【正文】 ricks,liquefaction, and piping within the hydraulic fill areattributed to such failures [2]. Therefore, permeability ofthe hydraulic fill in the stope is a critical parameter inthe design。 continuous e?ort is made during mining toensure that it is kept above a threshold limit in thevicinity of 100 mm/h [3]. Larger permeability leads toquicker removal of water from the stope, thus improving the stability of the fill contained within the stope.Permeability tests for mine fills and barricade bricksare discussed by Rankine et al. [4]. The constant headand falling head permeability tests carried out on thehydraulic fill samples give permeability values in therange of 7e35 mm/h. In spite of having permeabilityvalues much less than the 100 mm threshold suggestedby Herget and De Korompay [3], each of these hydraulicHorizontalaccess drivePorous barricadebrick wallHorizontalaccess driveDecant waterHydraulic fillSlurry entersstope1169Production 14 (2020) 1168e1175Fig. 2. An idealised stope with two sublevel drains.fills has performed satisfactorily. Anecdotal evidencesand back calculations using the measured flow inthe mine stopes suggest that the permeability of thehydraulic fill in the mine is often larger than what ismeasured in the laboratory under controlled conditions.Kuganathan [5] and Brady and Brown [6] proposedpermeability values in the range of 30e50 mm/h, whichare significantly larger than those measured in thelaboratory for similar fills. These values are much lessthan the threshold limit prescribed by Herget and DeKorompay [3], suggesting that it is a conservativeremendation.. Stability considerationsThe stability of the hydraulic fill stope during andafter the drainage period depends on several parametersthat determine the strength and the sti?ness of thehydraulic fill mass. These parameters can be measured inthe laboratory using reconstituted samples or in themine using in situ testing devices. Due to the di?cultiesand high costs associated with carrying the in situ testingrigs into the underground openings, laboratory tests arethe preferred alternatives.Strengthandsti?nessaredirectlyrelatedtotherelativedensity of the fill. When the hydraulic fill is denser, therelativedensityandfrictionanglearehigher,andthusthefill is more stable. In geotechnical engineering, there areseveral empirical correlations relating relative density totheYoung’smodulusandfrictionangleofagranularsoil.. Maximum and minimum dry density testsA larger void ratio does not always mean a loosergranular soil. Relative density is a good measure ofthe density of the grain packing, and depends on themaximum and minimum possible void ratios for the soilwhilst still maintaining intergranular contact. Theminimum void ratio is generally determined by pouringthe dry tailings from a fixed height so that the grains areplacedataveryloosestate[7].Themaximumvoidratioisgenerallyachievedbysaturatingthetailingsandvibratingthem to attain dense packing [8]. These two extreme voidratios provide the lower and upper bound for the voidratios, and, depending on where the current void ratioof the hydraulic fill is, the relative density is defined as:DrZemaxC255ecurrentemaxC255emin!100% 240。1222。Laboratory sedimentation exercises at JCU laboratories, during which hydraulic filling processes weresimulated, showed consistently that when slurry settlesunder its selfweight, the relative density of the fill is inthe range of 40e70% (Fig. 3). Similar observations weremade by Pettibone and Kealy [9] at selected mines in the1170 N. Sivakugan et al. / Journal of CleanerUnited States. The in situ measurements showed relativedensity values ranging from 44 to 66% at four di?erentmines.The laboratory exercise also showed that the hydraulic fill slurry settles to a dry density (g/cm3) of timesthe specific gravity (Gs) for a wide range of tailings withspecific gravity values ranging from to . Drydensity (rd) and void ratio (e) are related by:rdZGsrw1Ce240。2222。This implies that all the hydraulic fills settle to a voidratio of and porosity of 40%. The laboratorysedimentation exercise verifies this.. Oedometer testsOedometer tests are carried out on hydraulic fills todetermine the constitutive modelling parameters for theCam Clay model e one of the constitutive models thatcan be adapted for hydraulic fills when analysed usingnumerical modelling packages such as FLAC, FLAC3Dor ABAQUS. In addition, oedometer tests are useful indetermining the constrained modulus (D) from which,Young’s modulus (E) can be estimated for an assumedvalue of Poisson’s ratio using the following equation.EZ240。1Cy222。240。1C2552y222。240。1C255y222。D 240。3222。Young’s modulus is a crucial parameter in deformation calculations using most constitutive models. TheRelative Density, Dr ( )Void ratio, e Densest possiblestateLoosest possiblestate 0100Hydraulic fillsin mines50eminimumemaximumFig. 3. Relative density of the hydraulic fills sedimented in thelaboratory.Production 14 (2020) 1168e1175oedometer tests on the hydraulic fills showed significantproperty test, as in a paction test, is to determineoptimum water content. In Fig. 5, the optimum waterFig. 4. Scanning electron micrograph of a hydraulic fill sample.creep settlements that took place on the pletion ofconsolidation settlements. This has yet to be verifiedquantitatively and on a fullscale stope.. Direct shear testDirect shear tests are carried out to determine thepeak and residual friction angle of the hydraulic fill. Thetests are carried out on reconstituted hydraulic fillsrepresenting the in situ grain packing in the stope, whichcan be at relative densities of 40e70%. Since there isno clay fraction, cohe