In China, seasonally-frozen soil affects 54% of the land area [1
]. The subgrade established in a seasonally frozen region must undergo freeze-thaw (F-T) cycles. In seasonally frozen regions, frost heaving and thawing subsidence, which lead to negative effects on the performance of subgrade across the lifecycle, are common mechanisms of subgrade damage in the F-T cycles [2
]; further, the subgrade damage causes pavement damage diseases such as longitudinal and transverse cracks, dislocation of cement pavement slabs, breaking in the corner of pavement slabs, and uneven settlement of the pavement slab, and so on [4
]. Engineering research on seasonally-frozen soil subgrade concentrates on reducing frost heave and thawing subsidence, e.g., by using non-frost-susceptible base materials, limiting soil moisture content and fine grain content, lowering the groundwater table, and setting the thermal insulation layer [6
]. Extruded polystyrene (XPS) board is a frequently-used material when engineers install a thermal insulation layer [8
]. In Finland, since the 1970s, the XPS board and expanded polystyrene (EPS) board have been used as the frost insulation for tracks in railway structures; however, due to the poor moisture resistance of the EPS board, it was discontinued in 1980 [8
]. Cai et al. [9
] and Zhao [10
] researched the engineering performance of XPS boards and EPS boards on compressive strength, heat insulation, moisture resistance, and their effects on subgrade performance. The results showed that the XPS board was feasible in technology and reasonable in the economy.
Oil shale ash (OSA) is a byproduct of shale oil production [11
]. Fly ash (FA), produced during coal combustion, is composed of the particulates and flue gases from coal-fired boilers [12
]. The Jinlin province, one important energy consumption and production region, with 27.1743 million residents and Gross Domestic Product (GDP) of 1.5289 trillion yuan in 2015, expended 94.95 million tons of coal and 94,000 tons of shale oil in 2015, and is projected to expend 92.75 million tons of coal and 200,000 tons of shale oil in 2020 [13
]. While the exact data for OSA and FA is scarce, based on shale oil and coal consumption, it is projected to be huge. For instance, the Wangqing Oil Shale Industrial Park, Jilin province, generates 105 tons of OSA per year, and the Jilin province possesses another three Oil Shale Industrial Parks of the same scale as that of Wangqing Oil Shale Industrial Park (Figure 1
The leachates of OSA and FA landfills are considered to be highly contaminated because they are rich in potentially toxic trace elements, which is congealed from the flue gas, and thus have attracted much focus and study on the environmental problem [12
]. For reducing the environmental problems by OSA and FA and reusing them, soil modified by them is an effective measure [18
]. Horpibulsuk and Phetchuay [20
] modified silty clay with fly ash and calcium carbide residue, and they classified soil stabilization by calcium carbide residues into three zones: deterioration, inert, active, and research showed that, by adding FA, the strength of the inert zone was increased significantly. Sridharan [21
] optimized expansive clay soil using different Classes of FA (ASTM Class C and F) and proportions by weight. The results showed that the swelling potential, compaction properties, and consistency limits of expansive clay soil, improved by FA, were significantly upgraded. It was also observed that 40% of FA content was the best value to optimize the plastic characteristics of expansive clay soil. A few attempts have also been made to stabilize soil with OSA. Mymrin and Ponte [22
] undertook an experimental study on the physicochemical interactions for oil-shale fly ash (OSA) with different natural clayey soils. The results indicated that, by changing the percentage of OSA, the strength of the soil could be increased significantly, and modified soil showed very high frost resistance and water resistance as their coefficients equaled or exceeded the 1.0 level. Turner [18
] proved that the silty sand, which was processed with OSA, obtained significant improvement in strength, resilient modulus, and F-T durability. In previous literature, research on soil stabilization with both OSA and FA is inadequate, and it focuses on the author’s previous research [11
]. Wei et al. [11
] modified silty clay (SC) by OSA and FA (In the following interpretation, the silty clay, modified by OSA and FA, will be referred to as MC for short), and conducted a battery of laboratory tests to research the mechanical characteristics and environmental impacts of MC; research results showed that the modified SC was suitable for road construction in seasonally frozen areas. Li et al. [15
] proved the feasibility of using MC and the XPS board as a road subgrade thermal insulation layer, by the numerical modeling and environmental evaluation on an experiment road located in the Jilin province; however, in their numerical modeling, they viewed the specific heat capacity and thermal conductivity of the MC, XPS board, and SC under F-T cycles as a constant value, so their results could not show the real changes of their thermal insulation performance under freeze-thaw cycles, and thus, their research results exaggerated or belittled the thermal insulation effect of the MC and XPS board. Therefore, research on the effect of F-T cycles on the specific heat capacity and thermal conductivity of the MC, XPS board, and SC is of great significance to accurately evaluate the thermal insulation performance of the MC and XPS board.
Based on the above information, this paper aims to (1) present experimental research on the effect of F-T cycles on the specific heat capacity and thermal conductivity of the MC, XPS board, and SC, (2) show that for MC, XPS board and SC, the improved calculation functions of specific heat capacity and thermal conductivity, considering the effects of F-T cycles, are established, (3) show that, by the numerical method of coupling moisture-temperature calculation, which considers the effect of F-T cycles on the specific heat capacity and thermal conductivity of the MC, XPS board, and SC, the thermal insulation performance, which utilizes the MC and XPS board as the subgrade thermal insulation layer, was identified.
Based on the above context, the research is novel in several ways based on the following: (1) at different levels of dry density and moisture content, we identified the variations on thermal insulation capability of MC in consideration of the effects of freeze-thaw (F-T) cycles reached 20 by laboratory test for the first time; (2) for the MC, XPS board, and SC, the improved calculation functions of specific heat capacity and thermal conductivity, considering the effects of F-T cycles, are established; (3) by the numerical simulation of coupling moisture-temperature considering the effects of F-T cycles, the thermal insulation capability of MC board and XPS board were studied quantitatively. The specific conclusions are as follows:
According to the test results of thermal conductivity of testing samples with different levels of dry density and moisture content, the MC, which is utilized as the subgrade thermal insulation layer of the experimental road, possesses a certain thermal insulation property compared with SC, and as the number of F-T cycles increases, the thermal conductivity of MC and SC gradually decreases, and the rate of decrease also decreases gradually, but the thermal conductivity of the XPS board is a strong positive correlation with the increase of F-T cycles, and it reaches the maximum value of 0.61 W/m/K when the number of F-T cycles reaches 17.
According to the test results of specific heat capacity of testing samples with different levels of dry density and moisture content, the specific heat capacity of MC, SC, and XPS board doesn’t change regularly as their moisture content and number of F-T cycles change, and their variations are in the range of test error (2%), indicating that the specific heat capacity of solid particle of MC, SC, and XPS board is almost unchanged even though the different levels of moisture content and number of F-T cycle are considered, and thus it is reasonable to cite their respective average value for the numerical simulation.
Compared with previous research [15
] by the authors of this paper about the thermal insulation capability of MC and XPS board used in Structures I, II, and III, based on the testing data and cited literature, the numerical simulation of the coupling moisture temperature calculation is improved by the fitting equation of thermal conductivity of SC, MC, and XPS board, and by fitting equation of specific heat capacity of water and ice, and the average value of specific heat capacity for respective solid particle of MC, SC and XPS board. According to the simulation results and compared with previous research [15
], the differences are mainly as follows: (1) The seasonal frost depths of Structures III and II are 0.55 m and 0.21 lower than Structure I, respectively, and the time to reach the seasonal frost depths is 50 days later than Structures I and II. (2) Structure III can protect both of SC and sand gravel of the experimental road from the frost heave damage. The reasons for the above difference between this paper and previous research [15
] are that with the thermal conductivity of MC, SC, and the XPS board and the specific heat capacity of water and ice, occurs changes with the increased number of F-T cycles, and the numerical model in this paper considers those changes.
The research method and results are of great significance for accurately evaluating the thermal insulation capacity and the sustainability of MC and the composite layer consisting of MC and XPS board, strengthening the stability of subgrade and increasing the availability of industrial waste, and meantime show good sustainability using the XPS board and MC as the subgrade thermal insulation layer in seasonally frozen regions, because the thermal conductivity of MC and SC is negatively correlated with F-T cycles, and despite that of XPS is positively correlated with F-T cycles, it is close to a constant value of 0.061 W/m/K after 17 F-T cycles.