Cold Regions Engineering 2006: Current Practices in Cold Regions Engineering

The Australian Antarctic Division developed an ice runway near the Casey base in 2003 and its long-term viability is currently being explored. As opposed to a conventional runway, a runway on glacial ice is constantly moving and deforming, hence knowledge of this movement is paramount to comprehending the long-term viability of the runway. The movement of the Casey runway was estimated through an examination of regional measured and modelled data, and then two years worth of GPS measurements were assessed to provide an indication of the actual runway movement. The position of the runway was then forecast incrementally out to the year 2050. Although the runway is moving to the South West at almost 12 metres per year, and also experiencing rotation, compression, and a change in elevation, it is not expected that this movement will significantly compromise runway operations. With adequate maintenance of the constructed snow pavement, and appropriate management techniques, it is expected that the runway at Casey could remain operationally viable for many years into the future.

The Cold Regions Research and Engineering Laboratory (CRREL) has been contracted to design and build an Autonomous Real-time Remote Observatory (ARRO) instrument shelter for deployment in Antarctica. The challenge is to maintain an environmental temperature within the shelter not significantly lower than 0 °C in the extreme Antarctic environment where temperatures drop to –80 °C and the average temperature is below –50 °C. Greatly compounding the difficulty of this task is the reliance entirely on renewable energy sources. Energy is stored both electrically in batteries and thermally in water. In order to evaluate alternatives CRREL has constructed an energy model of the system. This model accounts for all the major heat losses from the shelter including heat loss through walls, optical domes, instrumentation cable access ports, and infiltration heat loss. The heat balance on the interior components includes heat flows to/from the thermal storage media (water), to/from the batteries, and waste heat from the instrumentation and controls contained in the shelter. A simple model of the wind turbine generator output has also been included in the system model. A full-scale ARRO was constructed, instrumented and tested in one of CRREL's environmental chambers at –50 °C. This paper describes the design approach, full-scale test, and presents results of model calculations compared with measured values obtained during testing. A prediction of the shelter performance in Antarctica will be presented.

Sustainable engineering in remote cold regions requires use of readily-available natural materials for engineering applications such as building foundations. Construction of a large structure at the South Pole provides a unique opportunity to observe settlement of a compacted snow and firn foundation. Our settlement data from the new Elevated Station show that many factors enter into the determination of settlement rate. Because construction occurs over time, components of the foundation experience varying loads and different load rates. The addition of temporary snow access ramps, which were not part of the original design, introduced very large loads onto the snow foundation during construction. Survey measurements of the columns support the argument that a majority of the initial differential settlement concerns were associated with the snow ramps. The relatively short intervals between surveys during the summer seasons illustrated a random pattern of column settlement behavior. Upon elimination of any gross survey errors in data collection, we attribute this behavior to the (intended) load sharing design of the station's rigidly connected grade beams. Long-term settlement of the structure is dependent upon properties of the firn, and future investigations will address stress-strain relationships for firn appropriate at the engineering scale for this site.

Freezing-thawing damage is a major degradation mechanism for concrete structures in cold regions. This damage becomes more severe when coupled with deicing salts application. This paper presents results of an experimental study on the evaluation of concrete damage when exposed to freezing-thawing cycles and salt application in a programmable environmental chamber. The induced damage based on the degradation of the dynamic modulus of elasticity was evaluated by two non-destructive testing methods: pulse velocity and resonant frequency methods. This exposure condition resulted in a substantial deterioration in concrete, which was captured by the two methods. Two sets of concrete mixes were investigated, one had the same w/c ratio but different aggregate content, and the second had the same aggregate content but different w/c ratios. Quantification of environmental deterioration is important for the evaluation of concrete performance and service life prediction models.

Working environments at fishery ports and harbors during the winter in cold, windy regions are very severe for workers. To examine the necessity of facilities for the improvement of working environments such as the wind and snow shelters, we should quantitatively evaluate in terms of human sensation based on thermal indices, which can indicate bodily sensation, including thermal sensation, thermal comfort, pain and other factors, at given thermal environments. In this study, we conducted sensory experiments on subjects in cold room and in open-air under various conditions to examine the practical application of some thermal indices. We found that there were no differences between sexes and between cold room and open air regarding the relationship (regression line) between the values calculated by each thermal index and the responses given by the subjects. Of five thermal indices examined in this study the WCI (Wind Chill Index) seems to be the most useful index from a practical point of view, as its method of calculation is very simple and few input data are required (only wind speed and atmospheric temperature).

In colder geographical regions, such as areas in high latitudes and altitudes, care must be taken when constructing concrete structures so that the concrete will cure successfully. Moreover, in cases of the construction of large bridges in deep valleys, it is usually difficult to cure the entire bridge. The grouting for introducing pre-stress is the most important task regarding pre-stressed concrete bridges because the grouting stretches the stress to the whole of the bridge. If the concrete quality is not satisfactory, the designed concrete power cannot be obtained. Furthermore, the durability of the bridge life might be limited. This paper proposes a grout concrete curing method which compares the bridge body temperature under both heated and unheated conditions.

The problems of reconstruction of aged buildings in Siberia are linked to climate peculiarities of cold regions, such as wide annual range and extremely low winter temperatures. The longevity of construction and energy savings are the main priorities of reconstruction projects. The applications of effective façade heat insulation and protection solutions are considered in the paper.

Jarvis Creek flows as a wide braided channel in glacial moraine deposits through the Donnelly Training Area (DTA) of the U.S. Army's Fort Wainwright near Delta Junction, Alaska. The U.S. Army Alaska is in the process of designing a $55 million Battle Area Complex (BAX) and Combined Arms Collective Training Facility (CACTF) within the DTA for increasing training opportunities for infantry and mechanized troops. While open water flooding events are easily contained within the channel corridor, massive accumulations of aufeis generated during severely cold winters result in almost total channel blockage at several locations. Spring snowmelt events send water out of the channel and into the floodplain, eventually passing out of the Jarvis Creek basin. The U.S. Army Alaska contracted the Engineering Research and Development Center - Cold Regions Research and Engineering Laboratory (ERDC-CRREL) to determine whether these floodplain flows would effect the proposed locations for the BAX and CACTF within the DTA. HEC-RAS modeling was used to simulate aufeis blockage of the channel and floodplain water elevations at the proposed site locations. While the Eddy Drop Zone site was initially thought to be an optimal site due to the absence of permafrost and sensitive wetlands areas, the modeling has shown that significant overbank flow passes through the area. This paper describes Jarvis Creek, its flooding history, HEC-RAS modeling and the impacts of spring snowmelt flooding events on the proposed BAX/CACTF location at the Eddy Drop Zone site within the DTA.

In cold regions, ice covers on rivers, lakes, seas are often used as temporary roads, bridges, airfields, and construction platforms. For these uses, it is important to understand that there must be a sufficient margin of safety between the breakthrough loads and the actual loads placed on a floating ice sheet. Therefore, an inclusive study of bearing capacity of floating ice sheet is necessary and important. Based on Devinder Sodhi's experimental study, the wedging action hypothesis was forwarded to further explore the bearing capacity of ice sheet due to concentrated load. However, it is difficult to approve this hypothesis by means of experimental measurements. In contrast, the numerical simulation may be the best way to approve this important assumption. In this paper, a series of Finite Element models simulating the cracking progress of ice sheets due to a concentrated load on a circular floating ice sheet were created and Contact element was used to represent the cracking condition. Simulation of the cracking progress was performed. The wedging action force and moment generated in ice sheet were calculated. The successful research results in this paper can support the wedging action hypothesis well which could provide a relative accurate and simplified method to calculate ice loads acting on ice sheets. The results can be applied by the practical engineers in cold regions.

Cold regions cover more than half the United States. Engineering impacts of cold regions include those caused by temperature (e.g., freeze-thaw cycling, atmospheric icing), river and lake ice (e.g., ice jams, operational impacts of ice), and snow. Damages include the potential for loss of life and property due to flooding, structural damage, bed and bank erosion and scour, riprap failure, increased flood fighting and assistance costs, and environmental damage. Ice jams and atmospheric icing also prompt local, state, and federal emergency operations and can affect operation of power and flood control projects. Research and development sponsored by U.S. Army Corps of Engineers (USACE) in the area of cold regions engineering and science has resulted in numerous new and improved methods and techniques available to the engineering community for use in effective damage reduction activities. This paper describes current and future research by the USACE in cold regions aspects of flood and coastal storm damage reduction.

River ice jam formation and release both have the potential to cause flooding and present a threat to human safety. From a flood forecasting perspective, it is highly desirable to be able to predict the effects of ice on streamflow hydraulics. Although a number of models have been developed over the years that do consider ice effects, the majority of these are proprietary. This paper reports on recent developments in incorporating ice processes into the public domain River1D hydrodynamic model. The first objective in this effort was to incorporate river ice formation and melting processes and the application of these components is illustrated for the Peace River in western Canada. The model employs an Eulerian frame of reference for both the flow hydrodynamics and the ice processes and uses a Petrov-Galerkin finite element method to solve the primary equations. Model calibration and validation results with historical data are presented; these indicate that the present model adequately simulates water temperature and ice front progression. However, further enhancements are required to include certain dynamic freeze-up processes, in order to refine the ice front results. Ice jam formation and release components have also been added to the model to facilitate the ice jam flood forecasting application and these capabilities are demonstrated by modeling the large ice jam release event that occurred on the Athabasca River, AB, in 2002. Total ice and water mass and momentum equations are solved in an uncoupled sequence with ice mass conservation, with ice momentum effects considered empirically. The model is found to do a good job of modeling release wave speed and peak magnitude. Further enhancements to consider ice momentum effects more explicitly are underway.

Water resource managers are faced with increasingly complex issues arising from large-scale environmental changes due to climate dynamics. Current water resources planning and management procedures do not address the effects of climate impacts or the fragmentation in water resources management caused by balancing competing water uses. This is particularly true for transboundary watersheds in cold regions, where the effects of climate variability and long-term environmental impacts on extreme hydrological events (flood and drought) have been recognized for a number of years, yet are not accounted for in water management planning or policy. We recognize that the considerable uncertainties that do exist in global climate modeling are partially responsible for the lack of comprehensive planning that includes climate impacts. However, temperature-related effects of climate variability that affect floods and droughts in particular are quite predictable. These impacts should be considered in planning new projects and in the evaluation of existing projects. This paper presents an example where climate impacts can be considered to develop drought criteria in a transboundary situation, and suggests the development of policies to enable climate impacts as an integral component in planning and operations functions. However, implementation of water management plans at the basin level may be severely hampered by jurisdictional fragmentation.

The effectiveness of riverine flood damage reduction projects is assessed through the use of rainfall-runoff models for the open-water case. However, since no analytical model exists to describe the effects of complex hydrological and meteorological conditions that lead to ice jams, no similar method exists to assess the benefits of an ice control structure. In 1981, the New England Division of the U.S. Army Corps of Engineers completed an ice control structure (ICS) located about 0.5 miles upstream from the center of Lancaster, NH. In recent years, the ICS has fallen into disrepair and much of the impoundment area has filled with gravel, reducing its frazil ice storage ice capacity. The structure's current condition, combined with operation and maintenance safety issues and concerns about fish passage have prompted the town to pursue the option of its removal. This paper reports the results of a study is to determine, within the budget constraints and based on existing data, whether the ICS has had any impact in reducing ice-affected floods in Lancaster, NH.

A series of safe harbors are being established along the north shore of Lake Superior from Duluth to Grand Portage, Minnesota. Violent storms can arise quickly in this part of the lake, and the rocky shoreline offers little in the way of natural protection to small craft. At the marinas, dock structures, which represent a major portion of the project cost are typically left in the water all winter, and therefore subject to ice damage. At the Silver Bay Marina, an ice boom designed to maintain an intact ice cover on the harbor and keep out wind-driven lake ice failed after just one season. This study investigated ice processes in Lake Superior as they relate to the design of the safe harbors. From the analysis, we concluded that constructing a boom capable of withstanding the expected ice and wave action along the north shore would be difficult and costly and, in the specific case of Silver Bay Marina with its south-facing entrance, the additional protection offered by the boom would probably not merit the expense.

The 125^th anniversary of the First International Polar Year (IPY) will be commemorated during 2007–2008. Extensive international planning over the past several years has lead to the development of more than 200 coordinated projects, based upon over 1100 individual projects. The IPY includes data management, outreach and education activities. Four coordinated permafrost programs have been identified dealing with thermal state of permafrost, Antarctica, coastal observations, and carbon stocks. Long-term monitoring is an important element of each activity. Acquisition of a "snapshot" of permafrost temperatures in as many as 500 existing and new boreholes from both hemispheres is a major goal. Initial results will be reported at the Ninth International Conference on Permafrost in summer 2008. The International Permafrost Association (IPA) is facilitating the development of these four programs and the coordination with other related programs and activities of the IPY.

Excessive sediment is the number-one cause for listing impaired or threatened rivers on the U.S. Environmental Protection Agency's 303(d) list of impaired waters. Bed erosion under ice covers can be a significant cause of sediment in rivers and can destabilize contaminated sediments. Real-time measurements of scour under ice using Time Domain Reflectometry (TDR) scour monitors at three sites indicated that ice covers could significantly accelerate bed erosion in rivers, especially when the ice cover is restrained from responding vertically to changes in discharge. Three surface conditions were simulated in a recent laboratory investigation: open water, a floating ice cover, and a restrained ice cover with hydrostatic head conditions simulating increased discharge. The underside of the simulated ice cover was either smooth or rough to simulate thermally grown ice or consolidated brash ice, respectively. The laboratory study produced similar results to the field observations and concluded that increases in hydrostatic head increased the maximum velocity beneath the ice for the restrained cover case. The velocity profile shifts toward the smoother boundary, which is usually the bed for a rough ice cover case, thereby increasing the shear stress on the bed and subsequently increasing bed erosion. This paper is an overview of the field observations and laboratory tests and summarizes the effects an ice cover has on the scour process.

The Schoharie Creek Bridge on I-90 along the New York State Thruway collapsed on the morning of April 5, 1987, killing 10 people in the five vehicles that plunged into the river. The failure was attributed to scour at the bridge pier foundations. However, the bridge was located just upstream of the confluence of the Schoharie Creek and the Mohawk River, both of which are known for ice jams and have extensive ice covers during winter. The influence of the river ice jams and ice covers on the bridge failure has yet to be explored. This paper discusses the potential impact of the ice jams and winter hydraulic conditions on the Schoharie Creek and Mohawk River that may have contributed to the collapse of the Schoharie Creek Bridge.

The effectiveness of stormwater best management practices (BMPs) in warm climates has been extensively researched. However, little research has been conducted to determine the effectiveness of these BMPs for flow control and pollutant removal during cold weather conditions. Freezing conditions and short growing seasons can adversely affect the pollutant removal efficiency of BMPs by decreasing infiltration through frozen soils, inhibiting biological uptake, and decreasing residence times. In addition, deicers may compete for positively charged soil binding sites, thus increasing metal mobility. The objectives of this study were to determine how well BMPs in cold climates such as eastern Washington control runoff volumes and flow rates, and how efficiently they remove heavy metals (Cu, Cd, Pb, and Zn). This will be determined by integrating the following three subcomponents (1) soil metal concentrations within various BMPs, (2) sampling and monitoring of runoff during storm events, and (3) quantifying metal transport as affected by deicing/anti icing agents in soil cores in a laboratory setting. Soil metals distributions within five eastern Washington stormwater treatment BMPs have been quantified by sampling at increasing distances and depths from culvert inlets. Results show a significant decrease in soil metals concentrations with depth in the upper 30 cm of soil. Samples and flow rates have been taken from six runoff events in eastern Washington, These events include two snow melt events, a rain on snow event, and the first flush of a rain event. Comparison of pollutant loadings and flow rates indicate that lower volume snow melt events contain elevated chloride, (road salt), and solids concentrations when compared to larger volume rain or rain on snow events. As expected significantly larger runoff rates are produced by rain on snow events than by isolated snow melt or rain events. The effect of road salt (MgCl₂) on metal mobility in a laboratory soil leaching test showed marked releases of certain metal species and organic matter when soils were exposed to high ionic strength salt solutions followed by low ionic strength deionized water. These results are similar to those found in other literature sources and illustrate some of the concerns faced by designers in regions of the country where deicing practices are necessary.

During the fall of 2003, the Maine Air National Guard designed, constructed and implemented a system for diverting Aircraft Deicer Fluid (ADF) from surface water run-off to the local wastewater treatment plant. The system employs automated data acquisition and basic on-site laboratory work to ensure that maximum volume of ADF and BOD5 levels are not exceeded. Several modifications to the original design have resulted from lessons learned through the operation of the system. This paper describes the ADF diversion system, the automated data acquisition and control, and the lessons learned from operation of the system over three winter seasons.

Between 2004 and 2006 a range of six Low Impact Development (LID) designs were tested and monitored over 11 storm events for cold climate performance including filter media frost penetration and resulting hydrographs, seasonal variations on contaminant removal efficiency, and attenuation of chloride pulses associated with melt events. LID systems evaluated include 2 types of bioretention systems, a surface sand filter, a gravel wetland, a tree filter, and porous asphalt. The LID performance data will be contrasted with conventional structural BMPs (swales, retention ponds), and some select manufactured stormwater systems. Winter monitoring includes both rainfall runoff data and diurnal melt events. Contaminant event mean concentration (EMC), and performance efficiency were evaluated for storms with varying rainfall-runoff characteristics. Runoff constituent analyses included total suspended solids (TSS), diesel range organics (DRO), nitrate (NO₃), and zinc (Zn). Several water quality parameters (temperature, dissolved oxygen, pH, conductivity) were monitored as real-time data. Performance evaluations indicate that LID designs have a high level of functionality during winter months and that frozen filter media appears not be a concern. Trends in chloride attenuation are complex.

In the colder regions of the United States the range in temperature of wastewater is larger than in the warmer regions. Hydraulic capacity in colder regions is therefore affected to a greater degree by temperature. This paper looks into the extent that this happens and some of the implications. Hydraulic capacity is affected by temperature because the warmer the wastewater the faster it flows but under laminar flow conditions. Laminar flow conditions at a wastewater treatment plant occur in settling tanks and clarifiers. In order to settle the suspended solids quiescent conditions are needed. If the plant flow goes through these units, they will control the hydraulic capacity of the plant. Most plants have some form of settling tanks be it primary or final clarifiers. Under laminar flow conditions, flow is directly proportional to the water viscosity and viscosity is proportional to water temperature. Therefore plant hydraulic capacity is temperature dependent and hydraulic plant capacity varies throughout the year and with tanks that are open to the air it varies during the day. Temperature variations were studied at the Farmington, Connecticut wastewater treatment plant and laminar flow identified in various units. Plant hydraulic capacity changed 41% from summer to winter. At this 6.0mgd (22,000m³/d) plant this is a swing of almost 2.5mgd (9,000m³/d). The temperature change in the open tanks resulted in an 8% capacity difference from influent to effluent on a given day, increasing the capacity during the summer and decreasing capacity during the winter. From the 26 pages of plant design data (1994), the author could not determine what temperature was used in the sizing of this plant or if it was the same for the different treatment units. The author proposes the use of a subscript in describing the hydraulic capacity of a wastewater treatment plant to denote the design temperature. The Farmington, CT plant capacity would be listed as 6_47F mgd (22,000_8Cm³/day). Presenting it this way would be helpful in comparing parameters such as plant cost for plants from different temperature zones. For example, the 6.0 mgd (22,000m³/d) Farmington, Connecticut plant would have to be 17% bigger (7.0mgd or 26,500m³/d) if built in Bangor, Maine and 31% smaller (4mgd or 15,000m³/d) if built in Miami, Florida to provide the same hydraulic capacity.

In rural areas, land application of treated wastewater is widely used as an alternative to direct discharge to a nearby water body. Winter land application of treated wastewater in the form of snow developed as a practical alternative to larger storage lagoons, subsurface disposal and overboard discharge. Snow disposal includes the spraying of snow onto the ground surface using methods first developed in the ski industry. Regulators require that the resulting meltwater not run off the application site onto adjoining land for both aesthetic and health reasons. Prediction of the amount of snow that can be applied to a parcel of land is therefore crucial to these facilities. A finite difference model developed in-house at S. W. Cole Engineering, Inc. uses a spreadsheet to assess infiltration rates and predict allowable application rates. The model incorporates Darcy's Law. Model variables include sublimation, condensation (from the atmosphere), precipitation, runoff curve number, evaporation, porosity, hydraulic conductivity of the subsurface strata, and horizontal and vertical gradients in groundwater. Observations obtained from operating sites indicate that the model successfully predicts the allowable application rates. The model has been successfully applied to several facilities now in operation.

Traditional welded steel water tank design per the American Water Works Association [AWWA] design standard D100 uses welded joint efficiency factors less than unity and allowable tensile stress values for steel shell plates that are less than half the yield strength of the steel. While this provides a seemingly conservative approach, it does not account for the toughness requirements of steel at low temperatures. Appendix 14 of the AWWA D100 Standard allows for an alternative design method that provides higher allowable stresses, and weld joint efficiency values of unity. This toughness-based design approach has safety benefits, as well as reduced shell plate thickness requirements that result in a lower overall tank weight and possible costs savings. Specifications written by parties unfamiliar with the benefits of the Appendix 14 design may be inadvertently costing clients more for their water storage needs than necessary and are ignoring the reduced risk offered from ductile design. Even though the liability of leakage for petroleum and chemical products is surely much greater than for water, this toughness-based design method has been the primary design basis for welded storage tanks in the petroleum industry and in ASME pressure vessel design for many years. This paper will examine the toughness issues surrounding AWWA D100 - Section 14 design and provide an economic analysis to evaluate the possible costs benefit of using a toughness-based design approach.

In-line disinfection with 254 nm ultraviolet lamps was tested in a flush tank system in a rural Alaskan village. UV lamps resulted in a marginal reduction (p = 0.108) in samples testing positive for total coliform by the Colilert® method. Water use from the system was also a low 1.4 gallons per person per day (g/pd, 5.3 L/pd) in the study village. In the larger community of Bethel where no UV lamps were installed water use was 32 g/pd (121 L/pd) and total coliform presence was significantly less than in the village with or without UV (p ≤ 0.022).

A site in Maine of one of the first major soil-nailed wall installations with insulation was monitored for three years to determine the effectiveness of the design during freezing conditions. Numerous soil-nailed walls have been successfully implemented in non-frost areas, but their advantages may be utilized as well in frost prone areas. The difficulty with soil nail walls in frost prone areas is that natural soil is utilized directly behind the wall, and this soil can be frost susceptible. The site is located in western Maine adjacent to a highway that was being widened in a steep sided valley. Frozen ground generally occurs in this region from late November to late March with frost penetration in excess of one meter. The widening excavation into the slope was mixed bedrock and frost susceptible glacial till having 20 to 35 percent fines. Groundwater flowed above the steeply sloping rock surface. A three-meter high insulated station and an uninsulated station of similar geometric and subsurface conditions were monitored to quantify the effectiveness of insulation and the performance without insulation. In addition a ten-meter high insulated station was also monitored. In addition to temperatures, measurements included stresses along the nail for ten nails, soil pressures on the shotcrete, and pore pressures behind the wall. These measurements together with a subsurface investigation, construction observations, and finite element analyses give a thorough picture of the wall performance during this time period.

Blasting is one of the effective excavating methods for permafrost. It can provide a number of volumes of frozen soil for earthmoving machinery in a shorter time. The principles, kept on the construction of road cut in permafrost, to protect frozen soil and reduce the thaw-frozen disturbance are expounded. Combined with an engineering practice this paper introduces the application of blasting method in road cut for Qinghai-Tibet Railway. Base the crater tests of short hole pocket shot, short and deep holes, the reasonable the drilling machinery, spacing parameters, explosive consumption by volume and blasting scale for the excavation of road cut in permafrost is decided, and the smooth blasting was used. In additional, the construction of road cut including the temporary protection of insulation work drainage is set forth. It provides a reference for the similar engineering in the future.

Alaska Railroad Corporation (ARRC) is actively realigning several portions of track in-between Anchorage and the Matanuska Valley. This project focused on a curve straightening at MP 133 near Birchwood, Alaska. The design was complicated by anomalous subsurface conditions consisting of relic massive ice, frozen soils and normally consolidated clays that were encountered throughout the proposed alignment. Encountering massive ice in this area is unusual and rare since the Anchorage climate can not sustain massive ice growth. Massive ice consisted of up to 95% ice interbedded with silty clay with thickness up to 35 ft. Various design options were considered to support the proposed 30 ft high track embankment including: excavating the ice, thawing, maintaining the ice, bridging and realigning the proposed rail.

It has been observed that the environmental variables including temperature and seasonal frozen soil could significantly influence the soil-foundation-structure interaction (SFSI) and seismic response of infrastructures in cold regions. This paper presents the results of more than one year's monitoring and analysis of the effects of environmental variables on the dynamic properties of a selected bridge system in Anchorage, Alaska. The seismic instrumentation of the selected bridge is described in details. The ambient noises, traffic induced vibrations, and earthquake ground motions have been recorded by these instruments for a time period spanning from November 1, 2004 to December 25, 2005. Based upon these data, the dynamic properties of the bridge, including modal frequencies, mode shapes and damping ratios have been identified by System Identification method. The fundamental frequency of the bridge is found to vary 15% due to environmental impacts within one year. The environmental variables recorded from a nearby meteorological station are used to obtain the temperature and estimate the frozen soil depth for modeling purpose. A multiple-input Auto-Regressive (ARX) model for structural health monitoring was built and used to predict the dynamic properties for two scenario cases. The results show that the environmental variables including air temperature and seasonal frozen soil can significantly affect the dynamic properties and hence the seismic behavior of a bridge structure and seasonal frozen soil changes the stiffness of the bridge foundation system. Further investigation is required for better understanding of the soil-foundation-structure interaction considering the effects of seasonal frozen soil which will provide input to design code provision.

Effective in all soil and ground conditions, artificial ground freezing has proven many times to be a viable and versatile construction alternative for both open excavation support and groundwater control. The technique is cost-effective where complete groundwater cut-off and in-built ground support are required, where the work is deep, occurs in difficult ground conditions, or where access restrictions limit the use of other displacement-type ground improvement techniques. Ground stabilization at great depths also lends itself well to the use of ground freezing. The technique also works well in ground conditions and situations that are sensitive to vibration or settlement with minor ground losses associated with conventional support methods. In some cases, the protection afforded to adjacent buildings by a monolithic, impermeable structural wall is a significant factor favoring the use of ground freezing. Liquid nitrogen ground freezing has been used successfully for emergency situations in disturbed, displaced, and loose ground conditions, where detailed soil delineation cannot practically be performed. This paper demonstrates the viability and versatility of ground freezing for construction purposes through the presentation of various case histories. In addition, given that the current interest in artificial ground freezing is becoming increasingly focused on environmental remediation, this important - and growing - application is discussed in more detail and illustrated with case histories.

The mechanism of lowering temperature using slope protection of crushed stone is introduced in this paper. Through the field test of subgrade for Qinghai-Tibet railway, the variation and the evolution law of ground temperature field at the observed section, before and after the construction are discussed. The observed result shows the slope protection of crushed stone has a certain effect on lowering temperature. In additional the effect of various dimension of slope protection on the stability of subgrade is analyzed and the optimal dimension of slope protection using crushed stone is set up preliminarily. Namely, for the subgrade with height of about 1.5m the thickness of the right and the left slope protection using crushed stones with the diameter particle of 10cm were 1.3m and 0.8m, respectively. With regard to correcting the offset of ground temperature, that the thickness ratio between the left and the right slope protection is 1.7:1 is optimal.

Crushed-rock embankment has been utilized as a method for avoiding thaw settlement of roadway in warm permafrost regions. Its cooling effect has been recognized. However, in field, the crushed-rock embankments are often covered by sand or snow. Therefore, to investigate the cooling effect of the crushed-rock embankment on the underlying soil in cold regions when it is fully covered, a case of porous media was examined through the laboratory test. The results show that the crushed-rock embankment with impermeable boundaries still has good cooling effect on the underlying soil, quality of alterable thermal conductivity, and thermal diode effect. Furthermore, by analyzing the changing correlation between the top temperature and the temperature difference through the top and bottom of the crushed-rock layer, taking consideration of the temperature fields in the crushed-rock embankments, we not only prove the existence of nature convection in the porous embankment but also find its convective law. Generally speaking, in highway/railway construction of permafrost regions, crushed-rock embankment can be an effective means for protecting the underlying permafrost even if it is covered by sand or snow.

A reformative ventilated embankment installed insulated doors at the ends of ventilated ducts, which can open and close automatically according to air temperature lower or higher a threshold (such as 0°C), respectively, is proposed and used in some high-temperature frozen soil regions in Qinghai-Tibetan railway project. The main aim is to prevent heat transfer into embankment on hot season through ventilated ducts and to enhance the cooling effects of ventilated ducts effectively. Using the finite element method, the three dimension temperature fields for ventilated embankment with or without insulated door are calculated respectively on condition that the annual mean air temperature in Qinghai-Tibetan Plateau will be warmed up by 2.6°C in the future 50 years and the mean annual temperature at the natural surface is about –1°C. The results indicate that the ventilated embankment with insulated doors can decrease efficiently heat absorbed by embankment, and lowest temperatures in embankment are lower about 1°C than those without insulated doors. Thus the thermal stability of embankment is ensured even more, at the same time it also provides guarantee for the safe running of Tibetan railway. In same time, author provided advice for design department basis on analysis results.

It has been known that chilled gas pipeline buried through a boundary between permafrost and non permafrost is likely to suffer bending deflection due to frost heave since the pipeline is hold tightly in permafrost whereas the other end in non permafrost is surrounded by growing frost bulb and is subjected to frost heave. For the frost heave estimation of structure buried in frost susceptible soil, various methods have been proposed so far. However, most of those require many parameters to explain this complicated phenomenon as a numerical model since the frost heave depends on freezing rate, permeability of the soil, amount of water migration and etc. The authors, then, propose to adopt Takashi's equation for the frost heave estimation. Introducing only three parameters obtained through indoor experiment, Takashi's equation relates the frost heave ratio with the overburden pressure at the freezing front and its freezing rate. In this paper, a practical estimation method of frost heave which combines Takashi's equation and a beam on elastic foundation model is presented and numerical examples are shown. Introducing Takashi's equation, numerical example demonstrates the possibility of frost heave prediction by this method with satisfactory accuracy for engineering design.

During the winter of 2004, several agencies interested in quick and easy methods for measuring the competency of frozen ground gathered for a field test in northern Minnesota. The objective of the experiment was to test various tools for quickly measuring frost depth and strength in natural areas ranging from forests to peatlands. Six tools were evaluated for effectiveness and ease of operation. The goal was to find the best tool for determining frost depth and strength. Foresters, loggers, oilfield personnel, land managers, and military personnel need better guidance to maximize operability and minimize site damage from timber harvesting activities, oilfield operations, and military training. The participants in the study included personnel from the University of Minnesota, Department of Forest Resources; the Minnesota Department of Natural Resources, Division of Forestry; the Alaska Department of Natural Resources, Division of Land; and the Cold Regions Research and Engineering Laboratory. The tools evaluated were: 1) Alaska Slide Hammer, 2) Hilti Drill, 3) Dor-Cone (drop cone penetrometer), 4) Dynamic Cone Penetrometer (DCP), 5) Russian Cone Penetrometer, and 6) Generic (Orange) Slide Hammer. The frost depth or strength was measured using each tool, efficiency was noted, and correlations between penetration and resisting force for frost strength estimates were evaluated. The fastest method to measure depth was to drill through the frost layer using the Hilti drill; however this did not result in any strength parameters. Each of the slide hammers and penetrometers yielded a strength index profile, but the speed of the measurement varied, depending on the nature of the frost (soil type and wetness) and the geometry and weight of the tool. In some cases, the measurement took much too long to be feasible, based on the slow rate of penetration and number of blows. For shallow frost depths, which would be most critical for limiting terrain disturbance, the small diameter penetrometers (Russian Cone Penetrometer, Generic Slide Hammer, and Alaskan Slide Hammer) give the most information in the least amount of time. These instruments could be calibrated against standard Civil Engineering tools for more meaningful strength values. Comments on design and comparison of the strength indices between instruments are also discussed.

This paper focuses on two aspects of the freezing behaviour of porous materials. First, the evolution of the liquid saturation with decreasing temperature is analysed. The progressive ice invasion of the porous medium can be explained by considering thermodynamic phase equilibrium and mechanical equilibrium of the ice-water interface at the microscopic scale. Using information on the pore geometry given by the capillary curve, the characteristic freezing curve tracing the unfrozen water content as a function of cooling can be predicted. In the second part of the paper, initiation of freezing in a circular crack and subsequent cryo-suction process are considered. By making simplifying assumptions, the pressure loading inside the crack during the phase of ice crystallization and during the cryo-suction phase are computed. The solution presented in this paper is limited to the case where ice does not penetrate inside the porous material. The more general solution, taking into account the poromechanic behaviour of the partially frozen material will be studied in future work.

A controlling effect for frost heave of stabilized soil with cement-based stabilizer is investigated by means of laboratory tests according to the Japanese standard testing method. A hypothesis is introduced, which is able to explain a non-linearity of the stabilization effect on frost heave of stabilized soil, assuming the roles of ettringite and cement-silica-hydration. A series of laboratory tests are carried out, and results are shown to ensure the hypothesis.

In Japan, the C-box (Culvert-box) is constructed to allow small rivers and other crossroad intersections to pass beneath roadways. In cold regions, it is reported that due to frost heave in the banking soil around the C-box, cracking damage may develop at the surface of the pavement. The mound-shaped frost heave at the upper part of the C-box may result in cracking on the side wall of C-box caused by the frost heaving force. The purpose of this study was to test and describe countermeasures for these frost heave damages. A 1/20 scale model bank containing a concrete C-box was constructed in a large freezing room. Under controlled room temperatures, the temperature in the banking soil and the shape of freezing front in the soil around the model C-box were measured.

The mechanical properties of frozen soils depend on the structure of and the interaction between four phases: solid particles, ice crystals, unfrozen water and gas. Laboratory investigations on coarse grained sands show that the thermal conductivity of the various phases control the advance of the freezing front and the growth of the ice crystals on a microstructural level within the soil for different temperature conditions. The consequence is that the structure of the frozen soil changes until the soil reaches steady state thermal conditions. One-dimensional freezing tests further showed changes in the growth of the ice crystals in saline pore water due to solute rejection during phase change of the water. Based on these laboratory investigations numerical modeling was performed using finite elements. The numerical modeling allowed for investigating the effects of the four phases and their heat conductivity ratios on one- and two dimensional freezing. The modeling further confirmed that depending on the temperature gradient and freezing rates, different structures within the frozen soil are to be expected. Even though these numerical investigations were solely based on thermal modeling, i.e. changes in stress due to freezing were ignored, qualitative conclusions can be drawn about the variation in strength and creep behavior of a coarse grained frozen soil based on its thermal history. Depending on the volumetric content of the components and the heat conduction ratios between the four different constituents a freezing rate exists at which the freezing front advances uniformly to freeze the sample. On the other hand, rapid freezing results in a heterogeneous structure of the frozen soils and pockets of unfrozen water must be expected, which will weaken the partially frozen soil.

Using the relationship between the soil bulk electric conductivity and the pore fluid salinity, the salinity redistribution during freezing can be evaluated as a function of temperature and time. Laboratory freezing tests were carried out with fine sand saturated with a saline solution to study the redistribution of the salinity as the sample freezes. Time domain reflectometry (TDR) probes, which have been widely used in soil science and water resource engineering to measure volumetric moisture contents and soil electrical conductivity, were installed in a 1.5m diameter and 1.0 m high soil sample to monitor solute redistribution during freezing. Initial TDR calibrations were carried out to create a correlation between soil electric conductivity of saturated fine sand and the soil pore water salinity at various temperatures. In addition, the influence of temperature and probe length of the TDR test apparatus was studied. The result showed that: (i) the relationship of the apparent soil electrical conductivity, versus the soil solution salinity is temperatures dependent, (ii) this relationship is influenced by the length of the TDR probes, (iii) unfrozen volumetric water content can be overestimated by as much as 35 percent as a result of the influence of salinity when using the TDR technique, and (iv) the substantial solute rejection can occur for temperature gradients of 0.3 °C per day or less during radial freezing of saline soil.

The strength of soil depends on time, temperature, soil type, soil parameter and, last but not least, on its salinity. Whereas in scientific literature a lot of information is available concerning the influence of time and temperature, only little can be found concerning soil parameter and even less on salinity. The few published investigations on frozen soil behavior mainly focus on non-cohesive soils with a salinity of S = 3 %, which is the salinity of ocean water and higher salt concentrations. Former tests on non-cohesive frozen soil material, performed at CDM's frozen soils lab in Bochum, Germany, indicated a major impact on the frozen strength and deformation behavior at a salt concentration of S = 0.8%, whereas other tests on soil samples with a higher content of fines and nearly the same salinity showed a minor impact of the salinity on the frozen soil properties. Therefore, CDM decided to perform an additional test series in order to determine the relationship between the strength properties of various frozen soil types and salt concentrations. The results of the strength and deformation tests at different saline soil layers show a decrease of the strength properties with increasing salinity, whereas a significant influence of the amount of fines could be established. With an increasing amount of fines the impact of the salinity on the strength behavior is decreasing.

Of soils discharged from construction sites, those that have insufficient strength or are otherwise unsuitable for use as foundation material have been disposed of until recently in Hokkaido. With the growing interest in effective recycling of such soils, they are now being modified with solidifying agents and are often used as filling material. Construction projects have grown in scale, and to create employment opportunities, construction often extends well into the winter. Recycled soils modified with solidifying agents are becoming used in cold weather, bringing concerns over the effects of low temperature on the final strength of the material. Curing temperatures are known to affect strength when unsuitable soil is modified with cement or lime. The temperature is low during winter construction, but it rises in spring, and this change in air temperature could affect the strength of modified unsuitable soil. However, few studies have examined the effect of temperature changes on the strength of modified soil. This research obtained unconfined compressive strengths in modified soils that were cured under various temperatures and curing durations. The results have revealed that the strength varies according to the curing temperature; even with low initial curing temperatures, the final strength is greater when subsequent curing temperatures are higher; and the strength does not increase with time under freezing conditions.

Soil strength depends on the state of the ground as well as soil type. We have developed both a 1-D and pseudo 3-D SVAT (Soil-Vegetation-Atmosphere Transfer) model, FASST (Fast All-season Soil Strength). FASST predicts soil temperature, moisture, ice content and strength as a function of depth as well as snow accumulation/depletion and surface icing. We implemented a series of parametric tests with the 1-D model to develop the understanding necessary to predict soil strength as a function of soil type, soil hydraulic properties and strength parameterization.

Conditions of ice lens segregation when a new one is start to grow is explored in experimental approach in this paper. Two consecutive open-system frost heave tests have conducted with the same Diluvial Silt of which mean tensile strength is 0.75MPa. Four predominant ice lenses have segregated in 114.2mm high specimen during the first heave test. Then, applying positive temperature to pedestals: i.e. thawing, the second heave test has conducted under the same test condition as the first one. Ice lenses have appeared exactly the same location as the first test. Therefore, it is revealed that ice lens initiation takes place where lower tensile strength remains. The temperature of ice lens initiation during the second frost heave test which have been interpolated by the boundary temperatures have found to be higher than those of the first ones. The difference is almost the same as 0.95 °C. Expected pore ice pressure at the initiation calculated with generalized Clausius-Clapeyron Equation using the temperature difference is 1.05MPa. The calculated pore ice pressure is little higher than the sample tensile strength. Therefore, it is expected that pore ice is pushing structure of mudstone against its tensile strength and overburden pressure to fracture via unfrozen film water.

The formation processes of clathrate hydrate in porous glass powder saturated with a THF-water mixture were directly observed using directional cooling apparatus. As the cooling rate decreased, the THF hydrate grew while excluding glass particles and formed a lens like layer of hydrate in the same manner as in the phenomenon of ice lens formation in the freezing of soil. The growth rate of the hydrate lens was proportional to the degree of supercooling of the growth surface and was about 1/20th the growth rate of the hydrate layer in a THF-water mixture without glass particles. When sodium chloride was added to the system, the growth of the hydrate lens was inhibited depending on the salt concentration.

During the freezing of a fine grained soil, ice lens formation changes the structure of the soil and results in frost heave. The formation of the ice lenses is very complex and dynamic. The results are presented from laboratory freezing tests on saturated Devon silt, a frost-susceptible soil. A variety of pore-water salinities, vertical pressures, and temperature gradients were used to investigate the different effects on the freezing process and the formation of the ice lenses. Using a novel experimental methodology, ice lens growth at the pore scale was observed. Fluorescein was dissolved in the pore water, which allowed to locate unfrozen water under UV light. In this manner it was possible to visually observe and measure the ice lens growth ahead of and behind the frozen fringe. It is visually shown that the thickness of the ice lenses, the distances between the ice lenses and the thickness of the frozen fringe change with changing temperature gradient, vertical pressure and salinity. In addition, the ice structure within the saline soils became more three dimensional and irregular compared to the non-saline samples, where the ice lenses develop over the entire cross section of the sample.

Solid waste disposal units (i.e., landfills) are highly regulated engineered facilities, and are specifically designed to account for a multitude of location-specific considerations. In addition to conforming to the numerous federal, state, and local regulations and design guidelines, landfill designers, operators, and owners must also address climate-related factors such as temperature, precipitation, wind, and growing season. For landfills located in cold regions such as central Maine, additional climate factors include snow management, frost penetration, freezing temperatures, and ground thaw, which must also be addressed during landfill construction and operation. This paper identifies and describes several of the cold-region specific considerations that have had been incorporated into the design, construction, and operation of two landfills located in Central Maine.

One of the main degradation mechanisms of permafrost underlying transportation infrastructures is associated with the geometry of airfields and roadway embankments. The two phenomena that are associated with this mechanism are shoulder rotation and longitudinal cracking. These problems occur on the embankment side-slopes because of the important snow drift in winter, which causes a warming. This warming induces an accelerated thaw of the permafrost table under the shoulder causing a loss of support. In this context, the paper describes a new mitigation method, the heat drain, that was developed to counter permafrost degradation problem on the side-slopes of the embankment. The heat drain was developed by the Groupe de recherche en ingénierie des chaussées de l'Université Laval. This technique allows heat extraction from the embankment during winter. The heat drain consists in a highly permeable geocomposite placed in the shoulder. An air intake is installed at the foot of the embankment in order to allow the upward movement of air in the membrane. The paper summarizes the performance of this new technique. This new method was tested in laboratory and proved to be effective to reduce the ground temperature.

The effects of placement practices on decomposition of wastes were investigated at Anchorage Regional Landfill (Anchorage, Alaska) since 2002. Temperatures and gas concentrations of wastes placed at various seasons were monitored. Wastes were placed at sub-freezing temperatures during cold seasons. Waste temperatures generally increased upon placement. High variation was observed in waste temperatures near the surface whereas steady temperatures were obtained at depth. High maximum stable temperatures resulted from warm placement conditions. Steady temperatures between approximately –1 to +35°C were observed. The central portion of a frozen waste band (with a total initial thickness of 7 m at placement, currently between depths of approximately 8 m to 15 m) remains frozen 2 years after placement. Both the top and bottom regions of the frozen waste band have thawed. Heat Content (HC) varied between –8.2 (for 2-year-old waste at a depth of 11.9 m in frozen wastes) to +25.9°C-day/day (for 13-year-old waste at a depth of 32 m for waste placed in summer). The measured frost depths in waste ranged from 0.7 to 1.3 m and were less than that for native soil at the landfill site. Instantaneous thermal gradients ranged from –73 to +60°C/m. Gas concentrations were similar to air at the time of waste placement. Anaerobic decomposition conditions and onset of landfill gas production started within 3 to 4 years of placement for wastes placed during warm seasons. Virtually no decomposition or gas generation were observed in the frozen wastes. A 1-D numerical model was used to investigate distribution of temperatures for placement at varying temperatures and for varying lift thicknesses. It is recommended to minimize frozen lift thicknesses to obtain higher temperatures.

Polar mean annual air temperatures have been on the rise for the past few decades and are expected to continue to rise (Hassol, 2004). Because of this, mean annual ground temperatures are rising as well, and civil engineering projects in the circumpolar north are at risk due to the possible thawing of their frozen founding soils. The implications of the issue are not fully understood at this time, and studies are needed to describe the degree and nature of the changes that will take place, and what action the engineering community should initiate. It is also expected that engineering studies will be needed to learn techniques to remediate failing structures, learn new construction designs to withstand degrading permafrost, and methods to accurately predict areas at risk of degrading permafrost for project planning. This report details the anticipated research and technology needs to address this issue.

A small landslide was observed next to an existing large slide in the Travaillant Lake area in summer 2005. The area was burned by forest fire in 1998. Information available from air photos, field observations and recent photographs indicated that the observed slope failure is relatively new (probably within the last year or two). This paper describes field investigations conducted in the area in late summer 2005, and slope stability studies carried out to back analyze the slope failure mechanisms. Test pits were excavated in the area to investigate subsurface soil conditions and soil samples were taken for geotechnical laboratory testing. Field vane shear tests were conducted on the active layer to measure the shear resistance. Laboratory tests showed that the moisture content in the active layer increased with depth. The field vane shear tests indicated very low undrained shear strength of the fine-grained active layer soils. Slope stability analyses revealed that the slope would be in a critical condition when the vane shear strength parameters were used. The studies provided insights into understanding initiation of landslides in the region. The analyses also indicated that the field vane could be a useful tool for slope stability studies in fine-grained permafrost soils.

A site on Farmer's Loop Road in Fairbanks, Alaska, was the focus of considerable permafrost-related research from 1945 through the mid-1980s. Many field studies conducted at the "Fairbanks Permafrost Experiment Station (FPES)" included measurements of ground temperatures and depth to the top of permafrost. In light of recent accelerated climate warming in the Arctic, this historical information is of interest, and some key past work is summarized here for the convenience of researchers who may want to access this information. Initial studies conducted on FPES pertained to the influence of several factors on ground temperature distribution and permafrost stability, including ground cover and its removal, building foundations, types of pavement bases, insulated pavements, and various pavement surfaces. Other studies that incorporated a component of ground temperature measurement included measurement of forces exerted on various piles by frost heave and the reduction of frost heave by surcharges. Several research efforts that measured ground temperatures or generated results based on ground temperature data (e.g., frost penetration, depth to permafrost) are summarized here.

Two full-size piles with a spiral design on the lower part of the piles, augered and backfilled with slurry were tested in natural permafrost in Prudhoe Bay, Alaska to assess their performance as foundations for the Hot Ice No 1 Drilling Platform of Anadarko Petroleum Corporation. The platform was used to drill gas hydrates 40 miles southwest of Prudhoe Bay, Alaska in winters 2002 – 2004. The purpose of this paper is to present the pile displacement results and lessons learned during the testing procedure. The design requirements included a 667-kN ultimate pile capacity for a design life of 2 years, allowable settlement of 15 mm in 2 years and 25 mm in 40 years. The foundation system needed to assist in keeping the permafrost from degrading and to be removable so that the tundra would be left behind undamaged. The function of the hollow spirals was to aid in removing the piles by circulating a heated liquid down the pile, through the helixes and back to the surface, and to possibly increase the pile capacity. Three levels of compressive loads were applied using a load cell and a mechanical actuator. The pile displacement during the test was measured using a potentiometer and three types of secondary systems. Thermistors were used to monitor the ground temperatures during the freeze back period and testing. The displacement rate for primary and secondary creep in permafrost was obtained using the potentiometer and confirmed with the secondary monitoring systems. Dial gages proved to be the best secondary measurement system used. The test results indicated that the required factor of safety of ⩾ 1.5 was obtained for the pile capacity in regards of shear failure for the governing ground temperatures (from –4 to 0°C). However, the creep settlement rate for the temperature range exceeded the tight allowable rate. Recommendations for future testing programs are given.

The Minnesota Road Research Project (MnROAD) was constructed in 1990–1993 as a full-scale pavement testing facility. Several different cells were built with various materials, mix designs, and structural designs. Two different asphalt binders were used during the original construction: PG 58-28 and PG 64-22. The sections have all shown various degrees of low temperature cracking. In general the cells with stiffer binder (PG 64-22) experienced a higher number and greater severity of thermal cracks than those with the softer binder. The ride quality of the pavements has been adversely affected by the deterioration of the low temperature cracks. In 1999 three cells were reconstructed on the Low Volume Road as a study specifically examining low temperature cracking. These sections were designed using the exact same Superpave mix design except for the asphalt binder type, which differed at the low temperature performance grade. The performance grades for Cells 33, 34, and 35 were PG 58-28, 58-34, and 58-40 respectively. After several years in service these sections have begun to show marked differences in performance. Cell 35 has shown the most cracking, even though it has the softest grade at –40. The cracks on Cell 35 do not look like typical thermal cracks, while Cell 33 exhibits the expected typical thermal cracks. Cell 34 had virtually no distress after six years.

The Minnesota Department of Transportation (Mn/DOT) built the Minnesota Road Research Project MnROAD between 1990–1993. The 2.5-mile low volume road and the 3.5-mile mainline consists of a 2-lane roadway that originally contained gravel, hot mix asphalt, and concrete test cells designed for both low volume roads and interstate traffic. The mainline interstate cells are trafficked by public interstate traffic and the low volume road has a Mn/DOT 5-axle tractor-semi-trailer to simulated conditions of rural roads in two load configurations, resulting in the same equivalent axle loads or ESALS. MnROAD is located in a wet freeze zone that has affected both its base and subgrade materials with seasonal frost movements. This movement has slowly deteriorated each test cells ride over time. MnROAD has monitored the frost movements using frost pins and has measured the ride (international ride index - IRI) using high-speed profilers over for the life of the project. This paper investigates the loss of ride from both environmental and traffic loadings and how they have combined to cause the deterioration of ride over the last 10 years at MnROAD. The findings suggest that our current process to develop a mechanistic empirical design is currently missing the fact that seasonal differental frost movements play an important role in pavement performance in northern climates and need to be taken into account.

Compressive strength and elastic modulus on a new asphalt mixing material with different temperatures were tested with a universal material test machine and under cyclic frost actions, in which the temperature range was from –20 °C to 25 °C and the cyclic number was from 0 cycles to 30 cycles, in particular, the asphalt mixing material was applied on the road pavement in cold regions. The test results showed that the strength and anti-frost properties of the asphalt mixing material had a good improvement, and their Compressive strength and modulus were obviously increased with the temperature decrease, but their values with a same temperature would nearly be stable under the different cyclic number of frost actions, so it was a good pavement material and had a good application foreground in cold regions.

On a local gravel road at Leppävirta, Central Finland (about 62° 27' N, 28° E), test pavements were constructed in the fall of 1996 to strengthen a gravel road, suffering from thaw damage. The improvement methods studied were bitumen stabilization in situ of the base layer, and a macadam overlay reinforced with a steel net. In the primary monitoring in 1997–98 it was fond that the improvement of the surface modulus was significantly higher at the reinforced pavement than a non-reinforced one. The estimated material modulus of the reinforced macadam layer (200mm thick) during thaw was 300–500 MPa. According to the FWD measurements in the spring of 2005, the back-calculated, equivalent material modulus of the reinforced layer during thaw had decreased to about a half of the value estimated after construction. The modulus in summer state was not observed to have significantly changed. In the test pit excavated across the road down to the reinforcement, and the leveling of the reinforcement it could be observed a bending (rutting) of the reinforcement. The estimated rut depth was 25 – 85 mm, and it was increasing with decreasing minimum thaw modulus of the gravel pavement. The rutting of the pavement was relatively low. Thus, it seems obvious that the reinforcement should be treated, not only applying as stiffening of the overlay, but also as an anchoring element, preventing the horizontal and shear stresses below.

Test sections were constructed in two portions of Maine Route 9 to investigate the use of geosynthetics for reinforcement and drainage of 300 mm (12 in.) and 600 mm (24 in.) thick subbase courses beneath 150 mm (6-in.) of flexible pavement. Four types of test sections were constructed: geogrid reinforcement, drainage geocomposite, drainage geocomposite with geogrid reinforcement, and control. Test sections using reinforcement geogrid have strain gages attached to the geogrid to measure induced forces. Reinforcement sections have geogrid either on subgrade or in the center of the subbase to evaluate the effects of geogrid location. Drainage geocomposite and control sections have vibrating wire piezometers to monitor porewater pressure in the subgrade and subbase course. Thermocouples were used to measure the depth of frost penetration. This paper deals primarily with seasonal and long-term changes in the geogrid force.

Full-depth recycling in conjunction with cement stabilization is an increasingly popular method of reconstructing deteriorated asphalt pavements; in cold regions, cement treatment is especially useful for improving the strength and durability of frost-susceptible materials. The purpose of this research was to characterize the properties of cement-treated base materials constructed from reclaimed asphalt pavement (RAP) and aggregate materials obtained from several locations in New England. The aggregates and RAP were blended in approximately equal proportions by weight, and extensive laboratory testing was performed to evaluate the strength and durability of each blend in the untreated condition and after treatment with various levels of cement. Testing included determinations of particle-size distributions, moisture-density relationships, unconfined compressive strengths, and moisture susceptibility classifications for the materials. The results of the laboratory testing were used to establish design parameters for field test sections constructed in the summer of 2005. Field testing, including stiffness evaluations using a falling-weight deflectometer, was conducted to characterize the structural properties of both cement-treated and untreated sections subjected to early trafficking. This research suggests that modest amounts of cement can greatly improve the strength and durability of base materials, which should in turn increase their resistance to frost damage.

The use of cement stabilization in conjunction with full-depth reclamation (FDR) is an attractive pavement reconstruction method when economic, environmental, and engineering perspectives are considered. For reconstruction of Interstate 84 in Weber Canyon, Utah Department of Transportation engineers specified FDR together with the application of 2 percent Type II Portland cement by dry weight of aggregate for stabilization of the existing frost-susceptible pavement base material. However, due to the shortage of cement supplies in Utah, they accepted a value engineering proposal allowing the use of transition cement instead. This research compared the performance of transition cement to that of Type II Portland cement in several laboratory tests. Samples of base and reclaimed asphalt pavement (RAP) were obtained from the field site, and specimens were compacted using modified Proctor compaction energy and equal proportions of base and RAP by weight. Laboratory testing included particle-size analyses, moisture-density testing, 7-day and 28-day unconfined compressive strength determinations, 28-day resilient modulus measurements, and moisture susceptibility evaluations using the tube suction test. Three replicate specimens were subjected to each test, and the resulting data were analyzed using statistical hypothesis testing. The strength and stiffness of both sets of specimens exceeded the design values specified for the project, and the moisture susceptibility ratings for the two treatments were both determined to be "good." Statistical analyses performed on the data indicate that the strength, stiffness, and moisture susceptibility characteristics of specimens treated with Type II and transition cements are not significantly different between cement types at the standard level of significance of 0.05. To the extent that the sample of transition cement evaluated in this research is representative of the bulk supply of transition cement provided for the project, the data collected in this research suggest that the performance expectations for both types of cement are very comparable.

The construction of transportation infrastructures in permafrost conditions affects the thermal regime of frozen ground and may cause deterioration of the permafrost. If the pavement foundation is constituted of ice-rich permafrost, the structural and the functional capacity of the infrastructure will be affected. Moreover, pavements that have always been stable over permafrost now begin to become unstable as a result of the observed trend toward a global warming of the planet. This is becoming an important engineering problem for northern transportation infrastructures. Several techniques have been developed and experimented to protect pavements from permafrost degradation. However, few of these techniques have been widely adopted in current construction practice. The project is intended to provide guidance to Yukon Highways and Public Works and other interested Canadian highway agencies on possible adaptation scenarios. The objective was to propose and document cost-effective adaptation scenarios for Alaska Highway as well as other Yukon highways to permafrost degradation resulting from global warming. The project involves a comprehensive literature review on deterioration mechanisms and on protection techniques. It also involves the analysis of the thermal regime of an instrumented section of the Alaska Highway near Beaver-Creek YT. The paper will focus on the results on the thermal analysis and on recommended adaptation scenarios.

In Nunavik, permafrost degradation is now inevitable and it will eventually threaten the integrity of transportation infrastructures owned by Ministère des Transports du Québec (MTQ). This study was initiated by the MTQ in order to adapt its transportation infrastructures to the new climatic reality. The purpose of this study is to carry out a performance assessment of the Nunavik runways and access roads since their construction in order to determine the appropriate adaptation techniques to reduce permafrost degradation. Thereafter, initial and current base material thickness longitudinal profiles of various runways were compared to determine if settlement occurred. Comparisons between the initial and current runways condition were carried out using air photographs. The assessment made it possible to identify six unstable runways and two unstable access roads. This paper presents the three most unstable runways in Nunavik : Kangirsuk, Salluit and Tasiujaq. Recommendations for the problematic airports, such as permafrost characterization through deep drilling and implementation of mitigation methods, were given at the end of the study. Three mitigation methods are proposed to counter permafrost degradation at Nunavik airports : heat drain, air convective embankment and reflective surface.

Three products are presently used in Norway as thermal insulation in roads; lightweight clay aggregate (LWCA), foam glass aggregate and extruded polystyrene boards (XPS). As a evaluation on the present design guidelines for frost insulation and in order to compare the different insulation products, a field test is established in connection with the construction of a new main road (E6) south of Trondheim in Norway. The test site has four different sections. In addition to three sections where each of the three different insulation products are used, one additional section contains ordinary pavement materials produced from crushed rock.

Frost heave and thawing, resulting in deformations and reduced bearing capacity, cause major maintenance costs of the roads in the Nordic countries. The most common frost protection method in Norway is to increase the layer thickness of crushed rock. In addition has extruded polystyrene (XPS) been used since 1965, but this has gradually been reduced the last 20 years. Light expanded clay aggregate has regularly been used as frost protection for other civil engineering applications such as railways and buildings, and is occasionally also used for pavements. Lately also foam glass aggregate has become an option. To optimize the design, SINTEF has performed a number of numerical analyses and developed design charts that give the required layer thickness of frost insulation. The numerical analyses showed that the frost protection layer using expanded clay or foam glass aggregate should be 4–5 times thicker than a frost protection of extruded polystyrene.

During winter, frost penetrates in pavement materials and subgrade soils. When the frost front reaches frost susceptible subgrade soils, water is sucked toward the frozen fringe where ice lenses are formed. Heave of the pavement surface resulting from these phenomena can reach and even exceed 150 mm for climatic conditions prevailing in Canada. Frost heave in soils is the result of the combined action of heat and moisture transfer in freezing soils. Freezing soils in a pavement system are subjected to a thermal gradient. As a result, a negative pore water pressure gradient is created behind the freezing front, in a thin layer of partly-frozen soil known as the frozen fringe. Based on this understanding of the frozen fringe conditions, Konrad and Morgernstern (1980) have developed the segregation potential concept to model one dimensional frost heave in soils. The segregation potential is a mechanistic index allowing for the quantification of the frost susceptibility of a given soil in specific climatic conditions. The segregation potential concept has been used for more than two decades for several engineering applications. The paper will describe how the segregation potential can be assessed using a freezing test or estimated using various approaches. It will also describe how the segregation potential can be used to perform frost analysis of pavements subjected to cold climates.

Damages caused by frost heave as well as rutting caused by loss of bearing capacity in the spring thaw period are serious problems in most countries that experience frost in the winter time. In order to keep a road network in frost affected areas with minimum total costs for road users and owners, a proper use of materials which serves winter and spring thaw conditions is essential. The choice between extensive utilization of local materials and the use of high quality material, transported a long distance, is an important issue. In this context it is important to avoid severe differential frost heaves, and to keep a high bearing capacity throughout the year. For unbound material, the mineralogy, fines content and permeability are key factors to determine the frost susceptibility. In this project one grading of unbound material from three different rock types was investigated; one relatively weak mica-rich gneiss from Sweden and two typical hard rock types from Norway. The potential for frost heave was investigated by laboratory frost heave tests. Repeated load triaxial testing was used to study the resilient modulus and the permanent deformation properties under influence of different water contents. This investigation improves the knowledge on the behavior of unbound aggregate used in road construction in regions with cold climates.

Water vapor flow in response to temperature gradients within a pavement structure can be an important mechanism of moisture ingress in aggregate base layers during winter in cold regions. The objective of this research was to investigate the occurrence of vapor flow in a non-frost-susceptible granular material typical of base layers within the jurisdiction of the Swedish National Road Administration. A cylindrical specimen 145 mm in height and 123 mm in diameter was positioned 100 mm above a water table inside an insulated plastic tube, which was instrumented with thermistors to facilitate monitoring of the specimen temperature profile. A computer-controlled cooling element was utilized to freeze the specimen surface to a constant temperature of –1°C for 41 days, and the bath water was maintained at a temperature of 5°C. The specimen weight was measured weekly to quantify the vapor flow rate, and the specimen was then physically separated into four layers and oven-dried to determine the moisture profile at the end of the testing. Additional characterizations were also performed to facilitate material classifications, and numerical modeling was conducted to simulate the observed laboratory results. Analyses of the data indicate that the average vapor flow rate was 12.8 g/week and that the majority of the incoming water accumulated just above the freezing front. The water content near the freezing front was 10.8 percent, which corresponds to a degree of saturation of 115 percent, while the water content was only 3.9 percent near the base of the specimen. Although redistribution of moisture residing within the specimen prior to the start of testing could be partially responsible for the higher water content that developed in the vicinity of the freezing front, the overall increase in specimen weight is attributable only to the movement of water vapor from the bath into the specimen; the condensation and freezing of water vapor within the pores of the specimen caused localized super-saturation of the material. The data clearly show that water vapor can be an important source of increased water content within pavement base layers subjected to sustained freezing temperatures.

The Maine Department of Transportation (MDOT) maintains over 12,870 km (8,000 mi.) of urban and rural highways and 2937 bridge structures throughout the State of Maine. Maine has a diverse geology with seasonal temperature variations. Winter months have temperature extremes below freezing for extended periods. The temperature changes and geology can work in combination to heave flexible pavement and fixed structures. MDOT incorporates several methods in building the infrastructure to minimize the impacts of heave movement. Through the years, MDOT has adapted designs and construction practices to accommodate the affects of the changing seasons. These standard methods are used in several geologic conditions to correct abrupt bumps and deteriorated surfaces. Standard details and recommendations utilize transition zones, drainage, and undercuts to take advantage of the changed geology. This paper will discuss the background of typical conditions that cause road heaving in Maine. It will describe the various standard details and how these are applied to Maine highway construction. It will also discuss some of the new methods that have been implemented in the design and construction of highways.

Modulus (stiffness) is one of the primary inputs to mechanistic pavement design procedures and provides insight into long term pavement performance. Despite the importance of modulus, some aspects of pavement construction and management are still based on measurement of parameters that are not directly connected with long-term performance or on empirical based judgments. The goal of this research was to evaluate the effectiveness of the portable falling weight deflectometer (PFWD) in evaluating the support capacity of pavements during spring thaw. The performances of ten low volume roads, three of which were gravel surfaced, were evaluated through spring thaw and recovery. Instrumentation was used to measure frost penetration in the test sections on days when measurements were made. Comparisons were made to the traditional falling weight deflectometer (FWD). It was shown that the PFWD was able to follow seasonal stiffness variations and compared well with FWD derived moduli on gravel surfaces. Recommendations were made for using a PFWD to determine when to place and remove seasonal load restrictions.

The Maine Department of Transportation (DOT) spends a considerable amount of funds on construction, rehabilitation and maintenance of asphalt pavements in the state of Maine. Full depth reclamation without using any additive, or by using foamed asphalt or plant mixed recycled asphalt pavement provides a cost effective way of rehabilitation of old and distressed asphalt pavements. Generally, these reclaimed pavements consist of approximately 18 inch of subbase, 6 inch of reclaimed base and 3 inch of hot mix asphalt. One of the critical things needed for the durability of these pavements is the consideration of the entire structure and its resistance against the different kinds of distresses, namely, fatigue cracking, rutting and thermal cracking during design and construction. This study was carried out to model such pavements and simulate the effects of design traffic and environmental conditions. The scope of work involved collection of data for the pavement structures and materials, setting up models, and simulating the effect of different materials and structures with the newly available Mechanistic Empirical Pavement Design Software. Pavements reclaimed between 2002 and 2003 were surveyed to determine the prevalent type of distress. In relatively cooler areas with low traffic, the prevalent distress mechanism seems to be thermal cracking. The results of simulation with different binder grades show that a –34 grade binder is necessary to prevent thermal cracking failure (> 1000 ft per mile) within a design life of 15 years for pavements in projects located north of Bangor. To continue with the currently used –28 grade binder, these pavements must be constructed with 8 inch (or more) thick hot mix asphalt surface layers. In consideration of the initial cost, the use of a –34 grade binder seems to be the more attractive option to prevent premature thermal cracks in such pavements.

Development of seasonal pavement design models requires understanding of the various factors that can affect the models as well as a practical understanding of model development and pavement design. The New York State Department of Transportation (NYSDOT) desired a statewide seasonal model of pavement moduli to augment their current pavement design methods and to use with the new mechanistic-empirical pavement design approach being developed by NCHRP. Due to a limited project budget, development of a seasonal model for the entire state was unfeasible. Instead, the primary project goals were to develop a seasonal pavement design model for the limited inference space defined by the testing sites and to determine an efficient approach for development and expansion of such seasonal models. In working with NYSDOT, the development of the experimental design was influenced by economic considerations and practical time constraints. Rather than spend project funds on expensive laboratory testing with a low probability of use in the future, it was decided to perform more falling weight deflectometer (FWD) testing to define the spatial and temporal variations in pavement response. The results of this initial phase provided a series of lessons learned that resulted in a final approach for seasonal model building. In order for a seasonal pavement model to be developed using FWD testing, climatic analysis of the possible inference space must be completed to define climatic regions based upon expected pavement response. This analysis must define seasons in reference to how they will affect pavement design. In New York State, with its large variation in pavement response due to freezing and thawing of the unbound layers, five or six seasons are needed. In addition, the climatic analysis must determine the minimum number of FWD tests to determine the variations in the response of the pavement. The geotechnical and site investigation plan should use existing protocols as much as practicable without adversely compromising data needs. Prior to the FWD testing, a comprehensive, but flexible, testing and data analysis plan is critical. The FWD data must be analyzed as they are collected. Finally, the assumptions of the pavement model in backcalculation of the data and the forward calculation in the pavement design must be synchronized to allow proper comparisons. Model development must include spatial and stochastic variations in the real world and needs to be validated with real world data.

A 3-D Finite Element Model (FEM) was constructed of a loaded C130 tire rolling over a frozen unpaved runway. The runway was constructed of two layers, one frozen and one unfrozen, each of varying thickness, but the combined thickness was fixed. The material model used in the dynamic FEM simulation represents a frost-susceptible soil, which was used in full scale unpaved road tests at CRREL's Frost Effects Research Facility (FERF), and was calibrated with triaxial tests and validated against direct shear test data. The area of interest in both models is the interaction between the frozen (top) layer and the unfrozen (bottom) layer. Stress and strain data collected from the dynamic FEM simulation and a prior single point layered elastic (SPLEM) simulation were compared, along with the capabilities of each model. Initial findings show that the single point layered elastic model is much faster at obtaining results for a single point, but it can only solve problems with pure elastic material. The FE model can solve problems with any material and the results can be viewed at any location or point in time during the run. The FEM visually represents what is happening in the soil around the tire as it rolls along the surface, while the static SPLE model only predicts what is occurring directly below the tire.

The objective of this study was to evaluate and document claimed cost savings of a proprietary, pre-blended deicing chemical versus the conventional sodium chloride used by the New Hampshire Department of Transportation [NHDOT]. This supports the Department's continuing objective of instituting operational efficiencies. The NHDOT District 2 received approval for this proposed Federal Highway Administration [FHWA] funded Statewide Planning and Research [SP&R] project in April 2003. Contrary to the manufacturer's claims, some leaching of liquids in the test material stockpiles did occur, thereby affecting uniformity of product. However, partial remixing occurred as the material was loaded into the trucks. The materials were evaluated in 16 storms at Bristol and 12 storms at Sunapee during the winter of 2003–2004. In Bristol, the first 14 storms resulted in 1.04 tons less test material used per lane mile [17.4% savings], but it cost $51.67 per lane mile more because of increased material cost. In Sunapee, for the 12 events documented, the total materials saved was 1.03 tons per lane mile, but it cost an additional $48.81 per lane mile. These savings in the amount of material used cannot justify the additional cost of the test material of $17.86 per ton in this evaluation. However, if the cost differential is much closer, it may be beneficial to purchase small quantities to be used under those few favorable conditions documented in this evaluation.

Icing problems on roadway are very common in cold regions. It binders traffic and causes traffic accidents, and also damages roads after it thawed in spring breakup period. Many measures, such as blind ditch, pit for containing ice, ice fence/wall, insulated drain etc., have been used. But they were not always effective. This paper introduces the icing problems in China and presents two engineering practices on mitigation icing problems on road in seasonal frozen soil regions and permafrost regions respectively. The seasonal frozen soil regions are in the west of China (high elevation cold regions) and the permafrost regions is in the northeast of China (high latitude cold regions). Two methods of mitigating the icing problems on road are introduced. They are the blind ditch and the multi-outlets drainage system. The blind ditch was used in seasonal cold regions and the key is the imbedded depth of the ditch. The multi-outlets drainage system method was applied in the permafrost regions and the key of this method is the multi-outlets and the good cooperation among all the components of the system. Results show that they were effective on mitigating icing problems on road.

This study evaluates the effectiveness of snow fences in reducing crashes and road closures on an 80.4 km remote rural section of Interstate-Highway 80 in southeastern Wyoming. No snow fences were in place when this highway was first opened to traffic in 1970, but serious snow drifting problems necessitated a large-scale snow fencing program from 1971 to 1990. The proportion of road protected by snow fences has increased to a present level of 73%, consisting of 69.7 lineal kilometers of snow fence, ranging in height from 1.8- to 4.3 m and constructed at a cost of $1,910,000. Total crash rates per million vehicles, for October through April, have declined in proportion to the percent of road protected by snow fence at a statistically significant rate of approximately 20 crashes for every 10% increase in snow fence protection. This decline in crashes is equivalent to a 75% reduction in crash rates in areas protected by snow fences, and is attributable to the elimination of snowdrifts, improved visibility in blowing snow, and reduced road ice. At the 2003–04 seasonal traffic volume of 1.9 million vehicles, the existing snow fences prevented 275 crashes, 124 injuries, and 3.1 fatalities per year. This implies an annual return of $25,000,000 on the original capital investment of $1,910,000. If the fences were replaced at current costs, and traffic volume remained constant at the 2003–04 level, these benefits yield a benefit-to-cost ratio of 47: 1, with construction cost amortized within the first year. Reduced traffic delay is another important benefit of the snow fence protection—the present snow fence system reduces road closure time by an average of 8.3 days each year, representing an additional annual benefit in the millions of dollars.

The roads that connect remote communities with other communities, and particularly with urban areas, are crucial for the economic and social survival of those remote communities. They do not experience high traffic volumes, but must remain open to traffic continuously at all seasons even if the ride quality is somewhat diminished. Yet the revenue available to fund pavement maintenance and reconstruction is very little given the length of road to be maintained and the small taxation base. This paper therefore gives particular attention to overcoming and/or reducing the impact of the spring-thaw weakening problem, particularly as it results in rutting. Consideration is given to drainage needs and to sensing and predicting areas which are likely to be problematic and to justifying adequate funding for pavement maintenance on the basis of socio-economic benefits.

A significant percentage of roads in seasonal frost areas are not designed to carry normal loads during the spring thawing process. Determining dates to apply and remove seasonal load restrictions (SLR) is of great interest to many federal, state and local agencies. In a recent study for the USDA Forest Service, Berg (2004) developed a method which estimated SLR application dates at a site in Minnesota and another in Ohio. In this report, we apply the same method to estimate SLR application dates on a road in New Hampshire and an airport parking apron in Vermont. The calculated results are compared to dates the SLR was actually applied and computed results were nearly the same as the application dates which were established by state DOT personnel. The method is also extended to estimate the time that the SLR can be removed. The Modified Berggren Equation was used to develop a procedure, but a regression equation developed by Rutherford, Mahoney, Hicks and Rwebingira (1985) was more accurate when compared to measured data.