by permanently frozen material. It has
been determined that stresses developed in freezing ground may exceed 28,000
pounds per square inch. When ground conditions established by nature are
disturbed during construction, the thermal equilibrium must be restored as
quickly as possible. Groundwater may exist in frozen or thawed ground and,
as migrating water transmits heat, many of the problems of freezing and thawing
are related to it. In winter, drifting snow creates many problems and, there–
fore, is always considered when selecting sites for building or routes for
freighting operations. In the summer, snow remains on the ground in lowland
areas for short periods only.
Thickness of the ice on the rivers, on the lakes, and on the Arctic Sea,
is important in connection with transportation. In northern Alaska, ice begins
to form on fresh-water lakes early in September and on salt water usually the
latter part of the same month. Maximum thickness of ice is reached in early
May on both fresh- and salt-water bodies.

EA-I. Roberts: Work Feasibility

From even this brief description it can be deduced that the weather during
the major part of the year is such that detailed analysis of climatic factors
will be required before units can be committed to arctic field operations. The
principal factors to consider are as outlined: light, temperature, wind (the
combined wind-chill factor), precipitation (snow and rain), ceiling, visibility,
and the thickness of ice and other factors of sea condition. A comparison of
the above and of the work or mission to be accomplished is necessary. As an
aid in making such an evaluation, a chart of the factors may be of assistance
(see Fig. 1). On this chart are shown the factors of climate and the work or
mission to be accomplished, such as transportation, construction, and water
supply. Examining the month of February, the following factors are noted.
Daylight: 3 hours increasing to 9 hours with an additional 1-1/2 hours
of twilight per day.
Ceiling: 1,000 feet or lower, 19% of the month.
Visibility: one mile or less, 14% of the month.
Temperature: absolute maximum, 35°F; mean maximum −15°; mean tempera–
ture, −20°; mean minimum −25°; absolute minimum −56°.
Wind: maximum recorded, 71 miles per hour; average, 10 miles per hour;
direction SW. 43%, NE. 15%.
Wind chill: varying from B, fair for traveling, to E, extreme limit
for operations.
Precipitation: 4 inches of snow; trace of rain.
New ice thickness: ocean, 40 inches increase to 50; lakes and rivers,
48 inches increase to 54.
Reviewing these it is noted that over-ice freighting with tractors and
sleds can be conducted. Ground surface will support wheeled vehicles, tractors, Fig. 1. Work Feasibility Chart.

EA-I. Roberts: Work Feasibility

and sleds . can be conducted. Ground surface will support wheeled vehicles,
tractors, and sledo.
Planes can operate on skis throughout the area and on
wheels from prepared strips. However, the number of flying hours per day is
limited by daylight (moonlight can be used for supplementary purposes for
several days either side of the full, if the sky is clear). Shallow lakes
and rivers will be frozen but water can be taken from deeper lakes or obtained
from melting ice or snow. Due to poor visibility and cold, surveying is not
recommended. Earth work is not possible at this season but foundation work
can be accomplished by the use of explosives or by thawing methods. For
outside construction work, flood lighting will be required. The month of
February could be utilized to repair equipment in shops and accomplish inside
construction. All equipment to be used out of doors will required complete
winterization, arctic lubricants and fuels, and heated inside storage.
Personnel will require adequate protection, proper clothing, and indoctrination.
This same procedure, analyzing the climatic and operating factors, can
be followed for each month of the year. Charts of this type can easily be
developed for all parts of the polar, arctic, and subarctic regions and
should be made available for the planning of operations.
Palmer W. Roberts

Arctic Surveying

EA-I. (Ralph W. Woodworth)

ARCTIC SURVEYING

CONTENTS
Page
Introduction 1
Preparation s for Arctic Surveys 2
Local Transportation 3
Surveys 4
Geodetic Control Surveys 4
Astronomical Observations 4
Base-line Measurements 5
Horizontal Control 5
Level Net 5
Station Marks 5
Topographic Surveys 6
Hydrographic Surveys 7
Hydrography 7
Tides 9
Magnetics 9
Conclusions 10
Bibliography 10

EA-I. Woodworth: Surveying

LIST OF FIGURES
Page
Fig. 1 Alaska triangulation net 10-a
Fig. 2 Alaska hydrographic surveys 10-b

[EA:I. Ralph W. Woodworth]

ARCTIC SURVEYING
INTRODUCTION
The special problems encountered in surveys of the land and water areas
of arctic regions are due to the raw or undeveloped nature of the country. The
land routes of communication, except in the vicinity of the larger settlements,
are primitive and the meteorological conditions are rigorous. Both combine
to limit activities, at present, to less than six months of the year. Winter
work is reduced to a minimum.
Surveys in the Arctic are not difficult to accomplish. The subzero tempera–
tures are uncomfortable and the danger element is, of course, always present.
With careful planning, a knowledge of the country, experience with the limita–
tions of the various types of transportation, and the elimination of unneces–
sary risks, operations may be carried out safely and efficiently although at
high cost.
The over-all success of a project is dependent to a considerable degree on
the detailed plans made for the project, as under the comparatively harsh condi–
tions of the Arctic, nothing may be left to chance. The selection of proper
transportation equipment is perhaps the greatest single factor that will con–
tribute to the success of a mission.

EA-I: Woodworth: Surveying

Surveys in subarctic regions are planned and accomplished in much the
same way as surveys in the temperate zones. These areas are more accessible
and weather conditions are milder than those of the Far North. Field methods
and instrumental equipment are similar to those used in the temperate regions.
Preparations for Arctic Surveys
To achieve maximum efficiency in the short field season, the following
preparatory steps should be taken.
1. Prepare comprehensive and detailed plans to cover every phase of
supply, administration, and operation well in advance of the beginning of the
project.
2. Procure, test, and condition suitable rugged equipment and instruments
with spares of all items and a generous supply of spare parts.
3. Insure and adequate supply of communication equipment.
4. Ship the equipment to arrive well in advance of probable needs.
5. Prior to departure from the continental base, instruct all personnel in
the nature of the proposed su s rveys, the character o the country, the special
care required for the equipment, and the methods of operation.
Procurement, particularly for the specialized items, should be instituted
a year in advance of operations. Checks must be made to insure that items that
cannot be procured in the Arctic are in the hands of the survey party at least
two months before departure.
Personnel should be transported from the continental area to the Arctic
by airplane on the following schedule which may be modified as weather conditions
permit.

EA-I: Woodworth: Surveying

1. Mid-January: liaison group to complete preliminary field arrangements.
2. Early February: astronomic groups.
3. Mid-February: chief of party and staff, plus geodetic (triangulation
observing) groups.
4. Early June: balance of party.
Good organization requires that each subparty shall be as nearly self–
sufficient as practical; that one reliable Eskimo familiar with the country
shall be included with each unit; that each observing group shall include an
extra observer; and that the specialists, the heavy equipment, labor pool, and
emergency transportation shall be under control of the headquarters group.
Base camps should be selected at least two years in advance and supplemen–
tal camps one year.
Plans must provided that mobile parties carry a full allowance of rations,
sleeping bags, and sample communication equipment. Radio schedules must be set
up and these schedules maintained.
Local Transportation
Because of the undeveloped nature of the country, local field transporta–
tion requires limited support of both small and large airplanes, bulk-hauling by
tractor or sledge trains, and employment of small amphibious-type tracked vehicles.
Wheeled vehicles, unless ski-mounted, are nearly useless.
Dog teams are useful in an emergency. The light pay load and the neces–
sity of carrying or securing food for the animals definitely limit their value.
The dog sledge is a very desirable article of equipment and may be trailed
behind weasels” or jeeps.
Ship or launch transportation in the Arctic is limited to the few months of
open water.

EA-I: Woodworth: Surveying

SURVEYS SURVEYS (delete underlining)
Surveys in the Arctic comprise geodetic control surveys for the deter–
mination of geographical positions and elevations for use in charting and
mapping; topographic surveys from aerial photographs; and hydrographic surveys
for the determination of depths of water for navigation.
Geodetic Control Surveys
Geodetic control surveys include astronomical observations for determina–
tion of the position of the starting point and initial azimuth; base-line measure–
ments; and horizontal observations of directions for extension of triangulation
or traverse.
Astronomical Observations . Where possible, the starting point and initial
azimuth should be determined with precision. This can be done with standard
methods and equipment. Azimith may be obtained with a direction theodolite.
Where weight of instruments is a factor, satisfactory observations of latitude
and longitude may be obtained, using the “equal altitudes” method and the pris–
matic astrolabe.
The best time for observations, especially above the Arctic Circle, is
during the hours of darkness in the early spring or late fall. Spring obser–
vations are made ordinarily in less time and more economically than fall obser–
vations because the skies are clearer and radio reception is better due to the
absence of the northern lights, although the temperature is lower.
Above latitude 60°, special care must be taken in order to secure a satis–
factory value for azimuth because of the close spacing of the meridians. The
method outlined Hoskinson and Duerksen (7) for observations of this type will
yield excellent results.

EA-I: Woodworth: Surveying

The instruments used are sparingly oiled with a very fine watch oil and
tested to −50°F. in a cold chamber. Between observations they are placed in
unheated storage to prevent condensation and subsequent fogging of the lenses
and freezing of bearing surfaces. Glycerin rubbed on contact points will
effectively prevent freezing.
Base-line measurements are more easily made over flat ground or the flat
ice of lakes or lagoons with the line laid out in the direction of the pre–
vailing wind. First-order accuracy may be obtained with standard methods.
Stakes should be steam-jetted into the ice or ground using a portable boiler.
Horizontal control in the Arctic is extended by triangulation methods.
Triangulation is faster than traverse and will produce more points for the
same amount of effort.
Standard triangulation observing methods as given by Hodgson (6) and in
another manual of the U.S. Coast and Geodetic Survey (15) have been found
to be the most economical methods for daylight work in the Arctic.
The observing period begins when there is sufficient light to warrant plac–
ing the observing groups in the field and ends when the combination of spring
thaw, making travel dangerous, and melting of snow, which produces excessive
refraction, render operations uneconomical. North of the Arctic Circle this
period extends, normally, from February to May or June.
Level Net . Except for the lines of spirit levels along the Alcan Highway
and the Alaska Railroad, elevations have been determined by trigonometric
leveling. Such elevations may at times be subject to errors due to excessive
vertical refraction varying over the period of the level run.
Station Marks . In subarctic regions, including the Aleutian chain, stations

EA-I. Woodworth: Surveying

are marked with brass disks set into a drill hole in rock or into a concrete
block buried in the ground. In the higher latitudes where the permafrost
is close to the surface, the best marks are brass disks brazed to an iron
pipe steam-jetted into the ground.
Topographic Surveys
Aerial photographic methods provide the best means of obtaining topo–
graphic data. Ground methods are useful only for fill-in. The photographs
are controlled by a tie between ground features and stations established by
triangulation.
Photographic and field inspection operations are confined to a very short
period of the year. To secure accurate information, the ground must be photo–
graphed and the field inspection made when snow is at a minimum and the shore
line is clear of ice. In the Far North the best time has been found to be
during July and August and in the subarctic regions in the spring or late fall.
The long shadows cast by the sun interfere with and often obscure detail
on slopes. This condition is minimized by taking photographs, where possible,
when the sun is at its maximum declination but there is no satisfactory
solution to the problem. Elevations supplemental to those obtained by leveling
may be obtained by means of the phototheodolite.
The best and quickest method of spotting control for the field inspection
parties is to furnish them with low- or medium-al [: ] titude verticals and obliques
in the vicinity of the stations. Methods which involve the use of ground
transportation may be used when more economical. When topography is not car–
ried out simultaneously with control, each station must be inspected on the
ground.

EA-I. Woodworth: Surveying

Hydrographic Surveys
Hydrography . Standard field methods, as outlined by Adams (1), are
suitable for arctic surveys. The principal problems are not technical in
nature but have to do with the full utilization of the short working season
and the selection of the best available equipment. The plans should provide
that all preliminary work, such as preparation of the sheets and overhaul
and conditioning of the equipment and instruments, will be completed well
in advance of the probable date of arrival of open water, so that hydro–
graphic operations may be commenced without delay.
The hydrographic signals are built and located [: ] by the triangula–
tion part as it progresses. On a resurvey they may be located by graphic
means or photogrammetric methods, either of which yields satisfactory results.
Several types of boats have been used in the Arctic. The Navy rearming
boats are 35-foot diesel-powered single-screw wooden launches with enclosed
cabins and are capable of 12-15 knots speed. The landing craft are of various
types. The LCVP (Landing Craft Vehicle Personnel), a diesel-powered single–
screw wooden boat with open cockpit and ramp bow, is capable of 8-12 knots
speed, but is a very wet boat in a seaway. The LCM (Landing Craft Mechanized),
a 50- or 56-foot diesel-powered, twin-screw, steel boat with open cockpit
and ramp bow, is capable of 8-10 knots, and is a reliable craft.
Modifications to fit the boats for efficient use include: construction of
a shelter for the hydrographic crew on open-cockpit boats; removal, in some
cases, of the top of the ramp flush with the coaming of the boat to improve the
visibility; installation of strainers designed to keep slush out of the
engine on raw water cooling systems, or replacement of this type with keel
condenser systems; installation of metal sheathing along the water line for

EA-I. Woodworth: Surveying

wooden boats in order to protect the hull from “young” ice; replacement of
Au: OK? the magnetic compass with a magnesyn-type compass; and application of a coat
of orange paint for easy identification. In addition to hull and engine
modifications, provision must be made for installation of the sounding and
radio apparatus. Each boat should be fitted with two Fathometers, an out–
board and inboard projector, and a two-way radio. Emergency rations and
sleeping bags for all personnel should be carried at all times.
The best method of developing an area is by lines parallel to the beach.
Such a system provides maximum coverage with least cost. Every effort should
be made to complete one area before progressing to the next. For maximum
safety, boats must work in pairs and remain in close proximity, one sound–
ing out an offshore line while the other works along an inshore line. A
fully equipped spare launch should be available in each camp.
The launches should never enter a lead in the ice, but if caught, the
landing craft should pick out a shelving section of ice and allow the boat
to be pushed upon it. Conventional boat types should tow a light dory and
carry two outboard motors.
When ice is drifting on the shore, it will ground in water deeper than
the draft of the launch so that generally the boat may find open water close
to shore. In open ocean areas, if the wind picks up, the launches must get
through an inlet before the ice or breakers close it.
Experience to date has indicated that the period from mid-July to mid–
September is the most suitable for hydrography in arctic water, but this
period is subject to wide variations from year to year. In the Aleutians,
the period from May to October is best.

EA-I. Woodworth: Surveying

Tides . The tide gages must be established prior to commencement of
hydrography. The types in common use are described in a special publication
of the U.S. Coast and Geodetic Survey (16). The portable automatic gage is
the only practical type for these areas unless a plain staff gage is estab–
lished for a short series of observations.
At Point Barrow, the normal tide is small — a daily range of less than
a foot. The wind effect is much greater and has been recorded as reaching
3 feet. Winds blowing along the coast northerly and easterly from Point
Barrow lower the water level at Point Barrow, while winds from the opposite
direction raise it.
Because of the small range of tide, gages are visited frequently and
the marigrams changed every 48 hours. The staff is checked at each visit
and revealed when damage is suspected.
If necessary to build a gage on the ice, it may be done in the follow–
ing fashion: Erect a plain staff in a suitable position, make a hole through
the ice, and insert a pipe (4 inches in diameter or larger) filled with oil.
Pass a wire through the pipe and anchor it to the bottom, lead the free end
through a sheave on or near the top of the staff and down in front of the
staff. Secure a weight to it which will serve to keep the wire taut and to
indicate the depth. As the tide rises and falls the weight will register
on the graduated staff.
Magnetics . In subarctic regions where the magnetic field is sufficiently
strong to orient the needle quickly, standard methods, as outlined by Hazard (5),
may be used.

EA-I. Woodworth: Surveying

In the Arctic no comprehensive magnetic surveys have been made. It is
doubtful if any great coverage of this region may be made until air-borne
operations are feasible. In this area, the horizontal component (declination)
of the earth’s magnetic field is weak and displays complexities and variations
that are neither well mapped nor well understood.
Instruments must be especially calibrated for the magnetic latitude when
used in high arctic regions. Observations are impeded by the length of times
required to dampen the oscillations of the needle and by frequent unpredict q a ble
magnetic storms.
Conclusions
Progress in surveying and mapping in the Arctic will be accelerated by the
increasing interest in the region, by increased efficiency due to the probabl y e
introduction of more efficient methods of surveying such as Shoran, and the
longer field season that will result as knowledge of the region increases.
The att c ac hed sketches (Figs. 1 and 2) show graphically the progress that
has been made by the U.S. Coast and Geodetic Survey in Alaska in the fields
or triangulation and hydrography through 1948. It is apparent that although
the progress during the last 50 years has covered a large proportion of the
most accessible part, a vast area still remains to be done.
BIBLIOGRAPHY

1. Adams, K.T. Hydrographic Manual . Wash., D.C., G.P.O., 1942. U.S. Coast &
Geodetic Surv. Spec.Publ . no.143. Rev. (1942) ed.

2. American Society of Photogrammetry. Manual of Photogrammetry . N.Y.,
Chicago, Pitman, 1944.

Fig. 1 Fig. 2

EA-I. Woodworth: Surveying

3. k Drygalski, Erich von. Die Deutsche Südpolar-Expedition. Ber lin, Mitller,
(1902-03) 3v. in 2. Berlin. Univ.Inst. für Meeresk. Veröff . H.1-2, 5.

4. Great Britain. Royal Air Fo r ce. Empire Air Navigation School. Report no.45/24.
2 North Polar Flight of “Aries. ” 1946. 10 pts. Especially pt.10,
sect.5, “Terrestrial magnetism and magnetic compass in the Arctic.”

5. Hazard, D.L. Directions for Magnetic Measurements . Wash., D.C., G.P.O., 1921.
U.S. Coast & Geodet.Surv. Spec.Publ . no.166.

6. Hodgson, C.V., Manual of First-Order Triangulation . Wash., D.C., G.P.O., 1926.
Ibid . no.120.

7. Hoskinson, A.J., and Duerksen, J.A. Manual of Geodetic Astronomy. Determina–
tion of Longitude, Latitude, and Azimuth
. Wash., D.C., G.P.O., 1947.
Ibid . no.237.

8. International Boundary Commission. Joint Report Upon the Survey and Demarca–
tion of the International Boundary Between the United States and Canada
Along the 141st Meridian from the Arctic Ocean to Mt. St. Elias
.
Wash., D.C., 1918.

9. Leffingwell, E. de K. The Canning River Region, Northern Alaska . Wash., D.C.,
G.P.O., 1919. U.S.Geol.Surv. Prof.Pap . 109.

10. Madill, R.G. “The search for the North Magnetic Pole,” Arctic , vol.1,
T pp.8-18, Spring, 1948.

11. Mussetter, William, Manual of Reconnaissance for Triangulation . Wash., D.C.,
G.P.O., 1941. U.S.Coast & Geodetic Surv. Spec.Publ . no.225.

12. Rappleye, H.S. Manual of Geodetic Leveling . Wash., D.C., G.P.O., 1948.
Ibid . no.239.

13. Sverdrup. H.U. “Magnetic, atmospheric-electric, and auroral results, Maud
Expedition, 1918-24,” Carneg.Inst.Wash. Publ . no.175, vol.6, pp.309-524,
1927.

14. Swainson, O.W. Topographic Manual . Wash., D.C., G.P.O., 1928. U.S.Coast &
Geodetic Surv. Spec.Publ . no.144.

15. U.S.Coast and Geodetic Survey. Manual of Second and Third Order Tria gn ng ulation
and Traverse . Wash., D.C., G.P.O., 1929, Ibid . no.145.

16. ----. Manual of Tide Observations . Wash., D.C., G.P.O., 1935. Ibid . no.196.

Ralph W. Woodworth

Highways, Bridges and Protection from Ice Damage

EA-I. (Angelo F. Ghiglione)

HIGHWAYS, BRIDES, AND PROTECTION FROM ICE DAMAGE

CONTENTS
Page
Location 1
Bridges 2
Road Construction 5
Maintenance 7
Effluent Seepage Icing 9
Stream and River Icing 11

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

PHOTOGRAPHIC ILLUSTRATIONS
With the manuscript of this article, the author submitted
16 photographs for possible use as illustrations. Because of the high
cost of reproducing these as halftones in the printed volume, only a
small proportion of the total number of photographs submitted by all
contributors to Volume I can be used, at most one or two of the illus–
trations accompanying each paper, and in some cases none. The selection
of halftone illustrations for Volume I must await further progress in
publishing plans, for the number that can be used will be determined by
the publisher, and the choice of those used should be made in conjunction
with a representative of the publisher. All photographs are, therefore,
being held at the Stefansson Library until a selection can be made.

EA-I. (A. F. Ghiglione)

HIGHWAYS, BRIDGES, AND PROTECTION FROM ICE DAMAGE
The accumulated experiences in Alaska of a generation of engineers,
construction men, and prospectors are properly reflected in the present
practices and operations of the Alaska Road Commission. The most workable
location, construction, and maintenance procedures for the near-arctic and
subarctic conditions encountered have been evolved through years of practice.
Location
Most of the road planning in Alaska in pioneer location through virgin
country, which has not been surveyed or settled upon, and the route is,
therefore, not appreciably controlled by other than construction and sub–
sequent maintenance considerations. In addition to the usual location criteria
of conformance to the general road system planning, shortest route, standard
alig h nment and grades, reasonable economic design, and safety of traffic,
certain very important regional requirements must be considered.
Permanently frozen ground, or permafrost, underlies approximately
80 per cent of Alaska. Where at all possible, road location should avoid
these areas and the high construction and maintenance costs incident thereto.
The locating engineer soon learns to identify frozen ground by the type of
foliage and moss supported upon it.

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

A prime consideration is the development of locations on the south,
rather than the north, slopes of hills and mountain ranges. Several reasons
for such southerly exposure of location, all directly resulting from the
appreciably greater heat effect of the sun, are as follows. First, the
f g round is more likely to be thawed and free from permafrost. Second, the
construction will have the advantage of an earlier spring thaw and later
fall freeze-up together with more rapid thawing of the ground as it is
worked and exposed. Third, the maintenance problems are reduced [: ]
by the lighter snows, less slippery surface, less winter ground ice forma–
tion, and earlier spring thaws.
Wet sidehills or slopes, which indicate possible effluent seepages,
are avoided since spalling of slopes and major slides are to be expected
and winter ground icing will normally result. Likewise, the location must
be kept sufficiently above the known stream icing elevations of all streams
and rivers, either marginal to or crossing the line.
Other precautions essential toward minimizing the winter maintenance
problems include avoiding through cuts which induce drifting, and planning
slightly raised or fill sections which tend to be swept clear of snow.
Normal construction where permafrost cannot be avoided is to use an unbalanced
fill. Minimum grades and curvature are utilized to reduce winter maintenance
and increase safety of winter traffic operation.
Bridges
Bridge locations in subarctic areas require the additional consideration
of winter ground icing possibilities, stream ice break-up and flow, and the
channel-shifting to be expected of most glacier streams. In addition, numerous

EA p -I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

glacier rivers in Alaska are subject to extreme flash floods from the
bursting of glacier-dammed lakes or streams. Glacier floods in large rivers
such as the Nizina, Knik, and Kenai irregularly raise those rivers as much
as twenty feet and cause serious ice flow, bank erosion, and drift problems.
Such floods may occur in summer or in winter and, in the latter case, do
considerable damage by carrying heavy broken river ice against the bridges,
in addition to ice brought down from the glaciers, and ordinary drift com–
posed of trees, stumps, and other debris.
The bridge locations, to cope with the above conditions, are preferably
chosen well in advance of the road location and act as control points for
the route. A narrow crossing where the river is confined in one channel
by geographical features is preferred and a high level crossing, supported
by geographical features, is similarly desirable. Such a location would
naturally eliminate many usually maintenance problems resultant from the
above-listed subarctic conditions. In actual practice, however, it is
usually necessary to accept less than the ideal crossing, since the economics
of the over-all route location seldom justify allowing the bridge crossing to
control the line.
Bridge design must provide a minimum of restriction to river flow and
sufficient height to clear floods, ice flow, and winter ground and stream
icing formations. As a rule, clear spans are more satisfactory than trestle–
type structures, and mid-channel piers are undesirable since special icebreakers
and protective structures must nearly always be provided. On wide, flat,
glacier stream crossings where channel shifting is prevalent, the over-all
bridge length can often be reduced by utilization of dikes, wing dams, and
other channel control structures of restrict the flow to a single opening.

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

Such bank protective measure are costly and require continual maintenance,
but are often found economically justified by saving in cost, resulting
from decreased length of the bridge structure required.
Standard steel bridges have been widely employed for Alaskan roads.
These bridges are mostly simple truss and girder spans, though arches, sus–
pension bridges, and cantilever designs are freely used for special crossings.
Timber T t restles and truss spans have been used extensively in the past but
are being replaced with steel as rapidly as possible. The timber trestle is
used for small crossings and on secondary roads.
Concrete is used for bridge seats, footings, and some abutments, but is
seldom used for spans or piers because of the comparatively high costs involved.
No cement is produced in Alaska and the freight costs to interior Alaska for
this bulky material amount to several times its value. Most gravel and sand
available for concrete aggregate requires extensive washing and screening to
remove glacial silt which predominates in the deposits. In the isolated
locations common to most Alaska bridge construction, these factors, together
with the additional plant and skilled tradesmen required for the aggregate
preparation, form erection, concrete mixing and placing, and heating during
cold weather placing, make concrete construction exceptionally costly. By
comparison, steel structures can be easily handled, freighted, and erected with
a relatively small plant setup. Steel structures in the North do not deterio–
rate appreciably as compared to similar structures in southern moist climates.
Concrete, on the other hand, deteriorates faster in the North due to the
numerous cycles of extreme freezing and thawing.
Piers for the bridges in Alaska are predominantly of the steel “H”
piling bent type which have been developed by the Alaska Road Commission and

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

proved very satisfactory. Their cost is considerably less than comparable
concrete piers, the permanence is equal thereto, and the comparatively
simple handling and erection facilitates use in isolated locations. Their
cost is relatively low since they require none of the expensive excavation,
form work, cofferdam or caisson construction, or underwater work common to
concrete piers. In addition, they are well adapted to use in frozen ground
and can be placed as easily in winter as in summer. This factor is impor–
tant in Alaska where most bridge work is accomplished in the winter when
concrete operations require costly heating measures.
Road Construction
The construction of roads in Alaska involves special problems and is
limited, in general, to the summer season by the frozen ground and the
extreme cold temperatures of the winters. This restriction results in
long periods of idle equipment, seasonal employment of crews, and increased
costs as mentioned above. Bridge work is usually scheduled for winter con–
struction. This is done in order to provide longer employment for key
personnel and to provide utilization of the otherwise idle equipment.
Other measure for reducing costs resulting from the limited seasonal con–
struction period include winter freighting of supplies and materials to
advanced construction sites while frozen rivers and swamps aid rather than
hinder such operations, winter camp erection, and concentration of equipment
overhaul in the winter months.
Under normal conditions, the construction methods and equipment used
in Alaska are the same as those generally employed on large-scale road
construction projects in the United States. Discussion will, therefore,

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

be limited to the problems particular to the North where frozen ground is
one of the biggest obstacles to road construction. The swamp-covered
permafrost in Alaska is termed “tundra” and is generally a flat or rolling
terrain on which a thick moss insulates frozen ground beneath the surface.
The surface, however, thaws unevenly during each summer, and this produces
a continuous surplus of water. The moss and grass prevent adequate drainage
so that the water forms small pools and the frozen subsurface retards runoff
of this water, with the result that in the summer the ground is very swampy
and soft, even on relatively steep hillsides. The wet soil, practically
suspended in water, interferes with the operation of regular road-building
equipment, and the ice or frozen ground just beneath the wet cover impedes
grading operations.
The best procedure in such cases is to avoid disturbing the equilibrium
maintained by nature between the factors in the permafrost province, known
as the thermal regime. Any construction operation which changes or disturbs
this thermal regime will be subjected to the forces of nature as they endeavor
to reestablish this equilibrium. Climatic conditions and geologic changes may
also affect the thermal regime and, thereby, result in either the degradation
of aggradation of the permafrost. Where at all possible, therefore, fill
sections are used over frozen ground — the fill placed with the least possible
disturbing of the moss or overburden. Where thawing has made the area swampy,
it is general practice to place a corduroy mat for additional protection and
support.
Where the topography requires cutting into frozen ground, the work may
have to be programmed over several seasons to permit progressive thawing as
the exposed surface is worked off. Drilling the blasting of permafrost, as

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

in the case of rock, is resorted to in some cases in order to accelerate
breakdown of the impermeable material.
Maintenance
Summer maintenance of roads in the northern regions of Alaska requires
very much the same procedures as used in the Midwestern States. After the
spring thaw the roads dry out and become dusty, even though frozen ground
may be only a few feet below the grade. Particular local problems of
heaving, sliding, or subsiding ground, resulting from the uneven thawing
of the underlying frozen ground, require continual maintenance attention
and are quite common. Corrective measures include removal of slide material
or placing of additional surfacing material to reestablish uniform vertical
alig h n ment.
Brush cutting, ditch and culvert cleaning, spot graveling, and other
usual maintenance requirements are of necessity concentrated in the summer
working season. In addition, all possible preliminary steps toward facili–
tating winter maintenance are taken. Such measures include placing culvert
stakes and snow takes of sufficient height to be discernible above snowdrifts
to aid in locating the culverts and the edge of the road when thawing or
plowing; placing of snow fences and ice fences where know drifting or
icing occurs; flattening of the road crown in the fall to minimize danger
of sliding on icy roads; and building up stockpiles of sand and cinders
for winter surface sanding.
In the consideration of winter road maintenance, and Alaska Road
Commission has of necessity developed methods of preventing and coping with
winter ground ice formations that endanger highways and highway structures,

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

and constitute a serious hazard to normal traffic flow. Such ice forma–
tions are locally known and described throughout Alaska as “glaciers”
and any other designation would be confusing to the average maintenance
man. However, the term is incorrectly applied and since true glacial
conditions are seldom encountered on road maintenance, the term will be
used only when correctly applicable to masses of snow-formed ice. In this
article, winter ground ice formations particularly affecting winter highway
travel will be referred to as “icing.” Icing results from the successive
freezing of sheets of overflowing water into masses of ice attached to the
ground. For road maintenance considerations, these ice formations are
classified into two general types — the effluent seepage type and the river
or stream type. Ground icing of the effluent seepage type is experienced
throughout most of Alaska while the river or stream type is prevalent only
in the interior as differentiated from southeast and south-central Alaska.
As would be expected, the actual formation of the ice masses varies
unpredictably with the ground, water, and weather conditions. The severity
of the winter, the amount of snowfall and whether it falls early in the
season or late, and the nature of the season prior to the freeze-up, whether
wet or dry, all comprise contributing factors toward the resultant icing
possibilities. From past experience, certain generalities as to these
factors have been deduced and may be briefly summarized as follows:
( a ) A rainy season prior to freeze-up will increase the ground-water
flow in winter and usually cause more and heavier icing.
( b ) An early heavy snow will tend to insulate the ground and minimize ice
formation. Conversely, prolonged freezing weather with little snow results
in considerable icing.

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

(c) A severe winter with long periods of extreme cold may tend to
freeze back and stop the flow and formation of seepage-type ice but will
increase the formation of the stream and river type. The reverse is
generally true of mild winters.
These generalities cannot be considered infallible since various com–
binations of the conditions may produce opposite results. As a result,
methods of controlling the ice have been developed both as preventative,
where possible to forecast probable formations, and as control measures
where the ice forms infrequently or unpredictably.
Effluent Seepage Icing
The effluent ground seepage ice is the predominant form to be coped
with in winter maintenance operations in Alaska. When relatively warm
ground seepage meets the surface of a highway-cut slope in summer, it
normally may evaporate, flow down the slope and into the regular drainage
ditches, or may continue to flow just under the surface. However, in the
winter, with continued freezing temperatures, this flow freezes soon after
exposure to the air and forms successive l ev ay ers of ice which build up to
many feet in thickness or until hydrostatic equilibrium is reached. Since
this type of ice normally forms on sidehill cuts, the resultant sloping ice
surface built into the road increases the danger to traffic by crowding it
to the outer edge and tilting it toward the outside at increasingly steeper
slopes. This type of icing, besides building to many feet in depth, often
forms slopes too steep to permit passage of any kind of traffic.
A relatively inexpensive and very workable method of controlling this
type of icing has been developed by the Alaska Road Commission and has proved

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

so successful that its use has spread throughout the Territory and into
Canada where similar problems occur. Termed “ice fencing,” this method
employs the principle of damming of the seepage and controlling its freezing
before it reaches the road. Since the actual head of water to be diverted
is never more than the depth of the seepage film, the term “dam” is mis–
leading and the actual fence used may be of light temporary construction.
In operation, the ice fence is placed as soon as icing is observed, or,
in cases of annually repeating ice, may be placed before the freeze-up.
The fence consists of wire netting on pos e t s to which has been fastened to
a layer of canvas, roofing paper, or similar relatively impervious material.
The posts may be buried in the ground, if placed prior to the freeze-up,
may be frozen in holes picked into already formed ice, or may be temporary
tripods. The wire netting used is either fish-trap wire, chicken wire, or,
in some instances, standard slat-type snow fencing. The resultant fence,
with impervious facing for diverting the seepage, need only be of sufficient
strength to withstand the local winds. It will actually be increasingly
stabilized by the ice as the thickness of the controlled icing develops.
The ice fence is placed between the seepage and the road and actually
controls the ice by diverting the flow parallel to the road until it freezes.
The resultant vertical wall of ice may build up to considerable depth and
require a second and sometimes a third lift of the fence during the winter.
These fences may have proved very effective in keeping ice from forming on
the roads and are especially preferred to other control methods since only
infrequent attention of the maintenance crews is required.
While the ice fence has simplified control of forming ice, it is still
possible, in some instances, to avoid the icing problem entirely by striking

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

closer to the source. Interception ditches, graded to pick up the seepage
before it reaches the road out and to divert it into other drainage channels,
are possible. This type of control can be used effectively where the ground
water flows near the surface and can be intercepted at some distance from
the road.
Similar interception ditches to collect and channel the seepage flow
have been effective where such concentrating of the water preserves suffi–
cient latent heat to keep it from freezing until the water has passed through
the road drainage structures. While icing will probably still occur, it will
be below the road and of no danger to traffic. An added measure to insure
operation of this type of diversion consists of covering or insulating the
channeled flow against rapid freezing in order further to delay the icing.
A practical method consists of covering the ditch with brush of sufficient
thickness to support snow; the snow cover provides considerable insulation.
Where flow of ground water can be intercepted at some distance from the
road, it is also possible to cause or control the icing by exposing the
seepage at a desired location. Clearing and stripping or simply ditching
may provide sufficient interception and exposure to initiate the icing where
desired.
Stream and River Icing
An ice problem more easily coped with is that of stream or river icing —
formed when the stream freezes to its bed and then continually overflows
and freezes, forming successive ice layers, in many cases reaching depths
of more than 20 feet. This type of icing occurs where the stream flows
over permafrost, and is relatively predictable, since it occurs year after
year in the same locations. The bridges planned for crossing such streams

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

must be kept sufficiently high to prevent the overflows from reaching the
level of the deck with the resultant heavy ice loads built onto the
structure and danger to traffic. Where overflow has occurred, the bridge
is not structurally endangered until the period of spring thawing when
the stream cuts back down to its bed and undermines the ice, leaving many
tons of it hanging on the bridge. In such cases, it is often necessary to
remove large quantities of ice from the structure by chopping, thawing and
cautious use of blasting charges.
Besides planning high level bridges for icing stream crossings, it is
important that a clean understructure or clear span be used. As mentioned
above, use of trestle-type structures is avoided, if possible, since the
removal of the ice formations from the trestle supports is extremely diffi–
cult to accomplish without damaging the bracing.
Where an icing stream threatens to close the o ep pe ning of a large culvert
or a bridge, the stream may sometimes be kept flowing through the opening by
channeling and heating. The application of heat to the water above the
opening tends to delay its freezing until below the highway and also to cut
out a channel in the already formed ice. The heat may be supplied either by
means of a boiler through steam pipes laid above or through the opening, or
more simply by means of a 50-gallon steel drum heater placed on end in the
channel above the opening. This type of heater is merely a drum with one
end cut out and in which a fire is maintained by use of oil or wood. Usually
an infrequent fire, once a day or less, will provide sufficient heat to keep
the channel open.
Where icing streams and rivers parallel highways, the principal protec–
tive measure is again that of building the road sufficiently high to avoid

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

overflow. Where overflow of the road does occur, continual attention is
necessary to maintain traffic and to protect the road from subsequent
washouts during the spring thaw and runoff period. During winter mainte–
nance of such sections of road, it is necessary to provide a relatively
flat and level traveling surface for the safe operation of traffic by
picking wheel ruts, diver g t ing or chan n eling the overflow, and blasting and
blading the uneven ice. Numerous instances of traffic being maintained
over long sections of the ice several feet thick without the public being
aware of its being completely off the road surface have been the result of
proper maintenance provisions.
During the spring thaw and runoff, it is essential that culverts be
thawed open and the streams be controlled in their regular channels as
quickly as possible. Culvert thawing is usually accomplished by means of
steam from small truck-mounted boilers. Varying lengths of small-sized
pipe (usually ½' in.) are connected with sufficient steam hose to permit
working at considerable distance from the boiler position on the road. The
pipe is worked through the ice in the culvert as the steam thaws ahead —
usually starting from the lower or outlet end in order that the melted water
may escape. It is essential that culvert stakes be set prior to winter
maintenance in order to facilitate locating frozen culverts for thawing.
All culverts are staked by the Alaska Road Commission in a uniform manner
to simplify these operations.
In some instances the thawing of culverts by the method described above
is impracticable. Such cases are encountered in extremely long culverts as
under high fills, in culverts which have had several feet of ice form around
and above the inlet and, in some instances, also above the outlet, and in

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

old partially collapsed or deformed culverts through which a straight
thawing pipe is difficult to handle. Where the failure of such a culvert
might endanger a large fill or critical section of road, thawing may be
provided by means of permanently installed thawing pipes. Such pipes of
relatively small diameter are hung inside the culvert prior to the freeze-up,
carried out and up to easily accessible locations on each end, and capped
to prevent plugging by condensation. Any necessary thawing of the culvert
may subsequently be accomplished by attaching a line from a mobile steam
boiler to the permanently installed pipes.
In large fills where the culverts are extremely long, an added precaution
is sometimes taken by installing a “relief” culvert near the top of the fill
above the icing height. Such relief culvert will handle spring runoff until
the lower culvert thaws open.
The freezing back of thawed culverts during spring operations may be
partially controlled by use of rock salt — a quantity of salt deposited near
the culvert inlet in a burlap sack will tend to prevent such freezing until
the flow of the thawing runoff cuts a sufficiently large opening through
the ice.
Where roads are not kept open during the winter and icing has been
allowed to proceed uncontrolled, a considerable problem for traffic results
in the spring opening. In extrems cases, such ice has covered sections of
road several thousand feet in length and to depths exceeding 20 feet. Re–
moval requires blasting, cutting with tractor and bulldozer, and repeated
blading as the surface softens through thawing. Sprinkling dirt or ashes to
accelerate the thawing effect of the sun is effective, and also liberal use

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

of rock salt will expedite the ice removal. While traffic can be main–
tained over relatively flat thawing ice, it is sometimes hazardous and
requires the constant attention of the maintenance forces.
Numerous other ice problems are encountered in road work in Alaska,
but are of relatively minor importance and will, therefore, be dealt with
briefly.
Ice flow in rivers during the break-up period, known as debacle, con–
stitutes considerable menace to bridge piers and marginal road structures
and fills. Protective measures include ice shears, heavy rock riprap, and
steel piling icebreakers on the upstream sides of the piers. Timber piling,
as in trestle structures, must be sheeted with metal or planked for full
width of bend above and below ice line to prevent being cut off by passing
ice.
Frost mounding and frost heaving, while deforming the roads in places,
do not usually constitute serious hazards to winter traffic. Frost mounds
have, in instances, upwarped the surface as much as six feet in the middle
of the road and have again subsided leaving no apparent damage. The opposite
action occurs occasionally where roads are located over permafrost containing
ice lenses or ice veins. The uneven melting of the ice or the shifting of
underground water channels may cause serious subsidence of many feet over
relatively short periods of time.
Frost boils, resulting from uneven and accelerated spring thawing, are
a serious problem to proper road maintenance and result in rough and some–
times impassable roads. Maintenance procedures include covering with coarse
pervious gravel, removing the weakened surface material by digging or blasting
with ditching powder and backfilling with coarse gravel, or by - passing the

EA-I. Ghiglione: Highways, Bridges, and Protection from Ice Damage

traffic and permitting the boil to dry up naturally where road width will
permit.
With the every-increasing winter traffic flow over the principal
Alaskan highways, the problems incident to proper and safe maintenance
operations are becoming of greater importance in the over-all highway
program. Additional preventive measures are justified both in the initial
construction planning and in the maintenance improvements such as minor
changes in alignment and grade. As an accumulated result, the actual daily
winter maintenance problems are being simplified, the general winter travel
conditions are being steadily improved, and the highways are being made
safer for use by the general public.
Angelo F. Ghiglione

Drainage, Flood Control and Beach Erosion in Alaska

EA-I. (James Truitt)

DRAINAGE, FLOOD CONTROL, AND BEACH EROSION IN ALASKA

CONTENTS
Page
Lowell Creek 3
Tanana River and Chena Slough 5
Salmon River 7
Skagway River 8
Erosion 10

EA-I. ( Colonel James Truitt)

DRAINAGE, FLOOD CONTROL, AND BEACH EROSION IN ALASKA
Degradation of the land mass throughout Alaska is progressing at a
relatively high rate as the result of water action, despite the fact that
the major drainage systems are icebound and practically dormant for fully
50 per cent of the year. In general, the central portion of Alaska is not
subject to heavy precipitation or torrential rainfall normally associated
with serious flood conditions. Flooding along Alaskan water courses does
not conform to the general concept of disastrous floods in that the duration
and extent of inundation are quite limited. However, the steep gradients
and resultant velocities of the uncontrolled waters cause serious destruction
in their path and threaten the establishment of channel changes. Consequently,
stream bank erosion and flooding are matters of primary concern in the planning
of all large-scale developments in Alaska.
The principal Alaskan streams have their sources in the snow and ice
fields of the surrounding mountains, hence, floods of various magnitudes may
be expected during the warmest and driest period of midsummer. The intensity
of this type of flood may vary from a bank-full stage, resulting from an uninter–
rupted discharge of snow-melt, to out-of-banks flash floods caused by the forma–
tion of local ice jams or the failure of ice dams at or near the terminus of
hanging glaciers. Two outstanding examples of such flooding characteristics

EA-I. Truitt: Drainage, Flood Control, and Beach Erosion

are noted in the Knik River, discharging into the head of Knik Arm, Cook
Inlet, and the Copper River which empties into the Gulf of Alaska. Devas–
tating flash floods are practically annual occurrences on these streams
and are responsible for serious damage to bridges and other improvements
in their lower reaches. In 1926, the writer was directed to investigate
the source of these damaging floods in the Copper River drainage and to
determine what, if any, steps could be taken to prevent their reoccurrence.
The source was found to be at the foot of the Skolai Glacier in the head–
waters of the Nizina River, a tributary of the Copper. In this case, an
arm of the glacier abuts a vertical rock wall, forming a dam which impounds
a lake some 5,000 acres in extent. By midsummer, the snow- and ice-melt
from the glacier fills this basin and when the hydrostatic pressure reaches
about 120 feet, the gravel base under the ice dam blows out releasing the
entire lake down a channel where the gradient varies from 7 per cent
initially to about 2 per cent in the lower reaches. The top o graphy at the
abutment of the ice dam is such that a spillway and short tunnel through
the rock wall could be built to relieve this condition. However, such a
plan could not be recommended since the tunnel would in all probability
become inoperative because of ice. Furthermore, the high cost of construc–
tion in this extremely remote locality would have far outweighed the benefits
that would then accrue. A somewhat similar condition exists at the head of
the Knik River and there are undoubtedly many other instances that have not
been investigated or reported.
Floods in the Salmon and Skagway rivers, in the coastal regions of
southeastern Alaska, follow a much more conventional pattern, as floods in
these drainages generally occur in the early spring following heavy rains

EA-I. Truitt: Drain ing age , Flo o d Control, and Beach Erosion

and the accelerated melting of the snow cover in their basins.
A combination of flood causes exists in the case of Lowell Creek a
mountain torrent which discharges into Resurrection Bay at Seward. Floods
in this stream are unpredictable. It may flood in the early spring due to
unusually heavy precipitation, or at any time in midsummer by the release
of glacial impoundments at its source.
The Tanana River, in central Alaska, is subject to flooding and dis–
tributary channeling in the Tanana lowlands above Fairbanks as the result
of ice jams during the spring break-up. The condition in the Fairbanks
area is further aggravated by the general movement of the Tanana River
toward the right limit of its valley floor, due to the fact that the tribu–
taries from the Alaska Range which form the left limit of the valley have
extremely steep gradients, and, in consequence, are building up the valley
floor with heavy detritus, thus forcing the river toward its northern limit.
Each waterway presents its own characteristic flood pattern. The limited
development in Alaska to date has only warranted the investigation or the
a e d option of control measures on a few streams. A brief description of the
flood-control projects which the Corps of Engineers has been called upon to
provide in Alaska follows.
Lowell Creek. This is a Lowell Creek small stream discharging into
the head of Resurrection Bay at Seward, Alaska. It drains a 5-square mile
of glacier-covered mountain slopes through a deeply incise s d channel. Owing
to its extremely steep gradient, it transports immense quantities of glacial
detritus at flood stage. During ages past, this material has been deposited
immediately upon release of the stream from the confines of the canyon section
to form a delta, which now extends seaward from the mountain wall for a distance

EA-I. Truitt: Drain [: ] age , Flood Control, and Beach Erosion

of about 1,200 yards on a slope of 5 per cent. This delta became the
townsite of Seward, Alaska, and later the southern terminus of the govern–
ment-owned Alaska Railroad.
A series of unusually severe floods during the years 1923 to 1926,
inclusive, caused extended interruption of traffic on the Alaska Railroad,
with consequent effect on settlements throughout the Alaska Railroad belt.
The need for control measures on this waterway was apparent, and under
authority of an Act approved February 9, 1927, the Alaska Road Commission
constructed a rectangular timber flume, 12 feet wide, 7 feet deep, and
3,300 feet in length, to convey the debris-laden flood waters from the
mouth of the canyon across the townsite to tidewater. This improvement
was completed in 1929 and served adequately until 1935, by which time the
deposit of solid materials at the outlet of the flume had built up to such
an extent that free discharge was no longer possible and the flume became
choked with boulders and gravel during every flood stage. The volume of
solids transported is indicated by the effect of a flood in October 1935,
when the flume was completely filled with gravel from outfall to intake, a
total quantity of 10,000 cubic yards in 11 hours. The actual quantity of
solids transported considerably exceeded this apparent amount since the
flume functioned to a limited degree for the initial 3 hours of this 11-hour
period.
A revised plan of control was necessary and further studies were made
by the Corps of Engineers. As a result of this investigation, it was
decided that the only plan that would insure full protection to the town
of Seward and the terminal facilities of the Alaska Railroad would require

EA-I. Truitt: Drainage, Flood Control, and Beach Erosion

a complete diversion of Lowell Creek from the townsite area. This condi–
tion could be met most satisfactorily by construction of a low dam across
the canyon, about 1,500 feet above its mouth, and a discharge tunnel
2,100 feet in length, through the mountain forming the right limit of the
canyon. This plan would permit discharge of the flood waters at an ele–
vation 65 feet above sea level and about 1,000 feet away from any developed
area, thus providing a disposal site for about seven million cubic yards,
or a usable life of about 180 years.
The project was adopted by the Congress in August 1937, completed by
the Corps of Engineers in 1940, and provides necessary flood protection
to the community and the terminal facilities of the Alaska Railroad.
Tanana River and Chena Slough. The Tanana River has its source on the
northern slopes of the Wrangell Mountains near the Alaskan-Canadian border,
flows northwesterly about 500 miles, and discharges into the Yukon River,
140 miles west of Fairbanks. The river closely follows the northern limit
of the Tanana lowlands, its valley floor, where it has been forced by the
deposition of heavy glacial debris brought down by tributaries from the
north slopes of the Alaska Range which form the left limit of the valley.
The Chena River, a southwesterly flowing tributary of the Tanana, has its
source on the low divide between the Tanana and Yukon rivers and flows
through rolling hill country to emerge onto the valley floor of the Tanana
about 20 miles above its actual confluence with the Tanana River.
The city of Fairbanks and the military reservation that embraces Ladd
Field are located in the lowland area between the Tanana and Chena rivers,
about 10 miles above their confluence. The Tanana River lowland in this
locality are subject to overflow during the spring break-up, and the security

EA-I. Truitt: Drainage, Flood Control, and Beach Erosion

of Fairbanks and other improvements in the vicinity were further threatened
by the existence of Chena Slough, a distributary of the Tanana, which drains
from the Tanana at high water stage and flows westerly to join the Chena
River a few miles above Fairbanks.
A marked increase in the volume of water carried by Chena Slough with
each succeeding flood stage in the Tanana River indicated that a flow far
exceeding the capacity of the Chena River channel at Fairbanks could be
expected, and there was considerable evidence pointin t g to an eventual
channel change of the entire Tanana River.
Pursuant to Act of Congress, June 1936, the Corps of Engineers was
directed to investigate the condition and determine what action might be
taken to avert disastrous floods in the Fairbanks area. The investigation
revealed that the outflow from the Tanana River into the Chena Slough
channel was through a number of small chutes in a reach between 25 and 30
river miles above the mouth of the Chena River, and that ice james in the
Tanana were responsible for enlargement of existing overflow channels as
well as the creation of new distributaries in this area. The unstable
condition of the river banks and the length of the reach involved precluded
consideration of bank stabilization; hence, less conventional methods of
control were necessary. Field surveys developed that the Chena Slough
drainage was concentrated against a rock point projecting into the valley
floor from the north limit at a point about 15 miles east of Fairbanks.
The Corps of Engineers recommended that a rock-fill dam be constructed across
Chena Slough at this point and the barrier extended southwesterly as a
training dike across the valley floor for about 16,000 feet to the right
bank of the Tanana River, thus collecting and returning all overflows to
the main river.

EA-I. Truitt: Drainage, Flood Control, and Beach Erosion

The plan was adopted and authorized by Congress in 1938. The project
was completed in 1941 and has provided full protection against flood menace
from this source. It should be noted, however, that there is no assurance
against further attacks of the Tanana River along its right limit down–
stream from the improvement provided.
Salmon River . Salmon River has its source in the precipitous mountains
of the coastal range in southeastern Alaska and empties into the head of
Portland Canal at Hyder on the extreme southeastern border of Alaska. It
flows through a deeply incised valley generally less than one-half mile
in width and its total length is about 25 miles. It is fed by numerous
mountain torrents heading 3,000 to 5,000 feet above sea level in the snow
fields of the mountain slopes. The gradient of Salmon River is extremely
steep in its upper reaches and it retains a fall of 30 feet in the final
mile above tidewater; hence, velocities are capable of developing new
channels at any flood stage through the unconsolidated materials on which
the town and port facilities of Hyder are located. Hyder had been frequently
inundated, and destruction of the town and its port facilities were threatened
by the encroachment of Salmon River.
Rapid shoaling at the head of Portland Canal was seriously restricting
navigation at the head of this waterway. The port of Stewart, British Columbia,
just across the boundary line, was already deprived of navigable depths and
a recent change in the trend of the Salmon River at its mouth indicated the
delta would shortly include the wharf area at Hyder. Corrective measures
were necessary both in the interests of navigation and the control of floods.
In January 1931, Congress directed an investigation with a view toward

EA-I. Truitt: Drainage, Flood Control, and Beach Erosion

the control of floods. The corrective measures recommended consisted of
a stone-faced wing dike, 4,400 feet in length, extending from a rock point
on the left limit of the river above the townsite to a point on the fan
where the greater slope was toward the right limit of the delta. This
plan afforded flood protection to the townsite of Hyder and the wharf
approaches, and also diverted the discharge of debris away from the wharf
face toward deeper water in the Portland Canal.
The project was authorized by Congress in 1934, and constructed by
the Corps of Engineers in 1935. It has afforded the protection desired
and maintenance costs have been negligible.
Skagway River. This water course has its source in White Pass, a
relatively low saddle in the coastal range between northwestern British
Columbia and tidewater at Skagway. Its steep, narrow valley provided an
access route into the Klondike during gold-rush days and was later developed
for the location of the White Pass and Yukon Railway, a narrow gage line,
110 miles in length, connecting Whitehorse, Yukon Territory, with the
United States seaport of Skagway.
The total length of the river is less than 15 miles and it drains an
area of approximately 125 square miles of mountainous terrain. It discharges
into the head of Taiya Inlet, and in recent geological time, it has brought
down sufficient rock waste from the surrounding mountains to fill the head
of the inlet for a distance of about 3 miles. This deposit of unconsolidated
material became the site of Skagway but the building process is still in
progress; hence, the security of all developments in the lower reach of the
river was threatened at every flood stage.

EA-I. Truitt: Drainage, Flood Control, and Beach Erosion

Investigation of the condition with a view toward providing flood
protection for the Skagway townsite and wharf area was directed by Congress
in 1935. In response thereto, the Corps of Engineers recommended the con–
struction of a rock-faced levee along the left limit of the stream, extend–
ing 6,700 feet downstream from a stable abutment on the canyon wall through
the townsite area to tidewater, with an 1,800-foot rubble mound breakwater
extension over the tide flats to deflect the river discharge away from the
wharves.
The project was authorised by Congress in 1938 and completed in 1939.
Further improvements in the interests of navigation by extension of the
breakwater was subsequently authorized.
The growing recognition of the strategic and economic importance of
the Territory of Alaska will undoubtedly result in its rapid development.
Stream valleys will certainly be adopted as the site of future communities
and the information that has now been collected and recorded on the flooding
characteristics of Alaskan streams can be of inestimable value in the planning
for permanence and security.
The control of floods will, of course, become a matter of greater concern
as settlement and development of the Territory increases. However, it is felt
that future developments will be planned, taking into consideration the flood
characteristics of adjacent waterways. These conditions were not recognized
or carefully weighed in the siting of early settlements in Alaska, as
evidence s d at Seward townsite, laid out on the built-up delta of Lowell Creek,
where it emerges from the canyon section, and again, in the case of all Tanana
lowland developments along the right limit of the Tanana River, where continual
encroachment of the river is indicated.

EA-I. Truitt: Drainage, Flood Control, and Beach Erosion

Erosion of the coast line of Alaska is relatively nonexistent. Although
some beach erosion has occurred at Nome on the Seward Penninsula as the
result of severe storms, the coast line generally is building up rapidly
from the discharge of silt-laden rivers and streams. This condition is
readily noted throughout the entire coast line from the head of Portland
Canal in southeastern Alaska to the mouth of the Colville River, discharging
into the Arctic Sea.
The delta of the Stikine River in southeastern Alaska has practically
annexed an offshore island to the mainland within the recorded history of
Alaskan shipping. Shoaling of the tidal waterways at the head of Knik Arm
has progressed so rapidly that small trading schooners that served trading
posts in this region as late as 1910 can no longer navigate these waters.
The extensive de l ta l s of the Kuskokwim and Yukon rivers can be readily noted
from any map of Alaska, but the deposition of the silt load of these streams
is not confined to the delta areas, as indicated by the existence of extensive offshore
bars reaching for many miles from the river mouths in the direction of the
littoral drift.
James Truitt

Strength and Uses of Fresh and Salt Water Ice

EA-I. (Ralph Hansen and Kenneth Linell)

STRENGTH AND USES OF FRESH AND SALT WATER ICE

CONTENTS
Page
Uses of Ice in the Arctic 1
Strength Properties of Ice 3
General Characteristics of Ice 3
Published Test Results 4
Effect of Intracrystal Structure of Ice 4
Effect of Impurities in Ice 5
Effects of Horizontal Stratification in Ice 7
Effect of Vertical Cracks 8
Temperature during Ice Formation 8
Residual Stresses 9
Effect of Temperature during Test 9
Effect of Rate of Loading on Strength Tests 12
Effect of Orientation of Test Specimen 13
Size and Proportion of Test Specimens 13
Load-Bearing Capacity of Ice 14
Records of Actual Ice Loadings 14
Historical Military Rules 14
Moskatov’s Empirical Method 15
Elastic Theory Analysis 16
Preliminary Elastic Theory Ice Thickness Curves 23
Comments on Elastic Theory Method and Thickness Curves 23
Effects of Moving Loads 25
Methods of Increasing Supporting Capacity of Ice 26
Bibliography 28

EA-I. Hansen-Linell: Fresh & Salt Water Ice

LIST OF FIGURES
Page
Fig. 1 Bearing capacity of old sea ice for airplanes
with wheels
16-a
Fig. 2 Bearing capacity of fresh-water ice for airplanes
with wheels
16-b
STRENGTH AND USES OF FRESH AND SALT WATER ICE
Uses of Ice in the Arctic
The most important use of both sea ice and fresh-water ice in the Arctic,
in the present and in the past, has been in aid of transportation. The most
famous journeys of such explorers as Peary, MacMillan, and others were made
over frozen seas, on foot or with sledge and dog team. In many arctic areas
frozen rivers, lakes, and bays are regularly used for winter travel between
land points, frequently shortening distances by many times over those required
for overland travel. Ice makes possible the passage in winter of tractor–
drawn supply trains over water bodies, swamps, and bogs which in other times
of the year would be impassable.
Even railroads have been constructed on floating ice. The Russians have
been particularly persistent in developing this type of construction. A rail–
road line was, for example, constructed on the ice of Lake Ladoga in the siege
of Leningrad in World War II, supplementing motor truck supply routes also
laid out over ice. The value of ice in assisting military operations by pro–
viding relatively easy means of crossing rivers, lakes, swamps, and other water
areas is obviously very large in arctic areas.

EA-I. Hansen-Linell: Fresh & Salt Water Ice

The ice of the Arctic provides a surface to support floating weather
and scientific stations, a notable example of which was the Soviet Papanin
Expedition, which was established on the ice at the North Pole in May 1937,
and for nearly a year gradually drifted southward, until removed from the
ice at a point off the coast of East Greenland in February 1938.
Various possibilities for increasing the structural usefulness of ice
remain to be explored. A start was made in this direction during World War
II in an investigation carried out under the direction of Dr. C. J. Mackenzie,
President of the National Research Council of Canada (4). The object was to
devise a means of constructing floating airdromes of ice, which would save
critical materials required in other methods of creating landing facilities.
The successful prosecution of the war in Europe resulted in abandonment of
the project, but not before it had been found that by means of an admixture
of about 14% wood pulp, the compressive strength of ice could be increased to
1,100 p.s.i. and its tensile strength to 700 p.s.i. Other means for using ice
for various structural purposes have been suggested but have received little
or no actual trial.
Ice serves as a water supply source in the Arctic. In land areas, all
surface water sources will usually be completely frozen in the winter and the
only available source of water will be from the melting of ice or snow. In
salt-water areas, fresh water is obtainable by melting old sea ice; in summer,
melting of this ice forms pools of fresh water on the surface of the ice.

EA-I. Hansen-Linell: Fresh & Salt Water Ice

Strength Properties of Ice
General Characteristics of Ice . Ice may be described as a highly viscous
material having characteristics of a solid. Various observers have reported
that the range of stress magnitudes within which ice is truly elastic is quite
small. Although it fractures like a brittle material under breaking loads,
ice will flow or deform gradually under nonfailure loads, the rate of deforma–
tion under a given stress being higher at higher temperatures. Under favorable
conditions for “creep,” deformation may be quite rapid. Ice will also melt
under pressure; the effect is most pronounced with ice temperature at 32° F.
The pressure required to cause melting increases very rapidly with only small
drop in temperature below 32° F.
It will be apparent from these ice characteristics that it is not easy
to express the action of ice under stress by one or two simple figures. For
a material such as steel, working stresses may very easily and accurately be
established in relation to its yield point. For ice the yield point is some–
what obscure, and is apparently below practical working stresses; as yet, the
only really useful orientation point on the stress-strain curve is the failure
point. When a steel structure is designed, material of given characteristics
is specified with assurance that the actual steel used will vary an insigni–
ficant amount from the assumed properties. With ice, however, the material which
Nature has formed must be used, and its strength will vary considerably, in
accordance with the chance variations of the several strength-influencing factors
involved. Even cakes of artificial ice manufactured apparently identically
show wide spreads of results.

EA-I. Hansen-Linell: Fresh & Salt Water Ice

Published Test Results . Ranges of some ice-strength results are shown in
Table I. These results represent tests under a wide variety of test methods
and conditions on many types of ice (including artificial ice) and on speci–
mens of a wide range of shapes and sizes. Most of the figures are individual
maximum or minimum test values reported by the various investigators (1; 3;
5; 7; and original data). Except for strengths of single crystals, the minimum
Table I. Ranges of Ice Strength Test Results .
Type test Values reported, p.s.i.
Compressive strength 70-1,800
Tensile strength 36-223
Flexural strength 44-311
Shear strength 39-353
values should actually approach zero, corresponding to the condition when ice
is rotting or candled and is able to support virtually no load. Possibly the
compressive strength values show the greatest range, in part because more
investigators have used this relatively easy test than the other types.
Effect of Intracrystal Structure of Ice . When an ice sheet is formed on
the surface of a body of fresh water, the ice crystals form with their optic
axes vertical; that is, with their basal planes horizontal, and grow downward
with the added ice having its basal planes also [: ] parallel with the ice surface.
This influences the over-all strength since, as reported by McConnel from his
study of the way ice yields to stresses, the individual crystal behaves as if
it consists of an infinite number of indefinitely thin sheets of paper, normal

EA-I. Hansen-Linell: Fresh & Walt Water Ice.

to the optic axis, fastened together with [: ] some viscous substance which allows
one to slid e over the next with great difficulty. The sheets “offer no resistance
to bending, but utterly refuse to stretch except, of course, elastically” (9).
Thus, on the basis of the structure within the individual ice crystals themselves,
and apart from other structural characteristics, the properties of ice determined
from tests may be expected to vary according to the direction of orientation of
the applied stresses between the horizontal and vertical, relative to the
original surface of naturally formed ice.
Effect of Impurities in Ice . The presence of relatively small amounts of
dissolved or suspended matter in water usually results in ice that is weaker
and/or which deteriorates more readily under the effects of warmer temperatures
and sunlight. However, additions of considerable amounts , of such materials
as wood pulp, as noted previously, are reported capable of increasing strength,
and also may serve to slow up the melting (4).
In any water body found in nature, there will be found some dissolved
minerals and other impurities. Sea water at one extreme is an obvious
example, but even the clearest natural fresh-water body is not without impurities.
When each ice crystal forms, it tends to reject these impurities, which then
form a layer of concentrated impurities about the crystal. When adjoining
crystals meet, the impurities form a layer between the crystals of lower
melting point than the crystals themselves. This may greatly affect the physical
properties of the ice even though the amount of impurities is very small.
Melting will always be g in g at these boundaries between the ice crystals, and at
a lower temperature than 32° F. Fresh-water ice becomes rotten as thawing,

EA-I. Hansen-Linell: Fresh & Salt Water Ice.

starting at the inner crystal boundaries, separates the ice into separate
needles or columns. This ice is said to be “candled.” Thus, as a result
of this crystal structure in bulk, the physical properties of ice may vary
widely, especially when its temperature is near the melting point, depending
on the amount of impurities in the ice, the temperature, the age of the ice,
and the length of time that near-freezing temperatures have persisted.
In formation of sea ice, the effect of the salt content is, of course,
very pronounced. Freezing of the water portions does not begin until a
temperature of 28.6° F. is reached in undiluted sea water, and the structure
produced is porous, containing pockets of brine from which solid salt crystals
begin to precipitate when the ice cools to about 17° F. (8). The structure and,
consequently, the strength properties of sea ice improve with time, as the
concentrated brine drains and as the salinity of the ice becomes gradually
smaller. Newly formed salt-water ice of relatively high salinity is flexible
and elastic as compared to ice formed on fresh-water bodies (which is charac-
teristically brittle), and it does not candle. Salt-water ice more than one
year old is much tougher and stronger than young sea ice, but its surface is
much more likely to be rough. It is reported that when old sea ice becomes
sufficiently modified so that it approaches the composition of fresh-water ice,
it too will exhibit a tendency to candle.
From Soviet data, young sea ice requires about 1 2/3 times the thickness
of old sea ice to carry the same load, but this rule is very generalized. Lake
ice is usually assumed to be 2 to 3 times stronger than sea ice, although its
brittleness does not permit it to stand as much bending without cracking. River

EA-I. Hansen-Linell: Fresh & Salt Water Ice.

ice is generally not quite as strong as lake ice.
Any entrapped matter such as silt, clay, or plankton (2) tends to weaken
ice in the warmer periods because it absorbs sunlight and tends to cause
internal melting because of the heat produced thereby.
Ice containing much entrapped air tends to be weaker than clear essentially
air-free ice. Tests by the Soils, Foundations and Frost Effects Laboratory
of the New England Division, Corps of Engineers, in 1946-47, on 6 by 6 by [: ]
18-inch beams of artificial ice showed about 12% reduction in flexural strength
on beams cut from cloudy ice manufactured from ordinary tap water, as compared
with beams cut from clear air-free ice. Tests were performed at 10 ° to
13° F.
Effects of Horizontal Stratification in Ice . The strength of ice may be
affected by development of horizontal laminae of variable properties. Variations
in temperature, temporary thaws, rains, formation of snow-ice, flooding, and
driving by the wind of salt spray onto the surface of otherwise relatively
fresh ice or snow will all form horizontal layers in the ice sheet, the effect
of which may be important in relatively thin ice but will be insignificant in
heavy ice.
Another stratification development occurs in coastal areas where river
and stream inflows and tidal effects cause variations in the salinity of the
water during the ice formation period. An influx [: ] of relatively fresh water
under the ice surface will result in rapid freezing to this surface, which may
have cooled to as low as 28.6° F. when normal salt water was present under the
ice. The same effect occurs in the polar seas during the summer thaw periods

EA-I. Hansen-Linell: Fresh & Salt Water Ice.

when meltwater from the surface sinks down through holes and freezes onto the
cold undersurface of the ice pack. Such stratification in the lower part of
the ice sheet, where critical tensile stresses occur when the ice is loaded,
may be expected to have far greater effect on the over-all bearing capacity
than laminations which develop in the upper half of the sheet.
In the polar ice pack much less regular horizontal structure layering
occurs when pressure causes rafting and buckling of the ice sheet.
Effect of Vertical Cracks . Cracking of the ice sheet under the effect of
temperature stresses results in a pattern of vertical defects or discontinuities
distributed over the ice sheet. These do not ordinarily extend more than part .
way through the depth of the ice. Although they are conspicuous in the upper
surface of the sheet at low temperatures and may be readily visualized as a
possible cause of weakness, static-load tests by the Soils, Foundations and
Frost Effects Laboratory, in 1947, registered essentially no observable effect
of such cracks on the bearing capacity or failure action of the ice under a
load. Cracks formed due to the load showed very little tendency to follow the
existing partial cracks. However, these tests involved only single-load
applications, and it is reported that under heavy traffic, temperature cracks
do have an effect to reduce the load-bearing capacity of the ice sheet.
Cracks that pass all the way through the ice sheet seriously lower the load
capacity and are especially hazardous if they occur under a snow blanket, as
the snow may delay their refreezing and hide them from view.
Temperature during Ice Formation . When ice forms very rapidly at extremely
low temperatures there is less opportunity for dispersion of the impurities

EA-I. Hansen-Linell: Fresh & Salt Water Ice.

which are rejected by the ice crystals. Young sea ice formed at 14° F. has
entrapped about 5 parts per 1,000 salt, but that formed at −40° F. has 10 to
15 parts per 1,000. Differences in structure (and consequently of strength)
of the ice would be expected to result from such differences in actual volume
of included matter and the manner of its inclusion. It may be noted that the
more concentrated brine surrounding the freezing crystals during fast freezing
would tend to cause formation of the individual ice crystals at still lower
temperatures than in ordinary sea water, with possible effects upon the ice
properties.
Residual Stresses . Internal stresses in ice specimens due to unadjusted
temperature effects, to stresses introduced i [: ] obtaining and preparing test
specimens, or to external loadings of the ice sheet may well be in part re–
sponsible for variability of reported test results.
Effect of Temperature during Test . Temperature has a very pronounced
effect upon the strength properties of ice, as shown in Table II.

EA-I. Hansen-Linell: Fresh & Salt Water Ice.

Table II. Effect of Temperature on Strength of Fresh-Water Ice.
Temperature
of
test, °F.
Strength tests, p.s.i. Direction of load–
ing in relation to
crystal optic axes
Remarks Source
Compres–
sion
Flexure Shear
28 300 Perpendicular 3” to 5” square speci
mens of St. Lawrence Ri–
ver ice 5″ high; strees
increased at approximate
rates of 20 to 60 p.s.i./
sec.
(7)
14 693
2 797
Tests with load applied
parallel to crystal
axes have been deleted
28-30 156 Perpendicular 3″ wide × 2″ deep beams
of St. Lawrence River ice,
41″ span; loads applied
at 14″ from supports;
stress increased in in–
crements at various
rates averaging be–
tween 0.175 and 1.4
p.s.i./sec.
(7)
14-16 240
28-30 184 Parallel ( a )
14-16 214
25 196 Essentially
parallel ( a ) , ( b )
6″ × 6″ beams of air–
free artificial ice,
18″ span; third point
loading; stress in–
creased at constant
rates between 1.9 and
2.6 p.s.i./sec.
Soils, Fdns.
& Frost
Effects Lab.
13 203
−8 214
1 2

EA-I. Hansen-Linell: Fresh & Salt Water Ice.

Table II. Effect of Temperature on Strength of Fresh-Water Ice. ( contd. )
Temperature
of
test, °F.
Strength tests, p.s.i. Direction of load–
ing in relation to
crystal optic axes
Remarks Source
Compres–
sion
Flexure Shear
32 180 94 Parallel ( a ) Artificial ice flexure
specimens 2½″ to 6″ deep,
2½″ to 5″ wide, 15″ to
30″ long; shear specimens
2½″ sq. and 6″ long;
specimens failed in one
to three minutes
(15)
2 - 90
−9 256 -
−10 - 115
23 220 Parallel (nor–
mal to the
natural ice
surface)
5 cm. cubes of fresh –
water ice; lower platten
moved at constant rate
of 0.1 mm./sec.; yield
of upper load-measuring
platten unknown
(12)
14 252
5 354
−4 394
−13 488
−22 506
−31 584
−40 615
−49 665
−58 694
−67 725
−76 779
3

EA-I. Hansen-Linnel: Fresh & Salt Water Ice

Vitman and Shandrikov conclude from their test results, shown in Table II,
that the average compressive strength of ice increases more than four times
as the temperature decreases to −76° F. (12) . The data on change of flexural
strength with temperature also shows an increase with drop in temperature,
though not as rapid as for compressive strength. Finlayson (5) found that
temperature has very little if any effect on the strength of ice in shear,
which is in agreement with the results of Horeth and Wilson (15) given in
Table II. Under a given rate of increase of stress, the rate of deformation
of the specimen rises (while the ultimate strength drops) with an increase
in ice temperature; thus, ice, acting like a highly viscous material, flows
more readily at higher temperatures.
Effect of Rate of Loading on Strength Tests . Tests of ice specimens will
give different results depending on the rate at which the load is applied.
Numerous investigators have observed that ice will deform gradually under
a steady load. However, there appears to have been only little practical
investigation to determine the proper rates of load application for use in
correlating test values with the load-supporting capacity of ice.
From flexure tests reported by Brown in 1926, it was concluded that the
modulus of rupture did not vary much at the different loading rates (7).
However, when his results at 14 to 16° F. are plotted together with a number
of flexure tests at 10° F. performed by the Soils, Fdoundations and Frost
Effects Laboratory in 1946-47, covering a wider range of rates of loading,
up to 7.65 p.s.i. per second, there is shown an apparent marked loss in
ultimate strength with increase in the rate of application of stress. It
is reasoned that if the indicated trend should be correct, then possibly the

EA-I. Hansen-Linell: Fresh & Salt Water Ice

lower rates of stress increase permit plastic flow to occur at points [: ] of
initial stress concentration, thus allowing adjustment of stress distribution
and higher breaking loads. It must be expected that the critical range
of loading rates, over which maximum changes of results occur due to varying
degrees of stress adjustment within the specimen, will be strongly dependent
on temperature. Finlayson found in shear tests on ice, that at temperatures
below 0° F., greater values were obtained when the load was applied rapidly (5).
Effect of Orientation of Test Specimen . Since the bending stresses, which
act in an ice sheet under load and which control in part the bearing capacity
of the ice (the other factor being buoyancy), act in a direction parallel
to the ice surface, analyses of bearing capacity should, theoretically at
least, depend on tests in which specimens are stressed with their crystal
axes oriented correspondingly. Practically, this offers difficulties,
and if the bearing capacity of thick ice is to be determined from tests on
cores recovered from borings into the ice, it will probably be necessary to
establish an empirical relationship between strengths parallel to and at
right angles to the surface of the ice sheet.
Size and Proportions of Test Specimens . As in testing concrete or
soils, the size and proportions of test specimens may be expected to influence
the results of tests on ice. Certainly a test on a small portion of a single
“needle” would not give the same results as a test on a collection of full–
length , “needles.” What the minimum desirable size of specimen should be
would be expected to vary with the type and condition of ice. However, there
seems to have been little productive study given to this phase of ice testing
as yet.

EA-I. Hansen-Linell: Fresh & Salt Water Ice

Load-Bearing Capacity of Ice
Records of Actual Ice Loadings . Records of the innumerable, and in
many areas regular, actual ice crossings which have been made with motor
vehicles, tractors, sleds, railroad trains, etc., appear to be a rather
meager source of data on the supporting capacity of ice for several reasons.
First, when such crossings have been made, there appears to have been only
rarely any attempt to record the actual conditions of the crossing, such as
ice thickness and quality, weight of machine, temperature, size and shape
of loaded area, and presence of snow cover. Second, when such observations
have been made, they have been incomplete since all the factors involved
may not have been recognized. Third, there seems to have been little
interest in making such records generally available in publications.
Empirical rules have been developed on the basis of experience with crossings,
but it is difficult to trace these back to original records.
Historical Military Rules . There are a number of old rules relating to
the load-bearing capacity of ice which mostly have their origin in military
engineering and are the result of observation and experience in the movement
of troops, equipment, and supplies across frozen bodies of fresh water.
The data in Table III are a modernized and improved version of these rules
(11). The rules should be considered to apply to fresh - water ice of good
color and soundness, away from unsupported edges. The thickness measure–
ments should disregard any soft or poor quality ice on the top or bottom
of the ice sheet.

EA-I. Hansen-Linell: Fresh & Salt Water Ice

Table III. Load Capacity of Ice.
Load Thickness
of ice, in.
Minimum
interval,
ft.
Single rifleman on skis or snowshoes 1.5 16
Infantry in single file, 2-pace distance ( a ) 3 23
Infantry columns , ( a ) , single horses, motorcycles,
unloaded sleds, or motor toboggans
4 33
Single light artillery place, ¼-ton truck, 4 × 4 6 49
Light artillery, passenger cars, medium 1½-ton
trucks with total load of 3½ tons
8 65
2½-ton trucks, light loads 10 82
Closed columns of all arms except armored force
and heavy artillery ( a )
12 98
Armored scout cars, light tanks 14 115
20-ton vehicles 16 131
45-ton vehicles 24 164
4
Moskatov’s Empirical Method . Moskatov (10) has given a procedure for
determining required ice thicknesses for aircraft on skis which he calls the
“Method of Analogy.” Formulas derived by this method are given in Table IV.
The basic formulas for aircraft on skis have been adapted to the case of
wheels by use of a factor of 20%, which is used by the Russians. P is the
gross weight of aircraft in tons and t is the ice thickness in inches. The

EA-I. Hansen-Linell: Fresh & Salt Water Ice

formulas are said to apply only for ice formed and maintained at a tempera–
ture below 16° F. At higher temperatures 25% greater thickness is required.
Table IV. Minimum Required Ice Thickness in Inches.
Type of ice Aircraft on skis Aircraft on wheels
Lake ice <formula>t = (27/8)√(P)</formula> <formula> t = (81/20)√(P)</formula>
River ice <formula> t = (15/4)√(P)</formula> <formula> t = (9/2)√(P)</formula>
Old sea ice <formula>t = (27/4)√(P)</formula> <formula> t = (81/10)√(P)</formula>
Young sea ice <formula>t = (81/8)√(P)</formula> <formula>t = (243/20)√(P)</formula>
The formulas given in Table IV are simple and they indicate that the
carrying capacity of the ice varies as the square of the ice thickness;
this is confirmed approximately by elastic theory analysis. Design curves
from these formulas are shown on Figures 1 and 2, where they are designated
as “Russian (Moskatov) Empirical Analysis.”
Elastic Theory Analysis . If a stationary load is placed upon the surface
of a floating ice sheet which extends out to a large distance on all sides
from the load, a deformation of the ice sheet occurs which is a maximum under
the loaded area and which extends outward in all directions, decreasing in
amplitude with distance from the center. When the deformation occurs, a
buoyancy force is caused to act upward on the bottom of the ice sheet. This
force is at any point directly proportional to the deflection at that p o int
as long as the ice does not become submerged, with water flooding on top. Fig. 1.
BEARING CAPACITY OF OLD SEA ICE
FOR
AIRPLANES WITH WHEELS
JANUARY-1949 Fig. 2.
BEARING CAPACITY OF
FRESH-WATER ICE FOR
AIRPLANES WITH WHEELS
JANUARY-1949

EA-I. Hansen-Linell: Fresh & Salt Water Ice

At the same time, the bending strength within the ice sheet itself is
mobilized. The buoyancy and bending resistance forces combine to provide
the total supporting force for the applied load. (When, as may occur along
a shore line, the water drops away from under the ice leaving it suspended,
the ice sheet will have at that location greatly reduced load-supporting
capacity.)
Since under static loading the buoyant force of the underlying water
at any point acts in exact proportion to the amount of deflection of the ice
sheet at that point, the latter may be assumed to have a perfectly elastic
support. If we may also assume that the ice sheet has properties approximating
those of an elastic slab, then the theories of elasticity may be applied to
compute stresses and deformations in the ice sheet and to determine thicknesses
required for given conditions of loading.
As pointed out by Volkov, static loading of floating ice is more critical
than impact loading. Under an impact force, the ice is simply compressed
between the applied load [: ] [: ] and the underlying water; in order to develop
bending stresses, a definite volume of water must be displaced from under the
ice, near the area of load application, and under a sharply applied load the
time required for this movement to occur is insufficient. Shearing action
is usually not critical. Experience has shown that thin ice may be crossed
by a moving load when a stationary load would break through and also that
thin ice may sustain a stationary load for some time, then give away. Thus,
load-capacity studies may be based upon the static loading condition, except
as it may be necessary to apply modifications for wave action under moving
loads and for fatigue under repeated loadings.

EA-I. Hansen-Linell: Fresh & Salt Water Ice

For the condition of static loading, a comprehensive theoretical
analysis of the bearing capacity of floating ice has been developed by
Mr. H. F. Shea of the Soils, Foundations and Frost Effects Laboratory by
adapting to the case of an ice sheet formulas developed by Dr. H. M. Wester–
gaard for determining the stresses at the interior portion of an elastically
supported slab (14). In their general form, the formulas are not adapted
to giving quick answers since results must be computed. However, the
formulas permit analysis of a variety of combinations of loading patterns
and intensities which could not otherwise be approached.
For uniformly loaded circular areas, these formulas are as follows:
(1) Inside the area of loading:
<formula>fx and fy = (0.6517P/t2)[1.6713 + (3/4) loget – logeB + (0.012B2/t3/2)] + (0.1516P/t2)[(r2/450B2)(cos2Θ)]</formula>

EA-I. Hansen-Linell: Fresh & Salt Water Ice

(2) Outside the area of loading:
<formula>fx and fy = (0.6517P/t2)[1.6713 + (3/4) loget – logeB + (0.012B2/t3/2)] + (0.1516P/t2)[(r2/450B2)(cos2Θ)] )]</formula> <formula>c = √(1.6d2 + t2) – 0.675t when d < 1.724t</formula> <formula>c = d when d > 1.724t</formula>
For uniformly loaded elliptical areas (or where a uniformly loaded
ellpitical area may be reasonably assumed), the following stress equations
apply:
(1) Inside the area:
<formula>fx and fy = (0.6517P/t2)[5.0372 + (3/4) loget – loge (a + b) – x2/(a(a + b)) – y2/(b(a + b))] ± (0.1516P/t2)[((ba)/(a + b)) – (2ab/(a + b2)) ((x2/a2) – (y2/b2))]</formula>

EA-I. Hansen-Linell: Fresh & Salt Water Ice

(2) Outside the ellipse, for, y = o, x = a ,
<formula>fx and fy = (0.6517P/t2)[5.0372 + (3/4) loget – loge (x + √(x2 - a2 b2)) – x/(x + √(x2a2 [suspect missing symbol here] b2)] ± (0.1516P/t2) [((a2 + b2)/(b2a2)) – 4b2x/((b2a2) (x + √(x2 + a2b2)]</formula>
(3) Outside the ellipse, for, x = o, y = b ,
<formula>fx and fy = (0.6517P/t2)[5.0372 + (3/4) loget – loge (y + √(y2 + a2 - b2)) – y/(y + √(y2 + a2b2))] ± (0.1516P/t2) [((a2 + b2)/(b2a2) – 4a2y/((b2a2) (y + √(y2 + a2b2))]</formula>

EA-I. Hansen-Linell: Fresh & Salt Water Ice

In these formulas the notation is as follows:
f x , f y = bending stresses in p.s.i. at surface of ice at points
x , y on coordinate axes
P = total applied load in pounds
t = thickness of ice in inches
e = base of natural logarithms = 2.7128
B = ratio of radius of circular area to radius of standard
bearing plate (15 inches) = c/15
r = radial distance from center of circular area to point
of stress investigation in inches .
θ = angle between x axis and line drawn between point of
stress investigation and center of circle
a = minor radius of elliptical area in inches
b = major radius of elliptical area in inches
c = substitute radius of loaded area in inches
d = actual radius of loaded area in inches
a and b are further identified by the equation for an ellipse, <formula>(x2/a2) + (y2/b2) = 1</formula>
For example, the stresses induced by a 60,000-1b. load transmitted
to ice 50 in. thick by a B-29 main dual wheel unit, assuming load is dis–
tributed equally over two circular areas, was determined.
( 1 ) Given the following data:
Area (for one circular area) = 370 sq.in.
Radius of circle, d = 10.9 in.
Distance between wheel centers = 37.3 in.
<formula>c = √(1.6d2 + t2) – 0.675t</formula> <formula> P = 30,000 lb.</formula> <formula> t = 50 in.</formula>
P = 30,000 lb.
t = 50 in.

EA-I. Hansen-Linell: Fresh & Salt Water Ice

(2) Using formulas (1) and (2) for the case of the circular areas:
f = 37 p.s.i. when x = 0 and y = 0
f = 26 p.s.i. when x = 37 p.s.i. and y = 0
total f = 37 + 26 = 63 p.s.i.
In the above formulas, the following constants were assumed:
Poisson’s ratio for ice = 0.365
Modulus of elasticity = 1,294,000 p.s.i.
Observed values of Poisson’s ratio are reported variously from 0.25 to [: 0138 ]
0.38 and the value adopted above for the computations is, therefore, near the
higher limit. Stress-strain moduli have been reported by various investigators
from under 70,000 to over 1,500,000 p.s.i. Recent flexure tests by the Soils,
Foundation and Frost Effects Laboratory showed a tangent modulus value of
610,000 p.s.i. on a single specimen of artificial ice at 10° F. up to a
strees of about 25 p.s.i.; near the ultimate failure stress of 212 p.s.i., the
secant modulus (related to the origin of the stress-strain curve) was approxi–
mately 190,000 p.s.i. A value of approximately 400,000 p.s.i. was also
determined from the deflection of a naturally formed fresh - water ice sheet
8.4 in. thick in a 27-ft. tank when loaded at approximately 10° F. Brown has
also presented considerable data on effects of temperature, and of rate and
range of loading, on stress-strain moduli (7), which indicate generally lower
values than the one adopted above for computation of the equations. Reduction
of assumed value for either Poisson’s ratio or modulus of elasticity in the
equations will result in an indicated increase in load capacity. Thus, the
formulas, as given above, are on the conservative side with respect to these
basic values.

EA-I. Hansen-Linell; Fresh & Salt Water Ice

Once the thickness of ice required for a given load, size of area, and
factor of safety has been determined, the thickness required for a different
load on the same loaded area and with the same factor of safety may be
computed from the relation that the thickness required varies directly with
the square root of the load.
Preliminary Elastic Theory Ice Thickness Curves . Curves derived from
the elastic theory equations as presented are shown on Figures 1 and 2 for
aircraft with wheels and designated thereon by the phrase “Elastic Theory
Analysis.”
The curves of Figures 1 and 2 have been computed with the following ice
strengths, assumed as conservative values for ice under the temperature condi–
tions noted on the figures:
Hard, sound lake ice = 150 p.s.i.
Hard, sound river ice = 125 p.s.i.
Old sea ice (more than 1 year old) = 75 p.s.i.
Similar curves could be computed for other types of loadings, such as for
tractors, sleds, aircraft with skis, and even structure footings. Figures 1
and 2 show results of a few actual loadings by planes and in tests.
Comments on Elastic Theory Method and Thickness Curves . Although as yet
not complete k ly proved, the elastic theory equations which have been given
are considered the most accurate theoretical approach available. The constants
used in the equations are subject to considerable modification as more is
learned about the properties of ice. It may be that eventually more than one
set of constants will have to be used.

EA I. Hansen-Linell: Fresh & Salt Water Ice

For theoretical analysis purposes, the ice sheet has conventionally been
assumed to be in equilibrium under external forces and unstressed. This is
likely to be far from the truth in the upper part of the slab. The stress
pattern in the upper part of an unloaded ice sheet due to temperature and
lateral forces must be very complex. That the stresses may be high is shown
by the cracking which occurs due to temperature alone. Fortunately, however,
elastic theory analysis shows that under an applied vertical load, the critical
stresses act in the lower portion of the slab where temperature stresses
are at a minimum.
It should be emphasized that the curves apply at the interior of an ice
sheet; near the edge of a crack or open water, the load-supporting capacity
can be expected to be much less.
It is also emphasized that the curves plotted on Figures 1 and 2 from the
elastic theory are preliminary, and are necessarily over-simplified. It is
obviously not possible to [: ] adopt one strength value for fresh - water ice
and one for salt - water ice and expect that all ice in all latitudes and
locations will conform to either one or the other of those standards under
all the wide variety of conditions of salinity, temperature, and other factors
which will be encountered. Rather, the curves are guides to the approximate
answer, based on conservative strength assumptions. From the small experience
acquired to date, it is believed that the curves for factor of safety of 2.0
represent very safe thickness values under the limitations stated on the
curve sheets. Below the latter curves is a “zone of uncertainty,” extending to
below the curves from the Russian (Moskatov) Empirical Analysis. It is this

EA-I. Hansen-Linell: Fresh & Salt Water Ice

zone of uncertainty which future studies of the bearing capacity of ice
sheets should explore. Eventually, there will probably be not just two but
a considerable number of curve sheets to cover the many possible combinations
of ice strength, temperature, salinity, and other variables. With such curves
and with the results of quick field tests of the characteristics of the ice
in question, using test apparatus of types now beginning to be developed,
accurate numerical evaluations of the load-bearing capacity of ice should
be possible. However, for the present, prudent procedure in critical cases
is to weigh together the empirical rules and formulas, the elastic theory
equations, and the promptings of judgment and experience.
Effects of Moving Loads . Russian investigators have observed the
possibility of danger from development of resonance in waves created in the
ice sheet under moving loads and have actually encountered conditions in which
several machines in motion and passing each other in both directions set up
such wave action that the ice broke (6). They have reported that such occur–
rence becomes especially dangerous when the machines are traveling at speeds
corresponding to those of the propagation of the ice waves and that in such
a case a single vehicle can produce sufficiently strong resonant vibrations
to cause it to break. The Russians have prepared rules for speed and spacing
of vehicles to avoid this effect. On ice about 24 inches thick, over water
of the order of 16 to 33 feet deep, the Russians have found that at speeds
between 3 and 9 miles per hour a depression of the ice under the vehicle moves
at the same speed as the vehicle. At speeds above about 12 miles per hour,
wave vibrations are formed which spread far to the sides of the path, at speeds
which on this lake were found to be 19 to 25 miles per hour. They concluded
that on this lake initial speeds of about either 15 or 25 miles per hour should

EA-I. Hansen-Linell: Fresh & Salt Water Ice

be maintained and that speeds intermediate between these rates should be
avoided. Overtaking of one machine by another should also be avoided, and
spacings between vehicles varying from 200 feet to 650 feet and more were
recommended, depending on the speed of movement and the strength of the ice.
Particular caution was advised near shores where complex wave effects may
occur.
Repetition of loading under extensive operations causes ice to become
weakened. It is reported that one reason for this is the progressive
deepening of temperature cracks under the flexing action of heavy traffic.
Under such conditions, a desirable procedure is to provide sufficient area,
or alternate routes, so as to permit frequent changes in traffic lanes.
It is also reported that ice will break up much more readily under
vehicles with low-speed motors than under those with high-speed motors,
because of the differences in frequency and intensity of vibrations trans–
mitted to the ice. Traveling parallel to major cracks must be avoided.
Methods of Increasing Supporting Capacity of Ice. The bearing value of
an ice sheet may be increased by: (1) thickening the ice sheet as a struc–
tural member, and (2) distributing the load over a wider area of application.
The ice thickness may be increased on the top by flooding the surface
in shallow layers, allowing each to freeze completely before adding the
following one, or the increase may be obtained naturally on the bottom of
the ice layer, over a longer period, by either keeping snow cleared entirely
from the surface or by packing the snow cover down solidly so as to increase
the rate of heat loss from the underlying water and thereby speed freezing.
Because even tightly compacted snow has insulating qualities, rate of thickening

EA-I. Hansen-Linell: Fresh & Salt Water Ice

will be higher if snow is completely removed. If a heavy thickness of snow
covers the ice, it may sometimes be possible to assist the flooding procedure
by boring holes through the ice at intervals, the weight of the snow being
sufficient to submerge the ice and saturate the snow. However, flooding of
snow must carry through at least to the top of the snow cover, otherwise
the insulating effect of the unwet upper zone of the snow will permit the
underlying water-snow mixture to remain as slush for a long period without
freezing.
Although superficial wetting down and/or compaction of the snow cover
sufficiently to produce a hard vehicle surface may result in a higher bearing
capacity of the ice sheet, it is likely that this results mostly from the
load distributing characteristics of the snow layer and the increase in ice
thickness which develops from the increased thermal conductivity of the layer.
The load may also be distributed over a wider area by means of brush,
straw, plank, logs, or ice blocks laid on the ice, which may in turn be
covered with a layer of compacted snow to provide smooth travel and possibly
additional load distribution effect. Care must be taken to periodically
check ice thickness when such procedures are used, as the ice may melt on its
undersurface as layers are added on the upper surface, if the original ice
thickness has been in balance with the existing temperature conditions, if
the added materials have insulating qualities, or if currents are present.
At the weak zone which exists where ice is not frozen to the shore, or
at open cracks, adequate bearing capacity is usually best obtained by bridging
over with timber or other structural materials (13).

EA-I. Hansen-Linell: Fresh & Salt Water Ice

BIBLIOGRAPHY

1. Barnes, H.T. Ice Engineering . Montreal, Renouf, 1928.

2. Boorke, A. Morskie Ldy . (Sea Ice.) Moscow, Leningrad, Izdatelatvo
Glavsevmorputi, 1940. Polar Workers Library . Translation by E.
Olkhine in The Stefaneson Library, New York City.

3. Dorsey, N.E., comp. Properties of Ordinary Water Substance . N.Y., Reinhold,
1940. American Chemical Society. Monograph Series.

4. Finlayson, J.N. “Icebergs as ships,” Engineer , Lond. June 7, 1946,
pp.517-18.

5. ----. “Tests on the shearing strength of ice,” Canad.Engr . vol.53, no.1,
pp.101-103, July 5, 1927.

6. Ivanov, K. Ye., Kobeko, P.P., and Shulman, A.R. Deformatsiia Ledovogo
Pokrova pri Dvizhenii Gruzov . (Deformation of an Ice Cover under
Moving Loads.) Leningrad, Akademiia Nauk, S.S.S.R., Fiziko–
Tekhnicheskii Institutt, 25 July 1945.

7. Joint Board of Engineers on St. Lawrence Waterway Project. Report...November
16, 1926. Ottawa, Acland, 1927, Appendix F.

8. Malmgren, Finn. On the Properties of Sea Ice . Bergen, Grieg, 1927. Maud
Expedition, 1918-1925. Scientific Results , vol.1, no.5.

9. McConnel, J.C. “On the plasticity of an ice crystal,” Roy.Soc.Lond., Proc .
vol.49, pp. 323-43, Mar. 12, 1891.

10. Moskatov, K.A. “On posadke samoletov na led.” (Airplane landings on ice.)
Leningrad. Arktichaskii Nauchnyi-Issled.Inst. Trudy . T.110, pt.1,
art.5, pp.43-55. In Russian with English summary. Translated and
edited by Headquarters, U.S. Air Forces.

11. U.S. War Department. Operations in Snow and Extreme Cold . Wash.,D.C.,
G.P.O., 1944. Basic Field Manual FM 70-15.

12. Vitman, F.F., and Shandrikov, N. P . “Nekotorye issledovaniia mekhanicheskoi
prochnosti lda.” (A study of the mechanical strength of ice.)
Leningrad. Arkticheskii Nauchnyi-Issled.Inst. Trudy , T.110, pt.1,
art.7, pp.83-100. Translated and edited by Headquarters, U.S. Air
Forces.

13. Westergaard, H.M. Expedient Snow and Ice Roads . Wash.,D.C., G.P.O.,
26 September 1944. U.S. Army. Technical Bulletin TB Eng . 42.

EA-I. Hansen-Linell: Fresh & Salt Water Ice

14. ----. “Stress concentrations in plates loaded over small areas,”
Amer.Soc.Civ.Engrs. Trans . vol.69, no.8, pt.2, pp.831-86, 1943.

15. Wilson, J.T., and Horeth, J.M. “Bending and shear tests on lake ice,”
Amer.Geophys.Un. Trans . vol.29, no.6, pp.909-12, Dec., 1948.

Ralph Hansen and Kenneth Linell

Construction and Maintenance of Airfields in the Far North Regions

EA-I: (James D. Lang)

CONSTRUCTION AND MAINTENANCE OF AIRFIELDS
IN THE FAR NORTH REGIONS

CONTENTS
Page
Introduction 1
Types of Airfields 3
Reconnaissance 6
Planning and Design 8
Construction 12
Grading 17
Preparation of the Subgrade 18
Paving 20
Snow Removal and Maintenance 20

[EA-I: James D. Lang]

CONSTRUCTION AND MAINTENANCE OF AIRFIELDS
IN THE FAR NORTH REGIONS
INTRODUCTION
The engineering aspects of the subjects of airfields, pavements, roads,
and drainage as they apply to the arctic and subarctic regions readily lend
themselves to treatment as a single discussion. While the actual construction
of these items follows the usual pattern normally employed in more temperate
regions, the engineering aspects do not. Seldom, if ever, does engineering
play such an important part as it does in the Arctic in the conception and
execution of any construction program. For airfield construction, in parti–
cular, never is proper engineering as important as when such construction is
contemplated for the Arctic.
Assuming that for strategic, economic, navigational, or other reasons an
airfield must be built in a given area, flying characteristics of the site will
usually be the deciding factor in most areas of the world. In the Arctic, on
the other hand, more frequently engineering factors determine which one of
several sites will be chosen. Of course, both factors must be considered and
the best possible compromise be chosen, consistent with the mission.

EA-I. Lang: Construction and Maintenance of Airfields

Let us examine why engineering becomes so important. Arctic areas are
remote, relatively inaccessible, and devoid of facilities with which to per–
form any construction. The construction season is so short that any mistakes
or omissions are disastrous and may well result in the probable loss of one
full year in completion for each major error. The severity of the climate
creates problems of cold, wind, fog, snow, ice, permafrost, ice fields, and
other phenomena for which special measures must be taken. Any one or more of
the above conditions creates a requirement for extreme care in selecting and
examining sites and planning for the construction in minute detail. A firm
mental picture must be had of each and every move before any steps are taken,
otherwise valuable effort, time, and money will be lost, not only in determin–
ing the next measure to be taken, but probably in undoing false steps which
already may have been taken. Presolving problems and planning in detail so
as to permit completion of arctic construction at a minimum of cost in time,
[: ] effort, and expense involves engineering of the highest caliber. It
cannot be overemphasized that construction such as is discussed herein involves
an engineering problem requiring the utmost skill in planning and execution
and that it is best to make haste slowly.
Elsewhere in this encyclopedia are discussions of various natural phenomena
encountered in the Arctic. Before proceeding with this article it might be well
for the reader to familiarize himself with the general subject of permafrost in
order to understand what it is and how it behaves, also with the subjects of
surface and subsurface drainage. The engineer in the Arctic cannot undertake
any construction project involving excavation, drilling, pile driving, or even
just disturbing the earth’s surface without encountering either frost or

EA-I. Lang: Construction & Maintenance of Airfields

drainage problems. Briefly, permafrost is a condition of eternally frozen
soil as distinct from eternally frozen icecaps and glaciers or from soil
subject to seasonal frost. Permafrost is beneath the surface and may exist
either as a continuous layer, or in lenses and pockets. How far the perma–
frost table may be from the surface varies from a few inches to many feet and
depends to a great extent upon the type of surface material above it, the
classification of the soil, and the subsurface drainage pattern. The ground
above the permafrost table is known as the active zone since it alternately
freezes and thaws, however, not necessarily to the same degree each year.
Ground water may exist below, within or above permafrost, and may be either
under pressure or free flowing, depending upon whether there are nearby high
elevations to provide the hydrostatic head or whether plains or areas of weak–
ness exist to relieve the pressure. It is obvious that during thousands of
years of direct exposure to extreme cold, nature has set up a balance, or
thermal regime, which exists within narrow limits. Any disturbance or new
outside influence will upset that regime. Construction activity provides the
influence to disturb e or upset nature’s balance. It is the engineer s problem
either to avoid disturbing the balance or to devise ways and means of diverting
the disturbed condition to his own advantage.
TYPES OF AIRFIELDS
For military operations in the Arctic, three general types of airfields may
be involved. These are ice fields, light airfields, and heavy airfields , . Ice
fields, as the term implies, involves construction of a temporary nature and,
of necessity, such construction is employed when hasty measures and rapid

EA-I. Lang: Construction & Maintenance of Airfields

construction are required. In certain areas, of course, such as on an icecap,
these are the only fields possible. The principal requirements of the ice
field site selection are that the site must be able to support the mission for
which the field is being built; that the site must possess adequate flying
characteristics as to glide angles, wind direction, and grades; that the site
must be logistically supportable for the unit expected to operate therefrom;
that the terrain must be relatively level so as to minimize the amount of sub–
grading preparation; and that adequate quantities of loose snow must be avail–
able in order to provide surface materials for fine grading and compaction.
If a site can be selected which generally meets the foregoing criteria, the
problem of preparing the field becomes relatively simple. Methods of grading
and compaction of snow and ice utilize standard construction equipment and
techniques and present a problem quite similar to that encountered in grading
and compacting silty soils. An important element in selection of an ice field,
which has not heretofore been mentioned, is the selection of a site possessing
a suitable alternate landing area such that it may be developed into a permanent
field during the ensuing summer construction season should circumstances require
that flight operations be maintained for a prolonged period of time from the
general area. The ideal solution would be to select a location where this
permanent airfield could be built to utilize such structures and base facilities
as have been provided for the snow and ice field without the necessity of
extensive relocation of facilities or travel of unnecessary great distances
between the permanent airfield site and the base facilities.
The second type of airfield, the light airfield, is one which again must
be selected and constructed usually in a minimum of time. This type of field

EA-I. Lang: Construction & Maintenance of Airfields

involves the element of site selection to meet the requirements of mission,
flight characteristics, and logistic support. It also must meet the require–
men e t of enabling rapid construction of a base which can be used for year-round
operations. In general, the construction of a light airfield involves the
selection of a suitable flight area on which an airfield may be constructed,
utilizing nearby aggregate sources of free draining material, and spreading
and compacting these materials upon the undisturbed natur e al surface of the
ground in such a manner as to avoid cuts. This type of field requires rela–
tively short and narrow runways, and can be completed with conventional methods
and equipment with little added difficulty over that which could be expected in
a more temperate climate.
The third type of airfield, and the one which presents the greatest prob–
lem, is the heavy airfield. This installation requires a great deal of [: ]
equipment, time, and effort and should be built only after the greatest delibera–
tion and consideration of all factors involved. The site selected must, of
course, meet the requirements of the mission to be performed and the flight
characteristics of the aircraft to be employed therefrom. These criteria
are of secondary importance, however, to possession of desirable engineering
characteristics. This is not to say that the mission and flight characteristics
can be neglected, but it is intended to mean that fields of lesser desirability
from the standpoint of these two characteristics may be selected if the
engineering character i s tics are otherwise more advantageous.
Selection of any of the three types of fields discussed above naturally
rests upon proper reconnaissance of all available sites within the general
area from which the mission can be performed.

EA-I. L o a ng: Construction & Maintenance of Airfields

RECONNAISSANCE
The initial airfield reconnaissance can best be performed by air.
However, prior to such a move, a careful map study should be made to select
the general areas from which the mission can be performed and to eliminate
from those areas specific regions where mountainous terrain, lakes, sinuous
rivers, or other geographic factors obviously preclude consideration. A
systematic flight pattern should be set up and a suitable slow-flying
aircraft having good facilities for observation of the ground should be
employed. Experience in arctic areas has proved that observation of surface
condition will provide several excellent clues to the nature of subsurface
conditions which may be encountered. Areas covered by hummocky tundra, commonly
known as “nigger heads,” indicate the existence of a permafrost table near
the surface, poor drainage, and a general water-logged condition. This condi–
tion is true even where the existence of plateaus with sharply defined hill–
sides or terraces would otherwise indicate that there should be adequate
drainage from the area. The existence of a growth of birch trees indicates
thawed ground, i.e., an area which does not freeze between the permafrost and
the surface active zone. The existence of spruce trees indicates a permafrost
table relatively near the surface of the ground and poor drainage. On the
other hand, the occurrence of pine and fir trees indicates a well-drained
granular soil which in turn means that water flowing through this pervious
layer will have had a tendency to melt and lower the permafrost table. Such
sites are worthy of close investigation by ground parties and are indicative
of the most feasible airfield site. Areas covered by willows generally are
unsuitable since willows indicate the existence of ground water and a site

EA-I. Lang: Construction & Maintenance of Airfields

likely to be troubled with surface ice during the winter months. Areas covered
by tree growth, which is irregular, and containing individual trees inclined
at odd angles indicate ground susceptible to swelling and frost mounds. Areas
devoid of trees and shrubs and which are covered by tundra and moss indicate
a high permafrost table, particularly such areas that have a “crackled” appear–
ance [: ] or the octagonal designs characteristic of a high permafrost table.
Exposures of rock outcroppings indicate areas where the seepage of water
in the planes next to the rock is apt to cause a melting of the permafrost
and the development of subsurface channels of flowing water, which may make
construction exceedingly troublesome.
Having selected a number of possible sites from the air, arrangements
should be made for further investigation of [: ] these areas by ground
parties. Preferably a base of operations should be selected for the ground
parties from which several of the sites may readily be investigated. The ground
parties should consist of individuals who are experienced in the Arctic.
Parties should be equipped with normal survey instruments for complete and
detailed topographic surveys of the sites in question, also with light drilling
equipment and excavation tools for the digging of numerous test holes and
test pits, and also meteorologic equipment to initiate weather observations
throughout the entire period the party is in the area. It is extremely impor–
tant that a large number of test holes and test pits be dug to determine the
depth of the permafrost table; the characteristics of subsurface soils; the
existence, depth, temperature, volume, and direction of the flow of water; and
the nature of the natural insulating material which is found on the surface.

EA-I. Lang: Construction & Maintenance of Airfields

Detailed notations should be made of all natural phenomena, carefully logged
and plotted on an area sketch. Test holes and test pits should be care–
fully back-filled with material as nearly similar to the original material
in order to avoid disturbance of natural conditions. It is important that
insulating materials be left on the surface so as not to induce subsequent
thawing in the area investigated. When conditions are encountered which
indicate the unsuitability of a given site for airfield construction, a care–
ful verification should be made and then further efforts suspended on that
particular site in order to conserve time for additional observations and
tests at other sites which have not been definitely eliminated. At best, the
time available for reconnaissance is limited so that as rapidly as site
selection can be narrowed to one particular location it should be done in
order that the maximum information may be developed on the final site prior
to committing the construction forces.
PLANNING AND DESIGN
Following selection of a site, planning the construction operation must
be thoroughly, completely, and exhaustively pursued. Every scrap of engineer–
ing data accumulated during the reconnaissance must be reanalyzed and evaluated.
Maps, sketches, geological and geophysical reports, verbal reports, photographs,
long-range weather observations, ice tables, communication routes, available
equipment (both qualitatively and quantitatively), manning tables, shipping
capabilities, and the available time must be carefully reviewed until a firm
visualization of the construction problem is grasped with confidence. The
planner must then carefully fit these bits and pieces together, matching his

EA-I. Lang: Construction & Maintenance of Airfields

capabilities against his minimum requirements. One of the major considerations
he must face is the provision of adequate living accommodations and opera–
tional facilities for the construction force. Construction of these items
is a major operation in itself and is a heavy drain on manpower and resources.
It follows that not one square foot of building construction should be pro–
vided, which is not absolutely necessary, and not one man should be permitted
on the construction site who is not fully qualified in his field and whose
full-time presence is not needed. No equipment should be brought to the site
unless the exact need for that equipment is anticipated beforehand.
A second major consideration to be given during the planning phase is
that of establishing minimum requirements. It is a constant temptation to
add to the scope of any project. The planner must resist these tendencies
and enlist every means at his disposal to minimize and simplify his problem.
It is better to complete a spart o a n-like installation on time than it is to
half complete an ideal plan. Time, manpower, and mat e é riel devoted to a non–
essential may mean failure to complete one element of the over-all project;
such failure might mean the loss of an entire year in obtaining final use of
the installation.
To return to some of the engineering aspects of the planning, it is
essential that standard plans be selected or engineering designs be prepared
in advance of construction operations. This is where the data accumulated
during the reconnaissance and study phase of the project can be invaluable.
The drainage pattern, both surface and subsurface, is the key to the design
of the airfield. The landing strip, taxiways, parking aprons, and roads must

EA-I. Lang: Construction a Maintenance of Airfields

be located and designed so as to create the least possible disturbance to
the natural drainage pattern. Where disturbance cannot be avoided, then a
solution must be adopted which will take advantage of, rather than attempt
to defy, the natural arctic phenomena known to exist. Deep and narrow drain–
age ditches must be planned to by-pass and reroute surface drainage from
construction areas. Subsurface patterns must be broken at critical points
several hundred feet away by providing deep ditches to be cut into the perma–
frost so as to create planes of weakness which will rupture and permit sub–
surface drainage either to be diverted away from the building areas or im–
pounded in induced ponds or ice fields away from the critical areas. Plans
should contemplate that natural insulating materials be left intact wherever
possible or be restored using similar natural materials stripped from other
areas remote from the airfield when disturbance cannot be avoided. The
grades, both longitudinal and latitudinal, of the runways and its appurtenances
should be set such that fills and not cuts are required in establishing final
surfaces.
The building area for camp structures should be located downhill from
the landing field. Downhill, in this instance, is in terms of subsurface
drainage and not necessarily in terms of surface elevations. Construction
operation buildings, such as shops and warehouses, should be located between
the airfield and the camp. All structures should be located a minimum of one–
fourth mile away from the airfield, and preferably farther, up to one mile.
These distances are desirable to provide room for ex ap pa nsion, to minimize the
effects on the runway of ice fog that will form in the vicinity of the heated
structures at temperatures below approximately minus 40 degrees Fahrenheit,

EA-I. Lang: Construction & Maintenance of Airfields

and to minimize the effect on the subsurface drainage pattern both from the
standpoint of the effect of the airfield drainage pattern upon the building
area and vice versa. Structures themselves should be designed to minimize
shipping space required and to utilize local materials where possible. It
also is highly advisable to utilize prefabricated or precut designs to
minimize the manpower required at the site for erection.
The design of the airfield landing strips, taxiways, aprons, and road
system will depend upon the type and nature of the aircraft and vehicles
which may be expected to operate upon them. It is beyond the scope of this
discussion to deal with the technical aspects of how this design is done, but
it is sufficient to note that standard engineering practice is followed to
determine the basic dimensions of the construction required and the nature
of the materials employed therein. It follows that maximum utilization must
be made of locally available materials, the nature and location of which were
determined by the reconnaissance. It also follows that the location of the
field must have been predicated upon the availability of suitable building
materials within a reasonable hauling distance of the final location desired.
One general aspect of the design, however, should be emphasized, and that is
that the subgrade materials to be employed in any kind of a fill must consist
of freely draining or highly pervious material, and that these materials must
extend all the way out to the edge of any such fill. There must be a means of
readily disposing of surface and subsurface drainage which may accumulate at
the edge of the fill. It is mandatory that the subgrade be designed to contain
well-graded gravel in its top and bottom courses in order to prevent mud,
either above or below, from working into the central portion of the sub g rade

EA-I. Lang: Construction & Maintenance of Airfields

and thus lessen its free-draining characteristics.
The discussion on planning and design has been devoted primarily to the
construction of heavy airfields since this type of field naturally presents
the greatest problem. However, the basic principles of planning and design
are also applicable to the construction of a snow and ice field, or of a
light airfield, or a road or highway.
CONSTRUCTION
The initiation of actual construction operations is best approached by
moving to the site, either over land or by air, an advance construction party
consisting of personnel to complete more detailed investigation of the site
than was possible during the period of reconnaissance, and sufficient per–
sonnel to complete construction of base-camp facilities to be occupied by
the main body when it arrives.
During the reconnaissance phase a number of drill holes and test pits
were dug to determine the general nature of the terrain and drainage pattern.
The number of such holes and test pits must now be multiplied many times.
For example, test holes must be drilled along the center line of each projected
runway, taxiway, apron, and road at intervals not exceeding 500 ft. and
preferably as close as 250 ft. Additional test holes and test pits must be
dug throughout the area intended for building construction. This general
area will have been selected during the planning phase, but it is a different
matter to locate each specific building on an exact site. The variation in
the permafrost structure may be so erratic that the shifting of a building 20
ft. one way or the other may avoid permafrost in the foundation excavation.

EA-I. Lang: Construction & Maintenance of Airfields

When major structures such as hangars are to be located, it may be necessary
to drill on 50-ft. centers over an area of many acres to determine precisely
the most advantageous location. Additional drill holes must be driven at
each break in the terrain, on slopes having a variation of solar exposure,
along streams, edges of ponds, and particularly throughout areas expected
to be sources of borrow mat e é riel. The depth of test holes will vary. In
general, the rule to follow is that in areas where the permafrost is not to
be dist ru ur bed, the holes will be driven only to the surface of permafrost. In
other areas where removal of a certain amount of permafrost cannot be avoided,
holes should be driven at least 4 to 5 ft. below the lowest elevation of
the proposed foundation. For large structures these holes should be driven
to a depth equal to the distance from the hole to the nearest edge of the
foundation. Test holes along the center line of runways, taxiways, etc.,
must periodically be d e riven, say every 1,000 ft., to a depth of 50 ft. or
more to ascertain the thickness and continuity of the permafrost zone which may
lie beneath the runway. Where uncommonly large quantities of free-flowing water
are encountered, extra holes should be driven to determine the direction of
flow and to attempt to gain a more adequate understanding of the rate of
flow.
Let us turn now to the other group of the advance construction party
assigned the mission of preparing base-camp facilities for the main body.
The economics of such a move may not be at first apparent, but experience has
proved that time can be gained in the long run by holding to a minimum the
personnel so employed since they must exist initially without any base living
facilities. At the same time, the manpower thus saved and retained with the

EA-I. Lang: Construction & Maintenance of Airfields

main body back at the base of operations can well spend its time in further
organization, planning, and making ready for the intensive effort which is
to follow. Holding the size of the advance party to a minimum also has
another advantage and this is to permit the establishment of well-defined
traffic lanes through which all traffic must thereafter be required to pass.
These traffic patterns must be rigidly [: ] enforced. This is particularly
true on a site covered by muskeg or tundra since any disturbance to the surface
may induce a melting of the permafrost and the creation of a quagmire which
may have a disastrous effect on later operations. Where construction opera–
tions are confined to naturally exposed gravel areas, this matter of traffic
control is not important. A second carefully controlled situation must be
adopted at this time and rigidly enforced throughout the entire duration of
the construction operation. This control has to do with fire-prevention
measures. In the early phase of any construction operation in the Arctic, men
are not acclimated to the cold and will require additional heat which will
later not be necessary. Initial construction will be of a hasty nature and,
therefore, more hazardous from the standpoint of fire. Adequate fire-fighting
equipment will be lacking and conditions, in general, will be ripe for a major
conflagration which can be disastrous to subsequent construction operations.
The writer has learned through bitter experience that complete stocks of spare
parts, food, clothing, and construction data can be destroyed in a matter of
minutes. Full awareness of the disastrous effect of fires and the institution
of immediate measures to eliminate them can be a major factor in insuring
the successful accomplishment of the construction operations to follow.

EA-I. Lang: Construction & Maintenance of Airfields

Normally, it is most advantageous to move the advance construction party
to the site of operations during the time when the ground is still frozen.
Some equipment and heavy items of materials and supplies can be successfully
transported across the country without the necessity of providing special
measures for stream, lake, and swamp crossings. Clearing and surveying
operations can be accomplished more readily on ground in a frozen condition.
Construction of shelter and communal facilities for the main force, and
construction of access roads to the principal sites for borrow mat e é ri a e l are
items of useful work that can be performed in the relatively mild weather
before the annual spring [: ] break-up. Completion of these items before
the arrival of the main body means that a minimum of time will be wasted
upon arrival of that body and that full advantage may be taken of the best
construction season.
The principal item of work and the greatest time consumer on an airfield
project is the runway itself. It follows, therefore, that this feature s should
be the first item of major work undertaken by the main construction force.
During the planning phase an estimate of capabilities will have been made and
a decision will have been rendered relative to the advisability of completing
the subgrade and the pavement in one season or whether the paving can be
deferred one year to permit consolidation of the runway fill. The latter is
the better solution, but regardless of which solution is chosen, everything
must be subordinated to the one all-important element of completing the
scheduled amount of runway construction during the first season. Some curtail–
ment of elements of the project may be occasioned by encountering unexpected

EA-I. Lang: Construction & Maintenance of Airfields

difficulties. The runway must be the last item to suffer a reduction in
effort.
Taxiways and parking aprons, being closely allied to the runway, and being
also major items of work, should be given second priority and attention. The
grades established and the drainage pattern developed for the runway will
govern grades and drainage characteristics of the taxiways and aprons. These
problems may be mutually aggravated but should be resolved as items of impor–
tance second only to the runway.
The provision of basic utilities, such as adequate water supply, sewage
disposal, power generation, roads, and fuel storage, have third priority.
Living accommodations, including barracks, messes, latrines, washrooms,
and dispensaries, have fourth priority.
Airfield operational facilities, including control tower, operations
building, crash rescue station, minor repair and shop facilities, are fifth
in importance.
Administrative facilities, including warehouses, cold storage, maintenance
shops, major repair facilities, communications, and headquarters, have sixth
priority.
Last priority should be given to schools, training facilities, chapels,
theaters, recreation, and other items of building construction which, while
necessary, are not of a vital nature such as to jeopardize failure of the
airfield mission.

EA-I. Lang: Construction & Maintenance of Airfields

GRADING
As previously discussed, it is preferable to avoid making cuts into
undisturbed areas susceptible to permafrost. However, upon occasion, certain
areas will contain humps or pinnacles which will have to be removed in order
to preserve reasonable longitudinal and latitudinal gradients without such
heavy fills as to make the task impossible to accomplish within the time
allotted. One of the best times to accomplish this grading is when the air
temperature is slightly less than 32 degrees Fahrenheit. Heavy rooters may
be used to loosen the surface material or blasting may be employed, and the
resultant material is removed by conventional hauling equipment. Later in the
season after daytime air temperatures have risen above freezing, it may be
advisable to accomplish grading in alternate areas on approximately three–
day cycles. The material can be loosened by ripping or blasting on the
first day. It can be left to melt and drain the second day, while the ripping
operation is conducted in a second area. Then, on the third day, the melted
material may be removed from the first area, and the cycle is repeated.
Building areas may be thaw e d to greater depths if time permits by using a
method commonly employed in gold dredging operations in the Arctic. This
method consists of driving a large number of one-inch pipes into the ground
on a grid pattern on 15- or 20-foot centers and pumping cold water through
these pipes until the ground has become sufficiently thawed to permit pipes
to be driven still farther into the ground. After the thawing has been accomplished
to the depth required, the loose material may be excavated by conventional
methods. Steam points may also be employed in lieu of cold water to hasten
the rate of thawing if the time schedule must be shortened. It is important to

EA-I. Lang: Construction & Maintenance of Airfields

remember that after excavation in thawed ground has been completed, it is
necessary to restore immediately a layer of insulation to prevent the melting
process from continuing, for the site might become such a morass that it must
be abandoned.
PREPARATION OF THE SUBGRADE
Assuming that a section of natural terrain is available which is a suitable
site for a runway, or that an area has been prepared by grading in the manner
described in the proceeding discussion, then the runway may be started by plac–
ing subgrade materials thereon, using conventional equipment and methods of
hauling in much the same manner as airfields or roads are prepared in more
temperate climates. The main difference to remember is that the undisturbed
frozen soil lying immediately beneath the subgrade has great bearing value only
so long as it remains frozen, and that a means must be provided of retaining
this high bearing value by sealing the frozen material from heat and flowing
water. Sealing against heat may be accomplished by preserving the natural
tundra or moss in its undisturbed condition or by covering the frozen base
with a much thicker layer of highly impervious gravel, crushed rock, volcanic
ash, or other similar material. It is worth while to note that a 3-inch
layer of moss or tundra is equivalent to approximately 36 inches of inorganic
material. This emphasizes the importance of retaining the former if at all
possible. Sealing against flowing water may be accomplished by providing
diversion channels for subsurface flows and by guarding against retaining
moisture within the subgrade by virtue of constructing in it a course of gravel
or crushed rock of granular size such that no moisture will be absorbed

EA-I. Lang: Construction & Maintenance of Airfields

through capillarity. As previously discussed, the upper and lower layers of
the subgrade should consist of finer gravel to prevent mud from working
into the center section of the subgrade either from above or below. It is
important that no frozen material be placed in the subgrade. For this reason,
it is normal that subgrade preparation be accomplished during the warmer
months of the construction season.
To enable the reader to have a better understanding of the thickness of
subgrade required, and hence of the work to be accomplished, a wheel design
load of 15,000 pounds or less will require approximately 30 inches of gravel
in the subgrade. A wheel design load of 60,000 pounds will require approxi–
mately 48 inches, and a wheel load of 150,000 pounds will require approxi–
mately 72 inches. Modern aircraft requirements tend to approach the largest
of these figures. This factor, taken in combination with the necessity for
employing fills rather than cuts, means that, throughout the length of the
entire runway, considerable settlement may be expected by compaction of the
subgrade during the first year or two of its existence. The subgrade will
be of varying depth, which means that there will be a differential settlement
unless extreme care is taken in placing the fill material in lifts of not to
exceed 6 to 8 inches during construction, and unless the material is
rolled with heavy sheep’s-foot and pneumatic-tired rollers. If proper com–
paction has been obtained during construction, and if time will permit, there
is no reason why paving should not be accomplished the first year. However,
limitations of time or equipment usually do not make this possible, so it is
best to treat the surface with a temporary asphaltic seal at the end of the
first construction season, and then the following year to regrade the surface
and proceed with the final paving.

EA-I. Lang: Construction & Maintenance of Airfields

PAVING
As a general rule, runways, taxiways, and aprons built to accommodate
present-day aircraft should be paved. Rocks and gravel picked up by pro–
pellers and jet engines are damaging to the engines and the airplanes them–
selves, as well as being dangerous to personnel in the area. Paving of arctic
airfields is best accomplished by using either portland cement or asphaltic
concrete. Limitations of time, equipment, and transportation, to say nothing
of plain economics, almost invariably rules out portland-cement concrete.
Also, since differential settlement may be anticipated, it is best to select
a pavement of the flexible type, of which asphaltic concrete is the most
suitable. Emulsions, cutbacks, and road oils are not recommended because of
the low temperatures which do not permit proper curing. Asphaltic concrete,
being a hot mix, can be prepared and laid even in temperatures below freez–
ing. It is economical, quick, and simple to place, with standard equipment
and with crews of moderate skill.
SNOW REMOVAL AND MAINTENANCE
The problem of snow removal in the Arctic is obviously one which cannot
be neglected. However, it is not nearly as difficult as the uninitiated be–
lieve. Equipment for snow removal, such as is employed by highway departments
in our northern states, is entirely satisfactory. Snow crews must be organized,
trained, and kept in readiness, day or night, during the season when snowfall
may be expected. Snow removal operations are conducted [: ] promptly so as to
avoid an accumulation of packed snow greater than three inches at any one time.

EA-I. Lang: Construction & Maintenance of Airfields

Aircraft move with difficulty on loose snow, but without any appreciable
trouble on compacted snow. The limit of three inches of compacted snow is
established in order to avoid excessive slush and drainage problems during the
spring. Deep slush might necessitate the closing down of the runway for an
extended period. The melting of three inches of compacted snow, on the other
hand, presents a simple problem which can be handled without difficulty.
Runways must be regularly inspected and maintained at all times. Major
repairs should be held over until the summer season when patching can be ac–
complished as is normally done in the United States. There will be occasions,
however, when emergency repairs must be made during the winter months. In
these instances, only temporary expedients should be employed, utilizing logs,
brush, landing mat, or built-up layers of snow and ice when no other means
are available.
James D. Lang

Excavations and Foundations

EA-I. (Al T. Donnels)

EXCAVATIONS AND FOUNDATIONS

CONTENTS
Page
Gravel and Rock 3
Drainage 4
Disposal of Spoil 6
Airfield Foundation 6

EA-I. (Al T. Donnels)

EXCAVATIONS AND FOUNDATIONS
Excavation and foundation problems in Alaska vary greatly; however,
there are certain basic criteria peculiar to the country with which the planning
or construction engineer should be thoroughly familiar for the successful
and economical accomplishment of a contemplated project. The subject
matter of this article is proposed to encompass large-scale excavation and
foundation factors encountered in the construction of airdromes and heavy
structures. These factors, however, are applicable to all construction
problems and should receive thorough consideration in the scope of planning.
The specific location of a project immediately introduces or eliminates
certain problems for investigation. For example, south of the Alaska Range
no peculiar problems may be expected other than those usually encountered
in the northern United States. Permafrost conditions do not exist south
of the range. Depth of seasonal frost varies considerably depending on
ground conditions, vegetation, and snow coverage. Temperature conditions
are not severe and construction operations can be successfully carried on
from the first of March until December, frequently throughout the year. The
coastal regions (Prince William Sound area) have heavy snow and rough weather
from December through March and April. Average annual precipitation in this
area approximates 160 meters inches ; snow depths of 30 feet occurred in the winter
of 1947-48.

EA-I. Donnels: Excavations and Foundations

North of the Alaska Range, particular attention is required for the
investigation of permafrost conditions and subsurface ice lenses. Contrary
to popular belief among persons not accustomed to the country, snow precipi–
tation north of the Alaska Range is comparatively light. Snow is dry and
powdery and little or no wind occurs in the interior. Blizzards, such as
are experienced in Montana, the Dakotas, and many other states, are not known
in central interior Alaska. Severe temperatures may be encountered from
December through February and March, with seasonal thaws occurring through
April and May. Construction weather is ideal from May to November, with
24 hours of daylight during the summer months.
Excavation operations for air-strip construction may be started with the
spring thaws, usually in May, although frost is still retained in the ground.
This may be done by surface stripping with tractors and carryalls in systematic
passes over a large area, cutting down to the frozen ground, thus permitting
penetration of continued thaw. The operation is repeated upon return passes
of the equipment at such time intervals as thawing will permit. In this manner
the frost is followed downward to the desired grade in the gravel. Apparent
hopeless mud conditions will follow until free-draining gravel is reached.
Use of “pusher cats” should be anticipated for these mud conditions and difficult
traction on the frozen ground.
Permafrost may be encountered anywhere, under the seasonal frost line and
near the surface, particularly where the ground is covered by tundra vegetation.
It is essential that all tundra, silt, and frost-bearing material (fine mate–
rial subject to frost action) be removed down to sound gravel base. It must be
remembered, however, that the original thermal regime should be restored
whenever possible. (See also “Engineering Problems and Construction in
Permafrost Regions.”)

EA-I. Donnels: Excavations and Foundations

Upon reaching the final grade for stripping operations, it is advisable
to excavate test pits strategically located to search out sand pockets and
frost-bearing material. These pits should be located on a grid pattern not
exceeding 200-foot intervals and depths of 3 to 5 feet. Unfavorable condi–
tions may require their location at more frequent intervals and slightly
greater depths to locate the existence of ancient slough courses which are
prevalent in the Territory and which have deposited fine sand and frost–
bearing silts. The slough courses may or may not be discernible from the sur–
face after stripping operations, but are readily traced by indication of
excessive mixture of fin d — e material in the gravel. All such material should
be carefully removed and replaced with sound gravel. Removal of this
unsuitable material is accomplished by tracing out the slough courses with
tractors and scrapers.
Gravel and Rock
Good gravel deposits are prevalent in practically all of interior Alaska
and are usually found at shallow depths under the surface. Natural gravels
are not too well-graded but are conducive to excellent subcase material.
Gravel found in the upper Tanana Valley is relatively small — too small for
satisfactory crushing operations. However, the gravel found closer to the
north slope of the Alaska Range contains a fairly high percentage of coarse
gravel (4 inches or more) that is suitable for obtaining high-quality crushed
rock. It has been determined that the size of the gravels found is dependent
upon the distance carried by glacial action from the bedrock of the Alaska
Range, the gravel becoming smaller with continued grinding action beyond the
rim of the ice.

EA-I. Donnels: Excavations and Foundations

Other rock deposits in the Tanana Valley area are basalt, granite, and
Birch Creek schist, of which the last mentioned is the most prevalent and
usually the most accessible. The Birch Creek schist varies considerably in
quality and, although it is not considered to be of the superior quality of
the basalts and granites, is usable for crushing and subgrade material. It
has a tendency to break up under compaction, forming a silty substance, and
will not obtain the interlocking ordinarily desired for good subgrade or
foundation construction. Economical availability may dictate its usage,
however, but it should receive careful consideration and analysis for its
contemplated use. It is not recommended as an aggregate in asphaltic or
portland-cement concrete.
Drainage
The problem of drainage is one of the most important and frequently
one of the most troublesome aspects encountered in the North. The water
tables throughout the Territory have widely varying characteristics and
apparently have no known relationship to the river systems. The Tanana
Valley has a characteristically high water table, found 2 to 5 feet below
the gravel surface, varying slightly throughout the year, and flowing slowly
in a southerly or southeasterly direction. The water table in the upper
Nenana Valley (mile 373 on the Alaska Railroad) was found to be approximately
50 feet below the gravel surface and flowing in a northeasterly direction,
although the Nenana River is above the ground surface elevation at this
location and within 2 miles distance. The water elevation rose considerably
in the exploratory drill holes, indicating hydrostatic pressure, although
only sand and gravel were encountered down to the water level.

EA-I. Donnels: Excavations and Foundations

The river systems have built up their silt beds in such a manner that
natural surface drainage is away from, rather than toward, the rivers and
their tributary sloughs. It has been observed that the gradient away from
the rivers frequently occurs in the ratio of about 1 foot per 1,000 feet.
Extreme meandering of rivers and sloughs is a typical characteristic.
Further drainage and flood difficulties may be experienced in the
spring of the year when rising rivers may wash above the surface of the ice,
thus progressively building up thicker surface ice and higher water until
the final ice break-up occurs. Surface drainage normally disposes itself
rapidly through the free-draining gravel which has been stripped of the
surface tundra. During the spring thaw, however, free drainage is prevented
due to the ground frost and, unless other provisions for surface drainage
are made, an accumulation of surface water may be troublesome until the ground
frost disappears.
Excavation for structural foundation may necessarily be carried to
elevations below the normal water table, which adds materially to the con–
struction problems and costs involved, for which field conditions should be
studied in the development of plans and establishment of cost estimates. The
proposed procedure is to confine the operation to as small an area as possible
commensurate with available equipment, reducing the water table in this area
by the use of well points and pumps, and repeating the operation in the next
foundation section. The planning of subsurface installations such as base–
ments or other adjuncts should be based on full knowledge of the specific
subsurface and water-table conditions.
No specific criteria exist for the solution of drainage problems. Each
development must be worked out individually. Sloughs and marshes afford

EA-I. Donnels: Excavations and Foundations

disposal of drainage water, but the problem of flood and backup water is
present due to the usual flatness of the terrain. Freezing of drainage
culverts and outfalls may present difficulties in blocking the runoff during
the spring when alternate freezing and thawing occur. These should be
located to the best advantage of exposure to permit thawing during the day.
Sewerage and effluent trunks require installation in heated utilidors
as being the only effective means of protection against freezing during the
winter.
Disposal of Spoil
The matter of disposal of spoil is ordinarily not considered of major
consequence as it falls into the routine of ordinary earth-moving operations.
However, in the North conventional procedures frequently result in costly
effort. Planning of operations should, therefore, carefully include the
disposal of spoil to avoid its being handled more than necessary, or to
prevent spoil piles from freezing, thereby forming obstructions or detriments
to drainage or interference with other proposed operations. Removal of
frozen spoil may be difficult and expensive and unnecessarily adds to the
effort and cost of a construction project over the short construction season.
Spoil piles that require subsequent removal should be removed prior to the
onset of freezing weather.
Airfield Foundation
Pavement evaluation was made on a concrete military runway constructed
during World War II, including the removal of slab sections for subgrade
bearing analysis. Large voids were disclosed between the subgrade and concrete
pavement due to settlement of the foundation. At certain locations the

EA-I. Donnels: Excavations and Foundations

unreinforced pavement was unsupported for several feet and amazingly
withstood the impact of heavy planes without failure. Whether the settle–
ment of the foundation was due to faulty backfill containing frozen material,
or was caused by shrinkage of permafrost under the subgrade, or a combination
of both, is a matter of conjecture. The results, however, clearly point to
the importance of thorough investigation in dealing with permafrost and
placement of sound foundation material.
In connection with the stripping operations previously described, a
major factor in determining the final grade of the stripping operations is
the determination of insulating coverage of nonfrost-bearing material to be
placed over permafrost areas with due consideration given to the possible
rising of the permafrost table under the fill. That is, it may be necessary
to strip into the permafrost and backfill with sound material to a depth
sufficient to obtain adequate insulation to prevent thawing of the permafrost
from heat absorption of the runway. It may be pointed out that recession of
the permafrost in the Tanana and Yukon valleys areas is a natural process,
and once thawed will not return. It is understood, however, that this
condition does not necessarily apply in some areas north of the Arctic Circle,
in which return of permafrost may occur after thawing.
Extensive information is available from experiments and studies conducted
by the Permafrost Division of the Corps of Engineers for various types of
runways and other construction in connection with heat penetration into the
ground, depth of insulating coverage, etc. Excavation and foundation problems
encountered in the North are closely associated with permafrost problems and
engineers contemplating such operations are urged to avail themselves of this
complete source of information and visit the Permafrost Experimental Project
near the city of Fairbanks.

EA-I. Donnels: Excavations and Foundations

It is entirely practicable and frequently necessary to construct
airfields over permafrost owing to the existing pattern of the permafrost
areas and the usual limiting circumstances in the location of airfields.
However, in a certain selected location for an airfield in the Nenana Valley,
the runway and development site were found to be entirely free from permafrost.
When such conditions are encountered, it is indeed fortunate. Persons ex–
perienced in the characteristics of the North are able to ascertain with
reasonable accuracy the likely existence or nonexistence of permafrost by
observation of the forests and vegetation, sloughs, marshes, water table,
and general drainage.
“Drunken forests” are the result of frost heaving, which causes the
trees to assume wild angles from vertical growth, appearing to have the tendency
of falling. Profuse growth of the native spruce stands occurs over permafrost
areas, especially where high water tables exist. The native spruce has no
tap root, which permits the tree to live within permafrost areas. A know–
ledge of forestry and plant life in the North may be of much assistance to
the engineer who is engaged in making preliminary examinations of possible
site locations.
Sloughs and marshes usually indicate the presence of permafrost near
the surface. Areas subjected to good subsurface drainage are obvioualy
desirable in all respects and are less likely to contain permafrost. However,
such observations alone are not sufficient for determining the desirable
location of a “site.” A likely appearing location should be further explored
by test holes and other means applicable to the particular type of construction
contemplated.

EA-I. Donnels: Excavations and Foundations

Where extreme conditions of frost or ice action are observed, such as
ice or frost mounds, “drunk en forest,” and areas known to contain ice lenses,
these areas should be carefully avoided for any construction operations. An
upset of the natural thermal balance of the permafrost table by construction
operations may introduce insurmountable problems. The engineer confronted
with such a task should reassure himself with the necessary field research
and be familiar with the techniques of compensating for any upsets that would
appear probable in dealing with permafrost conditions.
The importance of site selection and foundation construction must be
emphasized where permafrost exists, for there is always the possibility of a
building filling with ice and breaking apart due to peculiarities of permafrost
and ground water; upheaval and collapse of a building due to shifting of the
permafrost table caused by the building itself; or foundation and structural
failure, as occurred to the Federal Building at Nome and a major building at
Northway and similar occurrences to dwellings and other buildings throughout
the North.
Subsurface ice lenses may be treacherous, for obvious reasons, if subjected
to thawing under the foundation of an air strip of building. For an engineer
who is involved with foundation problems a visit to the hydraulic mining
operations along the Steese Highway, north of Fairbanks, affords an excellent
cross-sectional view of subsurface ice lenses; here he can readily visualize
the potential hazards involved should he risk foundation construction without
full knowledge as to the presence of ice lenses.
The total importance of determining the permafrost and subsurface ice
conditions in connection with foundations lies in the simple fact that heat
penetration into the ground may eventually reach the permafrost, causing

EA-I. Donnels: Excavations and Foundations

thawing and subsequent settlement. In the case of airfields, sufficient
insulating earth coverage may readily be constructed to compensate for the
heat absorption at the surface. As previously described, the techniques of
such construction and required depths of insulation are obtainable from
precise experimental data available from the Permafrost Division, Corps of
Engineers.
In the case of building foundations, the heat penetration into the
ground is normally continuous and depths of heat penetration are dependent
upon the size of the building and other factors. For example, a small building
or narrow building permits the influence of exterior temperatures under the
foundation to a greater extent than would occur in the case of a building
covering a large area. Also, the problem may be complicated by a shifting
of the permafrost table, which may be expected to drop on the south side of
the building due to the sun’s rays on the wall surface, and rise on the north
side which does not receive the sun’s rays.
The extent to which subsurface explorations may be necessary greatly
depends upon the type and permanence of the buildings contemplated. Thawing
temperatures may be expected to penetrate the ground beneath a building at
the rate of from 3 to 5 feet per year under average conditions for depths
within the limits of average building construction. Thermal penetration
becomes slower with increased depths. Where permafrost is know to exist at
20- to 3 0 -foot depths, no serious foundation settlement would be expected
within the usable life of temporary or semipermanent installation. Numerous
military installations constructed over known permafrost areas during the war
have shown no indication of settlement up to the present time. It is probable
that sufficient time has not elapsed to permit thawing temperatures to reach

EA-I. Donnels: Excavations and Foundations

the permafrost and cause settlement to a perceptible degre — e .
The element of time is usually a major factor in the construction of
temporary or semipermanent military installations. For a construction
program of this nature, extensive foundation examinations would be imprac–
ticable and unwarranted.
Heavy structures or buildings of a permanent character, such as hangars,
power plants, or comparable major buildings, to which settlement would
seriously endanger structural or mechanical alignments, should not be con–
structed over permafrost. The foundation conditions for such structures
should be thoroughly investigated and the structure located in an area
devoid of permafrost.
The following test was made to ascertain whether or not appreciable
settlement would occur upon thawing of permafrost-bearing gravel and to
determine to what extent such permafrost-bearing gravel could be used for
foundation.
The ground was stripped of all surface material to sound gravel within
the permafrost table over an area sufficiently large to conduct the test
properly, and depressed sufficiently to contain water. Two 1-inch diameter
pipes about 5 feet long were welded to the inside bottom of a 55-gallon steel
drum in such a manner that the pipes were parallel, about 2 feet apart and
extending out of the open top of the drum. The drum was then filled with
concrete to obtain substantial weight. Welding the pipes to the bottom of
the drum prevented transmission of movement to the pipes from shrinkage of
the concrete. A heavy steel leveling plate was placed on the frozen gravel
surface and the concrete-filled drum placed thereon. Two pipes were securely
embedded into the permafrost on each side of the drum in vertical position.

EA-I. Donnels: Excavations and Foundations

A rigid steel angle was secured between these pipes. The steel drum was
adjusted to position so that the pipes extending therefrom were aligned with
the angle beam. Two micrometer dials (Ames) were secured between the angle
beam and the pipes extending from the drum in such manner as to register
movement of the drum. Water was placed in the depression, partially submerging
the drum, and heated by steam from a portable boiler. Care was exercised to
apply heat evenly around the drum to obtain uniform settlement. The results
of the test indicated a settlement of approximately 1 inch in 5 feet of thawed
ground.
Upon examination, the frozen gravel might give the impression that thawing
of the frost crystals in the voids would have little or no effect on the
shrinkage of natural compacted gravel, thus leading to the false assumption
that such material would be conducive to good foundation.
The results of the above test, however, indicated that a settlement of
6 inches or more may be anticipated in event of thawing of a permafrost lenso
of average thickness encountered in this particular area. This a mount of
settlement could obviously be disastrous to a large rigid structure.
Such ground would afford suitable foundation for light frame structures
of the temporary or semipermanent type to which no serious damage would result
that could not readily be repaired in event of slight settlement. This effect
should be considered in the installation of plumbing and utility connections.
The selection of a site location for a major structure entails the primary
consideration of locating an area void of permafrost. If such an area cannot
be located, the permafrost under the foundation area must be thawed which
involves a difficult and time-consuming operation. The exterior foundation line

EA-I. Donnels: Excavations and Foundations

line should not be less than 20 feet horizontal distance from a permafrost
body. The permafrost occurs in the form of islands or “kidneys” interspersed
with void areas. The determination of these void areas involves a considerable
amount of subsurface exploration. In the course of master planning, the plan–
ning engineer should allow adequate flexibility to permit the shifting of site
locations, if possible, to such a void area.
Exploration for the existence or nonexistence of permafrost by drilling
alone is exceedingly time-consuming. In order to speed up this operation,
the following method was satisfactorily utilized for establishing the location
for a hangar having 90,000 square feet of floor area.
The proposed building site was surveyed and s taked out (exterior foundation
line plus 20 feet). At each corner and at quarter points on each side and at
the center, railroad rails were drive d n with a drop-hammer pile driver. Explora–
tion was carried to a depth of 60 feet, as it was considered that no permafrost
would exist beyond this depth, or if it did, it would not be reached by heat
penetration. Upon striking permafrost, no further penetration of the rail
could be obtained. Permafrost was struck along one entire side of the building
site and partially along the two adjacent sides. No permafrost was found else–
where on the building site. This indicated that the permafrost lens out somewhat
diagonally across approximately one-quarter of the building site area. It then
became necessary to locate and plot the edge of the permafrost lens. The dis–
tance between the rail points which straddled the edge of the lens was bisected,
and a rail driven at that point. This procedure was repeated on a grid pattern
until the edge of the lens was located to an accuracy of approximately 5 feet.
Concurrently with the driving operation, well drills were used at strategic
locations to penetrate through the permafrost, measuring depth and thickness

EA-I. Donnels: Excavations and Foundations

of the lens, and also to verify the conditions where indeterminate rail
driving was encountered due to striking dense material or rocks. The tops
of the lenses were found at 15- to 30-foot depths and were from 20 to 30 feet
in thickness and usually tapered off toward the edge of the lens.
Establishment of the exact location of the permafrost lens afforded
knowledge of the required distance and direction to shift the building site.
Shifting the building site off the permafrost area also introduced the possi–
bility of moving the opposite side of the building onto another permafrost
island, which condition required investigation. Further possibilities as to
existence of permafrost islands within the building area also required investi–
gation. This method required comparatively few drill holes and the entire
work was accomplished in about two weeks.
The rail-driving method had not been previously attempted and some experi–
menting was required to effect efficient operation. Rails were cut from 10-
to 15-foot lengths, which was the maximum length that could be placed in the
driving leads. Upon driving to ground surface, a new length of rail was placed
in the leads and spliced to the driven rail by welding. Some difficulty was
experienced in the welds breaking or the rail bending under impact of the hammer.
The development of a more efficient splice and the improvision of a lead guide
to prevent bending corrected this difficulty. The driving method could be
greatly improved and the operation speeded up considerably with the planning
of an efficient equipment setup and manner of splicing. Several pile drivers
with swinging leads and well drills are required to cover efficiently the extent
of the work.
The use of geophysical equipment was considered for determining the location
of permafrost. However, the time involved for procurement of special equipment

EA-I. Donnels: Excavations and Foundations

and experienced operating personnel, together with the fact that such an
operation would be experimental with questionable reliability of results,
precluded its use d for the above operation. It is possible that the use
of geophysical equipment for such purpose on extensive operations might
prove of extreme value in saving time and labor, and introduce further
knowledge of subsurface conditions.
In the event that it is not possible to obtain a site location entirely
devoid of permafrost and it becomes necessary to thaw out such an area, the
site should be located to minimize the quantity of permafrost to be thawed.
The method of thawing permafrost by percolation of cold water as observed in
connection with mining operations in the Fairbanks and Nome areas would
ordinarily not be considered practicable for construction operations owing
to the time involved, unless a long-range, extensive construction program was
contemplated.
Although the experience of the writer did not entail the necessity of
performing thawing operations, as it was easier to move off permafrost areas,
a planned operation was conceived by driving steam points on a grid pattern of
10 to 15 feet over an area of not more than 100 feet square and connected to a
header system to which steams would be supplied from portable steam boilers.
It was calculated that this interval of steam points would adequately raise
the ground temperature to permit rapid thawing and permit continued penetration
of the steam [: ]
points by hand-driving similar to the method used on cold-water points, which
consists of a weighted mandrel slipped over the projecting pipe similar to
that used in driving steel fence posts.

EA-I. Donnels: Excavations and Foundations

The primary subject of excavation and foundation requires a complete
correlation with all other factors in planning and construction. Such
planning may be required in considerable detail for the effective operation
of the work. An extensive amount of preliminary field investigation may be
necessary for the development of master and detail planning. The procurement
and transportation of equipment, materials, and personnel involve the thorough
planning of logistics, particularly in connection with the heavy equipment
necessary for excavation and foundation construction like power shovels, tractors,
carryalls, dozers, rock crushers, and trucks. Shipping and unloading facilities
for quantities of such equipment at Alaska ports are limited. The ports of
Seward, Whittier, and Anchorage are the only territorial ports with railroad
connections. These ports at present (1949) have limited dock facilities. The
port of Anchorage is usually icebound during the winter months. Inland trans–
portation is subject to limited highway facilities to off-rail points.
In addition to the solution of transporting such equipment, it becomes
necessary to set up adequate shops and self-sufficient facilities for equipment
repair and maintenance. Excessive equipment deadlines have been evidenced on
past operations for lack of adequate planning for this phase of work.
Concurrent with the above requirements are those necessary for the housing
and care of personnel. These further tax the facilities for mat e é riel procurement.
Air transportation in the North is vital. In remote sections of the
territory, the primary operation of grading off a temporary airfield for use by
medium-sized planes will be found to enhance greatly communications and security
in providing the means of obtaining much needed supplies when starting and con–
ducting operations on a project.

EA-I. Donnels: Excavations and Foundations

All of these considerations require precise planning and coordination
to effect full advantage of operations during the relatively short construction
season. The failure of any one major factor may easily result in the partial
or total loss of a year’s operation, which delay may be the determining
element of success or failure of the project.
Al T. Donnels

Emergency, Temporary, and Semi-permanent Housing for Polar Areas

EA-I. Palmer V. Roberts

EMERGENCY, TEMPORARY, AND SEMIPERMANENT HOUSING FOR POLAR AREAS

TABLE OF CONTENTS
Page
Emergency Housing 2
Principles of Design of Temporary and Semipermanent Buildings 11
General 11
Insulation 14
Condensation 20
Methods of Condensation Control 23
Ventilation 28
Temporary Housing 30
Semipermanent Housing 35
Prefabricated Housing 35
Modern Prefabricated Buildings 40
Bibliography 47

EA-I. Roberts: Housing

LIST OF FIGURES
Page
Fig. 1. Steps in building a snow cave 3-a
Fig. 2. Steps in domeshaped snowhouse construction 6-a
Fig. 3. Steps in domeshaped snowhouse construction-continued 6-b
Fig. 4. Steps in earth and wood house construction 30-a
Fig. 5. Typical section Quonset hut (Barrow) 40-a

EA-I. Palmer W. Roberts

EMERGENCY, TEMPORARY, AND SEMIPERMANENT HOUSING FOR POLAR AREAS
It is a fundamental principle in building design that the best buildings are
those so constructed throughout that they meet the basic requirements of the region
while also serving the operational requirements of the users. The type of design
selected depends greatly on local soil, drainage, and climatic conditions. The
details of the design depend greatly on the materials locally available or trans–
portable to the area. Therefore, standards for the construction of shelters and
buildings in the Arctic and Subarctic, and particularly where permafrost is present,
will vary greatly from those applicable to the temperate and tropic zones. In
areas where permafrost does not exist, the construction of housing may be similar
to that of any region of cold winters where seasonal frost action is severe.
The selection of an appropriate type of housing in the Arctic is perhaps best
approached by the consideration of three factors: the length of time the build–
ing is to be occupied, the time required for construction, and whether or not
local or transported materials are to be used in the construction.
Various classes of housing may be defined on the basis of these factors as
emergency, temporary, semipermanent, and permanent. The following definitions have
been arbitrarily established for the purpose of providing an orderly approach to
the subject of arctic housing:

EA-I. Roberts:

Emergency: to be occupied for a period ranging from overnight to possibly
as long as one year, and capable of being completely constructed in less than
half a day from materials available on the site, or with lightweight materials
brought in for construction purposes.
Temporary: to be occupied from one to five years and requiring approximately
one day’s effort for construction, using local or imported lightweight building
materials.
Semipermanent: to be occupied for at least 15 years, requiring more than
one day’s construction effort, and utilizing lightweight prefabricated materials
brought into the area for the building.
Permanent: to be occupied for at least 25 years, requiring considerably more
than one day’s effort for construction, and using high-quality prefabricated or
on-site fabricated imported building materials of a standard nature.
Discussions here will be concerned only with the first three classes of
buildings, which are generally of one-story construction. Permanent buildings,
of course, follow the same principles; but as this type of construction largely
conforms to standard construction in the northern United States, it will not
be dealt with in this discussion of arctic housing.
EMERGENCY HOUSING
In an emergency, many [: ] types of shelters and windbreaks can be constructed
from materials at hand. One may first utilize the materials with which one has
landed. If one came by plane, any of the easily removable sections, such as the
cowling, will serve for construction. The cabin of a large plane will serve as
an excellent temporary shelter. Similar use may also be made of stranded vehicles
and boats. Any large piece of fabric may be used as a shelter roof, and the sides

EA-I. Roberts: Housing

can be built up with snow blocks, sod, etc. Snow, soil, stone, bones of large
mammals (whales), and wood, if available, will furnish building material for
housing once the immediate crisis is over.
The types of shelter that may be improvised by one or more men are numerous.
Tents may be made from shelter halves or parachutes. If deep drifts are found,
snow caves may be built to good purpose. Walls of snow blocks will serve as wind–
breaks. Standing or fallen trees may be most useful for individual survival.
Within a shelter, some form of insulation must be placed under a fire if the
floor of the shelter consists of snow; good insulation is required under a sleeping
bag or other bedding.
Snow Cave . One of the simplest forms of shelter is the snow cave. Such a
cave can often be built by two men in about two hours. The only tools required
are shovels. A square of canvas or other fabric may be useful to cover the door.
A site is selected on a reasonably steep slope where the snow has lain
in place long enough to have become compacted. ( See Figure 1. ) A working shelf is
cut into the slope or drift, and then a small cave two feet by two feet is ex–
cavated back for a foot or two; this becomes the doorway. The cave proper is then
started, working gradually upward, as the floor of the cave should have an upward
slope , away from the doorway. The easiest procedure may be to cut the snow into
blocks and cast them out through the doorway and down the slope. The ceiling should
be shaped into an arch or done. This allows condensation to run down the walls,
provides standing room in the center, and serves to strengthen the shelter.
The final cave size, to accommodate three or four people, should be about five
feet high, nine feet long, and eight feet wide. A sleeping place is arranged by
having the level of the bed platform at least one foot above the floor, which FIGURE 1

EA-I. Roberts: Housing

should be terraced. After this, the shovels and all personal equipment are brought
into the cave, a ventilating hole is cut upward through the roof, and, if the
entrance is not low enough, a square piece of fabric is placed to form a door.
If the bed platform is some two feet higher than the top of the entrance, it
will not be necessary to close the door, unless there is a strong wind that
forces gusts into the cave. For, cold air being heavy, it will remain at a low
level so long as the upper parts of the cave are filled with warm air.
A snow cave has many advantages over the tent, among them that the material
to be transported is lessened by the omission of the tent and stove. Once the
cave is occupied, the air within rises to a temperature above freezing and remains
there as long as the cave is occupied, no heating being required except that fur–
nished by the bodies of the occupants. In the daytime no lighting is needed,
for sufficient light filters through from overhead; at night a single candle
gives enough light, because of the high reflecting qualities of walls and ceiling.
Snow caves have been successfully used in the Arctic as well as in high
mountains. Instruction in their construction and use should form part of the
normal training of polar travelers.
The dome snowhouse is built from domino-shaped blocks of wind-pressed snow.
These blocks are cut four to six inches thick, twenty to forty inches long, and
from twelve to twenty inches wide. Such blocks, depending on the density of the
snow, weigh from fifty to a hundred pounds and are strong enough to support not
only their own weight but also that of the blocks resting on them. It is of prime
importance that the snow be of the proper consistency. This can be determined
by driving a blunt but slender rod into a snowbank with a steady shove. If it
sinks in slowly with an even pressure, the snow is of the proper consistency for

EA-I. Roberts: Housing

use in construction. Snow that is too soft crumbles in handling; if layers are
of uneven hardness the blocks may split; if the snow is too hard the blocks will
be poor insulators and the house will be cold.
In the building of a snowhouse the only tools required are a knife with a
blade fourteen to twenty inches long for cutting blocks, a shovel for piling snow
on the completed house, and a rod for determining the consistency of the
snow. For cutting blocks from very hard snowdrifts, a carpenter’s saw is good.
Three or four men make a desirable building team: one cutting the blocks, a second
carrying the blocks and chinking cracks, and the third working inside the house,
building. A fourth man, if available, may work outside the house, chinking cracks
between the blocks, but not following the builder too closely, for snowhouse
walls are fragile until the soft snow pressed into the cracks has had time for
setting into snowcrete.
The best location for building a snowhouse is on a level part of a well–
compacted drift where the snow is more than three feet deep. This depth provides
ample insulation and permite tunneling the entrance under the walls to a well in
the floor, which, as later explained, is the preferred entrance construction. If
this desired depth of snow is not available, then an entrance porch may be built
from snow blocks aboveground, whereupon it will be necessary to close the house
door at night, as a gravity differential will not be available for keeping the
cold air out. If the house is to be built on bare ground, a floor of at least
a five-inch layer of compacted snow must be placed to form an insulating layer;
for bare ground radiates cold.
The usual snowhouse is dome-shaped, with a five- to seven-foot ceiling, and
has a diameter equal to twice its height. A house can be built to accommodate ten

EA-I. Roberts: Housing

people, if the snow is good and the builder skillful; but such a house will have
to be so high that it will necessarily be cold, since the warmest air will gather
up in the dome where it is of no comfort to the occupants and threatens to melt
a hole in the snow roof.
The first step in building a snowhouse is to lay out a circle. For a
twelve-foot house this will have a six-foot radius. ( See Figure s 2 . and 3) The first row
of blocks are stood on edge, fitted together around the line of the circle, and
tilted slightly inward. The first course is tapered in height by cutting with
the knife, and succeeding blocks are laid in a continuous spiral, tilting farther
and farther inward until there is only a small opening left at the top of the
dome. Then the builder carves what is named the king block, and, from the inside
of the house, thrusts it out edgewise through the aperture; he then levels the
block and lets it sink into place. The builder on the outside of the house chinks
the openings between the blocks, gently but firmly, with loose snow, which begins
to set into snowcrete. After all blocks are set in place, the final chore is to
throw loose snow over the completed house with a shovel, taking care to get none
struck on the very top of the dome. This snow slides down and forms, as it
finds its natural angle of repose, a banking perhaps three feet thick at the
base, thinning upward.
The door is provided by tunneling under the wall to the well inside the house.
A bed platform, occupying about two-thirds of the interior floor space, is built
of snow. The level of this platform is at least eighteen inches above the top
of the entrance, as this is required for gravity control of the temperature
within doors. This eighteen-inch minimum difference relationship between the
level of the bed platform and the entrance is always strictly adhered to, other- FIGURE 2 FIGURE 3

EA-I. Roberts: Housing

wise a closure for the door would be required.
For sleeping, the bed platform is covered with some good insulator, such
as a double layer of skins, with the fur side down on the bottom skin and up on
the top one. This arrangement prevents body heat, which would cause thawing,
from passing to the snow underneath.
Warmth for the snowhouse is provided by the use of an alcohol or seal oil
lamp, or by a kerosene or gasoline stove. The colder the weather the hotter the
house may be kept without melting the roof, for the cold from outside penetrates
into the snow blocks to meet the heat from within and prevents melting. To glaze
the inner wall surface, the temperature in the house is raised high, immediately
after bedding has been placed in position; the inner surface of the walls, which
absorbs moisture like blotting paper, now becomes spongy. Then the fire is put
out, or a large hole is made in the roof, and the inner wall is allowed to freeze.
This glazes the inside with a film of ice, provides a hard protective inner sur–
face which does not crumble when the occupants touch it, and greatly increases
the strength of the dome.
Ventilation is controlled by the gravity method. A wooden chimney or
vent is placed in the dome, a little off center. When the lamp or stove is not
burning, the vent is closed by stuffing mittens or a cloth in the vent, but when
fires are in use the vent is opened as required. The upward movement of warm
air through the vent allows cold air to enter from below through the door passage.
Through this method the rate of change of air and the temperatures of the room
are controlled.
The interior of a snowhouse is sufficiently illuminated to permit the carry–
ing out of routine tasks during daytime hours by daylight passing through the

EA-I. Roberts: Housing

translucent walls. The full moon in a clear sky also gives enough light inside
the house for dressing and even for cooking. During periods of darkness, the
inside of the house can be illuminated by one ordinary candle, for the light is
reflected from the many crystalline surfaces of the snow dome, diffusing uniformly.
The principal advantages of the snowhouse are that the transportation of
building material or tent is not required, no additional equipment is needed for
construction, four men can build [: ] a comfortable camp big enough for six in a
little more than an hour, any adaptable man can be taught snowhouse-building in
one day, a minimum of heating is required to maintain a livable temperature,
clothes stay dry, and the dwelling is practically soundproof to such noises as
men shouting or dogs barking, permitting more restful sleep.
Though the dome snowhouse was probably an Eskimo invention, it was never
universally used by them. It seems likely that, before the coming of Europeans,
more than half the Eskimos had never heard of a dome snowhouse, or had learned of
it only by remote hearsay. Possibly less than twenty per cent of them had lived
in such houses, and then only during periods of intense cold. Perhaps another
ten per cent knew how to build snowhouses, but used them only in emergency.
Geographically, the dome d snowhouse was formerly unknown, unless by hearsay, to
Eskimos in northeastern Siberia, in all of Alaska, and in all of Greenland except
the extreme northwest corner, north of Melville Bay. Nowadays all Eskimos know
of the snowhouse through reading, photographs, and movies. The best snowhouses
are built in the area between King William Island and Coronation Gulf, in the
Canadian archipelago , and on the mainland.
We cannot, therefore, expect that Eskimos in every area will be able to

EA-I. Roberts: Housing

construct proper snowhouses. Accordingly, travelers and intending northern
residents should learn the art of constructing this type of temporary housing
themselves. In Sweden, since 1933, dome-shaped snowhouses have been successfully
used by skiers who tour the mountains, and they were early introduced to the
Swedish army; since about 1940, the construction of snowhouses has been taught
the armed forces of most countries that engage in exercises in polar and subpolar
regions.
Tents have long served as dependable shelters, and both fabric and skin tents
have been widely used in the Arctic. While the warmth of a tent, like clothing,
depends on air spaces, it does not seem practical with fabric tents to have these
air spaces within the fabric itself. The desired insulation quality is therefore
secured by the use of lightweight windproofed double tents. Skin tents derive
their warmth largely from the air spaces between the hairs, and, in the case of
caribouskin tents, additional warmth is provided through the hairs themselves,
these being hollow and filled with air. Under extreme cold conditions, double–
layer skin tents are used and so erected as to allow about a one-inch layer of
air between the skins. This is accomplished by placing the fur sides of the skin
toward each other.
Dome or umbrella-shaped tents are to be preferred over peak-shaped ones such
as the tepee (tipi), ridge, or pyramidal forms. Peak-shaped tents allow the
warmth to gather at the top of the tent where occupants do not have the benefit
of the heat. However, peaked tents are greatly improved in warmth if a false
ceiling of fire resistant fabric is placed not more than seven feet above the
floor, to prevent the heat from rising to the peak. Floors for tents, if provided,

EA-I. Roberts: Housing

should be furnished as a separate piece, or pieces, from the rest of the tent.
Floors attached to the walls of tents soon become covered with frozen condensa–
tion and snow, and cannot be cleaned easily. Conventional flap-type openings,
with ties in the front of the tent, provide a more satisfactory entrance than the
door which ties shut like the mouth of a duffle bag. The latter works only when
the material is either not frozen or else frozen and dry. When moisture gets into
the entrance, at below zero temperatures, the mouth of the bag becomes increasingly
difficult to tie and finally becomes entirely inoperative.
All tents for per s onnel should be provided with stoves, especially where
driftwood is abundant on a coast or along a river, or where bushes and trees grow.
Asbestos insulation should be placed around the aperture in the overhead layers
of canvas or skin through which the stovepipe passes. This asbestos should extend
at least four inches beyond the stovepipe. For wood-burning stoves the top–
most length of stovepipe should have a medium-mesh wire screen to keep sparks
from falling on the tent and burning holes in it; or a fireproof material should
be used.
Guy-rope fastening tabs should be reinforced. Guy ropes should be made of
nylon and should be provided with a toggle at the end of each rope. These toggles
may be buried and tamped into the snow or frozen into a trench out into the ice.
The use of deadmen is often advisable. If tent pins are used, a light weight
metal tent pin is preferred to wood or other such material, as the metal pin
may be driven into ice or moderately hard earth without damage.

EA-I. Roberts: Housing

PRINCIPLES OF DESIGN OF TEMPORARY AND SEMIPERMANENT BUILDINGS
General
Important factors which will influence the design of buildings are the
materials available; the temporary or permanent nature of the construction; the
time available in which to construct; the purpose for which the buildings are to
be used; and the location and orientation of the structures. The operational
requirements will, in many cases, determine the type of building that will be used,
but the fundamental principles that influence the design will apply to all build–
ings. All housing should provide adequate space, utility, heating, insulation,
ventilation, and , if required, portability. All building construction in the
Arctic and Subarctic must meet certain basic - requirements peculiar to these
regions. These requirements are fixed by the special climatic, geological, and
topographical conditions, such as low temperatures, high winds, drifting or
deep snow, long periods of cold, absence of direct sunlight in midwinter and of
darkness in midsummer; foundation conditions affected by permafrost and related
phenomena; and poor drainage. (See “Work Feasibility.”) As a consequence, all
polar building design and construction possess certain common factors and
characteristics.
Foundations for buildings to be erected in permafrost or adjoining areas
should be carefully designed after considering all factors. (See “Engineering
Problems and Construction in Permafrost Regions” and “Excavations and Foundations.”)
Surface and subsurface drainage under the buildings must be avoided entirely
or [: ] else it must be adequate in all seasons for conveying surface and groundwater
away from the site.

EA-I. Roberts: Housing

Building frames must be designed and constructed to resist stresses caused
by high winds. Resistance to high wind velocity, where required, may be pro–
vided by designing for a 20 lb./sq.ft. wind load, and, in areas of extremely
high winds, roofs should be anchored by guys. Consideration should be given to:
2. Anchoring walls or structures to foundations by bolts, reinforcing
steel, deadmen, cables, or cleats.
4. Tying roof trusses and rafters securely to the frame.
6. Using knee-braces to prevent the frame from leaning sideways.
8. Using diagonal sheathing for rigidity.
Roofs must be able to withstand strong winds and heavy snow loads. Both
may occur at the same time. Roof framing should be designed for a snow load of
30 lb./sq.ft. and strengthened if partial clearing of snow is not possible.
The outer shell of the building must be impervious to wind, rain, and snow.
High winds have a destructive effect on roof coverings, particularly on the low–
pressure leeward side. The best roofs are metal, laid from the leeward to wind–
ward side and coated with hot asphalt. Another satisfactory roof is the built-up
type, consisting of a base sheets of 30- to 90-lb. bituminous-saturated felt
nailed with large-head roofing nails spaced not more than 6 inches apart horizon–
tally and 12 inches apart vertically, followed by one or two layers of 15-lb.
bituminous-saturated felt mopped on to the base sheet with hot asphalt, over
which battens or wood strips must be nailed to hold down the covering.
Eaves . Where there are extremely high winds, eaves should not be used.
Insulation must be complete throughout walls, ceilings, and floors to insure
proper heat retention, and to prevent the thawing of the permafrost under the
building.

EA-I. Roberts: Housing

A ventilated air space must be provided between the building floor and the
ground, or an insulated cover must be placed on the ground, before the floor
is laid, to preserve the thermal equilibrium of the ground and prevent thawing
below the foundation base.
Positive ventilation and circulation of air must be maintained at all times.
This may be accomplished by the proper use of windows, doors, and vents. Window
and door openings should be small and as near the ground as is possible, to
minimize the loss of heat, and aid in controlling ventilation. All ext r erior
doors should open inward, as outward-opening doors may be blocked by snow, barring
egress, or blown off by high winds. Double doors or a storm vestibule must be
provided for each entrance, unless gravity control of air is used.
Heating . All personnel, administration, and shop buildings, and many
storage buildings, require heating. Due to the cost of construction of distribu–
tion systems, and heat losses incident thereto, individual heating units are
preferred, except for large installations. In the Arctic, due to the scarcity
of wood, coal- or oil-burning stoves are used; south of the tree line, coal-,
oil-, and as well as wood-burning stoves also may also be provided. (See Mechanical-Electrical
section for discussion on this subject.)
Fire-fighting measures and fire prevention should be carefully considered
in regard to all buildings and installations.
Clothes-drying facilities are required for all buildings housing personnel.
These may be provided near the entrance to sleeping quarters; where possible, warm
air used to dry the clothes should be exhausted from the building.

EA-I. Roberts: Housing

Insulation
By “insulation” is meant material used, in addition to ordinary building
materials, for the purpose of retarding the flow of heat through the walls,
floors, or roofs of buildings. (Insulation is also used to reduce the transmis–
sion of sound through walls, partitions, ceilings, or floors, but sound insula–
tion is not considered in this article. ) The primary purpose of thermal insula–
tion is to permit the maintenance of the desired interior temperature in an
economical manner, to keep buildings warmer in the winter by reducing heat
losses, and to keep them cooler in the summer by reducing the penetration of heat
from the outside. While maintaining the desired interior temperatures, insula–
tion is a major factor in retarding or preventing condensation and [: ] decreasing
or eliminating drafts.
Obviously, good insulation results in fuel economy during cold weather.
Heat losses occur in a building by conduction and by air passing through cracks,
doors, and windows. Insulation retards this flow of heat.
Ordinary building materials have some insulating value, but are not nearly
as effective as the special insulating materials we shall discuss. For example,
1 inch of insulation board retards the flow of heat as much as 2 to 3 inches of
lumber, 15 inches of brick, 23 inches of cement plaster, or 37 inches of concrete
made of sand and gravel.
Types of Insulation. Various substances are used as insulation material.
The substances most commonly used are vegetable or mineral products. As a general
rule, greater insulation is obtained from the lightest weight materials. Ordinar–
ily, lightness in weight is attained by entrapping tiny air cells in and between
the insulation substances. Heat travels relatively slowly through these air spaces,

EA-I. Roberts: Housing

since air is a very poor conductor. Insulation materials are manufactured in
various forms and shapes for case in application. They fall into the following
general classifications.
1. Flexible Insulation. This type of insulation is made of materials
manufactured into flexible blanket, quilt, or bat form, usually with a nailing
flap along each edge for fastening between ribs or studs. The materials most
commonly used for flexible insulation are: aluminum foil (see under “Reflective
Insulation”); animal hair, available as 100% pure hair or mi s x ed (felted) with
other fibrous materials, such as asbestos or jute, and stitched between sheets
of kraft paper; cotton, flameproofed and backed with water-resistant paper;
eel grass, a marine growth, processed and stitched between kraft paper; glass
fibers, interlaced into a wool d -like mass and backed with water-resistant paper;
mineral wool, felted from rock wool fibers and encased in water-resistant paper;
rock wool, blown from molten rock or molten furnace slag and encased in kraft
paper; wool fibers, manufactured into cre e p ed individual plies which are stitched
together, treated for fireproofing, and encased in water-resistant paper, or
asphalt-treated and flameproofed.
2. Loose Fill or Blow-on Insulation. This type of insulation is made in
loose form. Installation is accomplished by filling the spaces between ribs or
framing members or by blowing it onto the inner surface of the outer covering
and on structural members. The most commonly used materials for loose or fill
type insulation are: cellulose fibers, mixed with adhesives when blown on;
cork, granulated or in powder form; redwood bark, ground into fleecy fibers;
rock wool, granulated or in loose form for either fill or blow-on; vermiculite,
mica ore, highly expanded by heat; wood pulp, processed to resist fire, moisture,
and vermin.

EA-I. Roberts: Housing

This type of insulation is not recommended for use in walls where the outer
or cold wall surface is more resistant to the passage of moisture vapor than the
inner or warm wall surface of the building. This insulation is pri m n cipally used
on horizontal surfaces, as above ceilings and for filling wall cavities where
access is a problem.
3. Rigid Insulation. Materials are manufactured into rigid shapes of
varying thicknesses and marketed as corkboard, insulating building board, insulat–
ing lath, insulation plank, insulation sheathing, insulation tile, insulation
plaster or concrete, and roof insulation. The materials most commonly used are:
cane fibers, waterproofed and felted together; cork, compressed under heat; glass
fibers, compressed under heat and parcels coated with a waterproofing compound;
v V ermiculite, mica ore highly expanded by heat and mixed with plaster or cement;
wood fibers, felted together and pressed into solid form, usually treated for fire,
moisture, and vermin resistance.
Rigid insulation gives good results when used in temporary and prefabricated
building, as discussed in this article. It insulate d s the inner flanges of the ribs
or studs, and, since it is rigid, generally serves as an interior lining as well
as insulation.
4. Reflective Insulation. This type of insulation is distinguished from
the other types in that its effectiveness is based on its ability to reflect
radiant heat waves. Reflective insulation, therefore, should be installed with
the reflective surface exposed to an air space. Reflective insulation is made
into thin sheet or foil form and generally applied between the structural ribs or
studs and rafters. Materials most commonly used for reflective insulation are:
aluminum foil (thin sheets) or foil backed with a kraft paper; foil on one or
both sides of kraft paper laminated with asphaltum; foil separated by air cells
which are formed by crimped paper; foil laminated to [: gyp ] gypsum wallboard,

EA-I. Roberts: Housing

lath, or sheathing; steel (thin sheets) coated with a lead-tin alloy; paper,
laminated with asphaltum, reinforced with cords or fibers, and coated with
aluminum paint.
This type of insulation generally provides a vapor barrier in addition to
its insulation qualities. Dust settling on the reflective surfaces reduces the
insulation values. It has been found that this type of insulation is most
effective when installed horizontally and when the flow of heat through the
insulation is downward. Therefore, this type of insulation is most effective
when used to keep out summer heat or if used to insulate the floor in winter.
Insulation Values . The heat conductivity of a material is a measure of
its insulating value; the lower the conductivity the more effective (see Table I).
Thermal conductivity, k , is the amount of heat in B.t.u. which will flow in one
hour through a layer of material one [: ] square foot in area, where the
temperature difference between the surfaces of the layer is 1°F. per foot of
thickness. Thus the insulating value of a layer of material depends upon the
thickness of the layer as well as the k value of the material. In general, the
various types and makes of insulation are rated by their ability to resist the
flow of heat. A resistivity value, r , has been established for the heat resis–
tance offered by a unit area and unit thickness of each type of insulation material.
This is equal to the unit thickness of a layer of material divided by the k
value of the material. The insulating ability of one material may be compared with
the insulating ability of other insulating materials by comparing their resis–
tivity values. The larger the r value, the greater the insulating ability.
When making such comparisons it must be remembered that only the insulating
materials themselves are being compared and no consideration is being given to
the effectiveness of the entire installation. After application, the r value

EA-I. Roberts: Housing

Table I. Thermal Conductivities and Thermal Resistivities of Building Materials.
Materials Weight,
lb./cu.ft.
Conductivity,
k
Resistivity,
r
Insulating
rating
Flexible insulation
Animal hair (felt) 12.0 0.26 3.84 Excellent
Cotton bats 0.875 0.24 4.17 Excellent
Eel grass in paper 3.40 0.25 4.00 Excellent
Glass wool bats 1.50 0.27 3.70 Excellent
Jute fiber bats 6.70 0.25 4.00 Excellent
Kapok bats 1.00 0.24 4.17 Excellent
Kimsul bats 1.50 0.27 3.70 Excellent
Mineral (rock wool) 10.00 0.27 3.70 Excellent
Peat moss compressed 11.00 0.26 3.84 Excellent
Wood fiber bats 3.62 0.25 4.00 Excellent
Loose insulation
Ceiba fibers 1.90 0.23 4.35 Excellent
Cork granulated 8.10 0.31 3.22 Excellent
Redwood bark 3.00 0.31 3.22 Excellent
Rock wool 10.00 0.27 3.70 Excellent
Sawdust (various) 12.00 0.41 2.44 Excellent
Shavings (various) 8.80 0.41 2.44 Excellent
Vermiculite 6.20 0.32 3.12 Excellent
Rigid insulation
Asbestos fibers 48.30 0.29 3.45 Excellent
Cane fibers 13.50 0.33 3.03 Excellent
Cork, compressed
asphalt binder
14.50 0.32 3.12 Excellent
Corn stalk fibers 15.00 0.33 3.03 Excellent
Glass fibers - 0.40 2.50 Excellent
Wood fibers:
Exploded 17.90 0.32 3.12 Excellent
Hard W w ood 15.20 0.32 3.12 Excellent
Licorice R r oot 16.10 0.34 2.94 Excellent
Various 15.00 0.33 3.03 Excellent
Reflective insulation
Foil with air space
with heat flow up:
1 air space - 0.32 3.13 Excellent
2 air spaces - 0.27 3.70 Excellent
3 air spaces - 0.17 5.88 Excellent
Foil with air space
with heat flow down:
2 air spaces - 0.10 10.00 Excellent
3 air spaces - 0.07 14.29 Excellent

EA-I. Roberts: Housing

Table I. Thermal Conductivities and Thermal Resistivities of Building Materials.
Materials Weight,
lb./cu.ft.
Conductivity,
k
Resistivity,
r
Insulating
rating
Other building materials
Asbestos millboard 60.5 0.84 1.19 Good
Asphalt roofing 55.0 0.70 1.43 Good
Brick, masonry (wet) 120.0 ± 5.00 0.20 Fair
Cement mortar 100.0 12.00 0.08 Poor
Concrete, typical 142.0 12.60 0.08 Poor
Concrete, cinder 97.0 4.90 0.22 Fair
Concrete (1:2:4) 143.0 9.46 0.11 Poor
Concrete, expanded
vermiculite aggregate
20.0 0.68 1.47 Good
Concrete pumice aggregate
(mined in California)
65.0 2.42 0.41 Good
Gravel, loose dry 106.0 7.50 0.13 Fair
Ice, fresh-water 57.5 15.10 0.07 Poor
Limestone 124.0 6.45 0.15 Fair
Snow, compacted 15.0 + 0 50.0 1.48 0.67 Good
Soil, dry-packed 94.0 2.60 0.38 Good
Woods (across grain):
Balsa 20.0 0.58 1.72 Excellent
Douglas fir 34.0 0.67 1.49 Good
Hemlock 22.0 0.73 1.37 Good
Oak 38.0 1.02 0.98 Good
White pine 31.2 0.78 1.28 Good
Yellow pine (short leaf) 36.0 0.91 1.10 Good
Yellow pine (long leaf) 40.0 0.86 1.16 Good

EA-I. Roberts: Housing

of the insulating material is but one factor in the over-all effectiveness of
the installation. The number of air spaces created, the resistances of the
various building materials themselves, and the surfaces exposed, all contribute
to the over-all effectiveness of the insulation provided for the building.
Condensation
By condensation is meant the moisture which is deposited by the air on the
inside walls or roof of a building. It does not necessarily occur at all times
in buildings but is the more rapid the larger the temperature differences,
especially during periods of relatively high humidity. Air is a mixture of gases
such as nitrogen, oxygen, etc., and water vapor, all of them normally invisible.
The gases occur in comparatively fixed relationships but the amount of water
vapor in air varies greatly. This variation depends on the source of water vapor
and upon the temperature of the air. Air at higher temperatures can carry more
water than at lower temperatures.
Condensation is caused when air, carrying water vapor, is cooled to a
temperature at which it no longer can carry all of the water which it previously
contained. This is the dew-point temperature. When the dew point is reached,
the excess water in the air changes from water vapor to visible liquid or solid
water. This transition is called condensation.
Table II gives the maximum weight of water which one pound of air at various
temperatures can absorb and carry.

EA-I. Roberts: Housing

Table II.
Temperat o u re, °F. Carrying capacity of 1 lb, of air
in grains* of water
−50 0.29
−40 0.55
−30 1.02
−20 1.84
−10 3.22
0 5.51
10 9.20
20 15.06
30 24.18
40 36.49
50 53.61
60 77.56
70 110.74
80 156.31
90 218.3
100 302.3
110 416.1
120 570.4
130 781.2
140 1,073.8
150 1,487.5 (0.21 lb.)
5
The moisture-carrying capacities shown are for the same air pressure so as to run in this para with preceding page
give comparative figures. Notice that a pound of air can absorb and carry
110.74 grains of water at 70°F. The same pound of air can absorb less than numbers changed in this paragraph to conform with numbers in table.
half of this amount at 50°, less than one-fourth at 30°, and less than one-tenth
(9.20 grains) at 10°F.
The table gives the maximum amount of moisture that air can carry at the
temperature shown. When carrying this maximum , the air is called saturated and
has a relative humidity of 100%. Usually air is not fully sat i u rated with water
vapor and therefore seldom has a 100% relative humidity. The ratio of the amount
of water vapor in the air to the amount which it can hold when saturated is called

EA-I. Roberts: Housing

the relative humidity. If the relative humidity is 50%, the air contains one–
half of the water which it can carry when fully saturated at the same temperature.
Outside air brought into a building and heated will have a drying effect, even
though rain is falling and the air outside is practically saturated. For when
saturated air is warmed, it is no longer saturated but becomes capable of carry–
ing more moisture.
At this point it is perhaps well to consider the sources of this water vapor.
Water may exist in three forms. We usually think of it in liquid form. It also
exists in solid form, which is called ice, or it may exist in a gaseous form,
which is called water vapor or steam. Steam is invisible; the white cloud
commonly called “steam” actually is condensed steam and consists of visible
particles of water.
With no water vapor in the air, condensation could not take place. One hun–
dred per cent dry air is difficult to obtain even in a laboratory and such air is
not desirable from the human living standpoint. All outside air, even that in
a desert, carries a certain amount of water vapor; the air of warm deserts is
seldom as deficient in water as that of a fifty-below-zero day of the northern
winter. During a rainstorm, the outside air is practically saturated. The amount
of moisture in outside air, therefore, varies greatly depending on the surround–
ings, the temperature, and the atmospheric conditions.
In addition to the moisture which is in the outside air, water vapor may be
liberated inside of buildings from the bodies of the occupants or through other uses
of the building. Because water vapor is low-pressure steam, any operation such as
cooking, washing, bathing, or drying clothes adds to the amount of water vapor in
a building. Potted plants give off moisture. Animals and humans also give off a

EA-I. Roberts: Housing

certain amount of moisture, depending upon how fast they use their energy. A man
at rest liberates 0.2 lb. of water per hour while a man at normal work liberates
0.6 lb. per hour. A man walking 4 miles per hour liberates 0.9 lb. per hour.
In addition to the moisture in outside air and moisture from building uses,
water vapor is also produced by damp surroundings, wet floors, etc. If condensa–
tion problems persist within any building, it is advisable first to determine if
the sources of water vapor can be controlled. If these are not subject to regula–
tion, it becomes necessary to control the water vapor itself.
Methods of Condensation Control . Condensation may be controlled or entirely prevented in buildings by main–
taining the inner surfaces at or above dew-point temperatures, or by reducing the
relative humidity. These two objectives are attained by various methods.
1. Insulation. From a review of the moisture-carrying capacities of air
at various temperatures, it is obvious that if interior air does not become
cooled when it comes in contact with the walls or roof of a building, it will not
lose its moisture-carrying capacity, so condensation will not occur. This method
of control requires insulation to keep the inner surfaces of the walls and roof
near room temperature. In such installations, the moist [: ] inte r ior air comes in
contact with the inner surfaces of the insulation and not with things like cold
metal in walls or roof. Condensation is prevented because the interior surfaces
of the insulation are near room temperature. Care must be taken to install a
vapor barrier on the warm side of the insulation, as described below, in order to
retard the filtering of water vapor through the insulation itself.
2. Ventilation. It is obvious from the above that condensation will not
occur if there is no difference in temperature between the inside and the outside

EA-I. Roberts: Housing

of a building. This sameness of temperature inside and outside is most easily
maintained by ventilation, merely by keeping doors, windows, louvers, or venti–
lators open. However, in certain instances, forced or mechanical ventilation
may be required. It is obvious that this type of control is a summer method; due
to the low temperatures involved in an arctic winter this method is not satis–
factory at that season if warm buildings are required.
3. Humidity Regulation. Condensation can be controlled by reducing the
amount of water vapor in the air (relative humidity), since condensation will
not occur if the amount of water vapor is kept within the air’s ability to carry
the vapor at the lowest air temperatures. Certain uses of such buildings as
kitchens, laundries, and bathrooms create a great amount of water vapor as a by–
product of operation. In such installations, the increasing amount of water vapor
has a tendency to saturate the air and finally to condense, if not removed. Fre–
quent changes of interior air are [: ] recommended to prevent the accumulation of
these vapors. It is desirable, in certain installations, to remove the excess
vapor at its source by installing special exhaust ducts. However, the excess can
be removed in most instances by opening windows, ventilators, louvers, or by the
use of exhaust fans. By such operation, the cooler outside air enters the build–
ing; when warmed, it is able to absorb additional moisture. As it leaves the build–
ing, in its warmed condition, it carries out more water vapor than was brought
in by the cooler incoming air. In the Subarctic and Arctic, where extreme tempera–
tures of −50°F. may be expected, the relative humidity may be as high as 80 or
90%. The actual moisture content of the air at this temperature is slight. Air
with a temperature of −50°F. if taken into a building and warmed to 70°F., without
moisture being added, will have a very low relative humidity, possibly less than 5%.

EA I. Roberts: Housing

The amount of water vapor in the interior air may also be reduced by dehydrat–
ing agents such as silica gel, etc., or by mechanical dehydration processes. De–
hydration by these methods is a relatively expensive procedure.
4. Vapor Barrier. If the water vapor, which is in the air, is kept from
coming in contact with the cooler walls and roof, no condensation can occur.
Keeping the moisture of the interior air from the cold walls and roof is accomplished
by installing a vapor barrier such as a waterproof membrane or diaphragm between
the interior air and the walls or roof. Obviously, the vapor barrier must be
kept at a temperature above the dew point or else the water vapor will condense
on the barrier itself. It is, therefore, necessary to place the vapor barrier
on the warm side of the insulation.
Table III lists the vapor transmission rates of the better-known materials;
the lowest transmission rates [: ] how the most effective vapor barriers.

EA-I. Roberts: Housing

Table III.
Materials Transmission of water vapor
in grains/sq.ft./hr./in. of
mercury pressure difference
Double-faced aluminum foils surfaces 0.085 - 0.129
Roll roofing (smooth surface);
40# to 65# per 100 sq.ft. roll 0.129 - 0.171
Asphalt-impregnated sheathing paper:
50#/500 sq.ft. roll 0.213 - 0.770
35#/500 sq.ft. roll 0.171 - 2.060
Insulation back-up paper, treated 0.660 - 3.420
Gypsum lath with aluminum foil backing 0.085 - 0.385
Plaster and wood lath 11.000
Plaster and 3 coats lead and oil paint 3.680 - 3.850
Plaster and 3 coats flat wall paint 4.280
Plaster and 2 coats aluminum paint 1.150
Plaster and fiberboard or gypsum lath 19.700 - 20.600
Plywood, ½ in. (5-ply Douglas fir) 2.670 - 2.740
Plywood, ¼ in. (3-ply Douglas fir) 4.270 - 4.620
Plywood and 2 coats asphalt paint 0.430
Plywood and 2 coats aluminum paint 1.290
Mineral wool, 4 in., unprotected 29.100
Blanket insulation, ½ in. and 1 in., between
coated papers
1.920 - 2.000
Brick masonry, 4 in 1.100
Insulation is recommended for keeping the inner surfaces of the walls and
roof at near room temperatures. Because water vapor filters through nearly all
types of insulating material, as shown in Table III, a barrier should be placed
on the interior warm surface of the insulation material. This will keep most
of the moisture from filtering through the insulation to the colder part of
the well.
Asphaltic lining or coating is effective as a vapor barrier. Thin sheets
of metal, well joined at seams, make excellent barriers. Many insulation materials

EA-I. Roberts: Housing

are manufactured with a vapor barrier on one surface; some insulation boards
have a vapor barrier on both surfaces.
Two coats of oil paint or varnish over a well-sized smooth surface is an
effective vapor barrier; aluminum flakes added to the paint [: ] make it still more
effective. It must be remembered that this type of barrier is effective only
as long as it is free from cracks and checks, and that filling cracks by repaint–
ing or otherwise will renew effectiveness.
Special vapor barriers are required in the design of buildings for the Arctic
because ordinary building materials are vapor permeable. Under extreme tempera–
ture conditions, creating wide differences of air pressures at temperatures like
−50°F. and 70°F., there is a natural tendency for the unbalanced to equalize.
The warmer interior, having the higher vapor pressure, loses moisture at a ratio
in accordance with the vapor permeability factors of the wall structure material.
To eliminate this condition a barrier consisting of a layer of waterproof membrane
or other highly impervious material should be installed on the warm side of the
interior, usually under a layer of wallboard or other interior wall finish.
The vapor barrier must be kept at a temperature above dew point or else the
vapor will condense on the barrier. Care must be taken to assure that bolts,
nails, or other metal fasteners do not extend through the barrier into the outer
wall or center air space provided in the wall section or roof. The absence of an
effective vapor seal will allow moisture to condense and freeze within walls or
under roofs. Alternate freezing and thawing under such conditions will cause
buildings to deteriorate rapidly.

EA-I. Roberts: Housing

Ventilation
Ventilat ing ion needs depend on the type of housing and the use to which the
housing is put. Wherever men live and operate within doors, the gaseous products
of respiration and combustion should be removed as quickly as possible. This may
be accomplished either by natural or by mechanical methods of displacement of
the vitiated air and its replacement by fresh air.
Natural Ventilation . The two natural forces for moving air into, through,
or out of buildings are the wind and the gravity forces. The incoming air mixes
with and dilutes the foul air to a point suitable for healthful respiration.
In considering wind forces, the velocity and direction of wind and the near–
ness of hills and other obstructions must be considered. However, the ventilation
of most of the housing discussed in this article is induced mainly by gravity —
that is, by the thermal head produced by the difference between the density of
the column of air in the outlets and in the outside atmosphere. The size,
location, arrangement, and control of the inlet and outlet openings of a build–
ing should therefore be such that these two forces act together to provide
positive ventilation rather than in opposition. The most effective gravity
ventilation is obtained by the use of roof ventilators. The largest flow per
unit area of openings is obtained when the outlets and inlets are equal, but in
the Arctic that condition is not desirable because too much loss of heat would
result.
In mechanical ventilation the movement of air is positively maintained by
means of various types of fans, driven by an engine or by an electric motor.
This type of ventilation finds wide use in the Arctic for garages, shops, hangars,
mess halls, kitche sn ns , theaters, and special buildings, requiring rapid air change.

EA I. Roberts: Housing

It must be remembered that ventilation imposes a load on the heating plant
of a building. When the heated air is replaced with air from the outside, the
incoming air must be heated. In the Arctic this is a major factor in the total
heating load. For example, under extreme arctic winter conditions for a building
20 ft. wide, 48 ft. long, and 8 ft. high, one complete air change per hour in–
creases the heating load something like 25%, so that four complete changes would
mean a 100% increase. It is, therefore, necessary to hold ventilation to the
minimum consistent with physiological requirements. The following absolute
minimum ventilation requirements should be considered.
Garages: supply or exhaust one cubic foot per minute per square foot of
floor space.
Shops: same as garages, except that paint shops should be provided with
supply and exhaust fans for year-round operation, and should be capable of
effecting 20 changes of air per hour.
Hangars: 1 change of air per hour supplied usually by infiltration.
Kitchens: 10 air changes per hour.
Mess Halls: 2 air changes per hour.
Residences: 1 air change per hour.
Theaters: 10 c.f.m. per person fresh air introduced. Chapels and auditoriums
same as theaters.
Barracks: 1 air change per hour or 10 c.f.m. per person if men are crowded.
Toilets: 10 changes per hour.
Laundries: 3 0 changes per hour.
Hospitals: 12 changes per hour.

EA-I. Roberts: Housing

TEMPORARY HOUSING
The preceding discussion has dealt with the general principles of design of
housing for arctic and subarctic regions with important considerations regarding
insulation, condensation, and ventilation. It must always be remembered that
restrictions imposed on the weight and the cubage of construction material by
transportation facilities act as a checkre in on elaborate design. Even more
critical is that building in the Arctic is accomplished in an area in which snow,
earth, and a meager supply of wood are usually the only local materials available,
though rock materials do abound in some places. Simplicity of design, therefore,
is the factor that influences all buildings and particularly those that are to
be temporary.
Temporary buildings can be constructed from snow and from other more stable
[: ] local materials. Such buildings are the domed snowhouse, and the house of
earth and wood. These two dwellings, with skin tents, are the basic types of
housing used by most of the Eskimos since ancient times. The domed snowhouse
and skin tents have been described under “Emergency Housing.”
The earth and wood house may be built with driftwood along the coast, and
along rivers, or with the use of felled trees in the northern forest. The size of
the house will be determined largely by the availability of wood; the Eskimos
usually have a long alleyway and one proper room from nine to sixteen feet square,
with a seven- or eight-foot ceiling at center. Where two rooms are used, the first
is usually the smaller and considered to be a storage room; the inner or larger
room is the house proper.
The site for the house should be on a rise, not in the lee of a hill, with
Fig. 5 4 drainage away from the building (see Fig. 3 4 ). Posts are erected forming the FIGURE 4

EA-I. Roberts: Housing

corners of the house. In selecting the posts, ones with crotches should be used.
Posts are set in place by thawing the ground through alternately building a fire
over the spot and then excavating. If equipment is available, holes may be sunk
by using a steam point, the steam supplied by a hose from a low-pressure boiler.
With either method, when the hole reaches the desired depth, the post is set;
thawed earth is packed around and tamped. It is sometimes necessary to brace
each post until it freezes in position. Logs are now placed between the crotches,
forming the plate or top of the walls. Sticks and split logs are set off the
vertical between the ground and the plate, leaning slightly toward the center
of the building and forming vertical sheathing. This is necessary, as the wall
framework leans slightly inward to support the earth that is later to be heaped
against it.
Four posts are now erected in the center of each room, forming a three–
to five-foot square. These posts should be longer than the posts used in the
walls, to provide a slope to the roof of fifteen to thirty degrees. The tops of
the posts are connected by logs which become the ridge members. The square
formed by the ridged logs becomes a skylight. A particularly strong log, the
hip rafter, is placed between the corner of the skylight frame and each corner
post. Other rafters are placed between the plate and the ridge members or the
hip rafters. Roof sheathing should be placed over the rafters to support the
earth covering. Small sticks, tall grass, moss, or fluffy vegetation may be
used for this purpose as only a two-, three- or four-inch layer of dirt is re–
quired on the roof.
An outside entrance or doorway is then excavated from a point about ten
feet away from the wall of the house. This trench should slope gradually to a

EA-I. Roberts: Housing

point just outside the wall of the house, and then run level into the house.
The walls and roof of the trench may be of light material. The doorway to this
entrance need not be airtight, for it is not intended that this alleyway be heated.
Broken sod and loose earth are then piled against the framework of the
building, sloping from about four to five feet at the bottom to about twelve
to eighteen inches at the cave. The rafters are then covered with earth to a
thickness of about six inches at the cave to about two or three inches at the
edge of the skylight. The floor of the house may be earth, split logs, or brush
matting.
In former times, the skylight had a cover made by [: ] sewing together strips
of seal, walrus, or bear intestines; nowadays a thin oiled cloth is used. During
the winter, an ice pane may be used for a subsidiary window, placed low down in
the wall, usually facing south. This windowpane may be cut from a fresh-water
stream or pond. Sea ice may be used, but then the pane will be milky instead
of clear and will not admit as much light.
The vent is installed to one side of the skylight. This pipe may be made
of wood or a hollow log and should be from five to ten inches in width, depend–
ing on the rate of change of air required. The control of temperature and venti–
lation can then be maintained by the gravity method by stuffing a wad into the
vent, thus varying the size of the opening, and controlling the rate of change
of the air that wells up through the trap door from the alleyway, which is never
closed, unless barred to keep dogs out.
For heating the house a small stove is nowadays used; formerly the heating
was by the same whale or seal oil lamps which lighted the house. When using a
stove, a pipe must be installed through the roof. If the house is in a forest,

EA-I. Roberts: Housing

where wood is plentiful, an open fireplace, a hearth of stones, may be built in
the center of the house. The stones absorb heat and aid in controlling the
temperature of the house. When the open fireplace is in use it is necessary to
remove the skylight. After cooking, the fire is quickly removed and the skylight
replaced, whereupon the stones become the active heating agency, giving out
warmth for several hours.
Various types of earth houses may be constructed; the larger bones of whales
and other large sea mammals, especially ribs, shoulder blades, and (whale) jaw–
bones, may be used in place [: ] of wood. In that case, however, the size of the
house will be much smaller than just described.
Sod houses, American prairie style, with skin or brush roofs, and caves ex–
cavated into hills, may provide adequate shelter when no other materials are
available.
Wanigans . For units such as reconnaissance and initial construction crews,
and for over-ice freighting parties, operating away [: ] from camps with heavy
equipment such as tractors, wanigans may be used to advantage. These mobile
structures possess great versatility, can be built of frame or prefabricated
materials, and lend themselves to use in isolated areas. Wanigans serve as
galleys or cook shacks, mobile hospitals, repair shops, water and fuel carriers,
sleeping quarters, offices, and warming shelters for personnel engaged in work
out of doors in cold weather.
A wanigan is a rugged structure, usually about eight feet by twenty feet,
which may be mounted on a sled or “go-devil” and towed by a tractor or other
prime mover. The walls, roof, and floor are insulated, with windows and doors
provided for ventilation and egress. Each unit is heated by a stove. The heavy

EA-I. Roberts: Housing

braced construction of the wanigan enables it to withstand damage from twisting,
vibration, and impact while being towed over a rough terrain.
Jamesway-Portabilt Huts . This type of portable insulated temporary building
has proved to be one of the most flexible and satisfactory prefabricated units
developed to date. The Jamesway hut is versatile in that it is light in weight
and, therefore, can easily be transported by air, sea, and over ice or land.
It can be erected or dismantled quickly by inexperienced personnel and provides
a comfortable building which can be used for temporary housing of personnel,
administration, or storage. These buildings were originally built for the U.S.
Army Signal Corps, during World War II, by the James Manufacturing Company of
S Fort Atkinson, Wisconsin (see Fig. 4). They are now known as the “Portabilt
Shelters” and are manufactured by Portabilt Structures, Inc., of Fort Atkinson,
Wisconsin.
These standard units are sixteen or twenty-four feet wide and may be erected
in any length that is a multiple of eight feet. The framework consists of a
series of laminated wood (glued with urea-formaldehyde resin) or steel arches,
which when erected are attached to the floor panels. Each floor panel consists
of an exterior plywood surface fastened to and supported by a braced plywood
framework. The underside of the panel is faced with a blanket of fireproofed
rock wool or fiber-glass insulation, one and one-half inches thick, protected
by a vinylite resin impregnated, surface-coated fabric. In extreme cold a double
floor is required. The outside layer or roof blanket coverings are made in
eight-foot widths and of the correct length to extend in a continuous covering
over the full arch from base at one side of the building to base at the opposite
side. The roof blankets and end enclosure blankets are reinforced to withstand

EA-I. Roberts: Housing

severe wind and snow-load stresses and consist of an inner and outer surface of
vinylite resin impregnated and surface-coated heavy fabric, which encloses a
one and one-half inch layer of fireproofed rock wool or fibre-glass insulation.
The protective coating windproofs and waterproofs the fabric and does not become
brittle or sticky under extreme cold. These coatings are furnished in any color
desired. Vestibules and storm doors are available and should be included if the
buildings are to be occupied for a prolonged period. Gasoline or oil-burning
heaters are provided with each building.
The site selected for the erection of a Jamesway hut should be one where
natural drainage is away from the building. A level gravel mat should be provided
and well tamped before the end wall base sections and the floor rafters are set
in place. Hardwood or steel stakes are required to anchor the building base
sections and guy ropes. A Jamesway hut may also be used as a wanigan by mounting
it on a sled, “go-devil,” or on timber or pipe runners.
SEMIPERMANENT HOUSING
Prefabricated Housing
Prefabricated housing or portable buildings of a semipermanent nature lend
themselves to future use and large-scale development for the Arctic. South of this
vast polar area, and within reach by air and water-shipping facilities, lie broad
virgin forests, where the demand for prefabricated housing will inevitably foster
the growth of a prefabricated building industry.
The use of prefabricated houses in the polar areas is not new. The first oc–
casion we find on which an arctic expedition equipped itself with a prefabricated
building was also the first modern attempt at colonization in North America. In

EA-I. Roberts: Housing

1578 Sir Martin Frobisher was directed by Queen Elizabeth to undertake a “Thyrd
Viage to the Northwest to fitch 2000 Toones of Oore and to vittal and keepe there
100 Men 18 Monethes.” Writing a “True Discourse” about this abortive excursion
after a gold ore which turned out to be iron pyrites, George Best records that
distributed among the fifteen ships “there was a strong forte or house of timber,
artificially framed, & cunningly deuised by a notable learned man here at
home, in ships to be carried thither, whereby those men that were apointed there
to winter & make their abode ye whole years, might aswel be defended from the
danger of ye falling snow and colde ayre, as also be fortified from the force or
offence of those Countrie people, which perhaps otherwise with too greate
companyes & multitudes might oppresse them.”
Frobisher, gathering his fleet in the proper place, started the miners to
work. “The ninth of August, the Generall with his Captaynes of his counsell
assembled togither, beganne to consider and take order for the erecting vp of
the house or forte, for them that were to inhabite there the whole years, and
that presently the Masons and Carpentees might go in hande therewith. First
therefore they perused the Bils of ladyng, what euery man receiued into his
shippe, and founde that there was arriued only the eastside, and the South–
side of ye house, and yet not that perfecte and intier, for many peeces thereof
were vsed for fenders in many shippes, and so broken in peeces, whyles they
were distressed in the Ise.”
The craftsmen were called in and asked how long it would take to build a
smaller house for Captain Fenton and only sixty men. They answered eight or nine
weeks. “Wherefore it was fully agreed vppon, and resolued by the General and
his counsell, that no habitation shoulde be there this years.” What remained of

EA-I. Roberts: Housing

this building was then buried beneath the gravel on Countess of Warwick Island,
where Hall found it two hundred and eighty-three years later. So, unable to
winter, Frobisher sailed his laden ships home.
Whether discouraged by the fate of Frobisher’s crucial house, or unwilling
to allow that Cathay could not be reached in one season, northwest passage
expeditions for the next three centuries did not supply themselves with pre–
fabricated buildings. True, it is written that Hans Egede brought a portable
house with him to Greenland in 1721, but he came to stay as a missionary. Nearly
every other expedition relied upon converting its ship into winter quarters,
or , like Barents, else, in an emergency , built a house from the boards of the ship . or from driftwood found on shore.
The Russian hunters, who exploited Svalbard after their abandonment by the
whalers, lived in European-style sod huts. As Greenland was Christianized and
slowly developed for trade with the Eskimos, whole villages were erected with
mater ai ia ls, partially prefabricated, sent from Denmark.
By the latter part of the 19th century the attainment of the North Pole
became a sporting venture involving a good deal of national pride. In 1877
Captain Henry W. Ho w gate, U.S. Army, proposed that the United States establish
a polar colony in northern Ellesmere Island which might serve as a base for
a dash to the North. Lacking support, his project failed; but not before a
prefabricated house had been sent as far as Godhavn, Greenland. There it was
picked up in 1881 by Greely’s Lady Franklin Bay Expedition, which took it north
to Discovery Harbor, Ellesmere Island, where it was erected as “Fort Conger.”
The Fort Conger building was 60 by 17 feet and divided into three rooms, a
large living quarters for the enlisted men, a smaller one for the officers, and
a kitchen. Double-walled by half-inch planks, a one foot nonconductive air space

EA-I. Roberts: Housing

was left between. Both inside and outside walls were covered by the heaviest
black tarpaper, the first by tongue-and-groove paneling, the latter by secure
battening. The peaked roof was treated in the same manner as the outside wall,
but beneath it was a ceiling of more tongue-and-groove boarding. The black tar–
paper served to absorb solar heat, while the air space reduced heat conductivity,
and, outside, brumal snowbanks effectively minimized heat loss through radiation.
A bathroom, replete with tub, abutted against the chimney, insuring comfort at
all times. Within two weeks the building was ready for occupancy.
One of the best prefabricated houses used by a late 19th century arctic
wintering party belonged to the Frederick Jackson Expedition to Franz Josef
Land, 1894-1897. A Russian log hut, it had been purchased by Jackson in Archangel
where, to check it, he had ordered it completely set up before taking it north.
Every piece was carefully numbered, and a Russian carpenter was brought along
to erect it.
The central living space of Jackson’s house was only about 13 by 12 feet,
and was surrounded by pony stables and two small bedrooms. The exterior of the
building was thoroughly calked with oakum, and the interior wall covered with
green felt. On the floor lay sheets of heavy brown paper overlaid by patterned
felt, and the seven-foot ceiling was covered by gray felt. The windows were all
double-glazed. Ventilated by two apertures, one consisting of a series of per–
forations low in the south wall and the other a 4 -1/2 ½ -inch hole in the roof,
each easily plugged, it was heated by a large Russian oven and a small slow–
combustion stove. In one corner stood a snow-melting tank, heated by the warmth
of the main room.
The chief drawback of the building was that a site was chosen for it beneath

EA-I. Roberts: Housing

a cliff where dwelt the god, Aeolus, and the continual winds and strong gusts
frequently got a fingerhold on the roof planking, flinging it far and wide.
By no means cast down, Jackson said of “Elmwood:” “Ours is without doubt the
most suitable, best, and most comfortable house ever put up in this latitude.”
To the retreating Nansen, who sought refuge there, it seemed like high civilization.
If “Elmwood” was the most successful of the early prefabricated arctic
quarters, then certainly, in an esthetic sense at least, Peary’s famous “Red - +Cliff House”
cliffe” house was the least so. With tarpaper crudely fastened helter-skelter
on the exterior and weighted down by rocks, surrounded by a low and broken turf
wall, and with various pipes and cans jutting out from unexpected places, it
resembled nothing so much as a kitchen midden. Yet it was shelled by heavy
cardboard and lined with red blanketing which created an intershell air space.
Warm and comfortable, it took but six days to erect.
These early prefabricated buildings for the Arctic left a great deal to be
desired in the way of efficiency. Often erected in permafrost areas, their
internal heat caused the ground to thaw beneath them, resulting in the tilting
or sinking of the structure. Insulation was inadequate, and though the houses
were acclaimed as cozy or warm, excessive amounts of fuel were required to main–
tain comfortable temperatures. None of the buildings were fireproof, an un–
warranted risk in regions where replacement was impossible.
Perhaps the most significant point, however, is that the prefabrication of
these buildings was only partial at best. More often than not it consisted of
supplying merely the required materials, cut and numbered, with erection instruc–
tions. Here the 20th century was to see progress. The first mass production of
prefabricated buildings was carried out by a German firm, Christophe and Unmack,
which shortly after the Franco-Prussian War manufactured dormitories, cottages,
and barracks, sending them to many parts of the world.

EA-I. Roberts: Housing

Modern Prefabricated Buildings
Today there are many prefabricated structures being utilized in the
Arctic to provide housing. Such buildings are in use mainly because they
are easily transported, quickly erected with a minimum of skilled labor,
require a minimum of shipping space, and, while not supplying the o c omplete
answer to the housing problem, have proved worthy under arctic conditions.
Any of these buildings can be improved or modified, depending on the need and
ingenuity of the user. The principal types are the Quonset Hut, the Stressed
Skin Plywood Building, and the Military Arctic Hut.
Quonset Hut . This building (see Fig. 5), which saw universal service with
the U.S. Armed Forces during World War II, was developed by the U.S. Navy and
resulted from the improvement of the British World War I M N issen hut. These
buildings are now manufactured by the Stran-Steel Division of the F g reat Lakes
Steel Corporation, Detroit, Michigan. The huts are primarily a metal structure
with an arch-rib framework supporting corrugated steel sheet covering material.
The basic arctic unit is a building twenty by forty-eight fest, interior
size, with a four-foot vestibule extention on each end of the building. Fixed
wood end walls were provided, framed with steel studs. The end walls now are
built entirely of steel. An additional end wall is required for each end to
form the vestibule. The units are provided with inner insulated walls and the
resistance value of these walls may be controlled to meet the design dictated
by the climatic factors.
The over-all resistivity, r , for a Quonset building may vary from 4.82 to greater than 9.70, depending on the insulating and building materials selected
to meet the design requirements. A typical section of Quonset buildings in use
at Point Barrow, and elsewhere in northern Alaska, provides a total resistivity FIGURE 5 TYPICAL SECTION QUONSET HUT (BARROW)

EA-I. Roberts: Housing

of 9.63 and a total conductive factor of 0.103 (see Fig. 6 5 ). The floor
consists of four- by eight-foot plywood double panels with insulation in
the space between the two thicknesses of plywood. Doors, windows, interior
wall partitions, adjustable vents, and end wall louvers are furnished with
each building, as well as such miscellaneous materials as foundation anchor
bolts, screws, bolts, nuts, washers, clips, and nails. All framing members
receive one coat of rust-resisting paint.
The Quonset hut is simple to build, requiring about one hundred man-hours
to erect the bare structure, with additional time required to provide either a
gravel mat or pile foundation and to install the insulating materials.
Larger buildings of the arch-rib design are available for use. These buildings
are forty feet wide and one hundred feet long, and lend themselves well to use
as storage or recreation buildings as well as for shops. They are easy to
develop into two-story units for administrative or dual-use buildings. The
walls are insulated tih with kimsul and wallboard to meet the design requirement.
A building of the arch-rib type may be erected on a gravel mat, concrete, or
timber pile or sills on top of block-type construction. Concrete or wood
floors may be used and should be insulated to eliminate the transfer of heat
from the building to the foundation or soil below. (See Civil Engineering
Section for subjects related to foundations and construction.)
Stressed Skin Plywood Buildings . Several types of these buildings are
manufactured by the Tower Company of Toronto, and have been set up on the
mainland and islands of the Canadian Arctic from Aklavik to Labrador. They
are light in weight, comparatively inexpensive, and can be quickly erected
by a small crew. Such buildings are prefabricated throughout and consist
of floor, wall, and roof panels, each highly insulated, which when nailed and
screwed together form a rigid frame structure. It is desirable to set the

EA-I. Roberts: Housing

building on top of a gravel bed, two or three feet in depth. The site
selected should be well drained and any vegetation should be left undisturbed.
The building panels consist of a stud frame on which exterior or marine-type
plywood layers are glued and nailed. Between the outer plywoods the panels
contain a sandwich of triple insulation of fiber glass aluminum foil, and
fiberboard to obtain the highest possible insulation against convection,
radiation, and conduction. The panels are held together by special connectors.
All outside joints are covered with felt-lined joint strips which help to
make the buildings airtight. The standard size of the panels is four feet
by eight, and they are light enough to be easily handled by two men. In
special cases, where only a small aircraft may be available to fly the building
components to the site, these panels have been made as small as two feet by
eight feet.
The outside of the building is painted with high-grade aluminum paint which
has proved very resistant to weather action in the Arctic. Aluminum sheet
roofing covers the roof. The gable ends and corners have specially made
flashings to keep the building snowtight. All overhangs, jut-outs, or sharp
corners are avoided to prevent finger-lift action by the wind; because of their
smooth exterior, the buildings withstand storms up to 65 m.p.h. without having
to be tied down. At locations where higher wind velocity and gusts are
expected, tie rods with turn buckles are used to anchor the buildings to the
ground.
The chimneys are either of precast, lightweight concrete, in which ground
mica is used, or galvanized pipe chimneys of the Yukon type. The triple-glazed windows
cannot be opened. Above them is a built-in ventilator with a refrigerator-type
door which provides the necessary ventilation. The buildings are heated with

EA-I. Roberts: Housing

oil-fired space heaters which are placed at a central point to insure
even heat distribution. The floor is covered with colored linoleum and
the walls and ceilings are painted with bright fire-resistant paint.
For larger structures of an industrial nature, like powerhouses, garages,
and warehouses, a different variant of the floating construction method is
used. These buildings generally have no panel floor. They lose, therefore,
the principle of the rigid frame as the floor panels no longer tie the buildings
together. Wherever possible, a reinforced-concrete slab is used as a floor.
Solid bedrock with no water veins or decomposed upper layer is the best
foundation for such a slab; but sometimes a slab can be laid on a gravel
fill which carries a special insulation of celboard (a two-inch thick concrete
insulation material made out of shavings which are sprayed with concrete).
Where the slab is not practicable, the gravel bed itself is used as the floor
of the building. In this case special ballasted outriggers have to be spaced
alongside the building to prevent it from moving or being blown over by the
wind. The outriggers consist of heavy timbers and tie rods which start at the
caves and spread about five feet away from the building. At the bottom they
are connected with horizontal wooden beams to the wall panels. They are spaced
eight feet apart and are ballasted with gravel, and are similar in action to
the outriggers used on South Sea boats. During the winter these outriggers
will catch the snow; the resulting snowdrifts on each side will keep the
building stable and warm.
Military Arctic Hut . A prefabricated hut, light enough to be air-borne,
has been designed to serve as a sixteen-man barracks, or for equipment storage,
in frigid zones, (Fig. 7). It will make possible, with a small outlay of S ✓
fuel, an interior temperature of 70°F. when outside temperatures are as low
as −65°F.

EA-I. Roberts: Housing

In recent tests, the hut was erected in one hour and forty-five minutes
by twelve Naval personnel (Seabees) using arctic mittens. Low temperatures
and the usual white-man type of arctic clothing will probably slow the
assembling operation to more than four hours. However, even under adverse
conditions, it will be more rapidly assembled than any type of hut so far
tested.
This hut, built by the Douglas Aircraft Corporation of California, is
made up of panels consisting of a resin-impregnated paper honeycomb core
sandwiched between two 1/50-inch aluminum skins. The wall panels are 3 inches
thick, the roof and floor panels 5 inches thick. Panel edges are of a plastic
fiber glass laminate, and the shiplap joints are fastened with pins approxi–
mately 6 inches long and 7/8 inch in diameter. The pins are of plastic
reinforced with fiber glass, to prevent conduction of heat. They are held in
place by metal wedges.
Twelve 4- by 8-foot panels form each side of the hut, which is 20 by 48 feet
(inside measurement) with a clear floor area of 960 square feet. The structure
weighs about 10,000 pounds or approximately 10½ pounds per square foot of usable
floor area.
The roof is slightly pitched, so the five panels at each end graduate from
8 feet 4 inches to 9 feet 4 inches in length. One panel at each end contains a
door, one being the entrance, the other for emergency use only. The corner panels
have mitered joints, and all joints throughout the hut are covered with felt
sealing strips. For permanent or semipermanent use, the joints are sealed
with mastic.
Except for the two center end panels, each panel contains a round window,
16 inches in diameter, consisting of two panes of plexiglas separated by a
1-inch air space to reduce heat lose through the windows. Plexiglas was used

EA-I. Roberts: Housing

in preference to ordinary glass to reduce breakage during shipment and
erection operations.
Because of the hut’s excellent insulation, the heating and ventilation
system should provide adequate heat and air intake with low fuel consumption.
Inside the building a bulkhead has been installed to form an entry
vestibule. This bulkhead may, however, be erected at any 4-foot station in
the building except at those stations where the roof beams are located.
The roof beams and other elements of the hut’s basic framework are made
of aluminum. A sectional roof bean along the longitudinal center of the hut
rests on a support at each end, and on three evenly spaced transverse beams
in between. The transverse beams in turn are supported by st e a nchions fastened
to the side panels. This structure rests on 2-foot-high foundation beams
which follow the outside dimensions of the hut, and has in addition a founda–
tion beam running longitudinally along its center.
Large holes have been provided in the foundation beams as closely as
possible without affecting beam strength, to allow plenty of air to circulate
under the building. This will prevent melting of frozen ground and consequent
undermining of the foundation. An added advantage is decreased weight of the
beams.
Assembly of the new Arctic Hut has been made extremely simple by the
designers. All parts, including the beams and supports, are easily connected
with pins and wedges. The only tools necessary for assembly are a mallet
and a socket wrench, the latter serving also as part of the door handle.
The largest panel is about 4 by 10½ feet and the foundation beams are
12 feet long. The three 20-foot transverse roof beams, the heaviest units

EA-I. Roberts: Housing

in the structure, weighs only 147 pounds. The shipping volume is 1,600
cubic feet.
Since the structure may have to be assembled and disassembled under
severe weather conditions, the parts are all large enough to be handled easily
with mittens; there are no blind holes from which ice cannot be removed.
The use of plastic and fiber glass on all parts which connect the outside
and inside of the structure prevents the formation of ice on the inner
surface and reduces heat loss.

EA-I. Roberts: Housing

BIBLIOGRAPHY

1. American Society of Heating and Ventilating Engineers. Heating,
Ventilating, Air Conditioning Guide . N.Y., The Society, 1948.

2. Browne, Belmore. “Let’s Build a Snowhouse,” Natural History Magazine ,
Vol.55, No.10, pp.460-65, 486.

3. ----. ----. Reprinted for The Boy Scouts of America by special
permission of Natural History Magazine . (n.d.)

4. Close, Paul D. Building Insulation . Chicago, American Technical
Society, 1945.

5. Esbach, Ovid W. Handbook of Engineering Fundamentals . N.Y., Wiley,
1936.

6. Gilder, William H. Schwatka’s Search. Sledging in the Arctic in
Quest of the Franklin Records . N.Y., Scribner, 1881.
Describes camping and residing in snowhouses among Eskimos
who, on some occasions, had no fuel for warming them.

7. Jenness, Diamond. “The Life of the Copper Eskimos,” Canadian Arctic
Expedition, 1913-18. Report . Vol. 12: The Copper Eskimos,
Southern Party — 1913-16. Pt.A. Ottawa, Acland, 1923.
Describes pre-white Eskimo housing in Victoria Island and on
the [: ] Canadian mainland opposite.

8. Koppes, Wayne F. Report on Characteristics of Snow Houses and Their
Precticability as a Form of Temporary Shelter. A Report to
the Subcommittee on Shelter and Clothing, Committee on
Sanitary Engineering and Environment, National Research
Council . Wash., D.C., Division of Medical Sciences, The
Council, 30 November 1948.

9. Murdoch, John. “Ethnological Results of the Point Barrow Expedition,”
U.S. Bureau of Ethnology. Annual Report , 9th, 1887-1888. Wash.,
D.C., G.P.O., 1892, pp.72-86. This publication, and the one by
John Simpson, describe Eskimo housing at Point Barrow, Alaska,
[: ] as it was before E ru ur opean fashions came in.

10. Peary, Robert E. Secrets of Polar Travel . N.Y., Century, 1917. This
book discusses snow houses, Eskimo houses of earth, bones , and — —
stone, and European-style winter quarters as built by Peary
himself.

EA-I. Roberts: Housing

11. Petitot, E É mile. Les Grands Esquimaux . Paris, Plon, M— or N?ourrit, 1887.
Describes pre-white Eskimo housing in northwestern arctic Canada.

12. Simpson, John. “Observations on the Western Esquimaux and the Country
They Inhabit; from Notes Taken During Two Years at Point Barrow...”
Great Britain. Admiralty. Further Papers Relative to the Recent
Arctic Expeditions in Search of Sir John Franklin, and the Crews
of H.M.S. “Erebus” and “Terror. ” London, 1855, pp.917-42.

13. ----. “The Western Eskimo,” Royal Geographical Society. Arctic
Geography and Ethnology. A Selection of Papers on Arctic
Geography and Ethnology. Reprinted, and Presented to the Arctic
Expedition of 1875, by the President, Council, and Fellows
of the Royal Geographical Society. London, Murray, 1875,
pp.233-75.

14. “Snow Caves,” The Polar Record , Vol.4, No.25, [: ] pp.26-27,
January, 1943.

15. Stefansson, Vilhjalmur. Arctic Manual . N.Y., Macmillan, 1945.

16. ----. The Friendly Arctic . N.Y., Macmillan, 1921. This and the
previous book give — by explanation, description, photograph ,
and diagram — explicit directions for the building and use of
snow houses.

17. ----. The Stefansson-Anderson Arctic Expedition of the American Museum;
Preliminary Ethnological Report . 1914. N.Y., The Museum, 1919.
Its Anthropological Papers Vol. 14, Pt.1. This book contains
descriptions and diagrams of pre-white Eskimo earth and wood
houses in northwestern arctic Canada.

18. ----. “Corrections and Comments,” American Museum of Natural History.
Anthropological Papers . Vol.14, Pt.2, pp.445-57. N.Y., The
Museum, 1916.

19. Teal, Jr., John H. Personal Communication November 30, 1949.

Palmer W. Roberts
Considerable parts of this
paper are based on the pub–
lished and unpublished writ–
ings of Vilhjalmur Stefansson.

Industrial Structures in Alaska

EA-I. (Al T. Donnels)

INDUSTRIAL STRUCTURES IN ALASKA

CONTENTS
Page
Docks 3
Warehouses 5
Power Plants 6
Storage and Repair Hangars 7
Nose Hangars 7
Utilidors 8
Fire Stations 9
Storage Tanks 10
Earthquakes 10
Construction Materials 11

[EA-I: Al T. Donnels]

INDUSTRIAL STRUCTURES IN ALASKA
It is conceivable that, through her great wealth and resources, Alaska
may some day be destined to realize the advancement of industrial development
far beyond average conception at the present time. Some of these potential ities
possibilities lie in the development of mines, mills, and smelters for the
mining and processing of practically every known mineral; oil production and
refineries; sea-food products; birch timber; gypsum and clay deposits; spruce
pulp mills; and numerous other industries. There are also parts of Alaska
conducive to agricultural and stock-raising development, such as the Kenai,
Matanuska, Homer, and Kodiak areas. Extensive developments of railroad, high–
way, and airway facilities are presently under way.
In visualizing Alaska’s possible future, the subject of industrial struc–
tures becomes an important consideration. The problems involved in connection
with industrial structures in Alaska are extraordinary and invite many improved
techniques in overcoming present excessive costs and proper adaptability to
varying climatic conditions. While some private industrial development has
occurred in Alaska in recent years, the major scope comprises military installa–
tions to which this article is principally confined.
The types of industrial structures which are of interest to the planning or
construction engineer in connection with military projects include docks, ware-

EA-I. Donnels: Industrial Structures

houses, power plants, storage and repair hangars, nose hangars, shops, fire
stations, storage tanks, utilidors, water supply installations, etc. Each of
these requires certain considerations that very considerably in various loca–
tions. Temperature, wind, and snow factors are important in selecting the
location of a structure, as are the materials for construction. For example,
in the Prince William Sound area, extreme wind and snow loads must be considered.
The annual precipitation approximates 160 inches; wet, heavy snow up to 30-
foot depths is not uncommon, requiring heavy structures and involving the prob–
lem of snow disposal. Temperatures are relatively moderate. Ice problems
exist but are of a different nature from those encountered in the vicinity of
Cook Inlet and north of the Alaska Range. Waterways and harbors are ice-free
during the winter. This coastal region is similar in many respects to the
Aleutian Islands except for the much heavier snow conditions.
The Cook Inlet area (vicinity of Anchorage), approximately 60 miles distant
overland from the Prince William Sound area, has an entirely different climate,
materially affecting design of structures. Extreme minimum temperatures in
this locality reach −35° to −40°F. with average winter temperatures within the
−20°F. range. Precipitation and winds are moderate. Average snow depths range
from one to three feet. Cook Inlet is usually icebound during the winter.
The occurrence of 30 to 40-foot tides in Cook Inlet causes severe ice movements
and ice loads on piling, involving a major consideration in waterfront con–
struction. As the tides recede, large ice accumulation remains frozon to t he
piling.
North of the Alaska Range, including the Tanana, Nenana, and the lower
Yukon Valley areas, precky precipitation is light. Snow load factors are
relatively negls negligible due to dry, powdery snow. Little wind occurs;