by JL TAYLOR · 1983 · Cited by 2 — Each filter has two rubber drainage belts, each belt with a central vacuum box and the two belts with a common head and a tail pulley (1,4 m in diameter). Each.

255 KB – 9 Pages

PAGE – 1 ============
Anoperationalreviewofthebelt-filtrationplantatChemwesLimitedbyJ.L.TAYLOR*.B.Eng.(Hons.).M.I.M.M..C.Eng.M.I.M..F.S.A.I.M.M.SYNOPSIS,Thebelt-filtrationplantatChemwesLimitedhasbeeninoperationforthreeyears.Theplantdesignisdiscussed.Operationalandengineeringproblemareasarereported,andwherepossiblethemodificationsorpracticesthatledtotheirsolutionarementioned.Capitalandoperatingcosts,togetherwithselectedgeneraloperatingdata,arealsoincluded.SAMEVATTINGDiebandfiltreeraanlegvanChemwesisalvir’ntydperkvandriejaarinwerking.Dieaanlegontwerpwordbe-spreek.Problemeondervindindiebedryf-eningenieursgebiedewordbehandelenwaarmoontlikworddieveranderingebespreekwatgeleihettotdieoplossingvandieprobleme.Gegewensoorkapitaal-enbedryfskoste,sowelasuittrekselsvanbedryfsdata,isookingesluit.IntroductionInlate1976anoptimizationstudywascompletedfortheStilfonteinUraniumProject(latertobecomeChem-wesLimited),whichrecommendedbeltfiltersinpre-ferencetocountercurrent-decantationthickenersordrumfiltersforthetreatmentofStilfonteinslimes-dammaterial.Asaresult,beltfilterswereincludedinthe’frozen’processdesign.Fourestablishedsuppliersofbeltfilterstenderedin1978and,afteradjudication,theorderwasplacedwithElginEngineeringforaFiltersGaudfrindesignplanttobemanufacturedinSouthMrica.TheChemwesplant,totreat270000tpermonthofreclaimedgold-plantresidueslimesfortherecoveryofuranium,wascom-missionedinJune19791(Fig.1).DesignParametersDoubtsaboutthemechanicalreliabilityofbeltfilterswerestillbeingexpressedduringthedesignstage,andaconservativeapproachwasadopted.Only85percentavailabilitywasassumedforthebelt-filtrationplant,which,forthepredicteddutyof12,5tpersquaremetreperday,requiredapproximately850m2offilterarea.Asolublelossof2percentwasplannedatawashratioof0,6:1(Le.0,6m3ofwashpertonofsolids),givingapproximately1,4m3ofpregnantsolutionpertonofsolids.Theconsumptionofaguar-basedflocculantwasestimatedat100gpertontreated.DescriptionofthePlantNinefilterswereinstalled,eachwithanareaundervacuumof94m2.Eachfilteroperatesasaseparateunitwithitsownvacuumpump,filtratesystem,andfilter-cakerepulper,andhasitsownlocalcontrolpanelshow-ingthedrivesystemandprocessalarmsassociatedwiththatfilter,togetherwithvacuumreading,belt-speedindicator,andspeed-adjustmentbutton!>.*ChemwesLimited,P.D.Box310,Stilfontein,2550Transvaal.(g)TheSouthAfricanInstituteofMiningandMetallurgy,1983.SAISSN0038-223X/S3.00+0.00.Thefiltersupportstructureisa316Lstainless-steelfolded-sectiondesignwithonlythreemajorpointsofsupport.Thisunclutteredstructureallowseasyaccessunderthefilter,andisneatandsimpleinappearance.Theholding-downboltsandnutsaremildsteelwithacorrosion-protectivecovering.Eachfilterhastworubberdrainagebelts,eachbeltwithacentralvacuumboxandthetwobeltswithacommonheadandatailpulley(1,4mindiameter).Eachbeltisapproximately60mlongand2,13mwide,andisconstructedoffive-plypolyesterfabricwithrubbercoversof20mmatthetopand6mmatthebottom.Thedrainagegroovesaretaperedtoasinglerowof15mmholesdrilledatthelowpointofeachgrooveinthecentreofthebelt,theholesbeingsituatedoverthevacuumbox.Thetwobeltsperfiltereachhaverubber-sidedshouldersoftriangularcross-sectiontoproducethenecesmryshallowtroughontopofthebelttocontaintheslurryandwashwater.Eachbelthasarubberwear-stripfixedtoitsunderside,whichslidesontopofthevacuumbox.Thebeltsareseparatedfromeachotherbytracking-guiderollers,andaresupportedonbedsofwaterincorrugationsofthebelt-supportdecking,whichismadeof304Lstainlesssteel.Thebelteffectivelyslidesoverthewetsurfacesofthestainless-steeldecking(Fig.2).Distinctivefeaturesofthepulleysarethetwoshallowchannelsrequiredtoaccommodatetheraised-vacuumwear-beltsectionsontheundersideofthebelts.Theheadandtailpulleysareofrubber-linedmildsteel.Thehead-pulleydrivesystemisviaahydraulicmotorthroughaplanetarygearbox.A75kWelectricmotordrivesthehydraulicpowerpack,deliveringpressuresofupto15MPa.Operatingpressuresareintherange6to11MPadependingonthematerialsusedonthecontactsurfacesbetweenthevacuumboxandthebelt,theconditionofthesurfaces,andtheeffectivenessofthelubricatingwatersupply.Beltspeedscanbeadjustedbetween5and25m/min,buttheoperatingspeedistypicallyabout15m/min.JOURNALOFTHESOUTHAFRICANINSTITUTEOFMININGANDMETALLURGYOCTOBER1983237

PAGE – 3 ============
~~~~LJ’/,.’.’,.”.'””‘”‘.,,~L.:.J.i’,ii~~J~-1IiIINJECmN),LL-~—.Fig.2-ThesystemofbeltsupportAfilterclothrestsontopofeachfilterbelt.Oncakedischarge,theclothseparatesfromthebeltandreturnsviaaseparatesetofreturnrolls,tensioningrolls,andtrackingrollersbacktothefeedendofthefilter.Trackingisachievedviaapivotedpneumaticallyactuatedguideroller.Cloth-edgesensingisbylimitswitcheswithwirefeelerssituatedoneachsideofthecloth,whichactuatethemovementoftheguideroll.Theguiderolloscillatesbetweentwofixedpositions,whichtherebyforcestheclothtomovefromsidetosidebetweenthelimitssetbythefeelerswitches(Fig.3).Dewrinklingofthefilterclothsisachievedbyabowedrollerimmediatelypriortotheclothreturningtothetopofthefilter.Cakedischargeandclothcleaningareachievedbyhigh-pressurewaterspraysthroughfinenozzlessetat150mmintervalsonpipesinfrontandbehindthecloth.Spraywaterjoinsthedischargeslimeintherepulpertoprovidethenecessarydilution.Vacuumisgeneratedforeachfilterbyasinglerotaryvacuumpumpwithacapacityof7000m3/h.(Onestand-byvacuumpumpfortheotherninewasalsoinstalled.)AIRSUPPLYSENSORFig.3-Thecloth-trackingsystemThevacuumpumpisconnectedthroughamistarrester(towhichspraywatercanbeappliedtoscruboutacidmistpriortothepump)tothecleananddirtyfiltratereceivers,andfromtheretothefilterviathevacuumlines.Thevesselsandpipesareallrubber-lined.Avacuumofapproximately55kPaistypical.Eachfilterhastwovacuumboxes,oneforeachbelt.Vacuumisappliedthroughthestainless-steelvacuumbox,throughtheholesinthecentreofthedrainagebelttothefiltercloth.Eachvacuumboxhaswearstripsfittedtothetopthatslideagainsttherubberwearstripontheundersideofthebelt.Waterisinjectedforlubri-cationandsealingthroughinletpointsinthestrips(Fig.4).Mostofthiswaterjoinsthefiltrate.FASTENINGSCREWSWATERINJECTION/Fig.4-AvacuumboxuuvUHVbeltgroovestojointhefiltrate.Foldinginoftheclothsattheedgesalsoledtowearand,finally,failureatthecrease.Also,iftheclothedgeremainedfoldedontheclothreturn,theedgescouldnotbesensedaccurately,leadingtoseveretrackingproblems.Problemswerealsoexperiencedwithfrayingoftheclothedges.Agreatimprovementinclothperformancewasachievedwithclothsofneedlefelt(alsoapolyestermaterial).Theseclothsareofatotallydifferentphysicalconstruction,andhaveathicker,stiffernaturethatpermitsmore-positivesensingandtrackingwithlessriskofedgefolding.Theseclothsarestillinusetoday,theaverageclothlifebeingabout74days.Thisincludesclothsthatarereplacedbeforebeingwornoutbecauseofphysicaldamage.Onedisadvantageofneedle-feltclothsisthattheyh~veatendencytostretch.Thishasbeenovercomebytheallowanceofaminimumamountofslackwhennewclothisfitted.Thestretchisthentakenupeitherbythenormalcloth-tensioningdeviceorbythefittingofextensionsonthecloth-returnrollerstolengthenthecloth-returnpath.Anotherdisadvantageofneedlefeltisthatittendstogivelowervacuumsthanwovencloths.Thismaybeduetotheeffectivelytighterweaveofwovencloths,butdifferencesinfilterdutyhavenotbeenobserved.Adefiniteadvantageofneedle-feltclothsisthatjointscanbemadebyhot-airwelding(melting)ofover-lappedsectionsofcloth.Onaverage,afilterrecloth,includingthetimefortheremovaloftheoldcloth,canbecompletedinabout3hours.Holesintheclothcanberepairedbythesametechnique,andrepairtimesaverageabouthalf-an-hour,thetimetakenobviouslyvaryingwiththecomplexityandsizeofthepatching.Experimentswerealsocarriedoutwithaneedle-feltclothproducedbyarivalmanufacturerinwhichthenecessarywidthwasachievedbythejoiningoftwoJOURNALOFTHESOUTHAFRICANINSTITUTEOFMININGANDMETALLURGYOCTOBER1983241″–

PAGE – 6 ============
sectionsofclothalongthecentre.Theseclothsalsotendedtostretchinuse,but,possiblyowingtouneventensionsresultingfromthejoins,therateofstretchwasnotthesameinbothhalvesofthecloth.Onehalfoftheclothwasstillintensionwhiletheothersideoftheclothhadaloose,wavyedgethattendedtofoldin,causingthewholeclothtobecomeuntracl):able.Thecloth-trackingarrangementisstillmaintenance-intensive.Asdescribedearlier,theclothisconstantlytrackedfromsidetosidebetweenthelimitssetbylimitswitches.Onsomefiltersthiscanresultintheactivationseveraltimesaminuteoftheswitchesandcloth-trackingroller,whichismovedpneumatically.Thisareaisalsosubjecttodrippingacidicsolutions,whichaddstotheproblemofmaintenance.Improvementsmuststillbemadeinthisarea.Clothrollersatthecakedischargewereinitiallycoveredwithrubber,butrapidfailureledtotheirre-placementwithstainless-steelrollers.Theserollershaveperformedwellinthisapplication,butthebearingstendedtofailprematurelybecauseoftheacidsprayfromthecloth-cleaningsprays.Initially,barrensolution(asforthecakewash)wasutilizedforthecloth-cleaningsprays,buttheresultantacidmistinthebuildingmadeconditionsveryuncom-fortableforstaffandwasextremelycorrosivetothesteelwork.Industrialwaterisnowused.Fine-meshfiltersinthecloth-spraylineshavegreatlyreducedtheproblemsofblockagesofthecloth-spraynozzlesbywoodchipsandothertrash.Thesophisticatedhigh-pressurecloth-washingsystemwithoscillatingspraypipesandthecausticcloth-wash-ingsystemprovedunnecessary,andmostofthesystemshavebeenremoved.VacuumPumpsandFiUrateSystemsInitially,thevacuumboxwasadjustedwithaclear-anceofabout2mmfromtherubberbelt,andvacuumhadtobeappliedquicklytopullthebeltdownontothevacuumbox.Significantdowntimewasexperiencedbecausethelargeclearancessettoimprovethelifeofthewearstripsresultedinproblemsintheapplicationofvacuum.Thiswasovercomebytheinstallationofarapid-openingbutterflyvalveonthemainvacuumline.Thefiltratepumpshavebeenhighmaintenanceitems,themodelusedbeingmountedonthesideofthefiltratereceiver.Breakingofshaftshasbeenaproblem,ashashighimpellerwear.Somecostsavinghasbeenachievedbythereconditioningofimpellers.Industrialwaterisusedforthesealingofthevacuumpump,butseriousscalingoccurredinthepumps,whichledtojammingoftheimpellers.Thishasbeenovercomebytheadditionofachemicaladditivetothewater.Thesituationisnotcompletelysatisfactoryasthesealwateroperatesonasingle-passsystem,andhencethevalueofthechemicaladditiveisimmediatelylost.Someimpellerpittinghasalsobeenexperienced.Ceramiccoatingoftheimpellers,shafts,andshellistobetriedtoreducethesealingandpittingproblems.Aregularproblemwiththevacuumpumpshasbeentheimpellercominglooseontheshafts.Reconditioningistime-consumingandexpensive.Impellersonshaftsare242OCTOBER1983JOURNALOFTHESOUTHAFRICANINSTITUTEOFMININGANDMETALLURGYnowshrink-fitted,althoughtheinitialvendorinformationindicatedthataslipfitwasadequate.RepulpersandDiscoorgePumpsTherepulpershavegenerallyperformedwell.How-ever,adesignfaultwasthepositioningofthemotorsandgearboxestooclosetotheagitatedslurry,wheretheyweresubjecttosplashandwerenotinaneasypositionformaintenance.Averticalfrothpumpwasinstalledtohandlethedischargefromtherepulper.Itisnowfeltthatthiswasnotagoodchoiceandahorizontalspindlepumpwouldbemoresuitable.Thebearingsareaweakfeature,andtoofrequentfailuresareencounteredwiththeimpellers.FiUerDrivesThefilterdrivesarehydraulic.Thehydraulicpumpshaveperformedextremelywellafterovercomingsomeinitialhydraulicfilterproblems,andminimalproblemshavebeenexperiencedwiththehydraulicmotors.However,theplanetarygearboxeshavegivenseriousproblems.Thedesignseemstobeinadequate.Theinitialproblemswereduetothefallingoutofbearings.Althoughthiswassolvedbyamodificationtothelockingplate,thelifeofthebearingcontinuedtobeshort(approximately1500hoursasagainstadesignof50000hours),andrapidtoothwearwasexperienced.Modifica-tionswerecarriedoutbythesuppliers,butfrequentattentionisstillrequired.Theconditionofthegearboxismonitoredcloselybytheexaminationofoilsamplesandbytheuseofmagneticplugsatthedrainpoints.Thelifeofthebearingshasbeenextendedtoabout4500hoursandthetoothwearreduced.However,crackshavebeenfoundincertaingears.SolubleLossThelossofdissolveduraniumhasbeenanintermittentproblem.Forlongperiods,acceptablelosses(closetothe2percentdesignfigure)wereachieved.However,atothertimeslossesashighas6percentforamonthormorewereexperienced.Thesehighlossesseemtobeassociatedwithperiodswhenveryfineslime(asfineas45percentsmallerthan10fLm)isbeingfiltered.Thecorrelation,however,wasnotconsistent.Samplingproblemswerealsoexperiencedinthatitwasnotpossibleforthehighsolublelossesfoundinrepulpersamplestobecorrelatedwiththesolublelossesinthefilter-cakedischarge.Frequently,negligiblesolu-blelosseswerefoundinthefilter-cakedischargewhensignificantvalueswerefoundintherepulper,andthesedifferencescouldnotbecorrelatedwithanyknownsolu-tionsenteringtherepulper.Itwasthoughtthatresidualfiltrateinthefilter-beltgroovesandfilterclothcouldbefallingintotherepulperattheheadpulleyafterthecakedischarge.Trayswereconstructedtocollectthesedrip-pings,andtheirvalueswerefoundattimestobehigh(similartothoseofthepregnantsolution),buthighunexplaineddiscrepanciesstillremainedbetweenthesolublelossesintherepulperandthoseinthecake,evenaftermostofthesedrippingshadbeenremoved.Asatisfactoryexplanationoftheproblemisstillbeingsought.Fig.5givestheresultsfortheyears1980and1981,indicatingthemonthlyaveragesolubleloss,to-

PAGE – 8 ============
performanceshiftbyshift.Changeintheadditionrateofflocculantisthemainparameterusedbytheoperatortoadjustthetonnagethroughput.Othermethodsareavailable,butflocculantadditionisthemostconvenient.Thismayattimesbeacontributorycauseofthehighsolublelosses.Theconsumptionofflocculantcanvaryfrom50toHOg/toftheguar.basedflocculantused.Thefilteravailabilityisgenerallyfairlyconsistent(about92percentonaverage),butthedutiescanvaryconsiderablyfromdaytoday(from9to14t/m2perday).InstrumentationNoneofthestaffofChemwesorthemanagementcontractorhadanyrealexperienceinoperatingbeltfilters.Thisgaverise,asmentionedearlier,toaconser-vativedesignphilosophy,togetherwithapessimisticviewofthethingslikelytogowrong,resultinginover-instrumentationonthefiltersandover.annunciationofalarms.Thelocalcontrolpanelsforeachfilterarecurrent-lylargelyignoredbytheoperatorssincetheydisplayflashinglightsforitemslongsinceredundantorofnointeresttotheoperator.Areviewofthesepanelsisunderwaysothattheycanserveasausefulmonitorofthefiltrationoperation.CapitalCostTableIprovidessomeideaofthecapitalcostsofthetotalfiltrationplant.Thetotalsumdoesnotincludethemanagementcontractfee,northecommissionsnormallyassociatedwithcapitalprojects.Thetotalcostcomparedveryfavourablywiththebudgetedcosts.Theheadingsoftheindividualitemsmaynotappearcompletelylogical,butcertaincostgroupingsaredic-tatedbythedivisionofthetotaljobamongthesub-contractors.TABLEIBREAKDOWNOFCAPITALCOSTS(1979MONEY)FORTHEBELT-FILTRATIONPLANT1.Filters,drives,repulpers,platforms,etc.2.Civils3.Piping,fabrication,anderection4.Buildingandpainting5.Vacuumpumps6.Instrumentation7.Filtratesystem,receivers,andpumps8.Cables,cableracking,erection,etc.9.Flocculantmake.up,feedpump,andflocculators10.Spillagehandling11.Cake-washtankandpumps12.ClothwashRX10347341016581529302147165133lOO6055367858OperatingCostsTableIlshowsestimatedoperatingcostsforthetwo.yearperiod1980-1981.Itshouldbenotedthatseriousinflationwasexperiencedoverthisperiod.Certainplansareinhandtoreduceworkingcosts(i.e.theuseofH.D.P.E.wearstrips),butitisexpectedthatinflation,togetherwiththeneedforthereplacementofrubberbelts,willoffsetanysavings.Certainofthecostsareestimated,buttheyarebelievedtobefairlyaccurate.244TABLEIIBREAKDOWNOFOPERATINGCOSTSFORTHETWOYEARPERIOD198()’-1981FORTHEBELT-FILTRATIONPLANTRCentpertontreated*10,618,9StoresFlocculantOther732000I302OOot203400065600028000039200018900055100029,59,54,15,72,78,059,5Power:j:Water:j:Whitewages§Blackwages§Renewalsandreplacements**4102000*6895000tonstreatedtIncludesfilterclothsandgeneralmaintenanceofallfilter-plantequipment:j:Estimated§Operatingandmaintenancelabour**Includevacuumwearstrips,drainagebelts,replacementofallvacuumpumpmotors(initialinstallationundersized)vacuum.boxmodifications,andrectificationofautomaticlubricationsystemCausesofFilterDowntimeTheinformationinTableIIIwasextractedfromdatathatwereaccumulatedduringthetwo-yearperiod1980-1981inclusive.Itshouldbenotedthatmanyitemsmoved,withtheaccumulationofdown.timedata,fromparticularfailuresorreasons,intotheGeneralMaintenancecategorybecauseplannedmaintenancepreventedthefailuresthathadbeenoccurringintheearlierpartofthereviewperiod.TABLEIIICAUSESOFFILTERDOWNTIMEFORTHEYEARS198()’-1981%oftotaldowntime1.Rubber.beltrepairsandtracking2.Operatingproblemsexternaltofilterplant*3.Patchingandreplacingcloths4.Nofeedtofilterincludingfeed.linechokes5.Generalmaintenance6.Powercutsandfailures7.Vacuumstripsandboxes8.Slimefeedvalves9.Cloth-correctingrollersandsensors10.Repulper-pumprepairs11.Vacuum.pumprepairs12.Power-packrepairs13.Belt-returnrollers14.Clothspraysandstrainercleaning15.Control.panelproblems16.Gearboxrepairs17.Cloth.rollerdefects18.Filtratepumpsandreceivers19.Otherfilterproblems15,313,212,610,88,27,67,24,23,12,92,52,42,42,21,31,10,90,91,2100,0*Includesproblemsoncontinuousionexchange(e.g.pregnant-solutiontankfull),residuedisposal,waterandcompressed-airshortages,andmaintenancestoppagesonothersectionsGeneralOperatingDataTheinformationinTableIVwasextractedmainlyforthetwo-yearperiod1980-1981andsummarizestheOCTOBER1983JOURNALOFTHESOUTHAFRICANINSTITUTEOFMININGANDMETALLURGY

PAGE – 9 ============
SELECTEDOPERATINGDATATABLEIV1.Ratioofpregnantsolutiontosolids2.Washratio(wash:solids)3.Dailyduty4.Solubleloss5.Slimegradingsmallerthan200mesh6.Filtratesuspendedsolids7.Flocculantconsumption(guar)8.Averageclothlife9.Filteravailability10.FilterrunningtimeH.Relativedensityoffilterfeed12.Temperatureoffilterfeed13.Averagevacuumperfilter(approximate)14.Vacuumcapacityperfilter15.Vacuumperm”offilterarea16.LifeofP.T.F.E.wearstrip.(approximate)17.Lifeofrubberwearbelt.(approximate)18.Consumptionofvacuum.boxlubricationwater(approximate)19.Filterspeed(approximate)20.Timetopatchholesinfiltercloth(approx.)21.Timeforfilterrecloth(approximate)1,47:10,5:112,3t/m”3,4%73,9%304p.p.m.69g/t74,1days92,4%90,8%1,49546°C5.5kPa7000m”/h74,5m3/h3000hourst3years7m3/ht15m/minthour::I:3hours.TheH.D.P.E.-polyesterwear-beltcombinationispredictedtoextendtheselivestoabout30000hourseacht8000hourshavebeenachieved:I:0,16m3permetreofvacuumboxoperatingdata,someofwhichwerementionedearlierinthispaper.ConclusionFortheChemwesliquid-solidseparationstage,beltfiltershaveprovedtobeviablemachines.Theycan,withsomeconfidence,nowbeaddedtothelistofliquid-solidseparationequipmentthatshouldbeconsideredforanewoperation.AcknowledgementsThecontributionsofthestaffofChemwestothecom-pilationofthispaper,andthepermissiongrantedbyGeneralMiningUnionCorporationLimitedforitspublication,aregratefullyacknowledged.Reference1.VILJOEN,E.B.,etal.Theevaluation,design,andconstruc-tionoftheuraniumplantforChemwesLimited.J.S.A/r.Inst.Min.Metall.,vo!.81,no.9.Sep.1981.pp.280-287.HydrocyclonesBHRA,theFluidEngineeringCentre,willholditsSecondInternationalConferenceonHydrocyclonesattheUniversityofBath,England,from19thto21stSeptember,1984.Offersofpapersareinvited.Hydrocyclonesarewidelyusedthroughoutthechemical,processing,mining,andwaterindustriesforliquid-solidandsolid-solidseparation.Recently,specialcyclonesweredevelopedfortheoilindustrytoseparateoilandwater,andtodegascrudeoils.Thesavingsinweightandspaceofferedbytheuseofthisequipmentareparticularlyimportanttoplatformoperators,andhavestimulatedfurtherinterestinthetechnique.However,althoughhydrocycloneshavenomovingparts,theflowwithinthechamberisextremelycomplex.Forexample,designersandoperatorsneedtounderstandwhateffectchangesingeometryhaveonthevelocitydistributionandturbulencewithinthehydrocyclone.Themeetingwillcoverallaspectsofthetheory,design,anddevelopmentofhydrocyclones,aswellaspracticalproblemsandapplications.Prospectiveauthorsshouldsubmittitlesandsynopsesofpapersassoonaspossible.ManuscriptswillberequiredbyMarch1984.Immediatelypriortothismeeting,BHRAwillrunanintensivetwo-daycourseforengineerswishingtofami-liarizethemselveswiththetheoryandpracticeofhydro-cyclones.Registrationforthiscoursewillbeindependentoftheconference.Offersofpapers,andenquiriesaboutthecourse,shouldbesenttotheOrganizer,Hydrocyclones,BHRA,Cranfield,BedfordMK43OAJ,England;telephone(0234)750422;telex825059.JOURNALOFTHESOUTHAFRICANINSTITUTEOFMININGANDMETALLURGYOCTOBER1983245

255 KB – 9 Pages