Journal of Cold Regions Engineering

A combination of thermal neutron radiography and time-of-flight neutron diffraction is applied to analyze ice melting kinetics in sand–heavy water mixtures. Two different temperature regimes are applied: (1) cooling down to 262 K followed by spontaneous melting at room temperature (neutron radiography) and (2) cooling down to 250 K followed by controlled linear heating (neutron diffraction). Coarse and fine quartz sand batches are tested with the diameter of grains within the range 630–1,250 and 100–630 m, respectively. The solid–liquid transition temperature is observed to be significantly higher for the tested mixtures than for the sample containing pure heavy water only. A similar increase in the melting point of the mixtures relative to pure heavy water was observed by both of the applied methods. The enhancement amounted to 3 and 6 K in the case of the coarse and fine-grained samples, respectively. Grain size–dependent change in heat transport through the volume of the tested sand–heavy water ice mixtures is suggested as the main physical process behind the observed effect.

High groundwater flow velocity is a major challenge in many artificial ground freezing (AGF) projects, especially in groundwater control projects. That condition may be unfavorable for the formation of a frozen body. This study aims to understand the effect of seepage flow on the heat transfer behavior and hydraulic conductivity variation of sand during the freezing process in AGF projects by using a vertically buried freezing pipe. The authors conducted a small-scale laboratory experiment that combined a flow test, freezing process, and infrared thermograph observations of thermal distribution. The authors prepared saturated Toyoura sand and a Toyoura–silica sand layer combination as the heat-transfer medium. The thermal distribution and hydraulic conductivity behavior were investigated by two series of freezing experiments with different sand layers and flow scenarios. Infrared thermograph observation results showed how water affects the thermal distribution characteristic, whereas flow test results showed the relation between hydraulic conductivity and frozen-body development.

The Department of Defense is planning more than $552M in military construction on Eielson Air Force Base (AFB) within the next three fiscal years. Although many studies have been conducted on permafrost and climate change, the future of our climate and any impacts on permafrost soils remains uncertain. This research focused on future climate and permafrost predictions for two sites on Eielson AFB. The most recent 2013 International Panel on Climate Change report predicts a 2.2°C to 7.8°C air temperature rise in Arctic regions by the end of the 21st century in the Representative Concentration Pathways, (RCP 4.5) emissions scenario. This study provides an explanation of the impacts of this temperature rise on permafrost soils. The University of Alaska’s Geosphysical Institute Permafrost Laboratory (GIPL) 2.1 Permafrost Model was used to model permafrost conditions from 1947 to 2015 using United States Air Force (USAF) recorded surface weather observations. Two emissions scenarios (RCP 4.5 and RCP 8.5) were then used to project permafrost conditions throughout the 21st century. Results show that permafrost conditions are continually deteriorating with anthropogenic activity and rising temperatures in the Fairbanks, Alaska region.

Hindcasting models can use historic ice jam observations and hydroclimatic data to identify conditions that form ice jams. However, historic ice jam records are often sparse or incomplete. New sources of historic ice jam data could improve hindcasting models, leading to better ice jam forecasting and flood warning systems. Because ice jams damage riparian trees, marker rings associated with historic scars include information about ice jam frequency and severity. This study examined marker rings from 56 trees along the Muskegon River to supplement the historic ice jam data on this system. The study team compared tree ring data to results from hindcasting models, which were independently validated with newspaper reports on 1,500 separate days. Logistic regression converted the marker ring data into annual ice jam probabilities. Ice jam dates from the dendrochronology data were too noisy to train or validate a hindcasting model. However, the marker ring data did confirm that ice jams on the Muskegon are nonstationary. Ice jams are significantly more likely now than they were before 1966. The marker rings also helped to distinguish between false-negatives and nondetects in the hindcasting model.

Mechanical properties of frozen soils (e.g., triaxial compressive strength, ${\sigma }_{tc}$ and Young’s modulus, $E$) are important in tunnel, shaft, or open pit excavation projects. Although numerous attempts have been made to develop indirect methods to estimate unfrozen soils’ ${\sigma }_{tc}$ and $E$ values, this has not been done with frozen soils given the difficulty of preparing and conducting relevant laboratory tests. In this study, the accuracy of artificial neural network (ANN), adaptive neuro fuzzy inference system (ANFIS), and support vector machine (SVM) models, developed to predict ${\sigma }_{tc}$ and $E$ for frozen sandy soils, was compared. To the best of the authors’ knowledge, no study has predicted frozen soils’ ${\sigma }_{t\mathrm{c}}$ and $E$ using these methods. Eighty-two poorly graded sandy soil samples from an urban subway borehole in Tabriz, Iran, were used to develop these models. It was found that temperature, confining pressure, strain rate, and yielding strain improved the accuracy of ${\sigma }_{tc}$ and $E$ prediction. Results indicate that SVM can successfully be used in predicting the ${\sigma }_{tc}$ and $E$ of frozen soils.

The annual cost of winter maintenance in the United States is approximately $2.3 billion, which makes measuring the performance of state departments of transportation (DOTs) a financially responsible and mission-critical task to improve services, strengthen accountability, and provide better information for effective decision making and resource allocation. Although state DOTs are increasingly interested in current trends for winter performance measurements, the data documenting performance are limited. This study synthesizes the current practices of winter maintenance operations and identifies the most commonly used performance measures and the impact of weather information on winter maintenance operations by conducting a survey of 31 state DOTs. The results indicate that (1) state DOTs are more dependent on weather information for planning rather than tactical purposes; (2) snowfall, road temperature, and freezing rain are the most important weather variables, and (3) labor/equipment hours and material quantities are the most frequently used performance measures. This study can help state DOTs discern a comprehensive picture of current trends in winter performance measurements and benchmarks for best practices.

It is well known from past studies that inclement winter weather conditions such as heavy snowfall and harsh temperature affect the variation of traffic volume. In order to quantify the change in traffic volume caused by winter weather conditions, a winter weather model explaining the relationship between weather conditions and traffic volume has been generally developed. However, none of the studies in the past have evaluated the performance of the developed model in terms of temporal transferability. In addition, there is no research showing that model specification for predicting correct traffic under winter weather conditions can be optimized for different types of vehicles. This paper presents an application of dummy-variable regression to develop winter weather traffic models. In an effort to fine-tune winter weather model specification by the vehicle class, two model validation tests were conducted. First, a temporal transferability test is conducted to determine whether the model developed is suitable to be used regardless of time. Second, a model specification test is conducted to determine whether the winter weather model specification initially structured is the most appropriate format or not. To pursue this, 6 years of traffic and weather data sets from weigh-in-motion (WIM) and weather stations were collected at Highway 2A in Alberta, Canada. The study results indicate that a dummy-variable winter weather traffic model that includes a temperature variable shows good performance in terms of accuracy and temporal transferability for passenger cars. In contrast, the naïve winter weather model specification that excludes a temperature variable is preferable in predicting truck traffic volume. It is very interesting to note that even if vehicles are traveling in the same traffic stream under the same environmental conditions, different model specifications are preferable for different vehicle types to predict more accurate traffic in winter season.

Snow loads in the western United States are largely undefined due to complex geography and climates, leaving the individual states to publish detailed studies for their region, usually through the local Structural Engineers Association (SEAs). These associations are typically made up of engineers not formally trained to develop or evaluate spatial statistical methods for their regions and there is little guidance from ASCE 7. Furthermore, little has been written to compare the independently developed design ground snow load prediction methods used by various western states. This paper addresses this topic by comparing the accuracy of a variety of spatial methods for predicting 50-year (i.e., design) ground snow loads in Utah and Idaho. These methods include, among others, the current Utah snow load equations, Idaho’s normalized ground snow loads based on inverse distance weighting, two forms of kriging, and the authors’ adaptation of the Parameter-elevation Relationships on Independent Slopes Model (PRISM). The accuracy of each method is evaluated by measuring the mean absolute error using 10-fold cross validation on data sets obtained from Idaho’s 2015 snow load report, Utah’s 1992 snow load report, and a new Utah ground snow load data set. These results show that regression-based kriging and PRISM methods have the lowest cross-validated errors across all three data sets. These results also show that normalized ground snow loads, which are a common way of accounting for elevation in traditional interpolation methods, do not fully account for the effect of elevation on ground snow loads within the considered data sets. The methodologies and cautions outlined in this paper provide a framework for an objective comparison of snow load estimation methods for a given region as state SEAs look to improve their future design ground snow predictions. Such comparisons will aid states looking to amend or improve their current ground snow load requirements.

Roadways over culverts experience severe distresses due to frost action in the subgrade. Thermal insulation placed near the culvert can reduce frost penetration in the soil surrounding the culvert. Three culvert sites backfilled with clay in Manitoba, Canada were instrumented with thermistors to monitor the changes in the soil thermal profile near culverts with and without thermal insulation placed below the culvert barrels. Temperature data from the instrumented sites were used to calibrate and validate a two-dimensional (2D) numerical model which was used in evaluating the performance of different placements of thermal insulation near culverts through a parametric study. The parametric study included evaluation of the use of thermal insulation above the culvert, below the culvert, and attached to the walls of culvert barrels to reduce the thermal disturbance in the soil. It was concluded that the least thermal disturbance in the soil was achieved when 75-mm thermal insulation was attached to the walls of the culvert. The outcomes of this study can be used to mitigate road roughness over culverts that is caused by frost action.