Lateral distribution of static loads in a prestressed concrete box-beam bridge - drehersville bridge, August 1966

Transcription

1 Lehigh University Lehigh Preserve Fritz Labratry Reprts Civil and Envirnmental Engineering 1966 Lateral distributin f static lads in a prestressed cncrete bx-beam bridge - drehersville bridge, August 1966 W. J. Duglas D. A. VanHrn Fllw this and additinal wrks at: Recmmended Citatin Duglas, W. J. and VanHrn, D. A., "Lateral distributin f static lads in a prestressed cncrete bx-beam bridge - drehersville bridge, August 1966" (1966). Fritz Labratry Reprts. Paper This Technical Reprt is brught t yu fr free and pen access by the Civil and Envirnmental Engineering at Lehigh Preserve. t has been accepted fr inclusin in Fritz Labratry Reprts by an authrized administratr f Lehigh Preserve. Fr mre infrmatin, please cntact preserve@lehigh.edu.

2 Prest:ressed Cncret:e B)(~Beam Bridges Prgress Reprt NOe LATERAL DS 11 F STATC l A A PRESTRESSE E BOX-BEA E DREHERSVLLE E by Wali:er J. Duglas David A. VanHrn Fritz.Engineering Labratry Reprt N04) 31501

3 PRESTRESSED CONCRETE BOX-BEAM BRDGES PROGRESS REPORT NO. 1 LATERAL,DSTRBUTON OF STATC LOADS N A PRESTRESSED CONCRETE BOX-BEAM BRDGE DREHERSVLLE BRDGE Walter J.. Duglas David A. VanHrn Part f an nvestigatin Spnsred by: PENNSYLVANA DEPARTMENT OF HGHWAYS U. S. DEPARTMENT OF COMMERCE BUREAU OF PUBLC ROADS RENFORCED CONCRETE RESEARCH COUNCL Fritz Engineering Labratry Department f Civil Engineering Lehigh Universi~y Bethlehem, Pennsylvania August 1966 Fritz Engineering Labratry Reprt N

4 TAB L E F CON TEN T S ABSTRACT NTRODUCTON,1.1 Backgrund 1.2 Object and Scpe 1.3 Previus Research TESTNG 2.1 Test Bridge 2.2 Gage Sectins and Lcatins 2.3 Strain and, Deflectin Gages 2.4 Test Vehicles 2.5 Lading Lanes 2.6 Timing and Psitin ndicatrs 2.7 Test Runs DATA REDUCTON AND EVALUATON Oscillgraph Trace Reading 3.2 Evaluatin f Oscillgraph Data, Strain Distributin and Lcatin f Neutral Axes Effective Slab Widths Girder Bending Mments Distributin Cefficients

5 Page 4. PRESENTATON OF TEST RESULTS Distributin Cefficients Distributin Factrs Sam.ple Girder Mments Girder Deflectins Transfrmed Effective Widths DSCUSSON OF RESULTS Experimental Strains and Neutral Axes Deflectins ' Transfrmed Effective Width Mdulus f Elasticity Cmparisn f Design and Experimental Live Lad Mments Distributin Cefficients Distributin Factrs SUMMARY AND CONCLUSONS 39 6,1 Summary Cnclusins ACKNOWLEDGMENTS TABLES FGURES, REFERENCES 95

6 A B S T RAe T This reprt describes the field testing f the first f five beam-slab highway bridges included in an investigatin f the structural behavir f bridges supprted with prestresied cncrete bx girders, and subjected t lading with test vehicles apprximating AA8HO H lading. The test structure was ne f the end spans f a three-span simply supprted bridge with a cast-inplace cncrete deck supprted by five precast prestressed cncrete bx girders laterally spaced at 7 ft. 2 in. The test span was 61 ft. 6 in. in length. The testing prgram cnsisted f the cntinuus recrding f girder deflectins and surface strains at varius lcatins n the girders, slab, curb, and parapet, as either ne r tw f the test vehicles were driven ver the test span at a speed f apprximately 2 mph. The principal bjectives were t develp infrmatin n the distributin f vehicle lads t the girders, and t evaluate varius field test techniques fr use in the fllwing studies. The measured strains were used t determine distributin factrs at a crss-sectin near midspan~ t was fund that the distributin factrs derived frm the field measurements were substantially different frm thse used in the design~ The test results indicated that full cmpsite actin was develped between the bridge deck, curb, and parapet. t was fund that the results f superimpsed single-vehicle runs clsely crrespnded t the results btained frm tw-vehicle runs. Als, it was determined that tests n a partially gaged

7 -2- crss-sectin culd be cmbined t simulated tests n a fully gaged sectin. This reprt cvers test runs at crawl speeds. n additin, runs were cnducted at speeds up t 34 mph. The results f the tests invlving speed runs will be included in a secnd reprt.

8 -3-1. N T ROD D, C T ON 1.1 BACKGROUND Ever since the cnstructin f the first prestressed cncrete bridge in the United States in 1950, the.walnut Lane Bridge in Philadelphia, there has been much develpment in the technlgy related t prestressed cncrete bridge cnstructin 0 One f these develpments was the design f girder and slab bridges utilizing precast, prestressed cncrete girders, alng with a cast-in-place cncrete slab. Over the past few years, the Pennsylvania Department f Highways has been develping bridge designs incrprating the prestressed cncrete bx girder. nitially, the bx shape was utilized in the cnstructin f the adjacent bx girder bridge. Lateral pst-tensining and shear keys between the adjacent girders aided in the lateral distributin f vehicle lads t the several girders cmpsing a sectin A wearing surface was applied directly t the tp f the girders t serve as the deck A later develpment f this initial design apprach was the cas~-in-place reinfrced cncrete slab The slab, acting cmpsitely with the girders, replaced lateral pst-tensining and shear keys as a methd f lateral lad transfer. The latest design is the spread bx girder bridge, in which the girders are equally spaced and spread apart t act as T-beams with the cast-in-place slab.

9 OBJECT AND SCOPE The current prcedures utilized by the Pennsylvania Department f Highways in the design f prestressed cncrete bridges are. 11 set frth in the PDH Bridge Divisin Standards 8T-200 thrugh ST-208, n these standards, the interir girders f the prestressed cncrete bx girder beams have been prprtined accrding t a live lad distributin factr f 8/5.5 where S is the average girder spacing. This factr is identical t the factr given in the AASHO:Specificatins,l Sectin 3, gverning the design f a cncrete slab laterally cntinuus ver interir steel -Beam stringers. Fr the exterir girder, the live lad is als distributed accnding t the general prcedure set frth in the AASHO Specificatins. This prcedure is based n the assumptin that the slab acts as a simple span between girders in transmitting wheel lads laterally. These design prcedures were instituted since it was bvius that the earlier criteria develped fr the design f the adjacent bx girders was nt applicable t the design f the spread bx girder bridges 0 Cnsequen'tly, the prcedures fr live lad distributin were based n a cnservative estimate f the expected distributin based n the prvisins gverning the varius members cvered by the AASHO Specificatins 0 n 1964, the Structural Cncrete Divisin f the Department f Civil Engineering at Lehigh University initiate:d an investigatin f the actual lad distributin characteristics fr this type f bridge 0 The results f the study are t be cmpared with current

10 -5- design practice, and a design prcedure reflecting actual behavir is t be develped. The verall investigatin has been divided int several phases. The first phase was the field test f an existing bridge, lcated near Drehersville, Pennsylvania. This phase was (1) t serve as pilt test fr additinal field tests, and (2) t prvide experimental data fr use in develping a methd f analysis fr use in design. This reprt will cver the respnse f the. Drehersville test structure t crawl-run lading. This type f lading is, explained in Chapter 2~ A secnd reprt will cver the behavir f the.drehersville structure under speed-run lading. The principal bjectives f this first phase were (1) t determine the actual live lad distributin factrs fr bth exterir and interir girders, and (2) t cmpare lading techniques, gaging patterns, and interpretatin f data in evaluating experimental results. Other aspects f the study were a cmparisn f experimentally determined bending mments with thse used in the design f the girders, the determinatin f girder deflectins, and the study f the behavir f bridge deck, curb, and parapet in resisting applied vehicle lads a Field testing was cnducted with the Bureau f Public Rads field test unit, cnsisting. f a lading truck and mnitring,trailer. T supplement this equipment, an additinal truck was prvided. Test runs acrss the bridge were made by directing a crawling truck alng ne f several lanes, apprximately equally spaced

11 -6- acrss the width f the deck. The centerline f each f these lanes crrespnded either t the centerline f a girder r t a line midway between girder centerlines. Data was btained frm gages lcated at tw lateral sectins. One sectin was lcated t measure maximum mment respnse, and the ther was lcated t measure respnse with nly the rear axle f the truck n the span. n additin t test runs cnducted with single test trucks, simultaneus runs were made with bth f the trucks side-by-side PREVOUS RESEARCH A number f field studies have been cnducted n bridges ~nstructed f cncrete slabs supprted by -shaped girders, either f steel r prestressed cncrete. Since the end f Wrld War, testing prcedures have becme mre sphisticated, especially in the areas f measuring and recrding structural respnse. n field tests cnducted by Hindman and Vandegrift 6 in Ohi in 1945, a study f lad distributin was based entirely n measured girder deflectins. Test lading was achieved by the upward thrust f a hydraulic 5 jack applied t ne girder at a time. A 1952 reprt by Fster n field studies cnducted in Michigan describes the use f SR-4 strain gages n the main girder flanges, and n small cantilever beams used t measure deflectins, Static strain gage measurements were btained with a prtable indicatr, while dynamic measurements were permanently recrded n pht sensitive scillgraph paper. Rather than a single cncentrated lading, a test truck, alng with a simulated truck laded t apprximately H20-S16-44 specificatins, were used.

12 n 1956, Hlcmb 7 reprted a series f field tests cnducted n tw bridges in wa, n this study, a 48-channel autmatic strain recrding device was utilized fr bth static and. dynamic runse Strains were measured n the web as well as the girder flanges. -7- Deflectins were measured by dial gages. The test vehicle, a semitrailer ttaling 98,000 lbs., did nt simulate AASHO truck lading. A 1957 reprt by White and PurnellDdescribed the field test f a cmpsite haunched girder bridge in Texas. One f the imprtant features f this test was the lcatin f SR-4 strain gages n the girder in rder t determine the psitin f the neutral axis. The ease and speed with which field measurements can be made and recrded was imprved with the Bureau f Public Rads field test equipment trailer, which huses autmatic recrding equipment. This trailer was first used in A tabulatin by Varney and Galambs1 4 f field tests cnducted between 1949 and 1965 includes many studies which have utilized the BPR recrding equipment. Studies f the dynamic respnse f a bridge in wa were reprted in 1958 by Prentzas,12 and describe 36-channel simultaneus recrding equipment used in the testing f tw bridges. n these tests, strain gages were munted n bth the tp and bttm flanges f the steel girders. n 1964, dynamic tests by Reilly, Guardia and Lney13in Maryland reprt the expansin f the BPR equipment t 48-channel capacity, with the additin f a truck simulating H20 S16-44 lading. A cmparable testing prcedure was reprted the same year fr dynamic studies cnducted in Virginia. 9

13 -8- Several recent tests have been cnducted n -shaped pre- s t ressed cneret e glr der spans. Hulsbs and Ll'nge~8,lO ~ ~eprted ~ a series f wa field tests f bridges, including ne cnstructed f five prestressed cncrete girders 0 Static lading was cnducted with the test vehicle crawling acrss the span instead f being parked n the span.. An extensive pattern f five gages was psitined n the face f a girder fr experimental neutral axis determinatin. The Maryland tests,13 mentined abve, included extensive testing f a nine girder prestressed cncrete span. The testing prcedures used in previus field wrk, especially thse utilized by Hulsbs and Linger,lO aided in the planning f the field tests included in the current investigatin at Lehigh University. The gaging, recrding, and lading prcedures were adpted frm the many recent tests described in the cmpilatin 14 by Varney and Galambs.

14 -9-2. T E 8 TN G 2.1 TEST BRDGE The final selectin f a bridge fr the pilt study necessitated the meeting f several structural and site requirements. The structural requirements were that the simply supprted span be f medium length, 60 t 70 feet. This span range is typical f the spread bx girder bridges built in Pennsylvania. The girders were t be at right angles with the piers and abutments, with minimum superelevatin f the radway. T allw fr maximum speed runs, minimum grade and tangent radway were required 0 The site requirements principally invlved the access t a nearby pw~r supply, and the existence f a space which wuld permit parking f the instpument trailer near t the test span. The structure which mst clsely cnfrmed t the abve requirements was the nrthwest span f the three-span bridge illustrated in Fig. 1. This bridge spans the Little Schuylkill River near Drehersville, Pennsylvania, and is lcated n Legislative Rute feet 6 inchese The test span was simply supprted with a length f The skew was 90, and the bridge deck was n a maximum grade f 0.2%, with nrmal crwn0 The sutheast apprach was n a superelevated curve, and slped dwnward tward the river, allwing speed runs up t 34 mph. Frm the nrthwest, the apprach was steep and shrt, and crssed ver a single railrad track apprximately 100 feet frm the nrthwest abutment f the bridge, Therefre,

15 -10- eastward speed runs were severely restricted. The crss-sectin f the bridge, alng with the girder designatin, is shwn in Fig. 2. The five identical prestressed hllw bx girders, which are 48 inches wide and 33 inches deep, are equally spaced at 86 inche?, center-t-center. Cast-in-place cncrete diaphragms, 10 inches in thickness, are lcated between the beams at the ends f the span and at midspan The reinfrced cncrete deck, prviding a radway 30 feet in width, was cast-in-place cmpsitely with the girders. The specified minimum thickness f the slab was 7-1/2 inches. Hwever, measurements taken near midspan indicate that the slab thickness actually varies frm 6.2 t 7.6 inches, with an average f 6.7 inches. The safety curb, which is indicated in Fig. 2 abve the lwer dashed line, is cmpsed f a S-in. wide parapet n tp f a 33-in. wide curb sectin. The jint between the slab and the curb was a cnstructin jint with a raked finish. Vertical reinfrcement fr the curb sectin extended thrugh the jint int the slab. Basically, the reinfrcement cnsisted f three N 5 bars laterally psitined acrss the jint, at a lngitudinal spacing f 15 inches. Fr typical details f the jint, as well as ther typical details f the bridge structure, see the PDH Bridge Divisin. Standards fr prestressed cncrete bridges. l1 n the design f the bridge the girders were basically prprtined fr AASHO H20-Sl6-44 lading. Fr the interir girders, a distributin factr f 8/5.5 = was used, as cmpared t the factr f used fr the exterir girders~ The impact factr

16 was The specified minimum 28-day cylinder strength f the girder cncete was 6000 psi; hwever, average 7020 psi fr the five girders. test cylinders indicated an All f the girders were pre-tensined with 46 7/16-in. seven-wire strands. 2.2 GAGE SECTONS AND LOCATONS Tw lateral crss-sectins were selected fr strain gage applicatinq These crss-sectins, specified as Sectins M and N, are designated n the elevatin view f the bridge, Fig Sectin M was lcated 3.55 feet west f midspan in rder that maximum girder mments wuld be prduced as the test vehicle drive axle passed ver this sectin when prceeding westward. Sectin N was lcated feet west f midspan. When the test truck prceeded westward, measurements culd be made as the rear axle f the vehicle passed ver this sectin, t study the effect f a single axle lading n the span. Tw gage cnfiguratins were used during field testing. The first cnfiguratin included gages at bth Sectins M and N At Sectin M, Girders C, D, and E were gaged as shwn in Fig. 3a, with tw additinal check gages per girder, n the bttms f Girders A and B, fr cmparisn with strains n Girders E and D Five additinal gages were applied at Sectin M, three n tp f the slab abve Girders C, D, and E, and tw n the bttm f the slab midway between Girders C, D, and, E. At Sectin N, Girders C, D, and E were gaged accrding t Fig. 3b, with tw check gages n

17 -12- Girders A and B, as at Sectin M Fr the secnd gage cnfiguratin, nly Sectin M was gaged. All five girders were gaged as illustrated in Fig5 3a. Additinal gaging at Sectin M cnsisted f three lngitudinal gages n tp f the slab abve Girders C, D, and E; fur transverse gages n the bttm f the slab perpendicular and adjacent t the tps f Girders C, D, and E; tw lngitudinal gages n the bttm f the slab adjacent t the sides f Girder D; and lngitudinal' gages n the utside f the nrth curb, tp f the nrth parapet, and tp f bth curbs 5 Bth gage cnfiguratins included five deflectin gages munted n the bttm f each girder at Sectin M~ The deflectin gages are shwn in Fig. 7, and will be referred t as deflectmeters 2.3 STRAN AND DEFLECTON GAGES Baldwin-Lima-Hamiltn SR-4 resistance-type strain gages, Type A-9-3, were used t gage all f the lcatins described abve, with the exceptin that the three gages applied t the tp f the slab were Type A-ge Each gage lcatin was grund and sanded smth, fllwed by thrugh cleaning with acetnee SR-4 cement was then used t seal the cncrete surface, and t prevent electrical grunding f the gagee After prper drying f the cement, strain gages were dipped in the same SR-4 cement, applied t the prepared surface, and allwed t drye Gages applied t the rain-expsed surfaces f the radway and safety curb, were waterprfed with Gage Kate-5. The waterprfing was then cured with a prtable heat lamp~

18 -13- The deflectmeters cnsisted f a strain gage bnded t a flexible, triangular aluminum plate. The aluminum plate was attached t a bar which, in turn, was clamped alng the bttm surface f the beam. A detail f the gage munting is illustrated in Fig. 7. A wire, cnnected t a OO-lba weight placed belw the gage in the river bed, was attached t the apex f the plate in a deflected psitin. The deflectmeter was calibrated s that the flexural strains f the plates culd be cnverted t girder deflectins. n additin t the active strain and deflectin gages described abve, there were crrespnding temperature cmpensatin gages lcated near each gage lcatin. Each active gage and temperature cmpensatin gage was cnnected t ne f the 48 channels f mnitring equipment in the BPR equipment trailer. Each channel frms a Wheatstne bridge cmpsed f a pwer supply, amplifier, scillatr, galvanmeter, and the tw types f gages described' abve.. As the galvanmeter respnds t the changes in resistance f the active strain gage, the path f a beam f light is recrded n light= sensitive scillgraph paper. Three variable-speed recrding machines are used t recrd the respnses f the 48 gages. 2.4 TEST VEHCLES The tw test vehicles, designated as trucks Tl and T2, are shwn in Figs. 8 and 9, respectively. Bth vehicles were laded with steel plates t apprximate AASHO Specificatins fr H20-Sl6-44 truck ladingg Truck Tl was prvided by the Bureau f Public.Rads, while truck T2 was prvided by Schuylkill Prducts, nc.

19 LOADNG LANES Lading lanes were lcated n the radway s that the cen- terline f the truck wuld crrespnd as clsely as pssible t a girder centerline, r t a line midway between girder centerlines The centerlines f the seven lading lanes were indicated n the radway with plastic tape. Figure 2 shws centerlines f the lading lanes t be spaced at 43 inches, with the exceptin f the uter tw lanes, Lanes 1 and 7, which were ffset 4 inches tward the bridge centerline t prvide better clearance between the curb and the Qutside face f the tires f a truck being run alng these uter lanes. n the AASHO Specificatins 1 it is specified that fr design purpses, the centerline f a wheel r wheel grup may be placed t within 24 inches f the curb face. With vehicle Tl running in Lane 1, this distance was 16.5 inches, while with T2, the clearance was 15.4 inches TMNG AND POSTON NDCATORS, Air hses were placed 75 feet east and west f gage Sectin M, in rder t mnitr the speed f the lad vehicle. A timer was actuated as the frnt axle f an appraching test vehicle passed ver the first air hse and was shut ff as the truckts frnt axle passed ver the fifth hse. Three additinal air hses, which served as psitin indicatrs, were placed at pints 50 feet east and west f Sectin M, as well as at Sectin M. Each axle passing ver ne f these hses caused an abrupt ffset frm the scillgraph trace representing these indicatrs. The ffsets were then used t crrelate the truck psitin with strain values in the data reductin.

20 TEST RUNS Befre and after a set f test runs, the gages were calibrated with n lad n the bridge t relate the deflectins f the scillgraph traces t base values. During the cnduct f static runs, 12 separate calibratins were made. Crawl runs at a speed f 2 t 3 mph, made with the truck engine idling, were cnsidered t represent the static cnditin. The scillgraph traces indicated that fr the crawl runs, the vibratin f the sprung mass f the truck had little effect n strains and deflectins. During the crawl runs, the test vehicle was guided by a helper t assure that the truck was centered in the specified lading lane. A view f a crawl run invlving the tw trucks is shwn in Fig. 6. As the frnt axle f the vehicle neared the first psitin indicatr hse, anther helper n the bridge ntified the persnnel in the trailer t start the recrding scillgraphs. The recrdings were cntinued until the trailer axle had passed ver the third psitin indicatr hse. Testing was divided int tw series, each crrespnding t ne f the tw gage cnfiguratins, as described in: Sectin 2.2. Series testing crrespnded t partial gaging at Sectins M and N, while Series testing crrespnded t the full gaging at Sectin M. Table cntains a listing f the runs cnducted in bth series Each single-truck run in a particular lane was duplicated by anther run in the same series, and by anther run, r runs, in the ther series~ The exceptin t this prcedure was that six tw-truck runs were made in Series with vehicles T1 and T2 passing ver the bridge simultaneusly.

21 D A T A RED U C T ON AND E V. A L U,A T ON 3.1 OSCLLOGRAPH TRACE READNG Data reductin began with the editing f the traces fr each test run. Editing required the crrelatin f trace numbers, each f which represented a particular strain gage, with the traces n the test recrd. The crrelatin was facilitated by the existence f trace breaks, crrespnding t 16 gage traces and 2 inactive reference traces n each f three scillgraph recrds frm a test run. Sme interpretatin was necessary where traces verlapped, r trace breaks did nt exist. n general, each trace had sme definite physical characteristic, such as line weight, which helped t assure the crrectness f the interpretatin. With editing cmpleted, all calibratin runs were evaluated and cmpared. The calibratin runs cnsisted f a base recrd and a calibratin recrd fr each gage trace. The distance between a reference trace and the gage trace was measured with an accuracy f 0.01 inch n bth recrds 0 The calibratin value was the difference, always cnsidered psitive, between the calibratin and base recrd readings. Since each series f tests extended thrugh a perid f several days, calibratin values fr sme gages did nt remain cnstant. Due t this variatin f calibratin values, nly six calibratin runs, f the twelve taken during crawl runs, were used The six calibratin runs used were thse immediately preceding a set f test runs,

22 -17- With the cmpletin f editing and the determinatin f calibratin values, the recrds f test runs culd be prcessed. A n-lad reading fr each trace was taken at the left side f each recrd. Mst lad readings were taken adjacent t the middle drive axle ffset, which crrespnded t the drive axle passing ver the air hse at Sectin MG Any ther vehicle psitins were lcated n the recrd by prprtining distances frm ne f the axle ffsets, as related t the knwn axle spacings and the distance between gage sectins. Therefre, by distance prprt~ning, the test vehicle c~ld be psitined at an alternate lcatin in which the rear axle was ver Sectin N. 3.2 EVALUATON OF OSCLLOGRAPH DATA Strain Distributin and Lcatin f Neutral Axes Fr the mst efficient methd f cnverting scillgraph traces t strains and deflectins, a cmputer prgram was written in the WZ language fr use with the GE 225 cmputer. Gage cnstants (cnsisting f gage resistance, gage factr, lead cable length factr, peratin attenuatin, and calibratin attenuatin), calibratin values, and recrd readings served as prgram input data. The prgram utput, cnsisting f strains and deflectins, was listed n prepared crss-sectins f the bridge in rder t prvide a visual check fr sizeable errrs. The results f several runs were pltted t determine the distributin f the strains alng a girder face, as in Figs~ 4 and 5.

23 -18- Anther WZ prgram was written t calculate the lcatin f the neutral axis at each girder face. The calculatin f each value was based n a linear distributin f strain as shwn in the strain plts in Figs~ 4 and 5. At Sectin M, with three strain gages, it was pssible t use the WZ prgram t calculate a neutral axis value frm each f three cmbinatins f tw strain gages. Each f the three neutral axes was listed as cmputer printut fr cmparisn, Neutral axis values varying mre than 2 inches frm the ther tw crrespnding values were eliminated frm the averaged value. At Sectin N nly ne value f neutral axis was btained frm the tw girder face strain gages. 3G2.2 Effective Slab Widths A third prgram was develped t calculate the effective slab widths fr the girders. Differences in cmpsite girder gemetry necessitated separate prgrams fr exterir and interir girders 0 Bth prgrams accunted fr the measured variatin f girder depth and slab thickness, n line with experimental results, exterir girders were analyzed as acting cmpsitely with the slab, curb, and parapet, while the interir girders were cmpsite with the slab~ The prcedure fr bth prgrams was similar, as the first sectin in bth prgrams equated the first mrrlents, with respect t" the averaged girder neutral axis, f the tensile and cmpressive areas f the cmpsite girder, based n the principle f the transfrmed sectin. n rder t equate these first mments f area, it was necessary t determine the transfrmed effective width f slab, r

24 -19- slab, curb, and parapet which was needed t balane the first mment f the tensile area. The transfrmed effective slab widths f the interir cmpsite gi~ders were calculated by the interir girder prgram. When the transfrmed effective slab widths had been calculated fr the interir girders, the exterir girder slab widths culd be calculated. The transfrmed effective slab width fr an exterir girder depended n the transfrmed effective slab width f the adjacent interir girder. f the slab width f the adjacent interir girder exceeded 86 inches, the calculated slab width f the exterir girder was less than 84 inches; whereas, if the interir girder slab width was less than r equal t 86 inches, the slab width f the exterir girder was cnsidered t be 84 inches. The dtted lines n the safety curb in. Fig. 2 indicate the prtins fr which effective widths were calculated. The sequence f calculatins fr the exter~r girder was (1) t check whether maximum slab width f 84 inches was required; if s, (2) t check whether the maximum curb width f 33 inches was required; and then if s, (3) t calculate the required width f parapet Girder Bending Mments After the gemetry f the cmpsite girder crss-sectins had been determined, the next step was t cmpute the mment carried by each f the girders. Fr each girder, the lcatin f the neutral axis at each f the tw vertical faces, tgether with the strain distributin n each f the faces, was used t determine bth the resultant mment and the inclinatin f the plane cntaining the

25 -20- resultant mment carried by each f the five cmpsite sectinse Finally, the cmpnent mment acting in the. vertical plane was calculated fr each girder. The mments cmputed shuld be termed mment.cefficients, since these cefficients wuld have t be multiplied by the effective mdulus f elasticity, E, f the cncrete t determine actual mment values. Reference t past studies has indicated that effrts t determine a value fr the effective B frm empirical relatinships related t T, r frm stress-strain infrmatin resulting c frm cylinder tests, have prved t be fruitless. Hwever, it was pssible t determine the effective value by equating the externally applied mment at the crss-sectin, which can be determined by principles f statics, t the internal resisting mment at. the crss-sectin, which is the prduct f the sum f the five mment cefficients multiplied by E. After E had been determined, the individual girder mments were calculated fr cmparisn with design values. The average effective mdulus used was btained frm all test runs cnducted in Series, with the truck psitined with drive wheels at Sectin M Distributin Cefficients Distributin cefficients were alculated t determine the percentage f ttal resisting mment distributed t each girder. The distributin cefficient f a particular girder is the mment cefficient fr that girder, divided by the sum f the

26 -21- mment cefficients fr all five girders 0 t shuld be nted that the mdulus f elasticity was nt necessary in this calculatin, since dividing the mment fr a girder by the sum f five girder mments wuld result in the same distributin factr Evaluatin f the distributin cefficients fr the Series testing, where nly three f the five girders were gaged, req~ired the superpsitin f data t btain mment cefficients fr all five girders 0 Syrr@etry f the bridge crss-sectin was assumed The fllwing example will illustrate this superpsitining prcedure 0 First, see Fig 2 t recall the relatinship f lading lanes t girders, A truck running in Lane 1 prduced mment cefficients in Girders C, D, and E, while a truck running in Lane 7 prduced mment cefficients in the same girders, which were equivalent, respectively, t the mment cefficients in Girders C, B, and A, with the truck, running in Lane 10 Other symmetric lane ladings in Lanes 2 and 6, and 3 and 5 were cmbined in the same manner. When lnment cefficients had been btained fr all five girders in this manner, a dis~ tributin cefficient culd be calculated fr each girder5 Distri= butin ceffi.cients culd nly be calculated fr ladings in the fur Lanes, 1 thrugh 4, t the right f the bridge centerline 0 The distributin cefficient evaluatin fr Series testing was greatly simplified because mment cefficients and distributin cefficients culd be calculated directly fr all five girders fr test runs in all seven lanes Since the lading lanes and girders were symmetrically lcated with respect t the centerline f the radway, it was pssible t calculate an average distributin

27 -22- cefficient by averaging the distributin cefficients f crrespnding girders fr trucks in symmetrical lanes 0 Fr example, it was pssible t average the distributin cefficient f Girder E, with a truck in Lane 7, with the distributin cefficient f Girder A, with a truck in Lane 1. n bridge design specificatins, prvisins fr lateral distributin f lad are cmmnly expressed as distributin factrs. These factrs are cefficients by which a line f wheel lads is multiplied in cmputing the design mment fr a girder. These distributin factrs are based n cmbinatins f lading which will prduce the maximum effect in each girder. T enable a cmparisn f maximum measured mments with the mments btained frm design specificatins, the measured distributin cefficients fr varius lane ladings were either cmbined t yield maximum distributin factrs, r taken directly frm the tw-truck runs cnducted in Series 110 in tw wayse The cmbinatin f distributin cefficients was dne n the first, the distributin cefficients resulting frm tw single lane ladings were cmbined, s as t crrespnd t tw-truck test runs~ n the secnd, the distributin cefficients resulting frm tw single truck runs were cmbined, such that the lateral spacing cnfrmed t AASHO Specificatin prvisins. l By the first methd, the distributin factrs, resulting frm the -maximum eccentric lading f the bridge, with ne f the vehicles in ne f the utside Lanes, 1 r 7 with the ther in the enter Lane 4, culd be calculated. Additinally, the distributin factrs, resulting frm symmetrical lading f the bridge with vehicles in Lanes 2

28 -23- and 6, culd als be calculated~ Als, by the secnd methd, the distributin cefficients, resulting frm the.aasho Specificatin prvisins fr lateral truck spacing, were cmbined fr trucks in Lanes 1 and 5 r Lanes 3 and 7.

29 PRE SEN T"A T ON F T EST R E.S U LT S 4. 1 DSTRBUTON COEFFC,ENTS Distributin cefficients are presented bth in graph and table frm. The graphs shw the magnitude f the distributin cefficient pltted fr each f the five girders. The lad distributin fr a particular lane lading is represented by a straight line plt cnnecting the resultant girder distributin cefficients. The laded lane is indicated t the right in each plt. T facilitate the interpretatin f the distributin cefficient graphs and tables, as well as ther tables, a lading key cnsisting f a diagram shwing truck lcatin and directin, and indicating the gage sectin frm which results were btained, is shwn abve each graph r table. Figures 10 thrugh 18 are graphs f Series single truck test runs fr the fur right-hand lanes, Distributin cefficients are pltted fr bth-sectins M and N fr three types f Tl lading indicated by the lading key: (1) west runs with drive axle at Sectin M, (2) west runs with rear axle at Sectin N, and (3) east runs with the drive axle at Sectin MQ Additinal Figs. 12, 15, and 18 pres~nt a cmparisn f distributin cefficients at Sectins M and N fr the te$t vehicle laded in Lane 1 and Lane 40 The slid graphs represent distributin cefficients at Sectin.. M, while the dtted graphs represent thse at Sectin-Nt Figures 19 thrugh 28 r~present the distributin cefficients resulting frm Series test runs. Plts are presented fr

30 -25- distributin cefficients resulting directly frm a single test vehicle in each f all seven lanes, and frm the runs in which the tw vehicles were run simultaneusly_,t check the validity f the distributin cefficients, the averaged distributin cefficients, as explained in,sectin 3,2.~ are cmpared with the directly calculated distributin cefficients resulting frm test ladings n the right-hand side f the bridge. Fr example, the averaged value f distributin cefficients fr runs in Lanes 1 and 7, and fr symmetrical girders fr the runs in. Lane 4, are cmpared with cefficients directly resulting frm runs in Lanes land 4, respectively. The direct distributin cefficients are pltted as slid lines and the averaged values are pltted as dtted lines. The distributin cefficients btained frm superimpsing runs in bth test series are presented in Figs. 29 thrugh 33~ A cmparisn f bth directly calculated and averaged distributin The distributin cefficient graphs were pltted frm the distributin cefficients listed in Tables 2 thrugh 6. Distributin cefficients at Sectins M-and N frm Series ~and cefficients is included fr the Series runs which were superimpsed. fr directly calculated and averaged values frm Series ~are listed adjacently fr ease f cmparisn &

31 DSTRBUTON FACTORS Distributin factrs were derived frm the distributin cefficients fr tw trucks prceeding west with drive axles at Sectin MG The values, either frm tw-truck runs r frm superpsitin f results frm single truck runs, are presented in Table 7. The values represent the placement f test vehicles which prdues the maximum mment fr each girder. Three grups f listings are shwn, crrespnding t three types f lateral psitining f the test vehicles. The first grup lists the factrs fr maximum eccentric lading with vehicles in Lanes 1 and 4 r Lanes 4 and 7, as explained in-sectin The secnd grup presents factrs fr trucks in lanes mre nearly cnfrming t,aasho Specificatin prvisins fr lateral truck spacing, that is, trucks in Lanes 1 and 5 r Lanes 3 and 7. Factrs fr all girders fr symmetrical' lading with vehicles in Lanes 2 and 6 are listed in the third grup, which als cnfrms t AASHO'Specificatin prvisins fr lateral truck spacing. All three grups include distributin factrs btained frm bth directly calculated and averaged distributin cefficients. 4.3 SAMPLE GRDER MOMENTS n Table 8, the maximum girder mment cefficients and mments fr single truck (T) lading are listed. Table 9 gives values fr bth the cefficients and mments resulting frm lading with tw vehicles. "The values in_the upper part were derived di~ rectly frm the results f the tw-truck runs (Tl and T2), while

32 -27- values in the lwer part resulted frm the superpsitin f values frm single truck runs (T). All girder mments were calculated using, a value f ksi fr the effective mdulus f elasticity f cncrete, as btained by the methd explained in Sectin n Table 10, design and experimental live lad girder mments taken frm Table 9 are cmpared fr the central, uter interir, and exterir girders. The design girder mments were calculated frm a free-bdy diagram f the bridge, using, a line f Tl wheel ladings and a distributin factr f 1.30 fr the tw interir girders and a distributin factr f 0.81 fr the exterir girder. 4.4 GRDER ;DEFLECTONS Experimental girder deflectins at Sectin M are listed in Tables 11 and 12 fr trucks in vari~s psitins. Values listed fr vehicle.tl runs, west and east, are the averages f fur runs west and three east, respectively, frm bth, Series and testing. The maximum girder deflectin fr single vehicle lading was inches, measured in exterir Girder E with test vehicle Tl running west in Lane 7. Deflectins fr tw-vehicle runs are listed in Table 12. The maximum girder deflectin fr tw-truck lading was inches, measured in interir Girder B with test vehicles in Lanes l.and 4.

33 , TRANSFORMED EFFECTVE WDTHS i Tables 13 thrugh 17 list the average transfrmed effective width f slab fr each girder fr each lane lading. The values listed are generally the averaged value f the results frm tw identical runs~ The effective slab width, as defined by the,aasho Specificatins fr interir girders, is 86 inches.

34 ~29-5 DS C U S S ON F RES U L T S 5.1 EXPERMENTAL STRANS AND NEUTRAL AXES The plts f typical strain distributin shwn in Figs. 4 and 5 are fr vehicle Tl in Lane 7 with the drive axle at Sectin M, where there are three strain gages alng the face f a girder. These plts indicate a linear relatinship f strains alng the girder faces and int the deck. Figure 4 shws a linear strain distributin between strains at the three gage lcatins alng the girder face f E and the three gages n the parapet and curb abve this girder. Figure 5 als shws a linear strain distributin between the gages n bth faces f Girder C and the gage n the tp f the slab at the centerline f Girder C fr the same test lad. The linear strain distributin shwn in bth figures demnstrates that full cmpsite actin between the curb, parapet, and slab was effective during the testing prgram. The cmputatin f a girder face neutral axis lcatin was based n linear strain distribl.j.tine Because there were sme small variatins between the measured strains resulting frm identical vehicle runs, the calculated lcatins f the neutral axis fr a particular face did nt always agreee At Sectin N, where there were nly tw strain gages n each girder face instead f three, the variatin f the neutral axis lcatin was especially nticeable 0 t is felt that in future tests, at least fur strain gages shuld be applied t the face f each girder in rder that variatins in

35 -30- the strains measured at each individual strain gage will have less influence n the determinatin f the lcatin f the neutral axis. There were tw variatins in the lcatin f the girder neutral axis, as determined by pints f zer strain n the separate girder faces. The first variatin was the inclinatin f the girder neutral axis between girder faces. The inclinatin f the girder neutral axis was especially evident in the exterir girders. The secnd variatin was in the lcatin f the average value f the girder face neutral axis lcatins~ Bth types f variatin depended n the lateral psitining f the test vehicle with respect t a particular girder The inc~inatin f a girder neutral axis was mre nearly hrizntal when the girder was directly laded than when the test vehicle was laterally displaced with respect t that particular girder. The distance frm the bttm f a girder t the average lcatin f the neutral axis was usually a maximum, when the girder was directly laded4 This distance generally decreased as the-truck was displaced laterally frm the particular girder cnsiderede Girder deflectins were generally quite small 0 With a single vehicle n the bridge in ne f the exterir lading lanes, the crrespnding exterir girder deflected mre than any f the ther fur girders~ Fr tw-truck lading f the bridge, with bth trucks as clse as pssible t each ther and t the curb (maximum )

36 -31- eccentric lading), the crrespnding first interir girder under the lads exhibited maximum deflectin. N attempt was made t crrelate the lateral distributin f deflectins with the distributin cefficients TRANSFORMED EFFECTVE WDTH The magnitude f the transfrmed effective slab width, as presented in the tables, crrelates with the amunt f the test lad carried by a particular girder. Since the rati between values f the mdulus f elasticity fr the cast-in-place deck cncrete and the precast, prestressed girder cncrete culd nt be determined fr the test structure, it was impssible t determine the actual effective width f the slab. t is pssible, fr example, if the mdulus f the cast-in-place cncrete were greater than that f the prestressed girders, that the transfrmed effective width f the slab might be greater than the actual effective width. f this were the case, the actual effective widths f the slabs f adjacent girders wuld prbably nt verlap- 5.4 MODULUS OF ELASTCTY As explained in Chapter 3, the girder mment cefficients were multiplied by an average effective mdulus f elasticity fr cncrete f ksi, btained frm 41 test runs cnducted in Series. The AC Cde 2 presents a methd fr the calculatin

37 -32- f mdulus f elasticity based n the frmula: E c = w ( f' (psi) c f ft C 3 w, the nrmal weight f cncrete, is taken as 145 lb/ft.and is taken as psi, which was the average 28-day strength f the cncrete used in the five girders, a value f 4839 ksi is btained fr the elastic mdulus f cncrete. This value fr the mdulus is 29% less than the experimentally. btained effective value. Tw f the factrs which may be respnsible fr the difference between the frmula value.. and the experimental effective value are lading rate and stress. range. Since an increase in ladihg rate results in a higher E, it wuld be expected that the crawl run speeds f the test vehicle wuld easily exceed the lwer rates f lading e~plyed in the develpment f the frmula. Likewise, since the maximum strains prduced in the field tests were relatively small, reflecting crrespndingly lw stresses, the stress range was lw. Therefre, the cnditin wuld parallel the cnditins fr which the initial tangent mdulus is defined, resuting., in a higher value f E. The field test results are in-line with the findings reprted in cnnectin with the test f a cmpsite beam (steel) and slab (cncrete) bridge in Califrnia. 3 n this reprt, values f E were cmputed frm strain distributins in the slab at bth

38 -33- midspan and quarter-span lcatins. The average values f the slab cncrete were fund t be 5980 ksi and 6670 ksi respectively. The previus explanatin might apprpriately accunt fr the majr part f the discrepancy between experimentally determined internal bending mments and the mments prduced by external lads, as described in studies reprted by Hulsbs and Linger,lO and by thers. 4,g 505 C0MPARSON OF DESGN, AND EXPERMENTAL LVE LOAD MOMENTS The design value f kip-ft. fr the interir girder live lad mment was based n a distributin factr f 8/5$5 specified by the PDH, and is equal t the effect at Sectin M f a line f T1 wheel lads, with the drive axle at Sectin M, multiplied by The distributin factr fr an exterir girder, as calculated by PDH Specificatins,ll is 0081, resulting in a live lad design mment f kip-ft. fr the exterir girders f the test structure. The experimental live lad mment values. listed in Table 10 were cmputed fr the tw-truck test runs and the cmbinatin f single truck runs, crrespnding t either the lateral truck spacing f the tw-truck test runs r the lateral truck spacing f the,aasho Specificatin prvisins $,The maximum mments in Girders. A, B, and C were achieved with vehicles Tl and T2 in Lanes 1 and 4~ The ex~ 'perimental mments in Girders E, D, and C, with vehicles in Lanes 4 and 7, were smaller than the mments in Girders_A, B, and C with

39 -34- vehicles in Lanes. 1 and 40 f the design-and experimental girder mments are cmpared with the experimental values as percentages f the design value, it can be seen that fr vehicles in Lanes 1 and 4, the experimental values fr interir Girders Band Care significantly less than the design values. Fr the center Girder C, the, average experimental value was 57% f the design value, while fr the uter interir Girder B, the average experimental value was 76% f the design value 0 Fr exterir'girder A, the maximum experimental value was 172% f the design. value. The secnd and third cmbinatins f tw-truck lading cmplied mre nearly with the AASHO, Specificatin prvisins fr lateral vehicle spacing. These lading cmbinatins were'the superimpsed Tl vehicle runs in. either Lanes. 1 and,s r Lanes 2 and 6, the latter lading cmbinatin crrespnding. t runs cnducted with tw trucks simultaneusly n the bridgeg Fr these tw lading cmbinatins the maximum experimental mment values in Girders A and B were achieved by the superimpsing f Tl vehicle runs in Lanes 1 and 5, while the maximum.. experimental mment in-girder C was. achieved with vehicle Tl in Lanes 2 and 60 Fr bth interir gi~ders the experimental mment values were nearly the same fr bth cmbinatins f lading, the experimental mment averaging 53% f the design mment fr Girder C and 68% f the design mment fr Girder B Fr exterir Girder' A, the,maxim~ experimental mment was 142% f the design. value 0 The maximum experimental girder mments were develped

40 -35- in the exterir girder when trucks were run in either Lanes 1 and 4 r Lanes 1 and 5. These cmparisns f experimental and design girder mments wuld be valid, in general, fr vehicles f the type utilized and fr medium span bridges. The vehicles used fr te$ting were apprximatins f AASHO specified H axle spacings and ladings. A test vehicle with a shrter wheelbase between the trailer and drive axles, such as the l4-ft. wheelbase specified by AASHO, and with axles laded t H lads wuld prduce slightly greater girder mments. Hwever, it is felt that the distributin cefficients wuld be affected very. ittle DSTRBUTON COEF,FCENTS The distributin f lad, fr Series. runs, was smewhat affected by the lcalized influence f the truck axles. The effect f this lcalized influence was especially nted fr single trucks running alng the center Lane 4. Fr either an eastbund r westbund vehicle with the drive axles at Sectin M, the lad distributin at Sectin N was mre unifrm than at Sectin M.,With the trailer axle lcated at Sectin N, lad distributin was mre unifrm at Sectin M than at Sectin N. When single, vehicles with the drive axle at Sectin-M were cnsidered in the lanes ffset frm Lane 4, the distributin cefficient fr the exterir girder was gen,erally larger at Sectin N than at Sectin M. Hwever, fr maximum mment lading, the applied girder mment at 8ectin N wuld be less. than

41 -36- that at Sectin M n general, there was n significant variatin between the distributin cefficients calculated near midspan and near the quarter-pint f the span A cmparisn f distributin ceffigients r~sulting frm single truck runs during Series enabled a check n the actual symmetry f the bridge, The slight lack f symmetry can be seen in the distributin cefficient graphs. Generally, girders n the right side f the sectin carried mre lad than thse n the left side. Fr instance, the sum-f distributin cefficients fr Girders A, B, and C with a vehicle in Lane 1 was abut 2 t 3 percent greater than the crrespnding sum fr Girders E", D, and C with a vehicle in Lane 7. Althugh there is a slight variatin f lad distributin fr vehicles in symmetrical lading lanes, the averaged distributin cefficients d nt differ-significantly frm the directly btained values. 'The distributin cefficients resulting frm the Series< tw-truck runs were greatest fr the exterir girder when the bridge was laded in Lanes 1 and 40 The averaged distributin cefficients fr the tw-truck runs were nt significantly different frm the directly btained, values 0 The distributin.cefficients resulting frm the superimpsed runs f single vehicles w~re_nt significantly different frm the crrespnding cefficients resulting frm the tw-truck runs,. The crrelatin f distributin cefficients fr tw-truck runs with vehicles Tl and T2 and fr the superimpsed single vehicle runs with the same vehicles was especially gd,

42 -37- Reasnable crrelatin was als btained fr the distributin cefficients resulting frm the superimpsing f runs with vehicles Tl and T2 and the cefficients resulting frm the superimpsing. f Tl vehicle runs. 5.7 DSTRBUTON FACTORS The design distributin factrs fr a truck-wheel lading were established as 1~30 fr the interir girders and 0081 fr the exterir girders fr the Drehersville test structure, The experimental distributin factrs listed in Table 7 were calculated frm runs made with tw trucks r tw superimpsed trucks prceeding westward with the drive axles at Sectin Me Bth the directly calculated and average distributin factrs resulting frm this maximum lading are presented t implement a cmparisn f values The tw~truck test lading and the superimpsed single vehicle ladings, with vehicles in either Lanes 1 and 4 r Lanes 4 and 7, result in the maximum experimentally determined distributin factrs fr any f the test structure girders,fr these critical lading cases, the distributin factrs fr exterir Girders. A and E exceed the design value f 0081, while the interir girder design distributin factr f 1030 exceeds the resultant experimental distributin factrs fr interir Girders B, C, and D The maximum experimental values fr this lading case were 1037 fr exterir Girder A and 0098 and Oa77 fr interir Girders Band C, respectively a The maximum experimental distributin factrs resulting frm tw-truck runs cnfrming t AASHO Specificatin prvisins

43 -38- fr lateral vehicle spacing were.l.. 22 fr exterir Girder A and 0.88 and O~70 fr interir Girders Band C, respectively.

44 -39-6.,S U M M:A R Y AND CON C L U,S ON S 6.1 SUMMARY The primary bjective f this pilt study was t determine the lateral distributin f the live lad t the precast, prestressed cncrete spread bx girders in a typical bridge cnstructed accrding t the bridge design standards f the Pennsylvania Department f Highways. A secndary bjective was the evaluatin f testing, and data reductin prcedures in establishing, the behavir f the bridge subjected, t design-lad vehicles. This reprt presents the results f a field test prgram cnducted n an existing bridge lcated near Drehersville, Pennsylvania. The bridge was basically cmpsed f a reinfrced cncrete slab supprted by five prestressed cncrete girders, and had reinfrced cncrete curb and parapet sectins. Tw test vehicles, simulating, AASHO H20-S16-44 lading, were driven acrss the span singly and in cmbinatin at crawl speeds f 2-3 mph alng seven apprximately equally spaced lanes. The testing was divided int tw series crrespnding t tw different strain gage cnfiguratins. n Series three f the five girders were fully gaged at tw different crss-sectins, while in Series,all five girders were fully gaged at ne crss-sectin. n additin, ther gages were munted at varius lcatins n the deck, curb, and parapet during bth series. The strain variatins resulting frm the varius truck runs were measured with cntinuus,strain recr~ing, equipment supplied by the

45 -40- U. S. Bureau-f Public Rads. The distributin f the strains was evaluated t determine the lcatins f the girder neutral axes, frm which the effective transfrmed areas f sl~b, curb and parapet prtins f the bridge deck culd be calculated. Girder mment cefficients were then calculated and cmbined fr cmparisn with design values. Experimental girder distributin-cefficients were then cmputed and maximum experimental girder distributin factrs were develped and cmpared with design distributin factrs calculated accrding t PDH design specificatins. 6.2 CONCLUSONS The fllwing cnclusins were made frm'the results f the pilt study crawl runs: 1. The test results revealed that the actual distributin.factrs fr the girders were sig~ificantly different frm the values used in the design f the structure. The magnitude f the difference is suffic,ient t cnclude that the testing f ther structures f the same type will yield similar infrmatin. The reader is cautined that this. reprt cvers the results f crawl run testing, nly, Hwever, a preliminary evaluatin f data frm speed runs cnducted as anther part f the same field test prgram indicated that the distributin factrs will nt be affected significantly by faster mving,_ vehicles. 2. The test results cnsi$ten,tly indicated full cmpsite actin between the slab and the curb sectins, and between curb and

46 -41- parapet sectins. This actin is undubtedly. ne f the primary reasns fr the difference between experimental and design values. f the distributin factrs. gnringcthe cmpsite actin-in. design des nt necessarily result in a cnservative design. That is, it is pssible. that higher stresses than thse cmputed in the design wuld be prduced in the exterir girder, even thugh stresses lwer than design values. wuld be prduced in the interir girders. 3. A review f the tables listing effective slab widths ~ndicates (1) that fr this structure, the center-t-center spacing f the girders is a reasnably accurate estimate f the effective slab width, and (2), that use f the same value f E fr bth beam and slab cncrete is als reasnable. 4. Fr testing at crawl speed, the superimpsing f the results f single truck runs t determine the effects f tw-truqk lading is a valid prcedure. 5. The strain 'measurements taken at a crss-sectin having half f the girders gaged can be cmbined t accurately represent measurements taken with all girders gaged. 6. At least fur strain gages shuld be applied t the face f a girder in rder t accurately establish the lcatin f the neutral axis.

47 ACKNOWLEDGMENTS j This study was cnducted in the Department f Civil Engineering at Fritz Engineering Labratry, under the auspices f the Lehigh University nstitute f Research, as part f a research investigatin spnsred by: the pennsyl1ania Department f Highways, the U. S. Department f Cmmerce, Bureau f Public Rads, and the Reinfrced Cncrete Research Cuncil. The experimental wrk was c01ducted under the directin f Dr. C. L. HUlsbs, frmerly Research Prfessr f Civil Engineering, Lehigh University, in cperatin with Messrs. R. F. Varney and C. F. Galambs, Structural Research Engineers, Bureau f Public Rads. The field test equipment, including the recrding equipment and the main test vehicles, were made available thrugh the cperatin f Mr. C. F. Scheffey, Chief, Structures and Applied Mechanics Divisin, Office f Research and Develpment, Bureau f Public Rads. The equipment was perated by Messrs. R. F. Varney, C. F. Galambs, and Harry Laatz. The secnd test vehicle was made available by Mr. Jseph Nagle, President, Schuylkill Prducts, nc The basic planning in this investigatin was in cperatin with Mr. K. He Jensen, Bridge Engineer, and Mr. He P. Kretzky, Engineer in charge f Prestressed Cncrete Structures, bth f the Bridge Engineering Department, Pennsylvania Department f Highways 0

48 -43- The fllwing Lehigh University persnnel were assciated with this study. Mr. A. A. Guilfrd was the Principal nvestigatr n the prject; Dr. J. M. Hansn, Dr. J. C. Badux, and Mr. Hward Brecht, assisted in the field testing; Messrs. T. W. Schaffer, S. H. Cwen, and R. J. Dietz assisted in the data reductin and cmputer prgramming; and Mrs. Carl Kstenbader typed the manuscript 0

49 8. T"A B L:E 8-44-

50 -45- Tab e 1 Listing f Test Runs Vehicle Directin Lanes Number Series Tl West 1 thrugh 7 14* Tl East 1 thrugh 7 14')'(' 28 Series West 1 thrugh 7 14')'(' T East 1 thrugh 7 7 T2 West 1 thrugh 7 14')'(' Tw Trt;Lck Runs Tl and T2 T1 an'd T2 Tl and T2 West West West 4 (Tl) 1 (T2) 6 (Tl) 2 (T2) 4 (Tl) 7 (T2) * T~ runs per lane

51 1 Table 2 Distrlbutin Cefficients fr Single-Truck Runs, Series Distributin Cefficient. c Mment CeffiC~, (100) Mment Cefficients Girder Girder E 0 C B.A E D C B A 'T i8j()j West - T ~west 0 0 M h 0 N h 0 Lane Lane Lane Lane ~ East rq"llt)it O~.East ~T ~ ~ ~ M O~ Lane Lane Lane Lane T ~west_ 0 O. Q ~7 M T ~west N Lane Lane Lane Lane ~ OJ

52 Table 3 Distributin Cefficients fr Single-Truck Runs, Series Distnbutin Cefficient =Mment Cefficient- (100) Mment Cefficients Directly Calculated Values Averaged Values Girder Girder E 0 C B A E 0 C B A 0 i 0 T QFd_ West M h Lane Lane Lane Lane Lane Lane Lane b tl ~west 0 M A Lane Lane Lane Lane La-ne Lane J Lane ~ -...j

53 Table 4 Distributin Cefficients fr Single-Truck ~ns, Series (cntinued). D t "b f C ff" t Mment Cefficient S rt ulan e lelen. = M ment C e ff" Cie nt s (OO) Directly Calculated Values Averaged Values Girder Girder E 0 C B A E 0 C B A b}gjt Q~ East. ~ M Lane Lane Lane Lane Lane Lane ~ Lane M 5 T1 ~West b,777777) Lane Lane Lane Lane Lane Lane Lane p. OJ

54 Table 5 Distributin Cefficients fr Tw-Truck Runs, Series Directly Calculated Values Girder E 0 C Mment Cefficient. Distributin Cefficient = ~ Mment Cefficients (100) Averaged Values Girder B A E 0 C B A T 0 dq}j ~west h M Lane 1 and Lane 2 and Lane 4 and M 0 T 0 12 L(jQy ~west 0 / Lane 1 and Lane 2 and ~ Lane 4 and ill

55 Table 6 Distributin Cefficients fr Superimpsed Single-Truck Runs E Lane 1 and 4 Lane 1 and Lane 2 and Lane 3 and Lane 4 and Dist buti C ffici nt: _=Mment Ceffident' (100'~ rl n e e. M t C ffi nt 'J ~ men e ae s Directly Calculated Values Averaged Values Girder Girder DeB A E 0 C B T d QJ b 12 1(iCd. we.sf M h A T 0 T 0 -d d ~west 0 ; h M Lane 1 and Lane 1 and Series Lane 2 and Lane 1 and Lane 1 and Lane 2 and Lane 3 and Lane 4 and Series Ul

56 Table 7 Experimental Distribu~in Factrs Sectin M - Truck Mving West DRECTLY CALCULATED VALUES GRDER AVERAGE VALUES GRDER LOAD LANE A B C D E A B C D E BASED ON MNMUM LATERAL TRUCK SPACNG Test T1,T2 (1,4) OQ75 (4,7) Super Tl,T2 (1,4) (4,7) Super T1,Tl (1,4) (4,7) BASED ON AASHO PROVSONS FOR LATERAL TRUCK SPACNG Super T1,T2 (1,5) (3,7) Super T1,T1 (1,5) (3,7) Test T1,T2 (2,6) ' Super T1,T2 (2,6) Super Tl,Tl (2,6) U1 ~

57 Table 8 Sample G~rder Mment Cefficients and M?me~t?, Single-Truck Runs EffeCtive MOdulus -f Elasticit, ksi Mment Cefficient (10 6 ft.- in 2 ) 'Nmen t (kip~ ft.) Girder -Girder E 0 C B A '. D 'C '3 A ~west T 0 0 M,h Lane i '58"~ l , La.ne , ~500 Lane ' ' 110 ~ 79tl' , Lane , : 1350~8P ' Lane Lane ~ Lane :316.42a Ln tv

58 Table 9 Sample Girder Mment Cefficients and Mments, Tw-Truck Runs and Superimpsed Single-Truck Runs Effective. Mdulus f Elasticity =6, ksi Mment Cefficient 00 6 ft.-in~).. Mment (kip~ftj Girder Girder E -D,",c-. B A E- D C,-e' '. A LQTl.dCQJ.. b ::2. ~. West ~.)- Lane 1 and i94' i ' Lane 2 and ' ~ Lane 4 and ' " a8~: ,190 h Lane 1 and 5 Lane 2 and 6 Lane 3 and 7 1Tlld:Ca <::>...T ~.' West.n..(Supe~psitin). ~ ~ ~39:.0g2" : ~ ' ' " 6.91 U1 LN

59 -54- Table 10 Cmparisn f Design and Experimental Girder Mments Sectin M DESGN MO:MENT kip-ft. EXPERMENTAL MOMENT kip-ft EXPERMENTAL DESGN NTEROR GRDER C Test T,T2 Lanes land Superpsitin Tl,T Lanes land 4 4, Superpsitin T,T! Lanes 1 and 5 Lanes 2 and NTEROR GRDER B Test Tl,T2 Lanes 1 and Superpsitin T,T Lanes 1 and Superpsitin T,T! Lanes 1 and 5 Lanes 2 and EXTEROR GRDER A Test Tl,T2 Lanes 1 and Superpbsitin Tl,Tl Lanes 1 and Superpsitin T,T Lanes 1 and 5 Lanes 2 and _

60 Table 11 Girder Deflectins,_ Sing1e-Tt'-u~k Runs (All values" in inches.) Girder Girder E 0 C B A E 0 C B A T ~ West. T2 ~ West M b- 0 n M - -- Lane \ Lane _ _ Lane " Lane Lane Lane Lane ~926 -~ Lane 1 Lane 2 _Lane 3 Lane 4 Lane 5 Lane 6 Lane 7 East Q&iT ~ (). - M ~ ~ O.Q5615 O.0~ O~ T' ~west 0-- M / O03155~ Oi D02818 O~ (Jl 11l

61 Table 12 Girder Deflectins, Tw-Truck -Runs (All values in inches) E Girder C B A Tlld:!=OJ 12 ~west.0 - M n Lane 1 and Lane 2 and Lane 4 and t T2 :Qy h T~West U M Lane 1 and Lane 2 and p Lane 4 and 7 0: ; : Ln OJ

62 Table 13 Transfrmed Effective Slab Widths, Series (All values in inches) Girder Girder -E c E c Lane 1 Lane 2 Lane 3 Lane 4 Lane 5 Lane 6 Lane 7 T i(3 d West Zl M $ * T t!;cy West { 6 N ~ East git O~. East CObJT O~ 0 0 M N Lane Lane Lane ' ltOO if... Lane Lane * 89,.28 84~OO Lane ', Lane $45 "';'( ~ Overlap f transfrmed effective width: Ul "

63 Table 14 Transfrmed Effective Slab Widths (cntinued), Series (All values in inches) E Girder c E Girder c 1 M T~West. T ~ West Q. N Lane 1 Lane 2 Lane 3 Lane 4 Lane 5 Lane 6 Lane ! ' ' k L1-3.61, r Sll ' *- O~erlap -f trans;ermed effective 'width U1 CD

64 Table 15 Transfrmed Effective Slab Widths, Series (All values in inches) "Girder E c B A T ~west ~ ---cr- - M h '~ane 1 Lane 2 ~Lane 3 '_Lane 4 ~Lane 5, i..ane 6 "Lane , ~(-.'"l * aj( Lane- 1.. 'Lane 2 ~Lane 3 Lane 4 :Lane 5. "Lane 6 Lane ,74.42 b T2.($CCY West a M q9 ~ * Q 37 lob.o * * : lq5~~8 60~~4 4~g26 ' : 0 00 *.. Ov~~l~p f transfrmed effective Width U1 LD

65 Table 16 Transfrmed Effective Slab Widths (cntinued), Series (All values in inches) E Girder C B A East 6JhlT U ~ M Lane 1. ilane 2 'Lane 31 'Lane 4' Lane S ~ane' 6 Lane 7 Lane 1 ;Lane 2. ~.Lane 3 :Lane 4' ~ane 5 -Lane 6 Lane cr, i: ' * ,28 7( '60.26 i *.,62 0 _ !'OO r T.~ West - ~/ * ~8~g~ ' , ' i * * _ * ' ~ 81-~95 ' ~ ;~ Overlap f tran~frmed effective width m

66 Table 17 Transfrmed Effective Slab Widths (cntinued), Series (All values in inches) Girder E c B A a T dcoj.. ~. West '0:' ~ - M. Lane -1 and 4 Land 2 and 6 Lea-ne 4 and ~2" S ld-qy - T ~ West 17 M ;Lane 1 and. 4 Lane 2 and 6 Lane 4 ancl.7, 84-; ".32 " fj J OJ }-J

67 9. FG U RES -62-

68 z ~:E z () ~ ~ ao 0 -ll :Ecn ll UJ ~ '), C 0 C 0 NORTHWEST SPAN gu 61'-S8 c.. t C. BEARNGS r.~ 62'-3" " " Fig. 1 Elevatin f Bx Girder Bridge m w

69 - - - / ~, v -""--- 1'-31~ 1"'6" " ROADWAY ~[:-==t=- ',,..._-- ---',::::::a :;::;::=: ~ c:j PARAPET CURB 1.5"12 1-0" 4 SPA. Q =28 1-8" 2 1-0" 11'-5" Fig. 2 Crss Sectin f Bx Girder Bridge OJ ~

70 -65- x ~~.~~ Gag.e Sectin M - type f Gage Symbl Strain Deflectin X Fig. 3a Gage Lcatin at Sectin M -~.t :: rt) j L2~211 Gage Sectin N - Fig. 3b Gage Lcatin at Sectin N

71 -\\ J:if $4 Q) "0 $4.,-.1 c.!) H $4 Q.l +J ~ ~ 4-{ 0 Q) CJ (1j j:::..j ~ Q) +J ::; 0 +J ct1 C J :::) ~ +J (J)., c.,...1 ct1 ~.j..j CJ) r-l ct1 CJ.,...1 p.. ~...j-. bo.,...1 ~ -66-

72 ~67- U f H (J) '"0 ~.r- \! t!j ~ 0,:.r- ~ 0).fJ Q H -.. ~ 0 Cl (J) () Cl3 \ ~ +J \ c1j d 0.r-f +J ~ A \.,-1 H +J Cl 'r- ~ c:l.r- t1j H.-J - - t/) r-l C'd U n.r- ~ \ ~ H J.l"'l b1) 'r- ~

73 -68- Fig. 6 Tw-Truck Crawl Run Fig. 7 Deflectmeters