Sai Pramod Singalreddy,Liang Cui,Kun Fang
Department of Civil Engineering,Lakehead University,Thunder Bay,Ontario P7B 5E1,Canada
ABSTRACT Cemented paste backfill (CPB) and rock interface interaction causes the formation of an interfacial loading and affects the thermal,hydraulic,mechanical,and chemical processes in bulk CPB and thus its in-situ behavior.In this study,a new meter-sized column model is developed to systematically investigate the multiphysics processes in CPB under interfacial loading.The obtained results discover that for the mechanical process,the interfacial loading leads to a reduced settlement and a weakened stress level in CPB.For the hydraulic process,lower matric suction and smaller moisture content coexist in CPB under interfacial loading.For the thermal process,the interfacial loading weakens the porosity-dependent thermal conduction and causes retardation in temperature variation relative to the ambient temperature.For the chemical process,weakened cement hydration with smaller electrical conductivity was observed in CPB under interfacial loading.Therefore,the obtained results reveal the linkage between the interfacial loading and multiphysics processes in CPB and thus contribute to an in-depth understanding of the multiphysics behavior of CPB in underground mines.
Keywords:Spatiotemporal evolution Multiphysics processes Interfacial loading Cemented paste backfill Underground mine
As an effective ground support measure,cemented paste backfill technology has been increasingly adopted in underground mines around the world [1-3].After backfill materials are delivered into the mined-out stopes,the relatively soft cemented paste backfill(CPB) body is surrounded by hard rockmass with irregular and uneven surfaces.Consequently,a mesoscale interfacial transition zone (ITZ) is developed between the soft CPB and hard rockmass [4].As shown in Fig.1,the ITZ is featured with a highly microstructural heterogeneity relative to the bulk CPB matrix.It is well known that microstructure plays a crucial role in the development of mechanical behavior and properties of CPB materials.Therefore,the mesoscale ITZ must be evaluated as an independent element in the CPB-rock system.Moreover,due to pore-water displacement in the massive CPB body,the primary consolidation process continuously proceeds and causes the formation of the relative displacement alongside the CPB-rock interface [5].Due to the rough interface characteristics,the vertical displacement on the soft CPB side unavoidably results in the development of interfacial resistance in the ITZ[6].The CPB-rock interfacial loading has the potential to lessen the influence of self-weight on the resultant vertical stress and thus leads to complex in-situ stress redistribution in CPB mass,especially in the narrow stopes [7].Therefore,the evaluation of CPB-rock interfacial loading must be integrated into the accurate and reliable assessment of the insitu mechanical performance of CPB mass.In this regard,extensive experimental and analytical studies on interface behavior have been conducted.For example,direct shear tests have been widely used to investigate the CPB-rock interface behavior and properties[8].It has been confirmed that the interface properties,including the angle of interface friction and adhesion,show distinctive values compared with the counterparts of CPB materials from early to later ages.Moreover,the linear interface shear behavior has been measured in the pre-peak regime [9],while the nonlinear hardening/softening interface response coexists in CPB materials [10].In addition,a series of analytical solutions [11,12] and interface constitutive models [4,5] have been derived to capture the stress redistribution in the mine backfill mass under the CPB-rock interfacial loading.The associated laboratory testing methods and mathematical tools significantly improve the understanding of the complex interface behavior and its effects on the mechanical behavior of CPB.
Fig.1.Formation of mesoscale CPB/rock interfacial transition zone.
However,it is noteworthy that the above-mentioned experimental studies and modeling focus primarily on the mechanical process (especially the stress redistribution) in CPB under interfacial loading.Based on previous studies [13,14],it has been confirmed that the coupled multiphysics processes govern the behavior of CPB and its in-situ stability.Specifically,for the thermal process,due to the combined effects of heat generation by cement hydration and heat transfer between CPB and rock mass,the field and laboratory studies [15,16] have confirmed that temperature varies in CPB materials with curing time.The resultant evolution of temperature affects the binder hydration rate and thus the hardening process of CPB.For the hydraulic process,porewater displacement and moisture loss by cement hydration cause a continuous transition from a fully to a partially saturated state[17,18].The associated evolution of pore water pressure and matric suction leads to changes in effective stress[19],which dominates the mechanical response of CPB.For the mechanical process,the CPB body is exposed to complex field loading conditions such as high geostress [20],dynamic loading induced by blasting operation and earthquake [21,22],and polyaxial stress state variation caused by the removal of surrounding rock mass.Consequently,the accumulation of deformation cannot only cause the degradation of mechanical properties [23,24],including elastic modulus,shear strength parameters,and material strengths,but also affect all porosity-dependent material properties [25,26],including hydraulic conductivity and thermal conductivity.For the chemical process,cement hydration contributes to the formation of interparticle bond strength at the mesoscale and the macroscale material strength.Therefore,to reliably assess the behavior and performance of CPB mass,the prerequisite is to understand the evolution of multiphysics processes in CPB under the interfacial loading.However,no studies have been carried out to discover the linkage between the thermal,hydraulic,mechanical,and chemical(THMC)processes and interfacial loading for CPB materials,which remarkably restricts the understanding of the in-situ behavior of CPB mass.To address this research gap,a metersized column model integrated with multiphysics sensors is developed in this study.Through the developed large-scale column model,the time-dependent evolution of multiphysics field variables was continuously monitored at different spatial positions inside CPB under interfacial loading over a time period of 90 d.The critical role played by interfacial loading in the spatiotemporal evolutionary trends of the complex THMC processes and the associated governing mechanisms were identified.Moreover,the implication of the obtained findings to the engineering design of CPB is discussed through numerical analysis as well.
2.1.Materials and mixture design
The materials used for CPB preparation consist of General Use(GU) Portland cement (mass fraction of Portland cement higher than 90%,calcium oxide in the range of 0.3%-3%,crystalline silica in the range of 0.1%-1.5%,chromate compounds lower than 0.1%,and nickel compounds lower than 0.1%),silica (quartz) tailings,and water.The silica tailings are composed of SiO2(≥99.7%).The chemically inert silica tailings can limit the uncertainties (such as internal sulfate attack by sulfide minerals) to a minimum level and are extensively adopted to experimentally investigate the material behavior and properties of CPB[27,28].Moreover,the previous study[10]has confirmed that the particle size distribution of silica tailings are comparable to that of rock tailings.The particle size distribution curve of the silica tailings is presented in Fig.2.Therefore,the adopted tailings are able to form the typical microstructure for CPB materials.To prepare CPB samples,the mix proportion is set to a cement content of 4.5% (mass fraction of cement relative to the total mass of solid phase) and a waterto-cement ratio of 7.6.The mixing procedures were performed in two phases: (1) a 5-min dry mixing is first conducted on quartz tailings and GU cement,and(2)then the tap water is added to perform a 10-min wet mix process.The adopted mixing approach can ensure that the GU cement is uniformly distributed inside the CPB matrix.The obtained CPB paste is cast into the meter-sized column models (a height of 1.75 m and a diameter of 0.20 m).Detailed information about the proposed meter-sized column models will be presented in Section 2.2.
Fig.2.Particle size distribution of quartz tailings.
2.2.Multiphysics monitoring program
The interfacial loading causes stress and strain redistribution inside CPB mass.Consequently,the microstructure and macroscale materials properties of CPB become sensitive to the changes in spatial positions,and it is essential to capture the spatial sensitivity of multiphysics processes in CPB materials.Therefore,a meter-sized column model was developed in this study.As shown in Fig.3,the proposed column model consists of three components,including a rigid fiber building tube with a dimpled membrane system,a water drainage system,and a multiphysics monitoring system.To form the rough surface,the inner surface of the tube was first coated with a layer of wax to prevent potential water leakage.Then,the dimpled membrane with diagonally oriented 8 mm dimples was coated to the inner surface of the building tube and thus created a rough inner surface.The artificial rough inner surface(Fig.3)was utilized to simulate the rough rock surface.The reasons for the usage of the dimpled membrane are twofold.First,the selected dimpled membrane comprises protruding dimples with a consistently conical cylinder shape,which can form a uniformly rough surface and eliminate the associated uncertainties in the development of interfacial loading.According to the previous research on the surface roughness of this type of dimpled membrane [29],it has been identified that the mean texture depth is approximately equal to 4.5 mm.Therefore,this type of inner surface can be considered to be a very rough surface and thus is able to mimic the in-stope rough rock walls.Second,compared with CPB materials,the surrounding rockmass possesses a relatively small hydraulic conductivity and is reasonably assumed to be impervious.Correspondingly,the selected dimpled membrane(high-density polyethylene) can prevent the moisture exchange between CPB and ambient environments and thus mimic the instope curing conditions.The tube coated with a dimpled membrane was fixed on a 3D-printed base ring.The water drainage system was composed of a perforated polypropylene plate covered by a porous stone and a filter paper,and an outflow pipe.
As shown in Fig.4a,the multiphysics monitoring system consists of multiple sensors and dataloggers.Table 1 summarizes the detailed information about the sensors and dataloggers adopted in the monitoring system.Moreover,a control model without the dimpled membrane is set up simultaneously to perform a comparative analysis.The monitoring program is implemented over a 90-d period.In addition,3 sampling columns with the interface membrane are fabricated to collect the samples at 7,28,and 90 d at the positions same to the monitoring points (Table 1).Therefore,5 m-sized column models are developed in this study.Additionally,it should be pointed out that the conventional centimeter-sized CPB/rock composite samples (Fig.4b) are widely used to study the interface behavior under direct shear tests [8],which cannot capture the spatial changes in multiphysics processes and their influences on the mechanical properties of CPB.Therefore,the proposed meter-sized column model is able to accommodate the differential spatial response of CPB and thus mimic its in-situ behavior.
Table 1 Multiphysics sensors and dataloggers.
2.3.Unconfined compressive strength test
To identify the influence of interfacial loading on the mechanical behavior of CPB,unconfined compressive strength (UCS) tests are performed according to ASTM C39[30].A professional dry electric hand-held diamond cutter (Model: Hilti DCH 300) is used to cut and disassemble the 90-d CPB in the high columns.Two sections of CPB samples are respectively obtained at the middle and bottom of the physical model.Then,a rig-based drilling rig with a 2-inch drill bit is utilized to cut the cylindrical sample.After that,a dual bevel sliding miter saw (model: Bosch GCM12SD) with a diamond masonry blade is used to trim the specimens with flat ends and a height of 10 cm for UCS tests.For the UCS tests,a displacement rate is set to 1 mm/min.The vertical displacement and the associated applied load are respectively recorded by a displacement transducer (nonlinearity: ≤0.03%)) and a laboratory load cell (load capacity: 450 kg).To maintain the reliability and reproducibility of the results,two replicate tests are conducted for each curing condition.
2.4.Hydraulic conductivity test
As a porosity-dependent hydraulic property,saturated hydraulic conductivity plays a key role in pore-water seepage.Correspondingly,the evaluation of saturated hydraulic conductivity can not only facilitate a better understanding of the hydraulic process in CPB but also reflects the changes in microscopic pore structure at the macroscopic level.According to ASTM D5084-16a [31],hydraulic conductivity tests are conducted by a Tri-Flex 2 testing system.Additionally,duplicate tests are conducted on the CPB specimens sampled from the high column at 90 d and thus ensure the repeatability of the test results.
Fig.3.Experimental setup of the meter-sized column model with rough inner surfaces.
Fig.4.Comparison of proposed meter-sized physical model with multiphysics sensors and conventional centimeter-sized CPB/rock composite sample for direct shear tests.
2.5.Scanning electron microscopy imaging
The strain redistribution in CPB under interfacial loading directly causes the changes in its microstructure and in turn,affects the porosity-dependent material properties at the macroscale.Therefore,scanning electron microscopy(SEM)observations were carried out to study the microstructure of 90-d CPB under interfacial loading.To prepare SEM specimens,CPB samples were placed in a laboratory drying oven at 45 °C for 24 h.Then,the oven-dried samples were further cut into small cuboid samples which are coated with a thin layer of carbon.Then,a Field Emission SEM system(model:Hitachi Su-70,with a resolution up to 1.0 nm)was utilized to obtain the SEM image of the CPB microstructure.
3.1.Mechanical process under interfacial loading
The mechanical behaviors of CPB play a key role in the mechanical stability of the in-situ CPB mass.In this regard,the settlement is closely related to the consolidation process and thus to the densification process.Since the granular materials with a higher density are able to possess a stronger interparticle interlocking and friction loading,the characterization of the settlement process can provide an insight into the evolution of material strength.Moreover,the interfacial shear loading can interfere with the stress distribution in CPB.As a result,the measurements of vertical stress can be used as a valuable indicator to assess the influence of interfacial loading.Therefore,the mechanical process can be clearly described by the evolution of settlement (Section 3.1.1) and vertical stress (Section 3.1.2) in CPB.In addition,the compressive behavior of 90-d CPB (Section 3.1.3) was also determined through UCS tests.
3.1.1.Evolution of settlement
Fig.5 demonstrates the comparison of settlement measured from CPB under the interfacial loading and control CPB without interfacial loading.It can be observed that the settlement mainly takes place during the early ages.There are two major contributors to the development of early-age vertical displacement,which include primary consolidation and chemical shrinkage.The former refers to the permanent deformation induced by water drainage.Specifically,the downward seepage occurs in CPB under the gravity effect.As a result,the local water loss reduces the pore space in the early-age soft matrix and leads to the settlement at the macroscale.The latter is induced by the reduced volume of chemically bound water during the hydration process.Although the mechanisms behind these two contributors are different,the magnitudes of permanent deformation stepped from both primary consolidation and chemical shrinkage are constrained by the formation of material stiffness and interparticle bond strength.This is because the hydration products gradually refine the pore space in the CPB matrix from early to later ages.Consequently,the stiffer and stronger CPB matrix forms at later ages and thus limits further skeleton deformation.Therefore,the CPB in both columns shows unnoticeable changes at later ages,irrespective of the interfacial loading conditions.
Fig.5.Evolution of settlement of CPB under interfacial loading.
However,through the comparison of measured settlement,CPB under interfacial loading shows a remarkable decrease in the vertical displacement.For example,the 90-d settlement under the interfacial loading reaches 5.5 millistrains,which is decreased by 58.3% with reference to the settlement (13.2 millistrains) without interfacial loading.It should be noted that except for the adoption of the rough inner interface,a consistent model fabrication process and the same curing environment were adopted for these two column models.Correspondingly,it can be confirmed that the interfacial loading is able to reduce the magnitude of settlement.Moreover,it can be observed that the settlement obtained from CPB under interfacial loading reaches the stable state at 26 d,which is earlier than that from the control CPB (stabilized at 42 d).This is because the interfacial loading applies resistance to CPB in the vertically upward direction and thus reduces the reduction in pore space inside CPB.Consequently,the relatively large pore space facilitates the primary consolidation to proceed at a higher rate,which can be confirmed by the early-age rapid settlement observed from CPB under interfacial loading (Fig.5).However,with the development of settlement,a stiffer and stronger CPB matrix can be developed and thus limit the further deformation in CPB.Therefore,the settlement of CPB under interfacial loading reaches a stable state earlier.The measured results clearly indicate that the interfacial loading can interfere with the evolution of settlement in terms of both magnitude and rate of change,which in turn indicates that the developed meter-sized column model can be utilized as a valuable tool to mimic the influence of interfacial loading on the development of deformation in CPB materials.
3.1.2.Evolution of vertical stress
As shown in Fig.5,the interfacial loading affects the settlement growth.Based on the stress-strain relationship,the vertical stress and the associated settlement in CPB are expected to evolve in a synergistic way.Fig.6 presents the vertical stresses measured at the bottom of CPB.It can be observed that the measured vertical stress shows similar values in these two columns at very early ages,which is consistent with the initial self-weight stress.
Fig.6.Evolution of vertical stress of CPB under interfacial loading.
where σswis self-weight stress,kPa;γCPBthe unit weight of CPB,kN/m3;andHthe backfilling height,m.Based on reported data[32],the unit weight γCPB=20.2 kN/m3is used to calculate the self-weight stress.Therefore,the interfacial loading exerts insubstantial influence on the stress redistribution inside the very early-age CPB materials.The discrepancy in vertical stresses becomes more evident in these two column models with curing time.For instance,the vertical stress in 90-d CPB respectively reaches 10.015 kPa under interfacial loading and 22.237 kPa without interfacial loading,i.e.,a 55% reduction in the vertical stress under interfacial loading.The stress reduction under interfacial loading will significantly affect the engineering design of CPB.This is because the self-weight stress is commonly adopted to determine the design strength and thus the mixture recipe of CPB.However,when interfacial loading is applied to CPB mass,the reduced vertical compared with selfweight stress indicates a weaker requirement on design strength.As a result,when the interfacial loading is considered,an optimum design in terms of backfilling cost management becomes feasible.After the curing time corresponding to the stabilization point(Fig.6),the vertical stress shows unnoticeable changes with curing time.Moreover,CPB under interfacial loading reaches the stable stress state earlier than that in control CPB,which is consistent with the evolutionary trend observed in the vertical stress (Fig.5).
However,through a comparison of settlement (Fig.5) and vertical stress (Fig.6),it is of importance to observe that the vertical displacement reaches the stable stage earlier than settlement under interfacial loading.This can be explained by the differential consolidation rate in the vertical direction.More precisely,the relatively low hydraulic conductivity of CPB significantly reduces the seepage rate of pore water.As a result,the excess pore water pressure cannot dissipate immediately inside CPB.However,due to the existence of a water drainage system at the bottom,a greater hydraulic gradient is generated between the inside and outside of the column base and thus results in rapid water drainage near the bottom monitoring point.Consequently,consolidation at the bottom is relatively fast compared with the counterpart inside CPB.Therefore,the consolidation near the bottom region first reaches a stable stage under the gravity effect.After that,no further changes in vertical stress are measured by the bottom pressure plate.However,the retarded consolidation inside CPB continuously contributes to the overall vertical settlement,which directly causes inconsistent stable points between vertical stress measured at the bottom and settlement measured at the top surface of CPB.
3.1.3.Stress-strain behavior of CPB under interfacial loading
The stress-strain curves of CPB at 90 d are plotted in Fig.7.Note that B stands for the bottom point;and M for the middle point.It can be seen that 90-d CPB demonstrates the strain-hardening behavior at the pre-peak stage and the softening behavior in the post-peak regime.The nonlinear mechanical response clearly indicates the development of permanent deformation and the associated accumulation of material damage under compressive loading conditions.Moreover,considering the progressive failure of CPB mass under extreme loading conditions,it is essential to incorporate the peak strength and residual strength into its mechanical stability analysis.However,from the post-peak branch of the stress-strain curves,it can be observed that the CPB materials present an extremely small residual loading.Precisely,residual stress is in the range of 100 to 200 kPa when strain reaches 5%.Compared with the corresponding peak stress,the residual strength(at a strain of 5%)accounts for 8% to 15% of UCS.Therefore,it is of practical significance to improve the residual strength of CPB through reinforcement techniques.Most importantly,as shown in Fig.8,there exists a discrepancy in the compressive mechanical properties between CPB under interfacial loading and control CPB.It can be seen that compared with the control CPB,the application of interfacial loading can yield a weaker UCS and lower elastic modulus at both sampling points.Precisely,when the interfacial loading is applied,the compressive strengths of CPB are respectively reduced by 16% at the middle point and 10% at the bottom point.Similarly,the elastic modulus shows a reduction of 13% at the middle point and 21% at the bottom point.The changes in UCS and elastic modulus under interfacial loading impose significant impacts on the engineering performance of CPB.First,the reduction in UCS indicates that the design strength through conventional cylindrical samples is unreliable.When the CPB undergoes interfacial loading,the laboratory design strength cannot be guaranteed and thus adversely affect the stability of the CPB body.In addition,the reduced elastic modulus under interfacial loading will weaken the immediate support from CPB subjected to finite deformation from the surrounding rock mass.Although interfacial loading can decrease the stress level and settlement,it can also lead to undesirable mechanical properties.Consequently,the interfacial loading complexifies the in-situ mechanical process in CPB.
Fig.7.Comparison of stress-strain behavior of 90-d CPB with and without interfacial loading.
3.2.Hydraulic process under interfacial loading
Compared with conventional cementitious materials such as concrete,CPB is commonly featured with a low cement content and a high-water content.Due to the seepage process and porewater loss through cement hydration,CPB materials undergo substantial changes in the saturation conditions and pore water pressure.Consequently,the hydraulic process is expected to play a more important role in the engineering performance of CPB mass.To identify the influence of interfacial loading on the hydraulic process,the evolution of matric suction,water content,and hydraulic conductivity will be discussed in this section.
3.2.1.Evolution of matric suction
Fig.9 illustrates the evolution of matric suction in CPB over a 90-d period.As expected,a lower matric suction was measured at the bottom of the CPB.This is because more water can accumulate near the bottom of CPB under the gravity effect.Moreover,through the comparison of matric suction,a lower level of matric suction was measured in CPB under interfacial loading over the monitoring period.This can be attributed to the cumulative effect of capillarity and pore water loss in CPB.For instance,the 90-d matric suctions are respectively reduced by 32.5% (to 204.02 kPa)at the middle point and 24.9%(to 191.124 kPa)at the bottom point compared with counterparts(302.399 kPa at the middle point and 254.443 kPa at the bottom point) measured from control CPB.Specifically,the pore-water loss induced by water drainage and cement hydration directly results in a continuous transition from a saturated state to an unsaturated state in CPB.Due to the capillary effect,the pore pressure jump across the curved air/water interface leads to the development of matric suction which is gradually enlarged with the continuous reduction in water content.However,the bulk volume changes of porous media and the associated changes in pore space cause changes in the capillarity of porous media [33].As discussed in Section 3.1.1,the reduced settlement was detected in CPB under interfacial loading (Fig.5),which indicates a larger pore size formed in the CPB.The correlation between matric suction and pore size can be evaluated by the Yong-Laplace equation:
where Ψ is the matric suction,Pa;Tsthe surface tension,N/m;α the contact angle,(°);andrthe pore radius,m.
An increased pore size leads to the reduction in the matric suction and thus a weakened capillarity.Consequently,a smaller matric suction(Fig.9)is generated in CPB under interfacial loading relative to the control situation.Therefore,the pore water loss causes the development of matric suction,while the weakened capillarity under interfacial loading conditions retards the development of matric suction.
Fig.8.Comparison of stress-strain behavior of 90-d CPB with and without interfacial loading.
Fig.9.Development of matric suction in CPB under interfacial loading.
Moreover,it can be seen that the early-age matric suction shows a rapid change in CPB under interfacial loading with reference to the counterpart in control CPB.This can be attributed to the large porosity resulting from the smaller settlement under interfacial loading.Consequently,a higher hydraulic conductivity is formed in CPB under interfacial loading and thus enhances the seepage rate during the early ages.A detailed discussion on hydraulic conductivity will be presented in Section 3.2.3.However,as the progression of the rapid early-age seepage under interfacial loading,more pore water flows out of the CPB matrix and thus significantly reduces the hydraulic gradient.As a result,a considerable reduction in the rate of change in matric suction was observed in CPB under interfacial loading,especially at the bottom monitoring point.Since matric suction affects the interparticle stress in the unsaturated CPB and thus the material strength,the obtained results imply that accurate evaluation of the material strength of CPB requires the full consideration of interfacial loading in stopes.
3.2.2.Evolution of volumetric water content
Due to the water retention capacity of granular materials,the development of matric suction is unavoidably accompanied by changes in water content.The evolution of volumetric water content(VWC)is plotted in Fig.10.It can be seen that VWC at the middle point is lower than that measured at the bottom point,which is consistent with the lower matric suction measured at the middle points(Fig.9).Moreover,the change in VWC exhibits a descending trend with curing time.As discussed previously,the pore-water consumption by cement hydration and water drainage is responsible for the water loss in CPB[34].However,with the progression of cement hydration,the hydration products form barrier shells around the unhydrated cement grains and thus prevent moisture diffusion inside the cement.Consequently,the hydration rate and the associated water loss slow down with curing time.Meanwhile,the transition from a saturated state to an unsaturated state gives rise to the generation of matric suction in the CPB matrix.The developed matric suction is counteracted by the surface tension alongside the curved air/water interface.The resultant surface tension maintains the pore water in the unsaturated pore space and thus prevents the potential pore-water migration.Consequently,the contribution of water drainage to changes in the water content becomes weaker with the development of an unsaturated state in CPB.In addition,through the comparison of VWC in these two column models,it can be found that compared with the control CPB,a lower water VWC was measured under the interfacial loading.For example,the 90-d water content respectively reaches 0.208 m3/m3at the middle point and 0.346 m3/m3at the bottom point of the column model,which is decreased by approximately 35%(0.318 m3/m3) and 12% (0.392 m3/m3) relative to the measurements in the control CPB.This is due to the fact that the interfacial loading limits the settlement in CPB and engenders large porosity in CPB.The resultant large porosity cannot only reduce the VWC for a given amount of pore water but also weaken the water retention capacity of the CPB matrix.Therefore,compared with the measurements from the control CPB,it is of great significance to find that lower water content (Fig.10) and lower matric suction(Fig.9) coexist in the CPB under interfacial loading.Therefore,the interfacial loading can result in a more complex hydraulic process in CPB.
Fig.10.Evolution of volumetric water content in CPB under interfacial loading and control CPB.
3.2.3.Evolution of hydraulic conductivity
Fig.11 presents the measured hydraulic conductivityKfrom 90-d CPB under interfacial loading and control CPB.It can be observed that compared with the control CPB,a consistent increase inKwas measured at both the middle and bottom points of CPB under interfacial loading.More precisely,theKincreases by 228%(from 2.58×10-6to 8.50×10-6cm/s) at the middle point and 221% (from 1.20×10-6to 3.86×10-6cm/s) at the bottom point.The changes inKcan be related to the development of local settlements inside CPB.This is because,as the key hydraulic property,hydraulic conductivity describes the ability of pore water to pass through the porous matrix and thus plays a critical role in the pore-water percolation in CPB.The magnitude of hydraulic conductivity,K,depends on the solid skeleton and fluid property and can be defined as:
where γwis the unit weight of water,kN/m3;η the water viscosity,Pa·s;andkthe intrinsic permeability,m2.For pore water seepage,the changes in its unit weight and viscosity are negligible.Therefore,intrinsic permeability governs the evolution of hydraulic conductivity.It has been confirmed that the intrinsic permeability depends on the solid skeleton and thus the porosity of the matrix[35].As discussed in Section 3.1.1,the interfacial loading constrains the settlement and thus results in a larger porosity.As a result,a larger hydraulic conductivity can be formed in CPB under interfacial loading and impose a significant influence on the seepage process in CPB.Specifically,more than 84% of tailings particles have a particle size less than 0.075 mm (Fig.2).The tailings can be classified as a fine-grained soil material according to the Unified Soil Classification System [36].For the fine-grained soils,the pore-water flow status can be evaluated by the Reynold number:
Fig.11.Effect of interfacial loading on the hydraulic conductivity of CPB.
whereReis the Reynold number(dimensionless);ρ the water density,kg/m3;Lthe characteristic dimension,m;v the pore-water flow velocity,m/s;and η the fluid viscosity,Pa·s.It has been confirmed thatReis typical in the range of 51.74-162.03 for finegrained soils [37].Therefore,pore-water flow in CPB belongs to the laminar flow regime.Correspondingly,the substantial evolution of hydraulic conductivity is able to cause a linear proportional change in the seepage velocity according to Darcy’s law,which directly facilitates the rapid pore water loss (Fig.10) and development of matric suction (Fig.9) in CPB under interfacial loading.
3.3.Thermal process under interfacial loading
Fig.12 shows the measured temperature in CPB over a 90-d period.It can be observed that a warmer temperature relative to the room temperature was measured in CPB at very early ages,which can be ascribed to the heat released by exothermic cement hydration.After that,the temperature decreases rapidly with curing time,which can be explained by heat exchange between CPB and the ambient environment.Specifically,the temperature difference between CPB and the ambient environment causes a temperature gradient,which leads to heat loss in CPB by thermal conduction.Meanwhile,the thermal advection induced by the pore-water loss through the bottom of CPB also leads to a reduction in temperature.Therefore,the thermal conduction and advection lead to a rapid change from the initial warmer temperature to the room temperature by approximately-two days.It should be pointed out that the initial temperature at the middle and bottom points of CPB under interfacial loading is almost identical(Fig.12),and thus the associated data points are overlapped.After two days,it can be observed that the changes in internal temperature are similar to the variation in room temperature,indicating thermal conduction plays a more critical role in the temperature variation inside CPB.
However,compared with the control situation (Fig.12b),the internal temperature of CPB under interfacial loading conditions shows a more apparent deviation relative to the room temperature during the monitoring period.More specifically,the retarded temperature changes become more obvious at the middle point of CPB under interfacial loading.This is mainly attributed to the weakened thermal conduction between the surrounding environment and CPB under interfacial loading.Due to the porositydependence of thermal conductivity,the smaller settlement under interfacial loading(Fig.5)causes a larger pore size near the middle monitoring point,which inevitably reduces the effective thermal conductivity of CPB.Meanwhile,the thermal conductivity of pore air (approximately 20 mW/m·K) is 30 times less than that of pore water (approximately 600 mW/m·K) [38].Therefore,the lower water content (Fig.10) at the middle point can further decrease the effective thermal conductivity of CPB under interfacial loading.Consequently,the thermal conduction between the surrounding environment and CPB slows down and thus causes variation retardation in the internal temperature relative to the ambient temperature.Therefore,the interfacial loading is able to interfere with the temperature variation in CPB.In other words,the interfacial loading has the potential to enlarge the temperature gradient in CPB.Considering the incompatible thermal properties (e.g.,coefficient of thermal expansion) of tailings and hydration products,the strengthened temperature gradient inevitably causes thermal deformation and the associated material degradation.Consequently,the complex thermal process induced by interfacial loading affects the engineering performance of CPB in underground excavations.
3.4.Chemical process under interfacial loading
Based on previous studies on cement hydration [39-41],it has been found that a mixture of cement with water causes the release of calcium (Ca2+),hydroxyl (OH-),and sulfateions into the water and results in an electrically conducting solution.However,due to the cement hydration,the precipitation of calcium silicate hydrate and hydroxide gels in pores gradually reduces the ion concentration in pore water.Correspondingly,the electrical conductivity (EC) evolves with the changes in ion concentration and cement hydration.Therefore,measurements of EC can be employed as an indirect indicator to describe the advancement of cement hydration [15,42].Fig.13 shows the measured EC in CPB from early to later ages.It can be found that(1)EC shows a declining trend with time.This is because cement hydration continuously consumes ions,including Ca2+,OH-,and,in the pore water[43,44]and thus weakens its electrical conduction.(2)The time rate of change of EC gradually reduces,which indicates that the hydration reaction slows down with time.This can be attributed to the formation of a hydration shell around the unhydrated cement particles[45].Correspondingly,the free water becomes more difficult to diffuse toward the anhydrous cement and thus reduces the hydration rate.(3) Through comparison with the control CPB,lower EC was formed in CPB under interfacial loading.This is because the interfacial loading reduces the settlement and gives rise to a larger porosity in CPB.As a result,the pore-water loss through the downward seepage can proceed to a greater extent and thus result in lower water content in unsaturated CPB.Correspondingly,the smaller moisture content weakens the water diffusion into the anhydrous cement grains and thus further retards cement hydration.Therefore,the interfacial loading can influence the chemical process through its effect on the changes in water content.
Fig.12.Evolution of temperature in CPB under interfacial loading and control CPB.
Fig.13.Comparison of electrical conductivity of CPB under interfacial loading and control CPB.
The continuous precipitation of hydration products refines the pore space [46] and thus changes the microstructure of CPB.Fig.14 shows the SEM images of the microstructure of CPB at the middle and bottom points of the column model.It can be seen that the large tailings particles are surrounded by the fine-grained hydration products and thereby forms a cohesive structural network,which clearly demonstrates the contribution of hydration products to the material strength.Meanwhile,a relatively coarse pore structure formed at the middle points (Fig.14a and c) compared with the counterparts at the bottom points (Fig.14b and d).As discussed in Section 3.1.1,the enhanced seepage near the bottom points can accelerate the consolidation process with reference to the middle points.Therefore,the resultant differential settlement inside CPB leads to spatial changes in microstructures.Most importantly,when the interfacial loading is applied to CPB,a coarser microstructure is developed inside CPB (Fig.14a and b)relative to the control CPB (Fig.14c and d).Therefore,the interfacial loading is able to interfere with the development of microstructure.As a result,the differential microstructure changes affect the macroscale porosity-dependent mechanical properties of CPB.Precisely,the measured UCS and elastic modulus (Fig.8)clearly show that (1) degradation of mechanical properties occurs at the middle point relative to the counterparts at the bottom points,and(2)a relatively large reduction in mechanical properties is discovered under interfacial loading.Since the macroscale mechanical properties are commonly adopted to evaluate the insitu engineering performance of CPB technology,it is essential to incorporate the effect of interfacial loading into the assessment and design of CPB materials.
Based on the monitoring results through the developed metersized column models,it has been found that interfacial loading affects the multiphysics (thermal,hydraulic,mechanical,and chemical) processes in CPB.However,it is noteworthy that the interfacial loading imposes both direct and indirect influences on THMC processes in CPB.Specifically,for the mechanical process,the effect of interfacial loading is straightforward.The rough interface resists the vertical settlement of CPB,which influences the stress redistribution in CPB.Therefore,the interfacial loading is able to directly vary the field variables,including stress and strain in CPB.However,the interfacial loading imposes indirect effects on the HMC processes through its influence on the porosity.Since the interfacial loading reduces the vertical settlement and gives rise to a relatively large porosity,the porosity-dependent material properties,such as hydraulic conductivity,will be affected as well.Correspondingly,a lower matric suction(i.e.,hydraulic process)and a retarded temperature variation (i.e.,thermal process) were observed under interfacial loading.The indirect effect of interfacial loading on cement hydration is more complex.The resultant large porosity leads to the low moisture content in the CPB matrix and thus weakens the water migration into the anhydrous cement particles.Consequently,the chemical process is indirectly affected by the interfacial loading.
Fig.14.SEM images of CPB under interfacial loading and control CPB at the middle and bottom points.
Moreover,it should be pointed out that the effect of interfacial loading on the multiphysics processes mainly takes place at later ages.This is because the interfacial loading consists of two components,including interface friction force and adhesion.However,there exists sufficient pore water in the early-age CPB matrix,which causes a strong lubrication effect alongside the interface and thus significantly weakens the interface friction loading at early ages.Moreover,due to the low cement content in CPB,the interface adhesion is extremely small at early ages.Therefore,the formation of interfacial loading is a time-dependent process and is gradually enhanced at later ages.Consequently,the multiphysics processes become sensitive to interfacial loading at later ages.Additionally,based on the conventional strength-based engineering design approach,the ratio between design strength and vertical stress is used to calculate the factor of safety for the engineering design of CPB.In terms of vertical stress,it has been confirmed that interfacial loading is able to reduce vertical stress (Fig.6).However,the design strength is commonly measured from the small cylindrical samples without the interfacial loading in the laboratory.As shown in Fig.7,CPB demonstrates a weak UCS when the interfacial loading is applied.Specifically,the UCS values of CPB are respectively reduced by 11.3% (from 1.449 to 1.285 MPa) at the middle point and 10% (from 1.511 to 1.359 MPa)at the bottom point.Consequently,the smaller UCS under interfacial loading leads to an overestimated mechanical stability of CPB mass through the design strength.Therefore,the obtained results indicate that the interfacial loading must be considered for the reliable engineering design of CPB.
A meter-sized column model was developed and successfully utilized to investigate the effect of interfacial loading on the multiphysics processes in CPB.The following findings can be drawn:
(1) The interfacial loading reduces the vertical displacement by 58.3% and vertical stress by 55% in CPB and thus directly affects its mechanical process.
(2) Lower matric suction with a reduction of up to 32.5% and lower water content with a decrease of up to 35% coexist in the unsaturated CPB under interfacial loading compared with the control CPB without interfacial loading.
(3) The interfacial loading condition leads to a retardation in temperature variation compared with the ambient temperature.
(4) Under the interfacial loading,the resultant lower water content further weakens the water displacement into the anhydrous cement and thus slows down the cement hydration.
(5) The interfacial loading gradually develops in CPB,and thus its effects on the THMC processes become more obvious at later ages.
Since the link between the interfacial loading and multiphysics processes was uncovered by the monitoring results,the developed meter-sized column model can be utilized as a valuable tool to mimic the in-situ behavior of CPB.Therefore,the findings from this study have the potential to enhance the understanding of the behavior of CPB under complex field loading conditions.
Acknowledgments
This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC).
推荐访问:hydro mechanical Chemical