Optimization of Roller-Compacted Concrete for Local Application

Size: px

Start display at page:

Download "Optimization of Roller-Compacted Concrete for Local Application"

Giles Sherman
5 years ago
Views:

1 TRANSPORTATION RESEARCH RECORD 1458 Optimiztion of Roller-Compted Conrete for ol Applition H.I. A-ABDU WAHHAB AND l.m. ASI Retly onstruted sphlt pvemts in Sudi Arbi re experiing high level of wheel trk rutting, whih is minly ttributed to the high mbit pvemt temperture, the high volume of loded truks, nd the visoelsti behvior of the sphlt pvemt. One proposed solution is to use reinfored onrete rigid pvemt. Beuse of orrosion of the reinforemt brs, ommon problem in Sudi Arbi, it is better to use roller-ompted onrete (RCC) pvemts. A omprehsive nd rtionl design of RCC mixes, suh s onvtionl onrete, is extremely omplex, beuse it is influed by numerous ftors tht depd on the soures of mterils nd their properties; methods of preprtion, plemt, omption, nd uring; nd, most importnt, previling vironmtl onditions. Inorportion of ll these ftors into RCC mix design not only requires the rduous seletion of pproprite proportions of the ingredits, but lso inuries our in the required properties of RCC tht re usully retified by tril-nd-error proedures. Effets of vrition of the wter to emt rtio, orse-tofine ggregte rtio, nd totl ggregte to emt rtio, on rollbility, dsity, nd strgth of RCC re investigted for three omptive efforts. The min objetive is to find optimum mix design proportions for lolly vilble mterils by vrying the mtioned vribles nd to develop model to predit the flexurl strgth of lol mixes. The inrese in the number of hevily loded truks using the Sudi Arbin rod network, nd the ountry's high mbit tempertures, hve resulted in wheel pth rutting. One proposed solution to this problem is to use rigid pvemt. However, beuse of humidity levels nd high wter-tble levels (espeilly in the ostl res of Sudi Arbi) nd n vironmt tht uses orrosion of reinforemt brs nd dowels, roller-ompted onrete pvemt (RCCP) my be n option. RCCP is mixture of ggregtes, binder, nd smll mount of wter, whih forms zero-slump portlnd emt onrete tht is spred nd ompted, first with n sphlt pver nd th with vibrtory roller, to obtin stisftory results. The resulting pvemt is of higher strgth nd lod-rrying pbility thn onvtionl onrete pvemt. Trditionl RCCP is like trl-mixed emt-treted bse (CTB) or soil emt. Some mjor differes tht exist betwe them re (1,2) Compressive nd flexurl strgths of RCCP re higher thn CTB, beuse of the greter ontt of portlnd emt (usully 12 to 14 pert emt for RCCP nd 6 to 8 pert emt for CTB. RCCP does not need protetive wering surfe wheres CTB needs to be overed. A true pver or lydown mhine normlly is used for pling nd finishing RCCP. RCCP is designed to resist brsion from trffi, to remin durble under severe wether exposure, nd to hve stisftory surfe finish nd stright-edge tolerne for the trffi served. Civil Engineering Deprtmt, King Fhd University of Petroleum nd Minerls, Dhhrn 31261, Sudi Arbi. The first highly publiized RCCP ws built in Cnd by the British Columbi Forest Produts Compny on Vnouver Islnd t Cyuse Dry og Sort in summer Beuse of the ese nd simpliity of onstrution, nd tht it's possible to sve one-third or more of the ost of onvtionl onrete pvemt onstrution (3), this new tehnology is thing on so fst tht, to dte, severl RCCPs hve be built in other prts of Cnd, the United Sttes, Spin, Swed, nd Austrli (4-7). Some dvntges of RCCP re its high flexurl strgth (25 pert higher thn CTB) nd its high resistne to brsion nd fuel nd hydruli spills. Tires turning in hot wether hve no effet on it, nd it is less ffeted by summer temperture. Finlly, it hs the dvntge of redued onstrution time nd ost (1-6). However, beuse RCCP tehnology is in its infny, stndrd proedures for proportioning nd fbriting smples re virtully nonexistt. Commonly used proedures re bsed on n rbitrry seletion of mix prmeters. Bthes re prepred using tril-nd-error proedures, nd djusted to hieve the required dsity, rollbility, nd strgth. Some smples hve be fbrited by onsolidting them on vibrting tbles (2) nd by using modified protor ( 8) nd Kngo hmmer (5). This study pplied ftoril experimt des_ign, whih filittes nlysis of the ombined effets of two or more vribles nd their intertions in terms of dsity, rollbility, n;l strgth-with view to optimizing mix designs. Smples were fbrited using Cliforni kneding omptor. MATERIAS The three min ingredits of RCCP re ggregte, portlnd emt, nd wter. The seleted mterils re desribed in the following setions. Aggregte Composition nd vrition of the ggregte is n importnt ompont of ll onretes, prtiulrly of roller-ompted onrete (RCC). Severl onrete properties, suh s dsity nd surfe texture, re ffeted by prmeters suh s the orse-to-fine ggregte rtio nd the orse-to-totl-ggregte rtio. If the ggregtes. re well proportioned, high dsity nd impermeble RCC is formed. RCC hs low wter ontt nd therefore is likely to segregte. To void segregtion, mximum ggregte size should not exeed 22 mm Ys in.) (9) nd should ontin high proportion of fines. However, if the pvemt is to be lid in lifts, the bse lift ould hve n ggregte size of up to 35 mm (1.5 in.). At lest 66 pert of the ggregte should be rushed stone ( 6 ). Aggregte grdtions used in Cnd nd Spin re indited in Figure 1. The Cndin limits pper to hve wider bnd, lthough

2 2 TRANSPORTATION RESEARCH RECORD 1458 z (.!) u:; 6 er: C 5 1- z ~ 4 :: U C o CANADA A SPAIN o USED GRAOAT ION C CA/FA= 1 ) * USED GRAOAT I ON C CA/FA= 1. 4) SIEVE OPENING (mm) FIGURE 1 Aggregte grdtions used in Cnd nd Spin. they pper to hve ommon limits, rnging from.75 to 19 mm (.3 to.75 in.). All grdtions were hos to minimize segregtion in the mix. In this reserh, orse nd fine ggregtes were olleted from lol res. Corse ggregte ws olleted from the Al-Summn re ner Riydh. The re is pottilly lrge soure of ggregte. The ggregte (limestone) is exvted nd rushed into vrious sizes to hve ngulr shpe nd rough texture. The fine ggregte ws nturl snd nd ws olleted ner Dhhrn. The two ggre- gtes were proportioned to form two grdtion velopes with differt orse-to-fine ggregte (CA/FA) rtios (Figure 1). Those grdtions were hos to fll within the speified limits shown in Figure 1 nd ontin ough fines to prevt segregtion. The ggregte ws subjeted to series of qulity tests in ord with ASTM stndrd speifitions for onrete ggregte ASTM C33. Test results nd ASTM limits re prested in Tble 1. Results indite tht the seleted ggregte stisfies ASTM requiremts. Cemt Contt Cemtitious ontt by weight of ggregte should be betwe 12 nd 14 pert (2). It my be mde up of portlnd emt nd dditives suh s fly sh. Beuse wer resistne is not ftor in the bottom lift of multiple-lift onstrution, redution of emt ontt from 6 to 8 pert by weight of ggregte is reommded (7). In ses where sulphte exposure is not problem, ny vilble portlnd emt, (with the exeption of Type III portlnd emt) or ombintion of portlnd emt nd pozzoln, or blded hydruli emt with pozzoln, n be used. If sulphte exposure is problem, Type II, Type V, or moderte sulphteresistnt blded hydruli emt should be used (2). Beuse sulphte ttk is mjor problem ffeting onrete durbility in the ostl res of Sudi Arbi, ASTM Type V emt ws used in ll the mixes. Wter Contt Wter ontt is proportioned in suh wy tht the resulting RCC hs zero-slump nd mximum dsity under omption. The TABE 1 Summry of Test Results on Aggregtes Aggregte Property Test Results Type Grding 3/4" mximum size grding speified in literture - Speifi grvity Bulk speifi grvity ov dry bsis = Bulk speifi grvity (SSD) bsis = Apprt speifi grvity = Corse Absorption 24-hour = l.26% - Abrsion loss 23.25% 4% ASTM imit Prese of ly lumps nd frible prtiles.74% 3% to 1% Soundness.58% 12% Grding Fine modulus = l Fine Speifi grvity Bulk speifi grvity ov dry bsis = Bulk speifi grvity (SSD) bsis = Apprt speifi grvity = Absorption 24-hour =.77% - Prese of ly lumps nd frible prtiles #8 =.45% #16 =.18% 3% Soundness 1.38% 1%

3 Whhb nd Asi 3 method outlined in ASTM D558 (ASTM 1988h) trets RCC s soil emt insted of onrete nd thus estblishes reltionship betwe moisture nd dsity for n pplied omptive effort. Abrms nd Jksh (1) hve suggested wter emt (W IC) rtio in the rnge of.3 to.4. Results of initil mix trils indite tht this rnge is too nrrow for lol ggregtes. The WIC rtio rnge ws therefore djusted to rnge of.3 to.5. In this reserh, potble wter vilble from lbortory fuet ws used in ll stges of testing, preprtion, nd uring of the onrete. EXPERIMENTA PROGRAM RCC for pvemts must ttin onsisty similr to soil emt yet still possess some fetures of onvtionl onrete, suh s strgth nd durbility. In the design of mix, prtiulr tttion should be giv to the three most importnt ftors ontrolling the qulity of onrete: W IC rtio, CA/FA rtio, nd totl ggregte to emt rtio (T A/C). To hieve the pproprite mix design for rollbility, dsity, nd strgth (omprehsive nd flexurl), trils usully hve to be onduted tht re not only lborious but lso time-onsuming. One objetive of this study is to investigte severl mix proportions in terms of rollbility, strgth, nd dsity, for three differt omptive efforts. To test for the differt vribles, five levels of WIC (.3 to.5), four levels of TA/C (5.4 to ), nd two levels of CA/FA (1. nd 1.4) were hos, s shown in Tble 2. By using these levels, ftoril experimt ws designed to explore the effet of the ftors on rollbility, dsity, nd flexurl strgth of RCC nd the intertions mong the ftors. Typil RCC mix proportions for four ountries re shown in Tble 3. The rtionle behind ountry's seleted the mix proportions reltes to mix's strgth nd serviebility, the vilbility of onstitut mterils, nd ost. A Hobrt eletrilly operted, lbortory mehnil mixer with pity of.12 m 3 (732 in. 3 ) ws used to mix the ingredits for RCC. Mixing ws done in bthes; wter ws the lst ingredit dded. The wter nd other ingredits were mixed vigorously for 3 min to redue the hne of segregtion. The Ronny Anderson method (5) ws used to mesure rollbility (the bility to roll or ompt the zero-slump onrete with resonble effort while mintining its homogeity without segregtion). In the Ronny Anderson method, 2.5 kg (5.5 lb) of RCC is pled into ylindril form. A 13.3-kg (2.3-lb) weight is lifted into ple, nd vibrting tble is run for 2 se. The rte t whih the weight sinks (the rte of omption) my be used s mesure of rollbility. Currtly there re no stndrd proedures for mking test speims. The uthors deided to use Cliforni kneding omptor to ompt the test speims beuse it is the best ms of simulting onstrution omption (similr to ASTM D setup). The kneding omptor ws librted to find suitble foot pressure for hieving the highest dsity with miniml degrdtion. This pressure ws found to be 1551 kp (225 psi). The RCC from eh mix ws st into three 4 X 1 X 1-mm (16 X 4 X 4 in.) bems using single lyer; the bems were th subjeted to three hos omptive efforts orresponding to 5, 1, nd 15 tmps, respetively, t foot pressure of 1551 kp (225 psi). Initil onsolidtion ws hieved by subjeting the smple to stti lod of 621 kp (9 psi) before it ws trnsferred to the omptor. Applying stti lod not only onstitutes initil omption but is lso ms of obtining reltively flt finished bem surfe. The speims were overed with wet burlp for 24 hr nd ured for 28 dys in uring tnk filled with wter, per ASTM Stndrd C192. RESUTS Within the rnge of WIC rtios, ll smples were within n eptble rnge of rollbility. Figures 2 to 7 show the effets of vrying W IC, C, CA/FA, nd number of tmps, on flexurl strgth nd mix dsity. Results indite tht for the WIC rtio used, the 28-dy flexurl strgth improves by inresing the W IC rtio to mximum vlue, th it strts deresing. This finding is ttributed to the ft tht, in the erly stges, emt mortr ts s filler mteril, filling the voids betwe the ggregtes until ll voids re filled. The sme is true for dsity. The mximum flexurl strgth ws 6.7 MP (971.5 psi), orresponding to WIC rtio of.45 nd CA/FA TABE 2 Shemti Represttion of the Experimtl Design evels of A evels of C evels of B (W IC) (CA/FA) (TAIC) b11 1b2C1 1b3C1 1b41 1b5C b1Ci,b2Ci,b3Ci,b4Ci 1b5C b13 1b2C3 1b3C3 1b4C3 1b5C b14 1b2C4 1b34 1b4C4 1b5C4 1b1C1 2b2C1 1b3C1 2b4C1 1bs b1Ci 1b2Ci 1b3Ci 1b4Ci 1b5Ci 5.8 1b1C3 2b 2~ 1b3C3 1b4C3 1b5C3 5.4 ib1c4 ib2c4 1b3C4 ib4c4 ib5c4 --rbree smples were fbrited nd tested for eh ell.

4 4 TRANSPORTATION RESEARCH RECORD 1458 TABE 3 Typil Mix Proportions for Four Countries Aggregte Portlnd Cemt Type I Fly Ash Wter CANADA 348 lblyd lb/yd 3 9 lb/yd lb/yd 3 CA/FA = 1.86 TA/C = 6.5 W/C =.35 Averge Compressive Strgth = 457 psi (31.5 MP) t 28 dys SPAIN Grvel 1 to 3 mm 39 lb/yd 3 CA/FA = 1.38 Grvel to 1 mm 275 lb/yd 3 Snd to 5 mm 4 lb/yd 3 TA/C = 8.5 ASTM Clss C Fly Ash 15 lb/yd 3 Ctegory P-45 Cemt 29 lblyd 3 W/C =.42 Wter 185 lb/yd 3 Averge Compressive Strgth = psi (2-25 MP) t 28 dys SWEDEN Grvel 16-3 mm 573 lb/yd 3 CA/FA = 1.5 Grvel -16 mm 276 lb/yd 3 Snd - 5 mm 691 lb/yd 3 TA/C = 5.26 Cemt 635 lb/yd 3 Wter 219 lb/yd 3 W/C =.34 Averge Compressive Strgth = 58 psi (4 MP) fter 28 dys Grvel Snd Cemt Type I Pozzoln (Clss F) Wter Aggregte Cemt-Pozzoln Type IP Wter U.S.A. [Portlnd Airport] 155 lb/yd 3 17 lb/yd lb/yd lb/yd 3 26 lb/yd 3 Averge Flexurl Strgth = 71 psi t 28 dys U.S.A. [Multnomh County Rd.] 339 lb/yd 3 55 lblyd lb/yd 3 CA/FA= 1.17 TA/C = 5.35 W/C =.43 CA/FA=.75 TA/C = 6.2 W/C =.44 Averge Flexurl Strgth = 6 psi (4.1 MP) t 28 dys 1 kg/m 3 = lb/yd 3 CA/FA = Corse to Fine Aggregte rtio T A/C = Totl Aggregte to Cemt rtio W/C = Wter to Cemt rtio rtio of 1.4, for mixture of 15 pert emt with 15 tmps. Flexurl strgth of norml onrete prepred using the sme ggregte grdtion t n optimum W IC rtio gertes vlue of 4.2 MP (612. psi). Tht indites n inrese of bout 37 pert in the flexurl strgth of RCC bove tht of norml onrete. The totl ggregte emt (T A/C) rtios used in the mixes were 5.4, 5.8, 6.4, nd. For ll WIC rtios, n inrese in the totl ggregte resulted in derese in strgth. At lower T A/C rtios, the volume of the emt pste is greter, nd it plys greter role in improving strgth. On the other hnd, inresing the CA/FA rtio inreses gps betwe the ggregte, llowing the emt to ot the ggregte effetively. Tht resulted in the mixes' inresed flexurl strgth. To hek tht the studied vribles (TA/C, CA/FA, nd WIC) signifintly ffet flexurl strgth nd dsity, n nlysis of vrine ws performed on ll test results. Anlysis of vrine determined tht the soure of vrition due to T A/C, CAIF A, nd W IC is highly signifint wh ompred with vlues of F (vrine rtios) t 95 pert probbility level (.5 signifine level). Tht finding indites the underlying influe of suh vribles on the strgth nd dsity of RCC. Sttistil methods were used to fit model betwe the obtined flexurl strgth of the RCC mixes nd the pplied vribles (CA/FA, TA/C, WIC, nd number of tmps). Figures 2 to 4 indite tht the reltion betwe flexurl strgth nd W IC rtio hs the shpe of sine funtion, but for the other vribles the reltion is liner. The resulting model is Flexurl strgth= (TNC) sine W/C - 3) 'IT] [( _ 15 * 2 +l.94 (CNFA) +.59 (no. of tmps) (1) The model hs oeffiit of determintion of 95.6 pert. A stepwise seletion nlysis indites tht the importnt vribles ffeting flexurl strgth re T A/C, W IC, CAIF A, nd the

5 Cemt o I J. Cemt.. ~ D. 14 Cemt o 15 Cemt :5 4.5 Cl..., x 3. u _3> FIGURE 2 Reltion betwe W/C rtio nd flexurl strgth for omption effort of 5 tmps. FIGURE 5 Reltion betwe W/C rtio nd dsity for omption effort of 5 tmps , :::E 5. :5 Ol 4.5.,., x 3. u )I( 12 lj D o (')E ' Ōl.:x. _3> )I( 12 o IJ D FIGURE 3 Reltion betwe W/C rtio nd flexurl strgth for omption effort of 1 tmps. 22-~~~--~~~~-~~~--'-~~~~'---~~~~~ FIGURE 6 Reltion betwe W/C rtio nd dsity for omption effort of 1 tmps , :::E 5. :5 Ol.,., x 3. u )I( 12 lj D ' Cemt o 13 Cemt D FIGURE 4 Reltion betwe W/C rtio nd flexurl strgth for omption effort of 15 tmps 22-~~~..-~~~-'--~~~-'-~~~-'-~~~--'-~~ FIGURE 7 Reltion betwe W/C rtio nd dsity for omption effort of 15 tmps.

6 6 TRANSPORTATION RESEARCH RECORD 1458 TABE 4 Fitted Model Sttistis Seletion: Forwrd Control: Mnul Stepwise Seletion Mximum steps: 5 Step: 4 F-to-ter: 4. F-to-remove: 4. R-squred: Adjusted: MSE: d.f.: 115 Vribles in Model Coe ff. F-Remove Vribles Not in Model P. Corr. F-Enter 1. TA/C 2. W/C 3. CA/FA 4. TAMPS Model Fitting Results Indepdt vrible oeffiit std. error t-vlue sig. level CONSTANT TAIC W/C CA/FA TAMPS R-SQ. (ADJ.)=.9564 SE= MAE= DurbWt = Previously: observtions fitt~, forest(s) omputed for missing vl. of dep. vr. + W/C=.3 C ~ 6 o W/C=.35 t:. :5 W/C=.4 Ol o 5 W/C=.45 _;; lle W/C=.5 x - 3 TI 2 u TI 2 C t Atul Flexurl Strgth. MP olle FIGURES Comprison betwe predited nd tul flexurl strgth. For the lol ggregtes, W/C rtio rnging from.3 to.5 produes mixes with eptble rollbility. Beuse of its lower wter ontt, the flexurl strgth RCC is bout 37 pert higher thn the flexurl strgth of norml onrete. Sttistil nlyses revel tht the effets of T A/C nd CA/FA rtios on flexurl strgth nd dsity re highly signifint for ll W IC rtios studied. Gin in strgth betwe 1 nd 15 tmps is not so signifint to justify the effort. The model is desriptive of the strgth dt developed for these prtiulr ggregtes within the rnge of the experimtl vribles. 7 ACKNOWEDGEMENT The uthors thnk the Ministry of Communitions nd the King Fhd University of Petroleum nd Minerls for providing vluble support. number of tmps, respetively. Stepwise nlysis nd model fitting results re shown in Tble 4. The model is pplible to lol mterils nd n be used for the rnges of ll vribles indited in Tble 2. Figure 8 shows the reltion betwe mesured flexurl strgth nd the model predited vlues nd indites tht the predited vlues re lose to the mesured ones. CONCUSIONS Chos lol ggregtes meeting the requiremts of ASTM C33 n be used for fbrition of qulity RCCP. REFERENCES 1. Al-Otibi, M. H., A. M. Smd, S. A. Shibt, nd F. F. Wf. Roller Compted Conrete Pvemt; Mterils Chrteriztion. Pro., 3rd Sudi Engineering Confere, Vol. 1, Riydh, Sudi Arbi, Nov. 1991, pp Pittmn, D. W., nd S. A. Rgn. A Guide for Design nd Constrution of Roller-Compted Conrete Pvem,ts. Finl Report prepred for US Army Corps of Engineers, Wshington, D.C., Nov Pittmn, D. W. Constrution of Roller-Compted Conrete Pvemts. In Trnsporttion Reserh Reord 162, TRB, Ntionl Reserh Counil, Wshington, D.C., 1986, pp Abrms, J.M., J.. Jksh, R~. Norton, nd D. J. Irvine. Roller Compted Conrete Pvemt t Portlnd Interntionl Airport. In

7 Whhb ndasi 7 Trnsporttion Reserh Reord 162, TRB, Ntionl Reserh Counil, Wshington, D.C., 1986, pp Anderson, R. Swedish Experies with RCC. Conrete Interntionl: Design nd Constrution, Vol. 9, No. 2, Feb. 1987, pp Jofre, C., A. Jos, nd F. Molin. The Pving of ow-volume Rods in Spin with Roller-Compted Conrete. In Trnsporttion Reserh Reord 116, TRB, Ntionl Reserh Counil, Wshington, D.C., 1987, pp Piggot, R. W. Roller-Compted Conrete for Hevy-Duty Pvemts: Pst Performne, Ret Projets, nd Reommded Constrution Methods. In Trnsporttion Reserh Reord 162, TRB, Ntionl Reserh Counil, Wshington, D.C., 1986, pp ogie, C. V., nd J.E. Oliverson. Burlington Northern Rilrod Intermodl Hub Fility. Conrete Interntionl: Design nd Constrution, Vol. 9, No. 2, Feb. 1987, pp Piggot, R. W., nd.. Ns. Roller-Compted Conrete Pvemts in British Columbi, Cnd. Pro., Symposium on Roller Compted Conrete, ASCE, Dver, Colordo, My 1-2, 1985, pp Abrms, J. M. nd J.. Jksh. An Airport Apron nd Country Rod. Conrete Interntionl: Design nd Constrution, Vol. 9, No. 2, Feb. 1987, pp Publition of this pper sponsored by Committee on Performne of Conrete.

CS 7700 FULL FUNCTION LOBBY TERMINAL REAR LOAD - 15" OR 19" DISPLAY

CS 7700 FULL FUNCTION LOBBY TERMINAL REAR LOAD - 15 OR 19 DISPLAY L0 LL -800--600 R S 00 FULL FUNTION LOBBY TERMINL RER LOD - " OR " DISPLY mm ( ") UL SFE ONSULT WITH DIEBOLD INSTLLTION/SERVIE BRNH FOR DDITIONL DETILS ND INFORMTION. PLESE SEE PLNNING ND SITE PREPRTION