Search

Increased construction activities in cold regions require a better understanding of the behavior of frozen soil. Although in situ testing procedures are being improved or developed, the highly variable and complicated conditions usually encountered in the field make it difficult to formulate interpretations that are useful in practice. Controlled laboratory tests on frozen soils are helpful, therefore, in determining behavior indicative of diverse field conditions and will assist in establishing design information. This paper is a review of the procedures used and the difficulties encountered in the transportation, preparation, and storage of naturally and artificially frozen soil samples required for testing in the laboratory. The effective control of sample disturbance by temperature variation and sublimation and the need for further investigation as to their effect on the mechanical properties of frozen soil are discussed.

ISCORD '91, Third International Symposium on Cold Region Development, June 16, 1991, Edmonton, Alberta, Canada

1991 CSCE Annual Conference, May 29, 1991, Vancouver, British Columbia, Canada

The ice lensing phenomenon in layered soils, a condition that simulates the process of ice growth in varved soils, has been studied experimentally. Under the limited conditions of layer thicknesses, soils used, freezing rates, and moisture conditions, ice lenses formed at the face of the fine-textured (clay) layer when it was encountered by the freezing plane. Only minor heaving occurred in the coarse-textured (silt) layer when its thickness was increased. Key words: frost action, ice lensing, frost heaving, varved soils, layered soils.

Finite element and finite difference methods for heat and mass transfer problems are usually based on two-time-level schemes for treating the time derivative. In freezing and thawing problems the equations are non-linear. The non-linearity arises because of the dependence of the material properties on both temperature and moisture fields. With two-time-level schemes it is usually necessary to iterate the solution in order to obtain accurate results. For non-linear problems of this type iteration can be avoided by using a three-time-level scheme with material properties evaluated at the intermediate time level. The most well known three-time-level scheme is, however, subject to oscillation which seriously limits its usefulness. This paper discusses a new three-time-level scheme and compares it with several other time marching schemes from the standpoint of accuracy and susceptibility to oscillation.

Values are presented for the thermal conductivity of frozen and thawed ice rich permafrost soils from Inuvik, Northwest Territories, measured under field and laboratory conditions with a cylindrical heat source. Samples were shipped to Ottawa, Canada, in the frozen condition at temperatures ranging from –5 to –70 °C. It was found that temperature of shipping and repositioning of the probe between measurements had no statistically significant effect on results. The dependence of thermal conductivity on bulk unit weight and dry unit weight is given for both frozen and thawed samples. Measured thermal conductivities were found to be in good agreement with other values obtained for similar soils and corresponding unit weight and water content.

X-ray photography has been used to locate the position of a growing ice lens in soil under laboratory conditions in order to establish the temperature of the actively growing face by means of its position in the thermal gradient field. It can be shown that in a general way the phase change temperatures are predictable from the Clapeyron equation. The structure of the ice phase also appears to follow the predictions of the heave rate equation proposed by Penner and Ueda (1978). When heaving occurs at low overburden pressures, there is a tendency for the ice lens to be very discrete and essentially soil free; at higher pressures and under similar thermal gradients it tends to develop in a more diffuse band and over a much wider temperature range.

The time-domain reflectometry technique was compared with the temperature measurement method for locating the frozen- unfrozen interface in water and sandy soils. This technique depends on the high-frequency (1-1000 MHz) electrical properties of water that change significantly and abruptly between the liquid and solid phases. Parallel wire transmission lines were inserted into the soil to guide electromagnetic pulses produced by a time-domain reflectometer (TDR). The frozen-unfrozen interface produced reflections measured by the TDR which were in turn used to locate the interface as it moved along the transmission line. In the laboratory it was possible to locate the interface using the TDR to within + or - 0.5 cm and in the field to within + or - 2.4 cm. These errors were equal to those associated with the temperature measurements.

This document contains the translations of some recent articles written by Soviet scientists on the strength and creep of frozen soils, applicable to the design of foundations in permafrost areas. These will be of interest to anyone engaged in northern engineering work.

Laminated frozen soils, consisting of alternate horizontal layers of dark brown clay and light brown silt, with ice lenses between the layers, are found in the northern area of ancient glacial Lake Agassiz. Laboratory tests confirmed the results of in situ ground anchor, penetrometer, and pressuremeter tests performed at Thompson, Manitoba, which showed these soils to be stronger in the horizontal direction, parallel to the layers, than in the vertical direction. Undisturbed block samples of these laminated ( varved) frozen soils were obtained and test specimens were prepared at various orientations to the direction of layering. Unconfined compression tests, using both naturally frozen and remolded frozen specimens, have shown the extent of strength anisotropy due to the natural anisotropic soil structure., Permafrost: 4th International Conference, July 17, 1983, Fairbanks, Alaska, USA

The authors of this paper have investigated the effect of total moisture content on the uniaxial compressive strength of three frozen soils (supes, suglinok and clay) under short- term loading at temperatures between -10 degrees C and -55 degrees C. Polycrystalline ice specimens were also tested under similar conditions.

One limitation of frost heave experimental research has been the necessity to use short soil samples to achieve one- dimensional heat flow. The temperature profile is then characterized by a high gradient and the soil has a very narrow freezing fringe. This fringe, a partly frozen band of soil, is the most important region to consider, since it is in this zone that suction potentials are developed, causing frost heave. This paper reports the design and testing of an experimental apparatus that permits a low thermal gradient and thereby an enlarged freezing fringe., 5th International Offshore Mechanics and Arctic Engineering Symposium, April 13, 1986, Tokyo, Japan

The processes that are operative during fine-grained soil freezing with no applied surcharge are described and related to the rate of net heat extraction. One-dimensional frost heave experiments were carried out on a Canadian silty soil under various freezing conditions. It was established that the segregational heave rate is strongly dependent on the rate of heat removal. No water flow was observed at heat extraction rates between 0.35 and 0.45 mW/mm2. Maximum segregational heave rates occurred at rates of about 0.1 mW/ mm2. For the same soil, however, significant variations in segregational heave rates were measured for a given heat removal rate during transient freezing. Frost heave predictions using freezing tests at rates of heat extraction representative of field conditions are also discussed.

The effect of a small alternating load superimposed on a static load on the displacement rate of piles in a frozen sand and a frozen natural soil has been studied at -2.2 degrees C. An alternating stress as small as 3 per cent of the static stress caused an increase in rate of displacement in the region where the displacement rate was decreasing with time, and doubled the steady-state displacement rate of both wood and concrete piles. Under dynamic loads the design life of pile foundations calculated on the basis of allowable settlement will be reduced or, alternatively, the bearing capacity of piles for dynamic loads will be smaller than the bearing capacity for static loads., Fourth Canadian Permafrost Conference, March 2, 1981

Baker, T.H.W., 1979. Strain rate effect on the compressive strength of frozen sand. Eng. Geol., 13: 223–231. Cylindrical specimens of fine Ottawa sand (A.S.T.M. designation C-109), compacted at the optimum moisture content and saturated before unidirectional freezing, have been tested in uniaxial compression at a cold room temperature of −5.5°C and strain rates between 10−7 and 10−2 s−1. The results agree with an extrapolation of data obtained by Sayles and Epanchin [1], but are much higher than those obtained by both Goughnour and Andersland [2] and Perkins and Ruedrich [3] at strain rates below 10−5 s−1. There is evidence that this may be due to variation in total moisture (ice) content, the conditions under which the specimens were frozen (closed system or an open system) and to the end effects at the platen—specimen interface.

It is possible to locate a growing ice lens in a frost heave cell in soil under laboratory conditions by means of a recently devised X-ray technique. If temperature distribution is measured in the sample at the same time, the temperature of the growing face of the ice lens can be determined. The objective of the study was to validate this method under a range of heaving conditions. It was done by determining ice lens temperatures at various overburden pressures and cold side temperatures; then, by adjusting the ice lens temperature for overburden pressure and pore water salt content, the remaining difference from 0 degrees C was expressed in terms of suction. Suction gradients in unfrozen soil under various freezing conditions were subsequently plotted as a function of rate of moisture flow and a proportionality was shown to exist. In addition the calculated permeability coefficient was the expected value for the material, Leda clay. These results lend credence to use of the X-ray - thermal gradient technique for studying thermal behaviour of ice lens growth., 3rd International Symposium on Ground Freezing; U.S. Army Corps of Engineers, CRREL: 22 June 1982, Hanover, New Hampshire, USA

Results are presented over a wide range of soil textures supporting the linear relation between the logarithm of frost heave rate and the cold-side temperature – overburden pressure ratio proposed by Penner and Ueda. The relation is simple to determine experimentally, permits heave predictions at various other temperature and pressure conditions for engineering purposes, and allows soil frost susceptibility comparisons to be made with the results determined by other laboratories.

The design of foundations in permafrost, excavation of frozen ground, and artificial freezing of ground to provide temporary support during construction of tunnels and other underground structures require understanding of the mechanical properties of frozen soils. This paper describes the equipment and procedures developed by the authors to determine the compressive strength of artificially frozen sands under confined and unconfined conditions at different temperatures and strain rates. A comparison of the results obtained under various conditions is made to assess the effects of testing conditions on the strength, strain at failure, and deformation behaviour of frozen sand. This comparison should assist in the standardization of laboratory testing procedures for the determination of the mechanical properties of frozen soils., Fourth Canadian Permafrost Conference, March 2, 1981

Uniaxial unconfined compression tests were carried out on frozen saturated Ottawa sand containing about 20% by weight of water, at temperatures between −2 and − 15°C, and at strain rates varying between 10−7 and 10−2 s−1. The compressive strength and the initial tangent modulus increased with increasing strain rate and with decreasing temperature. At −2°C, values of strength and modulus were considerably lower than those predicted by linear extrapolation of the values observed at lower temperatures, on a log–log scale. This could be due to the presence of unfrozen water in the samples at −2°C.

An alternating stress as small as 5% of the static stress was found to accelerate the displacement rates of piles subjected to long-term loads in frozen soils, as well as the creep rate of frozen soil samples subjected to uniaxial loading under compression. This effect essentially reduces the bearing capacity of piles in frozen soils, and has to be taken into consideration in the design of foundations in permafrost areas., 3rd International Specialty Conference on Cold Regions Engineering. Northern Resource Development, April 4, 1984, Edmonton, Alberta, Canada

Results of several long-term creep tests of piles in frozen soils are analyzed; the primary or attenuating creep is predominant in most cases. For effective design of pile foundations in permafrost areas, this aspect should be taken into consideration., ASCE Spring Convention: Foundations in Permafrost and Seasonal Frost, April 29, 1985, Denver, Colorado, USA

Frozen saturated quartz sand containing 20% moisture by weight, when deformed at -30 degrees C at various strain rates, showed various modes of behaviour such as visco- plastic, almost ideal plastic, and brittle with little plasticity with increasing order of strain rate. The values of peak strength observed for strain rates between 5 X 10[- 7] and 6 X 10[-3]s[-1] were in the range 15-42 MPa,, stress being related to strain rate by a power-law equation.

Stability of artificially frozen coal mining shafts in Germany has initiated interest in the mechanical properties of frozen ground. This article reports on laboratory uniaxial and triaxial loading tests on frozen soil samples taken from several geologicial zones through which a shaft has been driven. The effects of such factors as grain size, ambient pressure, temperature and rate of loading on the strength and deformation properties of frozen soils were investigated.

Wet rodding and multiple-sieve pluviation methods of controlling sand density in the preparation of frozen sand specimens for strength testing are compared. Multiple-sieve pluviation provides homogeneous, reproducible specimens at various dry densities. Unconfined compression tests indicate consistent strength data for specimens prepared by multiple-sieve pluviation, but significant scatter for compacted samples prepared by the wet rodding method. For the testing conditions reported in this paper the stress- strain behaviour of frozen sand specimens was dependent on initial dry density; at low densities, strain softening with constant peak stress was observed. For higher densities, the samples exhibited strain hardening behaviour, peak stress increasing linearly with density., 4th International Symposium on Ground Freezing, August 5, 1985, Sapporo, Japan

Several specimens of artificially frozen sand (ASTM C-109) were prepared to evaluate different testing procedures and techniques used to determine the mechanical properties of frozen soils. Two methods of preparation of fully saturated frozen sand specimens, described by other investigators, were compared. Compaction by a rodding technique and saturation under vacuum produced a uniform degree of compaction and saturation upon freezing. The most critical part of the preparation procedure leading to physical variability between specimens was found to be the time during freezing and machining. This time span must be kept constant to minimize the effects of sublimation.

Cyclic creep behaviour of frozen saturated sand and frozen clays was studied by superimposing alternating stresses over a mean static compressive stress on the samples. Cyclic loading increased the creep rate of these materials considerably, which could be due to the increase in unfrozen water content generated by the transient thermal energy produced by mechanical cycling. The strain increment per cycle was found to increase with increasing strain in frozen sand, whereas it decreased in frozen clays, showing a tendency for strain hardening., 4th International Symposium on Ground Freezing, August 5, 1985, Sapporo, Japan

The importance to the snow-ground thermal interaction of non- linear effects associated with temperature dependent soil thermal conductivity and with soil latent heat is discussed. Numerical model calculations are presented which indicate that the modification of mean annual ground temperatures associated with these effects can be of the same magnitude as the warming due directly to the seasonal snow cover., 3rd International Conference on Permafrost, July 10, 1978, Edmonton, Alberta, Canada

The processes that are operative during fine-grained soil freezing with no applied surcharge are described and related to the rate of net heat extraction. One-dimensional frost heave experiments were carried out on a Canadian silty soil under various freezing conditions. It was established that the segregational heave rate is strongly dependent on the rate of heat removal. No water flow was observed at heat extraction rates between 0.35 and 0.45 mW/mm[2]. Maximum segregational heave rates occurred at rates of about 0.1 mW/ mm[2]. For the same soil, however, significant variations in segregational heave rates were measured for a given heat removal rate during transient freezing. Frost heave predictions using freezing tests at rates of heat extraction representative of field conditions are also discussed., Reprinted from "ASME Winter Annual Meeting, New Orleans, LA, USA, December 1984, Heat Transfer Division, Pamphlet Paper No. 84-WA/HT-108, 7p.">

Serious problems due to frost heaving and thaw settlement can be experienced by structures founded in seasonally and perennially frozen ground. Various design and construction techniques can be used to overcome or reduce the severity of these problems. One approach is to prepare and improve the properties of the foundation soil in advance of construction such that it is less susceptible to frost action and to settlement due to thawing. Much work has been done in the U. S.S.R. on this subject and this recently published manual is therefore of special interest in its studies of construction problems in permafrost areas. The manual describes electrical and chemical methods and techniques used in the Soviet Union for preconstruction preparation of seasonally and perennially frozen soils. The physical and mechanical properties of the foundation soil are improved by electrical thawing methods and/or injection of various chemicals followed by compaction of the soil, if required, before a structure is erected.

Growth characteristics of individual ice lenses in natural soils have been studied in the laboratory using ramped temperature control. The growth rate rises rapidly after ice lens initiation and this is followed by a protracted period of decreasing growth. The temperature at the growing face of the lens decreases continuously during its formation.

The freezing point of soils is the boundary of change of physical state from a thawed to frozen condition. A paramount factor which affects the transition zone is the quantity and composition of the dissolved salt in the pore water. The marine soils used in the experiments were from the arctic coast with a range in the granulometric composition. The clay fraction consisted of 40% kaolinite, 30% montmorillonite, 25% chlorite and 5% quartz. Salts used were from soil extractions or from sea water. NaCl and CaCl[ 2] solutions were also used for artificial salination of the soil samples. The salt content used ranged from 0.2 to 1% by weight or in terms of the pore water was from 0.006 to 0. 055. Three stages were observed; a supercooling stage, a phase of comparatively constant temperature, and a gradual fall of temperature. The supercooling temperature was from 1.5 to 2 times the freezing point. Conclusions are as follows: (a) the grain size characteristics of the soil do not influence the freezing point when the moisture content is in excess of the liquid limit; and (b) when there are equal concentrations of the interstitual solution only the qualitative composition of the salts influences the freezing point. Within the range investigated the freezing point of the soil is inversely proportional to the molecular weight of the soluble salts.

This document contains translations of selected papers from the 1979, 1980 and 1981 issues of the Journal of Glaciology and Cryopedology published by the Lanzhou Research Institute of Glaciology and Cryopedology Academia Sinica. These papers were of special interest to Canadian permafrost workers. Papers provide information on the nature of permafrost studies carried out over the last thirty years, differences between permafrost in China and the Canadian North, geotechnical classifications of permafrost used in China, including such topics as the active layer, discussions on permafrost classifications, periglacial processes, frost heave, hydrogeological investigatory methods, ice content, mapping, pingos and other related topics. In all 20 papers have been translated into English.

A soil frost susceptibility test is proposed. The method involves the unidirectional freezing of a saturated sample by imposing a step freezing temperature at one end of the specimen and measuring the heave rate. The two most important conditions that influence heave rate during the test are the overburden pressure and the freezing temperature. Results show that the heave rate can be expressed as an exponential function of the applied overburden/freezing temperature ratio. The desirable features of the proposed freezing technique and the method of interpreting the results are outlined., 3rd International Conference on Permafrost, July 10-13, 1978, Edmonton, Alberta, Canada

The Inuvik airfield was constructed between 1956 and 1958 on a site underlain by frozen fine-grained soils containing considerable quantities of ice. Its design and performance were therefore of considerable concern. The airfield consists of an embankment of rock fill constructed on the undisturbed ground surface to a thickness sufficient to prevent, or minimize, thawing of the frozen subgrade soils: from 2.5 to 4.2 m, and averaging about 3m. The airfield was paved with asphaltic concrete in 1969. Ground temperatures were measured at several locations in the subgrade and the embankment from 1958 to 1974. All temperature observations showed that the permafrost table moved up at least 0.6 m into the fill after construction was completed in 1958 and remained at about the same level in subsequent years, even after paving. The airstrip has performed extremely well to date and has required little maintenance work., Fourth Canadian Permafrost Conference, March 2, 1981, Calgary, Alberta, Canada

The methods which are recommended in the handbook take into consideration, for the first time, the dependence of the physical, thermal and mechanical properties of frozen soils on their cryogenic structure. Frozen soils strongly manifest anisotropy and the heterogeneous distribution of properties which are usually evaluated as the averaged values of the characteristics. The completeness of the document cannot be overemphasized. A section on basic concepts and nomenclature of frozen soils, in particular the nature of the cryogenic structure, precedes the main chapters. Notes are also provided for the selection, packing, transportation and storage of samples of frozen soil. The main physical properties of frozen soils which are the total moisture content, ice content and specific weight depend on the cryogenic structure of the soil and special attention should be given to indications of these properties. The thermal properties for which test methods of determintion are given are heat capacity, thermal conductivity and thermal diffusivity, freezing point depression of saline soils and the construction and calibration of thermocouples, electric heaters and thermometers. The mechanical properties for which test methods are given include compressibility of frozen and thawed soils, resistance to normal pressure, long term value of cohesion, uniaxial compression, resistance to shear of both frozen and thawed soils, tangential heaving forces and finally viscosity of ice inclusions.

Compressional wave velocities at -10.0 and -3.2 degrees C and the unconfined compressive strength show a dependency on the dry density of frozen freshwater sand specimens in the density range 1550 to 1780 kg/m[3]. Shear wave velocities are independent of dry density at both temperature levels. Acoustic wave velocities and unconfined compressive strength are independent of dry density when salt is added to the pore water prior to freezing. Increasing the salinity greatly reduces acoustic wave velocities and strength. These reductions are most pronounced when the salinity is between 0 and 5 ppt. Ultrasonic elastic constants were compared with dry density and salinity. Dry density and salinity tests should be included in engineering site investigations related to artificial ground freezing and frozen ground construction and acoustic measurements should be used to determine the elastic constants of frozen soils used in numerical modelling., 4th Internaitonal Symposium on Ground Freezing, August 5, 1985, Sapporo, Japan

Four different platen configurations were used to determine the influence of some end conditions on the laboratory- determined uniaxial compressive strength of frozen sand. Cylindrical specimens of fine Ottawa sand (ASTM designation C-109), compacted at the optimum moisture content and saturated before unidirectional freezing, were tested at temperatures varying between -5 and -6 degrees C. Each specimen was subjected to a constant strain rate of 0.7 x 10[-3] min[-1] using an Instron, universal testing machine ( model 1127). A compliant platen designed to reduce friction between platen and specimen, to distribute the load uniformly, and to minimize stress gradients produced by eccentric loading, was found to be most desirable in determining the uniaxial compressive strength of the frozen sand specimens., 3rd International Conference on Permafrost, July 10, 1978, Edmonton, Alberta, Canada

Preliminary results of laboratory studies on the behaviour of model piles of wood, concrete, and steel, having different surface finishes, embedded in artificially frozen sand and subject to constant loads, are reported in this paper. The displacement rate [Formula: see text], was found to vary with the applied stress [Formula: see text]f, according to a steady state creep equation of the type [Formula: see text], as found for other viscoelastic materials such as metals at high temperature. The value of the exponent n varied between 6.7 and 9.1 for various types of piles.The results agreed favourably with field data obtained from pile pull-out tests carried out at Gillam and Thompson, Manitoba. The value of n from Gillam and Thompson tests, 8.05 and 7.5, respectively, were within the range of the present experimental values. Comparison with pile pull-out tests carried out in ice, and theoretical predictions for ice showed that the value of the exponent n was much larger for frozen soil than for ice. The bearing capacity at a particular displacement rate was also found to be almost an order of magnitude larger in frozen soil than in ice.

A series of field tests was carried out at a permafrost site near Thompson, Manitoba, to investigate the suitability of the Ménard Pressuremeter as a tool for in-situ determination of the rheological properites of frozen soils. Both short- term and long-term tests were conducted using the standard pressuremeter equipment with only minor modifications. Special care was taken in drilling and preparing the borehole for the tests to minimize the temperature disturbance. By using a newly developed method described in the paper, it was possible to obtain from the tests both the short-term and the long-term strength parameters. The first were obtained from conventional type pressuremeter tests, while the latter resulted from stage-loaded creep tests. In addition to the time dependent strength, the constitutive creep equation of the frozen soil was also established from the pressuremeter creep tests. The pressuremeter technique was found to be quite promising for creep testing of frozen soils. Its main advantages over laboratory methods used for similar purpose is that it is quicker, less expensive and involves a relatively large volume of soil. Its drawback is that it furnishes information on the soil behavior in only one direction, i.e., normal to the axis of the borehole which may be insufficient for foundation design if the permafrost stratum is anisotropic and contains horizontal ice lenses and layers., 2nd International Conference on Permafrost, July 1973, Yakutsk, Siberia

An important property in the solution of geothermal problems is the thermal conductivity of the ground. Data on conductivity of soils, particularly in permafrost areas, are extremely scarce owing in part to the fact that a relatively inexpensive and simple apparatus has not been available for measuring conductivity with confidence. This paper deals with the development of such an instrument for use in permafrost soils under both field and laboratory conditions. A series of tests has been carried out to calibrate the probe with the guarded hot plate using several granular materials.

Expansion of water when it freezes accounts for only a small fraction of total heave; a large percentage of the displacement in soils is attributed to the freezing of transported water from below the freezing plane. Previous experiments showed that the magnitude of soil moisture tension when the growth of the ice lens stops depends on pore size dimension. This paper is devoted to the determination of the effect of overburden pressure and suction. The material used was powdered quartz. Samples were molded in the Harvard Miniature Compaction apparatus. Saturation before freezing was achieved by permitting de- aired water to enter one end of the specimen. Freezing was done unidirectionally with the cold side temperature at -3 degrees C and a warm side temperature of +1 1/2 degrees C. Maximum soil moisture tension determinations when heaving stopped were carried out at overburden pressures of 23, 229 and 535 grms/cm[2]. All experiments were carried out at three different dry densities between 1.592 and 1.750 gm/cm[ 3]. The results showed that the state of saturation of the unfrozen soil was linearly related to the overburden pressure used; i.e. the higher the overburden pressure the higher the moisture content below the freezing plane at the point where ice lensing stopped and hence the lower the soil moisture tension. These results are supported by theoretical considerations using the Clapeyron equation.

Some aspects of the general theory of frost heaving are not consistent with the way water moves in saturated material, hence a new theory is presented based on the absorption forces of water on surfaces. Water molecules, being dipolar, arrange themselves on the surface of crystals to balance the electrical forces. The crystal surfaces become covered with several layers of water with the attraction force of each successive layer decreasing with distance from the surface. Because the arrangement of water molecules is different in ice the crystals are not attached to the mineral crystal surfaces with the same tenacity as individual water molecules. As more ice freezes, removing the outer layers of water from the sphere of influence of the mineral particles, water moves in from the surrounding area to replenish the water removed by freezing. Thr author has presented calculations to show that pressures in the water films can exceed 40,000 kp/cm[2] and thus explains the large heaving forces observed when ice lensing occurs. On the basis of this new theory one can assume that frost action can be changed using electrical fields. It has been found that an electrical current of 0.04 to 0.19 amperes caused considerable changes in the water films that develop on soil particles. However it may be too optimistic to assume that the electrical potential difference between metal surfaces of different temperatures would be sufficient to stop frost action.

Most of the existing theories of creep have been developed from two different viewpoints: micromechanistic and macroanalytical. The former deal with events occurring at the atomic level and provide knowledge of the processes that control creep. The latter are based on certain macroscopic experimental findings and represent, in fact, an extension of the theory of plasticity to include time and temperature effects. Both the micromechanistic and macroanalytical approaches lead to fruitful results and each can benefit from the other. However, although the former has the advantage of being derived from physical concepts, the use of the latter is often preferred in practice if it provides basic relations that are broad in scope and can lead to improved procedures for designing structures. In this paper, a macroanalytical view of the problem of creep of frozen soils is presented. The proposed theory of creep has been developed mainly with the purpose of being used as a basis for solving a specific soil engineering problem, i.e., the bearing capacity of buried footings and anchors. Since the problem is itself rather complex, it was endeavoured to present the creep information in a relatively simple mathematical form. The theory, while using certain concepts and data from the frozen soils literature, follows more closely, nevertheless, the methods usual in certain engineering theories of creep of metals.

The process of heat transmission in different kinds of ground takes place mainly by a conductive mechanism whereby thermal energy is not accompanied by transport of substance and by a convective mechanism in which heat transfer is effected by the movement of air and water in the pores of the soil. Freezing and thawing soils are always structurally heterogeneous. Heat may be propagated through the mineral skeleton by way of conduction and also through pores where transfer takes place by conduction, convection and radiation. If the soils are water unsaturated, thermally induced moisture movement exposes special problems which can be dealt with using classical thermodynamic processes. If the soils are saturated no redistribution of water is possible. In the transitional zone between thawed and frozen zones special consideration must be given to the development of water-ice phase changes. Systems of equations have been given that allow calculation of heat and moisture transfer based on fundamentals available in the literature. The convective mechanism of heat transfer for unsaturated soils has received very little attention. The equations in this paper are a first attempt to establish general qualitative laws but no solutions are given. The way soil parameters affect the values of heat and moisture characteristics is unknown.

The use of guyed towers for power transmission and communication systems in northern Canada has received increased attention in recent years. In many cases perennially frozen ground (permafrost) poses special problems. The resistance of frozen soil to uplift forces must be evaluated for the design of anchors of guyed structures built on permafrost. There is little experience or information available, however, concerning the behavior of frozen soil in response to uplift forces or the capacity of various types of anchors. A field study of anchors in permafrost was therefore undertaken by the Division of Building Research of the National Research counci l of Canada to evaluate, in particular, the displacements associated with long-term creep of frozen soil under load. initial appraisal of various kinds of anchors suggested that grouted rod anchors were a type that should be investigated. Early in 1967 a total of 18 such anchors (nine at each site) were installed in permafrost at Thompson and Gillam, Manitoba. This paper describes the installation and testing procedures and the results of the test programs conducted on the anchors at both sites. In the analysis of the results the creep behavior of grouted anchors is related to U basic creep parameters of frozen soil, taking into account the effects of temperature and normal pressure. The theory employed allows field creep data to be used for estimating the long-term adfreeze strength of frozen soil, information that is required in the design

In regions where there is deep penetration of seasonal freezing of fine-grained soil there is a serious problem of preventing frost heave of the foundations. A method considered here to provide a stable foundation for the structure is to anchor the footing in unfrozen soil. A foundation anchored in unfrozen ground resists heaving forces generated along the contact surface between the foundation wall and the layer of freezing soil. Before freezing, the load of the structure, the weight of the foundation and the weight of the soil above it are given by the stress acting along the foot of the foundation. When the upper layer of soil freezes heaving forces are generated along the lateral surfaces of the foundation wall which partially or fully counteracts the above mentioned load, thus decreasing or completely cancelling the footing pressure. A further consideration is the inclusion of swelling or "release pressures" beneath the footing when the foundation is being lifted up. Equations are provided to calculate all the forces and to also calculate a factor of safety. Numerical examples complete the presentation. An important factor not considered yet is the question of permissible foundation displacement which will undoubtedly depend on the type of structure involved.

Thermal conductivity measurements of two frozen soils, Leda clay and Sudbury silty clay, taken at temperatures between 0 and −22 °C by means of a thermal probe and a transient heat flow technique, compare favorably with estimates of thermal conductivity calculated by the DeVries method. Both measured and estimated values show a similar trend of increasing thermal conductivity as the temperature is lowered and the ice content grows. This increase is associated with the higher thermal conductivity of ice compared with that of water.

Whenever a soil of suitable texture and structure and with sufficient water content freezes slowly, the equilibrium between the amount of water in the film and the boundary- surface forces are disturbed in the film situated between the ice and the soil substance. The disturbance results in the loss of liquid water by the film to the ice and this increases the size of the ice lens. The disturbance in the film is immediately compensated, because the boundary- surface forces keep the amount of water in the film constant. The compensation is brought about by the movement of water from the unfrozen soil to the film. Because the liquid film, as such, retains a constant thickness the part which becomes ice must be displaced. Thus ice lens formation and growth, water displacement and ground heaving come about as different aspects of a single event. The mechanical work performed during this event is done substantially by the boundary-surface forces of the two boundary surfaces, that between the ice and the water film and that between the water film and the soil particles.

Thermal conductivities were determined for 10 fine-grained soils obtained from potential borrow pit areas along the proposed route of the Mackenzie Highway, N.W.T., and from the nearly completed section of the Mackenzie Highway, just south of Inuvik. The moisture contents and densities of the test specimens were based on the standard Proctor density curves. The measurements were made by the transient heat-flow method using a line heat source. The values were in good agreement with Kersten's data obtained by the steady-state heat flow method.Grain-size, mineralogical analyses of the clay-size fraction, Atterberg limits, and standard Proctor density determinations are given for all 10 soils.