Pore evolution and shear characteristics of a soil-rock mixture uponfreeze-thaw cycling

LiYun Tng ,ShiYun Sun ,JinGuo Zhng ,Long Jin ,YongTng Yu ,To Luo ,Xu Dun

a School of Architecture and Civil Engineering, Xi'an University of Science and Technology, Xi'an, Shaanxi 710054, China

b China Jikan Research Institute of Engineering Investigations and Design Co., Ltd., Xi'an, Shaanxi 710043, China

c State Key Laboratory of Road Engineering Safety and Health in High-altitude Regions, CCCC First Highway Consultants Co., Ltd, Xi'an, Shaanxi 710065, China

d China United Northwest Institute for Engineering Design & Research Co, Ltd., Xi'an, Shaanxi 710077, China

e Key Laboratory of Safety and Durability of Concrete Structures of Shaanxi, Xijing University, Xi'an, Shaanxi 710123, China

Keywords:Freeze-thaw cycling Soil-rock mixture NMR Pore change Shear strength

ABSTRACT The changes in pore structure within soil-rock mixtures under freeze-thaw cycles in cold regions result in strength deterioration,leading to instability and slope failure.However,the existing studies mainly provided qualitative analysis of the changes in pore or strength of soil-rock mixture under freeze-thaw cycles.In contrast,few studies focused on the quantitative evaluation of pore change and the relationship between the freeze-thaw strength deterioration and pore change of soil-rock mixture.This study aims to explore the correlation between the micropore evolution characteristics and macro-mechanics of a soil-rock mixture after frequent freeze-thaw cycles during the construction and subsequent operation in a permafrost region.The pore characteristics of remolded soil samples with different rock contents(i.e.,25%,35%,45%,and 55%)subjected to various freeze-thaw cycles(i.e.,0,1,3,6,and 10)were quantitatively analyzed using nuclear magnetic resonance(NMR).Shear tests of soilrock samples under different normal pressures were carried out simultaneously to explore the correlation between the soil strength changes and pore characteristics.The results indicate that with an increase in the number of freeze-thaw cycles,the cohesion of the soil-rock mixture generally decreases first,then increases,and finally decreases;however,the internal friction angle shows no apparent change.With the increase in rock content,the peak shear strength of the soil-rock mixture rises first and then decreases and peaks when the rock content is at 45%.When the rock content remains constant,as the number of freeze-thaw cycles rises,the shear strength of the sample reaches its peak after three freeze-thaw cycles.Studies have shown that with an increase in freeze-thaw cycles,the medium and large pores develop rapidly,especially for pores with a size of 0.2-20 μm.Freeze-thaw cycling affects the internal pores of the soil-rock mixture by altering its skeleton and,therefore,impacts its macro-mechanical characteristics.

1.Introduction

As a dual-element geological material,soil-rock mixtures have the characteristics of extreme nonlinearity and strong environmental dependence (Xu and Hu,2009;Chu,2014;Tanget al.,2022).It is different from soil and rocks widely distributed in cold areas (Peiet al.,2013;Cuiet al.,2014;Liuet al.,2022),and its mechanical properties and pore structure characteristics are highly susceptible to freeze-thaw actions.During freeze-thaw cycles,as the soil is affected by the ice and water phase change,it will first cause the creation and development of the pores,and this freeze-swelling effect is rapidly reduced during thawing.With water acting as a lubricant,this process,in turn,causes structural deformation of the soil.In particular,after undergoing freeze-thaw cycles,the internal micro-pore characteristics of the soil body undergo an overall adjustment,resulting in particle rearrangement and changes in the soil skeleton and ultimately affecting its mechanical properties (Konrad and Lemieux,2005;Xiaoet al.,2014;Wanget al.,2017).Consequently,it can compromise the stability of slopes in permafrost regions and cause landslides and mudslides (Dan and Liu,2013;Sunet al.,2016;Zhanget al.,2021).Therefore,it is necessary to study the shear and pore characteristics of soil-rock mixtures under freeze-thaw cycling to guide construction and disaster prevention and mitigation in cold regions.

The variation in mechanical properties of a soil-rock mixture depends on its complex internal structural parameters and susceptibility to environmental changes.Therefore,many studies focused on the mechanical properties of soil-rock mixtures via macro-tests,micro-tests,and numerical simulations.For example,the change in the strength of coarsegrained soil under freeze-thaw cycles was investigated using GDS triaxial tests,and it was found that the uniaxial strength of the coarsegrained soil decreased with an increase in the number of cycles and tended toward stability after nine cycles (¨Ozganet al.,2015;Quet al.,2018;Yanet al.,2019;Qiuet al.,2023).The maximum drop in the shear strength after thawing can reach almost 40%.In particular,Liu and Zhang studied the effect of uneven distribution characteristics of fine-grained soil on the mechanical properties of coarse-grained soil subjected to freeze-thaw cycles via triaxial shear tests,and they found that its strength decreased as the coarse grain content increased (Liuet al.,2017;Zhanget al.,2017).Liet al.(2021)investigated the influence of gravel content and dry density on the shear strength and stress-strain relationship of poorly graded gravelly soil and found that the internal friction angle rose and the cohesion increased first and then decreased with the increase in gravel content and sample dry density.To better understand the damage in shallow coarse soil,Liet al.(2022)studied the mechanical property changes of loose,broad,and coarse alluvial soil under low confining pressure(5-25 kPa).Li analyzed the influence of the sample particle graded on countershear strength index through large indoor shear tests and revealed the effects of specimen size and other factors on the mechanical properties of crushed stone(Liet al.,2004;Li,2012).Xinget al.(2018)analyzed the failure behavior of soil-rock mixtures with different rock contents after a freeze-thaw cycle,and the results showed that freeze-thaw reduced the strength of the soil-rock particles.Yanget al.(2015)explored the change in the elastic modulus of a soil-rock mixture after a freeze-thaw cycle via indoor triaxial tests.They showed that the formation of pores is the main reason for the decrease in the elastic modulus.Through laboratory experiments,Chenet al.(2014)found that the pore changes caused by freeze-thaw cycles reduced the rebound modulus of coarse-grained soil.Juet al.(2018) established a structural model of a soil-rock mixture by high-resolution computed tomography (CT).They explored the process of sprouting,evolution and aggregation of internal fractures under loading to clarify the mechanism of internal damage evolution of soil-rock mixtures with different rock content.Wanget al.(2022) discovered the internal microstructural changes in soil-rock mixture samples with varying stone contents under loading by triaxial and CT tests.In addition,Tanget al.(2021) used nuclear magnetic resonance (NMR) to analyze the influence of freeze-thaw cycles on the internal pore structure of a soil-rock mixture and established a relationship between the pore structure and shear characteristics under freeze-thaw cycles.Through direct shear and NMR tests,Liet al.(2022)analyzed the relationship between pore change and shear strength of a soil-rock mixture and suggested that the pore structure of a soil-rock mixture is more sensitive to freeze-thaw cycles when the water content is large.

Numerical simulations were also applied to study the internal structural changes and failure mechanisms of soil-rock mixtures.Guet al.(2014) conducted numerical simulations using the discrete element method(DEM)to investigate the microstructure of granular materials at different strain levels.They found that the greater the porosity of the damage zone,the more pronounced the sliding and rotation of the particles and the more significant the effect on their strength.Guet al.(2022)and Jiaet al.(2022)found that the content of fine particles plays a vital role in the shear strength of a soil-rock mixture and suggested that its failure was caused by the changes of contact force between particles.They investigated the failure mechanism and mechanical properties of the soil-rock mixture by the PFC numerical simulation.Zhouet al.(2018),via single-layer and double-layer embedded model simulations,found that the shear modulus of the soil-rock mixtures gradually increased with an increase in rock content.

The above analysis found that the existing studies primarily focused on the influence of temperature changes,stone contents,stone particle sizes,and water content on the mechanical properties and failure mechanisms of soil-rock mixtures,and most of these studies were qualitative.Only a few studies attempted quantitative assessment of the strength deterioration and investigated the interaction between shear strength and mesoporous characteristics of soil-rock mixtures under freeze-thaw cycle conditions.

Therefore,this study focuses on the relationship between shear characteristics and pore changes in soil-rock mixtures upon freeze-thaw cycling.Samples with a water content of 18% and rock contents of 25%,35%,45%,and 55%were selected as the influencing factors of the soil-rock mixture sampled from a permafrost region.A strain-controlled direct-shear testing machine and nuclear magnetic resonance (NMR)device were used to test the shear and pore characteristics of the samples after freeze-thaw cycles.The results from this study are important for understanding failure mechanisms and mitigation of hazards related to soil-rock mixtures in cold regions.

2.Sample preparation and test methods

2.1.Sample preparation

The test soil is typical silty clay sampled from the Qinghai-Tibet Plateau.Original soil samples were sieved to retain the portion under 2 mm.The silty clay has a liquid limit of 31.9%,a plastic limit of 14.6%,and a plasticity index of 17.3,and can be classified as frost-susceptible.The particle size grading curve is shown in Fig.1.

Fig.1.Gradation curves of crushed stone and soil.

Considering the significant size effect of geological materials in the soil-rock mixtures (Zhanget al.,2015) and the loading capacity of the testing device,the sample diameter was selected as 150 mm and the height 100 mm.According to the “Test Methods of Soils for Highway Engineering”(JTG 3430-2020),the maximum particle size of crushed rock should generally not exceed 1/4 of the sample height and 1/8 of the sample diameter.Therefore,the maximum size of the gravel used in this test was set to 18 mm,and the oversized gravel was removed using the equivalent replacement method(Xuet al.,2011).

The preparation process of remolded soil samples used in the test was as follows.(1)The test sample was prepared by the compaction method according to the specified requirements.First,sieved,air-dried finegrained soil was added to crushed rock according to designed proportions and then thoroughly mixed to ensure a uniform distribution of coarse and fine particles.(2) Distilled water was added according to the natural moisture content,and the sample was mixed well and sealed for 12 h to ensure that the moisture content of the soil sample was uniform.(3)The sample was divided into three layers for compaction.The soil at the boundary of each layer had to be fully shaved to avoid delamination during sample preparation.A plastic film was placed at the bottom to prevent soil particles from scattering and moisture from evaporating.(4)It was then compacted into a sample mold with a target dry density of 2.0 g/cm3.Fig.2 depicts the prepared soil-rock mixture samples with varying rock content(i.e.,25%,35%,45%,55%).The prepared samples were wrapped with cling film and labeled in a self-sealing bag to prevent water loss(Henry,2007;Qiet al.,2006)and were placed in an environmental chamber for freeze-thaw cycle testing.To ensure data quality,the water content test was conducted after 1,3,6,and 10 cycles according to the ''Test Methods of Soils for Highway Engineering'' (JTG 3430-2020).Considering the actual measurement results of the soil temperature in the field,the freezing temperature was set to -25°C,and the thawing temperature was 25°C during the freeze-thaw test.Note that this was a freeze-thaw cycle test under closed conditions without an external water supply.According to the''Test Methods of Soils for Highway Engineering''(JTG 3430-2020),the sample was frozen at -25°C for 12 h and then thawed at 25°C for 12 h to complete a full freeze-thaw cycle.Samples were subjected to 0,1,3,6,or 10 freeze-thaw cycles.

Fig.2.Prepared soil-rock mixture samples.

2.2.Testing methods

This study included a total of five groups of soil-rock mixtures subjected to freeze-thaw cycles.Direct shear and NMR tests were performed according to the ''Standard for Geotechnical Test Methods'' on samples with 18%moisture content and rock contents of 25%,35%,45%,or 55%.In the same batch with different rock contents and the same number of freeze-thaw cycles,there were nine samples for direct shear tests and three for NMR tests.The normal pressure on the potential slip zone is not large,as most slope failures involving soil-rock mixtures in cold regions due to freeze and thaw occur at shallow depths.Thus,the normal pressures were set to 100,200,or 300 kPa.Then,varying freeze-thaw cycles(0,1,3,6,and 10) were performed for each group of samples.Table 1 summarizes the test conditions.

The same batch of NMR samples was saturated per the''Specification of Soil Test''and''Test Methods of Soils for Highway Engineering''for 48 h(Stingaciuet al.,2010;Fenget al.,2018;Tanget al.,2023).Macro-MR12-150H-I NMR tester jointly developed by the Chinese Academy of Sciences and Suzhou Newmai was employed.The test involves the built-in phase stratification test method to perform layered T2relaxation(transverse relaxation time) using the NMR system for non-destructive,real--time monitoring of the saturated samples.The T2spectrum of the sample was used to determine the pore distribution of the soil-rock mixture samples.

To more accurately determine the mechanical characteristics of the soil-rock mixtures,this study utilized a strain-controlled direct-shear test system produced by the Nanjing Soil Research Instrument Factory with its shear box was customized and improved.The deformation during the test was obtained by a displacement sensor and recorded by an automatic data acquisition system.Fig.3 illustrates sample preparation and experimental procedures.The main control factors were the samples'rock content and the number of freeze-thaw cycles,and the test program involved a total of four groups,i.e.,240 samples.The effects of rock content and freeze-thaw cycles on the pore characteristics and strength of the soil-rock mixtures were examined by comparing the experimental results between groups.

Fig.3.Illustration of sample preparation and experimental procedures.

Fig.4.Effect of freeze-thaw cycles on pore size distributions.(a)Pore distribution for specimens with 25%rock,(b)Pore distribution for specimens with 35%rock;(c)Pore distribution for specimens with 45% rock;(d) Pore distribution for specimens with 55% rock.

Fig.5.Effect of variation in rock content on pore size distributions.(a)Specimens without freeze-thaw;(b)Specimens with one freeze-thaw cycle;(c)Specimens with three freeze-thaw cycles;(d) Specimens with six freeze-thaw cycles;(e) Specimens with ten freeze-thaw cycles.

2.3.NMR testing principle and method

NMR technology has gradually been applied to the field of geotechnical engineering.It provides substantial advantages for moisture detection of geotechnical materials under temperature changes,such as pore distribution and unfrozen water content change(Cowan,1997;Chuet al.,2021).Depending on the relationship between saturated water and porosity,the porosity of a water-saturated specimen can be measured by NMR technology.A spin magnetic field can be generated because of the spin property of positively charged hydrogen proton 1H in the sample pore water.Therefore,the hydrogen proton in the sample can be deflected for 1H under the action of a radio frequency(RF)magnetic field and alternating magnetic field provided by the NMR instrument.When the RF stops,the NMR signal begins to decay freely during the return of 1H to the equilibrium state(Ticeet al.,1982;Yanget al.,2014;Tanget al.,2023).The transverse relaxation time T2 is defined as the time when the transverse magnetization vector decays to 37%of the maximum and is the primary goal of NMR detection(Shenet al.,2020;Liet al.,2023).The spin-echo attenuation data is inverted togenerate a T2distribution,andthe areabelowthe T2curve is the porosity by appropriate scale.The T2 value of a single pore is proportional to the ratio of the surface area to the total pore volume,and the observed T2 spectra represent the pore distributions of the specimens(Liet al.,2014;Luet al.,2017;Duet al.,2019).The relationship between T2 value and specific surface area of pores is as follows：

where ρ2is the relaxation rate;is the specific surface area of pores.Assuming the pore is spherical,the above formula can be simplified as:

whereRis the pore radius.

3.Micro-pore evolution in soil-rock mixtures upon freeze-thaw cycling

3.1.Characteristic curve of pore distribution

3.1.1.Effect of freeze-thaw cycling on pore size distribution

This study considers that repeated freeze-thaw cycles will affect the internal microstructure of the soil-rock mixture sample,so the freeze-thaw cycle test under different working conditions is carried out to further clarify the response of the sample microstructure.First,based on the results of pore size testing using NMR and the definition of soil pores by previous scholars(Liet al.,2018;Zhanget al.,2019),the specimen pores were classified into three categories,namely small pore size pores(0.01 μm ＜r＜0.2 μm),medium pore size pores(0.2 μm ＜r＜2 μm),and large pore size pores(2 μm ＜r＜100 μm).Through the NMR test,the T2spectra of saturated soil samples under different freezing and thawing times were measured and are shown in Figs.4a-4d.As the number of freeze-thaw cycles increased,the changes in the second and third peaks were more prominent in the T2spectra,indicating more pronounced volume changes for medium and large pores.The changes in the first peak reveal that small pores continued to develop in the soil-rock mixture during freeze-thaw cycles.This was mainly by expansion and contraction owing to the freezing of the pore water when the soil-rock mixture underwent freeze and thaw.During thawing,the pores could not completely restore their volumes before freezing.As a result,the continuous development of some pores altered the arrangement of soil particles and its force transmission skeleton.

Moreover,with an increase in freeze-thaw cycles,the T2spectrum showed that the proportion of medium and large pore volumes decreased after three freeze-thaw cycles.Possible reasons for this phenomenon were proposed by Viklander(1998).The concept of freeze-thaw residual porosity suggests that freeze-thaw cycling will reduce the porosity of the soil to a certain extent,thus increasing the density and bringing the soil to a new equilibrium and steady state.Another possible reason is that with the development of pores during freeze-thaw,some fine soil particles may fall into large-sized pores,i.e.,some small particles fill the large-sized pores,occupying and reducing the space of large-sized pores.With the increase in freeze-thaw cycles,the shape change of the small pore T2spectrum generally shows a slight decrease in the peak value and peak width,whereas the large pores are more sensitive to the effect of the freeze-thaw cycle.Generally,the peak widths and heights in pore size ranges of 0.2-2 μm and 2-20 μm corresponding to the medium and large pores vary more significantly than other peaks,with the peaks shifting to the right overall.It shows that the change in the internal pore structure of the soil-rock mixture caused by freeze-thaw cycling is a dynamic process.

Freeze-thaw cycling has a more severe impact on the pores of the soilrock mixture with high rock contents.It can be seen from Figs.4a-4d that freeze-thaw cycling has little effect on the volume change of small,medium and large pores when the rock content is low.However,the pore volume changes gradually rise with the increase in rock content.

3.1.2.Effect of variation in rock content on pore size distribution

Considering the change in rock content significantly affects the microscopic pores of the soil-rock mixture after a freeze-thaw cycle,NMR tests of soil-rock mixture specimens with varying rock content subjected to different freeze-thaw cycles were carried out.The T2spectra of specimens with varying rock content are shown in Figs.5a-5e.From the specimens without freeze-thaw action,it can be found that with the increase in rock content,the peak corresponding to small pores in the T2spectrum varies slightly and shows a gradually decreasing trend,indicating that the change in rock content will affect its porosity.This is because,as rock content increases,coarse particles are dominant and control the pores.After ten freeze-thaw cycles,the T2spectra mainly showed three peaks.With the same number of freeze-thaw cycles,as rock content increases,the overall T2spectrum morphology showed a shift;that is,the small-pore T2peak shifted to the right toward larger pore sizes.Moreover,the proportion of large pores gradually dominates.This also shows that under the action of the water-ice phase change,new micropores were generated due to the continuous change in the internal structure of the soil-rock sample,as reflected by the newly generated micropores that gradually evolved into large-sized pores.

It can be seen that the change in the pore structure of the soil-rock mixture is sensitive to rock content.The proportion of small pores is higher than that of the medium and large pores.Freeze and thaw cycling causes the amount of small pores to continue to decrease and the number of medium and large pores to rise.The initial effect of freeze-thaw cycling on the soil-rock mixture structure is more significant.With the increase in rock content,the mixture's internal structure and various physical parameters tend to become stable,and the impact on its skeleton also tends to weaken gradually.

3.2.Variation of the pore proportion

The changes of pores in the soil-rock mixture during freeze-thaw cycling were studied based on laboratory experimental results.Fig.6 shows the pore volume distribution of the soil-rock samples with various rock content upon different freeze-thaw cycles.The volume proportion of different size groups changed with an increase in freeze-thaw cycles and rock content.As shown in Fig.6,the percentage of small pore volume was large,indicating that its number was higher than those of medium and large pores.For zero to ten freeze-thaw cycles,the percentage of small pore volume showed a downward trend overall,and the percentage of medium and large pore volume rose.At the same rock content,with an increase in freeze-thaw cycles,the small pore volume underwent a process of''decreasing,then increasing,and then decreasing,''accounting for approximately 63.8%-79.6%of the total pore volume.With the increase in rock content and under the same number of freeze-thaw cycles,the overall decrease in the proportion of small pore volume dropped.

Fig.6.Distribution of small,medium,and large pore volume ratios in the soil-rock mixture as a function of freeze-thaw cycles.(a) Samll pores (0.01-0.2 μm);(b)Medium pores (0.2-2 μm);(c) Large pores (2-20 μm).

The medium pore size pores accounted for 16.2%-25.1%of the total pores.After the first freeze-thaw cycle,the mesoporous pores slightly increased.Mesoporous pores gradually increased from 5.9% to 24.1%,indicating that rock content had a more significant effect on the pore structural change.During freeze-thaw cycling,large pores exhibited a different trend from that of small pores,i.e.,they first increased,decreased,and then increased,accounting for approximately 5.2%-26.3%of the total pore volume.With an increase in rock content,the change became considerably obvious and the maximum growth was 31%under the same number of freeze-thaw cycles.A comprehensive analysis showed that after ten freeze-thaw cycles,the number of small,medium,and large pores increased to varying degrees,among which the small and medium pores continued to expand and develop into large pores.As a result,the proportion of large pores continuedto rise,promoting continuous degradation inside the soil-rock mixture.There was a good correlation between the pores'physical structure and the mixture's strength.

3.3.Porosity changes

Porosity is an important indicator of soil's internal structure.This study uses the porosity change rate to describe the effect of freeze-thaw cycling(Liuet al.,2021),which is defined as:

where Δφ is the porosity change rate,φ1is the porosity of the specimen after freeze-thaw cycles,and φ0is the porosity of the specimen after previous freeze-thaw cycles.

Based on the NMR results for the soil-rock mixture samples with different rock content and freeze-thaw cycles,the porosity change rate is shown in Fig.7.With an increase in freeze-thaw cycles,the change rate of the porosity generally rose first,then decreased,and increased slowly for the same rock content.The number of freeze-thaw cycles significantly affects the micro-pore structure of the soil-rock mixture,especially after the initial freeze-thaw cycle,as indicated by the relatively large change rate.After six freeze-thaw cycles,the porosity change rate stabilized with only a minor increase.This was because,during the initial freeze-thaw cycle,the internal water gradually transformed into ice crystals to expand the volume and squeeze the surrounding particles.During thawing,the soil-rock mixture skeleton partially fell back and collapsed,destroying the original structure.The cementation between the coarse and fine particles caused them to shift or even deform.Because the soil body was not affected by external forces during thaw,the pore volume shrinkage during thawing could not fully compensate for the pore volume expansion during freezing.Therefore,the internal pore volume of the soil-rock mixture increased,as reflected in the porosity change rate after the quick initial increase in the freeze-thaw cycles.Additionally,as the number of freeze-thaw cycles increased,the freeze-thaw action may have changed the bonding form and interaction force between the coarse and fine particles,i.e.,the fine particles produced a cohesive effect,or the fine particles adhered to the surface of the coarse particles.This causes a certain degree of ''clumping'' inside the particles,which will cause the skeleton of the soil-rock sample to change,reducing the pore volume and decreasing the porosity change rate.After three freeze-thaw cycles,the porosity change rate gradually increased slowly.There may be two reasons for this phenomenon:1) The effect of the freeze-thaw cycle on the internal structure of the soil weakened,and the pore changes tended to be stable.And 2)repeated freeze-thaw cycles may have caused a certain degree of water content loss from the sample.

Fig.7.Porosity change rate for the soil-rock mixture as a function of freezethaw cycles.

4.Mechanical characteristics of the soil-rock mixture upon freeze-thaw cycling

4.1.Stress-strain curves

After repeated freeze-thaw cycles,the effect of rock content and freeze-thaw cycles on the mechanical properties of soil-rock mixture samples is very significant.According to the direct shear data with normal stress of 100 kPa.Fig.8 depicts the stress-strain curves for the soil-rock samples with different rock content subject to various freezethaw cycles (Tin the figure indicates the number of freeze-thaw cycles,andT=0 means no freeze-thaw action).The stress-strain curves demonstrate that freeze-thaw cycling significantly affects the shear behavior of the soil-rock mixture.At the same shear stress,the sample had a larger shear displacement as the number of freeze-thaw cycles increased.Additionally,the stress-strain curves of the soil-rock samples before and after thawing showed strain hardening.The reason for this phenomenon was primarily the dilatancy during the shear process.During shear loading,the positions of the particles shifted,and the interlocking effect was enhanced.During the initial stage of shear,the deformation was primarily due to the pores produced by the soil-rock mixture in the initial stage of freeze-thaw that were gradually compacted.As the shear deformation steadily developed,the slope gradually decreased and entered a stage of slow development;moreover,the curve gradually reached peak strength,and the yield platform began to appear.During this process,the shear stress generated at any point in the sample under normal stress was less than the shear strength of the soil-rock mixture,and the shear force was borne by the skeleton mainly formed by the coarse particles,which showed the elasticity of the soil-rock mixture sample.The modulus grew,and the mixture entered the elastoplastic deformation stage faster.As the shear displacement continued to increase,the mutual occlusion between the blocks and stones caused continuous rotation and tumbling between the particles,changing the internal structure and resulting in the growth of shear stress soon after entering the peak state.

Fig.8.Stress-strain curves of samples with different rock content subject to various freeze-thaw cycles.(a)Specimens with 25%rock content;(b)Specimens with 35%rock content;(c) Specimens with 45% rock content;(d) Specimens with 55% rock content.

After reaching the peak strength,the shear deformation continued to rise.At this stage,the shear stress at various points in the range of the shifted shear band reached or exceeded the shear strength of the soil-rock mixture.The source of the shear strength mainly depended on the friction between the particles at the failure zone,and the shear deformation was manifested by overturning,shifting,and breaking between particles.At the same time,the stress-strain curves of the specimens showed the same pattern with the increase in freeze-thaw cycles at the same rock content.The peak stress,initial tangential modulus,and peak point secant modulus of all specimens tended to decrease,then increase and decrease with the number of freeze-thaw cycles,and the overall stress-strain curve tended to be flat.The larger the failure strain,the more ductile the soil is,andvice versa.The area enclosed by the stress-strain curve and the horizontal axis eventually decreased after ten freeze-thaw cycles,indicating that the number of freeze-thaw cycles has a greater influence on the performance of the soil-rock mixture and that the ability of the soil-rock mixture to absorb energy becomes weaker.The stress-strain curves of the specimens reached their peak stress after three freeze-thaw cycles.This is due to the change in compressibility caused by the alteration in pore space as the blocks inside the specimens become more stable due to the adjustment of position after different freeze-thaw cycles.When the specimens were subjected to six freeze-thaw cycles,the stress-strain curves showed a clear upward convexity and downward concave shift in the yielding phase as the rock content increased.This also indicates that as the number of freeze-thaw cycles increases,the eventual damage to the soil-rock mixture becomes greater.

4.2.Shear strength

Fig.9 illustrates the relationship between the shear strength and the number of freeze-thaw cycles for the soil-rock mixtures with different rock content.The shear strength first decreased,then increased,decreased,and finally stabilized with an increase in freeze-thaw cycles.This can be attributed to the ice-water phase change of the pore water at the beginning of the freeze-thaw cycle,when the pores in the soil gradually developed,leading to changes in the arrangement of the particles within the soil-rock mixture,continuous deterioration of the structure,and reduction of coarse particles.The cohesion with fine-grained soil is represented by the macroscopic decrease in the shear strength of the soilrock mixture after the freeze-thaw cycling.After a certain number of freeze-thaw cycles,i.e.,three freeze-thaw cycles,the shear strength of the soil sharply increased to a peak value.This phenomenon was because the soil body is also subjected to a physical-chemical change during repeated freeze-thaw cycles.Due to ice crystal growth in the pores,the particles were squeezed,and colloids and clays were aggregated into microaggregates (Li,2012;Zhanget al.,2016),which increased the particle size of the soil and enhanced the agglomeration between the coarse and fine particles.The cementation and occlusal forces between the particles increased,and the cohesion of the soil-rock mixture rose.With the continuous increase in the number of freeze-thaw cycles,the effect on the internal structure of the soil by freeze-thaw cycling gradually weakened,and its structure became relatively stable;additionally,the shear strength showed a gradual decrease and then stabilized.

Fig.9.Relationship between shear strength and the freeze-thaw cycle for the soil-rock mixtures with various rock content.(a)Specimens with 25%rock content;(b)Specimens with 35% rock content;(c) Specimens with 45% rock content;(d) Specimens with 55% rock content.

Moreover,as shown in Fig.9,for the same number of freeze-thaw cycles,the peak value of the shear strength of the sample rose with an increase in rock content,and it then increased and decreased.This was because when rock content was between 25%and 45%,the structure of the soil-rock mixture behaved as a typical framework pore structure(Vallejo,2001;Wanget al.,2016).Under shear stress,the coarse particles started to make contact and bite into each other.As the shear deformation continued to increase,the compactness of the soil-rock mixture increased,causing the shear stress to reach shear failure.Therefore,the work increased.When rock content exceeded 45%,the peak intensity decreased.There may be two reasons for this phenomenon.First,this is primarily owing to the freeze-thaw separation of particles and the self-weight of soil particles caused by the freeze-thaw cycles.The fine particles could not entirely fill the pores between the rocks.In a loose state,the compactness was reduced,and the effect of freeze and thaw was enhanced.Second,when rock content exceeded 50%,coarse particles played a major role in the framework.At this time,the structure mainly takes the block stone as the skeleton,the porosity of the sample increased,and the indirect contacts of coarse particles within the unit volume decreased,causing stress concentration.The particles were more likely to break during the shearing process,thereby reducing the strength of the soil-rock sample.

4.3.Shear strength parameters

The shear strength parameters,i.e.,cohesion and internal friction angle,were obtained according to the Mohr-Coulomb strength criterion and listed in Table 2 and Table 3.The cohesion of specimens with different rock content decreased,then increased,and finally decreased gradually,whereas the internal friction angle showed a minor fluctuation without significant changes.The above phenomenon was primarily due to the continuous change in the internal structure of the soil body due to the effects of freeze and thaw during the first cycle.Part of the pores began to develop,destroying the chemical bonds initially formed by the indirect contact of the particles,and then the soil cohesion gradually weakened.This was a decreasing trend.Moreover,the freeze-thaw cycle was a strong weathering mechanism.During this process,the internal pores of the soil had both the effect of increasing and compacting,which made the arrangement of particles within the soil-rock sample more rough and random.

Table 2 Variation in cohesion of the soil-rock mixture under various freeze-thaw cycles.

Additionally,in the aqueous solution after the freeze-thaw cycle,because the clay particles carry negative electricity,the particles attracted each other by drawing surrounding cations to generate an electrostatic attraction,which increased the sizes of the aggregates and enlarged the contact points between the particles inside the soil.This,in turn,enhanced the cementation between fine particles and the adhesion between coarse and fine particles,thus increasing the cohesion of the soil-rock mixture.In addition,although the change in the internal friction angle has no obvious regularity under different rock content,it is found that the internal friction angle reaches the maximum after three freeze-thaw cycles.The reason for this phenomenon is that in the second to third freeze-thaw cycles,the skeleton collapse of the soil-rock mixture under freeze-thaw and strong weathering leads to decreased pores,thereby enhancing the overall structural stability of the sample.It is this compaction effect in the freeze-thaw process that leads to the increased soil particles wrapped on the surface of gravel particles,the increased contact points between particles,and the enhanced interlocking friction and sliding friction between soil particles.Therefore,the internal friction angle shows an increasing trend here.Meanwhile,as shown in Table 2,under the same freeze-thaw cycle,the cohesion generally rose first and then decreased as rock content increased.This was mainly because with an increase in rock content in the range of 25%toand 45%,because fine particles act as a filler between the crushed rocks,the crushed rocks were contacted by fine particles.The occlusion formed a skeleton,and the cohesion was significantly improved.At this time,both the coarse and fine particles controlled the deformation and strength characteristics of the soil-rock mixture.When rock content exceeds 45%,the soil-rock mixture exhibits a dense skeleton structure.Because the fine particles are greatly reduced,the overall viscosity of the soil-rock mixture sample is reduced,and the contact between coarse and fine particles is gradually reduced.Therefore,the cohesion of the soil-rock mixture as a whole decreased rapidly.

5.Mechanisms of freeze-thaw cycling influence on mechanical properties

The soil-rock mixture is a three-phase geological system mainly composed of coarse and fine particles,water,and air.The proportion of coarse and fine particles is dominant in forming the soil skeleton.Water and air fill the skeleton pores formed by coarse and fine particles that contact and cement each other.The mechanical properties of the soilrock mixture are primarily affected by the different characteristics of the three-phase medium during the intense and repeated freeze-thaw process.Fig.10 further reveals the material composition characteristics of the soil-rock mixture and the sample's microscopic NMR test and pore development process under the freeze-thaw cycle.

Fig.10.Failure mechanism of soil-rock mixtures under freeze-thaw cycling.

It can be seen from the test results of the above-mentioned shear and nuclear magnetic resonance tests that the soil-rock mixture can be regarded as a unique material composed of coarse and fine particles,and the change of the rock content rate has a significant effect on its deformation characteristics.During the freeze-thaw cycle,due to the ice-water phase transition of the in-situ water,the volume expansion of the ice crystals in the sample of the soil-rock mixture squeezes the particles,and the particles move and shift,which destroys the original soil-rock mixture.The cohesion and internal structure have gradually weakened the strength of the soil and the original degree of cementation.The mixture's skeleton is gradually formed with the increase in rock content.Under repeated freeze-thaw cycles,the squeezing effect of ice crystal growth within the specimen also leads to agglomeration between the particles,increasing the interparticle cementation(Changet al.,2014).In addition,during the shearing process,the coarse and fine particles are in contact with each other due to bite and friction,which in turn leads to a further increase in the cohesion of the soil and stone mixture and an increase in strength.The sample achieves its peak shear strength at a rock content of 45%,which is called the optimal rock content.In practice,the influence of rock content should be considered to determine the optimal soil and rock threshold.

As the number of freeze-thaw cycles increases,the distance between particles will change once the soil sample undergoes a freeze-thaw cycle.In general,the pore distribution trend of the sample did not change significantly,but the pore size and proportion showed different degrees of fluctuation after the freeze-thaw cycle;that is,the small pores gradually changed to medium pores and large pores.The pore size and ratio have increased,and the large pores occupy a large proportion of the shear strength of the soil-rock mixture so that after multiple freeze-thaw cycles,the pore characteristics of the sample change,the coarse and fine particles are rearranged,and the spatial distribution of the particles The structural characteristics of the formed force chain change,and the strength of the soil decreases again.As the rock content increases,the soil skeleton transforms from a pore structure to a dense structure,porosity increases,particle contact points decrease,stress concentration leads to the fragmentation of coarse particles,and strength decreases.Since the soil structure has gradually reached a stable state,its mechanical performance index is also relatively stable,the weakening amplitude is no longer obvious,and the changes in pores and strength tend to be relatively stable.From the above analysis,it can be seen that the freeze-thaw cycle effect significantly weakens the mechanical properties of the special geological material,such as the soil-rock mixture,and this deterioration effect can affect the slope stability.

According to the above analysis,freeze-thaw cycling significantly weakens the mechanical properties of the unique engineering geological material,and this degradation affects the safety of infrastructure in cold regions,such as embankments,tunnels,and natural slopes(Changet al.,2018).Therefore,to enhance the stability of the slope,we can proceed from two aspects:increasing the strength of slope soil and weakening the adverse effects of freeze-thaw cycles on the strength of slope soil.In terms of enhancing the strength of slope soil,anti-freezing and thawing materials can be used,and the slope can be closed by an anchor and shot in time to effectively reduce the effects of freezing and thawing.In terms of weakening the adverse effects of freeze-thaw cycles on the shear resistance of soil-rock mixtures,studies have shown that the occurrence of some freeze-thaw landslide disasters in high-altitude and high-altitude areas is primarily due to the frozen soil on the slope surface undergoing freeze-thaw cycles.The interior gradually melts,and melting water changes the loose soil layer into soil with a higher water content.With an increase in the number of freeze-thaw cycles,the internal structure of the soil-rock mixture continues to change,and more water accumulates inside the soil.It acts as a lubricant in the freeze-thaw dislocation zone,which significantly decreases the soil's shear strength.Under certain conditions,landslide damage will occur along the dislocation zone.Therefore,strengthening and effectively draining the slope in cold regions can significantly mitigate the adverse effects of freeze-thaw cycling on the mechanical properties of the soil-rock mixture and the stability of the slope.

6.Conclusions

This study systematically conducted cyclic freeze-thaw and shear tests on remolded soil samples with different rock content and investigated the correlation between pore characteristics and the shear strength of the samples under varying freeze-thaw cycles using NMR.The following main conclusions can be made.

(1) Based on the NMR experimental results,the effects of freeze-thaw cycling on the microscopic pore structural changes of soil-rock mixtures were studied,which provided an effective way to explain the changes in physical and mechanical properties upon freeze-thaw cycling.With an increase in the rock content,the large pore volume of the soil generally increased after the freezethaw cycling and continued to rise after a minimum appeared after three freeze-thaw cycles.The small pore volume of 0.01-0.2 μm accounted for the slight reduction in the ratio,while the medium and large pores changed significantly,and their volume increased overall.The pores with a size of 0.2-20 μm were most affected by freeze-thaw cycling.Therefore,this range involves freeze-thaw-sensitive pores.

(2) By analyzing the results of the shear test from a macro perspective,it was found that the effects of different freeze-thaw cycles and rock content on the shear strength of the soil-rock mixture were significantly different.The stress-strain curve of the specimen showed strain hardening.Moreover,for the same number of freeze-thaw cycles,when the rock content was 25%-45%,the peak shear strength of the sample gradually increased,the skeleton weakened,the porosity of the soil increased,and the peak strength decreased.Therefore,the soil-rock mixture with a 45%rock content can be approximated as the rock content threshold under freeze-thaw cycling.

(3) The effect of freeze-thaw cycles on soil cohesion is more significant than on the internal friction angle.As the number of freezethaw cycles increased,the cohesion showed a trend of first decreasing,then rising,and finally decreasing,and it reached a peak after three freeze-thaw cycles.There was no apparent change in the internal friction angle.For the same number of freeze-thaw cycles,the effect of freeze-thaw on cohesion increased first and then gradually decreased with increasing rock content.The freezethaw cycle changed the inherent skeleton and the pore size distribution of the soil-rock mixture and affected its mechanical properties.However,with further development of freeze-thaw cycling,the complexity was reduced,and the impact was gradually weakened.

Acknowledgments

This research was supported by the National Natural Science Foundation of China (Nos.42071100,42271144) and the Shaanxi Qin Chuangyuan ＂Scientists+Engineers＂ Team Construction Project (No.2022KXJ-086).