Effect Of Pozzolans Added To Sand-gravel Concrete Pavement

PEYTON,STINGLEY,MEYER:POZZOLANS I N C O N C R E T Eas against 1.36 for non-air-entrained sections.This same grouping produces an avei-age of1417 lineal feet of transverse and 160 linealfeet of longitudinal cracks per 100 half panelsin the air-entrained sections, and 875 linealfeet and 77 lineal feet, I'espectively, in thenon-air-entrained group. This trend holds truefor all classes of concrete and brands of cement analyzed individually with the exceptions of those sections containing fly ash andsections containing cement " H " , in which i tis I'eversed.2. Sections grouped on the basis of cementbrands were studied relative to all types ofcracking. One brand, cement " X " , was notused with air-entrainment so cannot be compared directly with the other brands. Cement" R " sections contain about half as much mapcracking as each of three brands " C " , " X " ,and " H " , and fewer cracks of all kinds.These field results do not parallel expansiontests in the laboratory prior to constructionwhich indicated greatest expansion for' cements " C " and " R " . Cement " H " had leastexpansion i n laboratory tests but sections inwhich this cement was used are second onlyto those sections using cement " N " in severityof map cracking.3. A t the present time none of the pozzolanadmixtures have inhibited map cracking on ascale comparable with results obtained f i o mlimestone sweetening.4. Flyash mixes, with air-entrainment, showsome improvement ovei- basic mixed aggregatesections, with regard to map cracking. M o w r yshale sections show no improvement overbasic aggj'egate, and jMontei'ey shale sectionshave higher map cracking factors than dobasic mixed aggregate sections. (See Table 1)5. A count of mudjacked and faulted jointsreveals that 75% of all joints i n the i-ight hand,or north bound lane are faulted, or have beenmudjacked to coirect a faulted condition.This compares with 47% i n the south bound,less heavily traveled, lane. The right lanealso contains 600 lineal feet of transverse and69 lineal feet of longitudinal cracks per hundred half panels, as compared to 547 linealfeet and 50 lineal feet i n the left hand lane.The right hand lane has a map cracking "fact o r " of 1.50, as compared to 1.3S in the lefthand lane.Test BeamsA four foot In- ten foot slab, nine inchesthick, was east near the right-of-way line i n305PAVEMENTeach experimental section. These slabs hadpartial depth wood dividers at six inch centei'sto sepai-ate the slab into test beams having,roughly, an " I " beam cross section. Theseslabs were cured i n an identical manner withthe pavement proper and have been given nospecial handling since then other than anoccasional shouldering up. They are tested atsix month intervals without soaking or otherpi'eparation. Twelve beams from each set havebeen tested to date, and the I'emaining eightwill be tested in the next five y-eais. The 10day beams on the charts are 6- by 6- by 36inch test beams.S U R V f l Y DATA B Y C O N C R E T E-Baste Mixed AggregateCLAS.SSectionsClass 1 and 2 beams, without and with airentrainment, made from mixed aggregate notutilizing limestone sweetening or pozzolanicadmixtures, have lost flexural strength steadilysince the first few months. These losses have,howevei', tended to diminish during the lasteighteen months and small gains were recordedat the end of year five, over the fifty fourmonth tests (see Fig. 5).Class 1 beams have average modulus ofrupture, at the end of five years, of 489 psi,although they averaged 648 psi at ten daysage and 845 psi at three months. Class 2 beamsfollow- the same trend.Based on the pavement condition surveythere is no indication however that these basicaggregate sections are i n any immediate danger. 'hile they are inferior to the limestonesweetened sections, they are in genei'al equalto or better than the "admixture" sections.MowryShaleTest beams from concrete utilizing M o w r yshale have declined in flexural strength sincethe twenty-fourth month but still averageover 750 psi, a gain of about 100 psi over 10day strength.The condition survey shows practically nodifference between M o w r y shale sections andbasic aggregate sections, on the basis of map,transverse, or longitudinal cracking.MontereyShaleOf all the test sections the ones containingMonterey shale have the poorest field recordto (late. The flexural strength tests are veryerratic, particularly with cements " R " and" X " , losses or gains of as much as fifty percent

1211843253183193101504628101272C'ement " C " . . .Cement " N " . ,("ement " H " .C'ement " R " . ,Cement " X " . .Non airAir entrAll panels91188Class 3 M R Y . .Class 4 M R V . .Class 679184Class 3 . M T Y . .Class 4 M T V . .Class 5909181163r23 l-'A4 551841219118879184909181163Rt.' 8472145395559585356615951Rt.Tr. Crackshong. Cr.Kaulted Jts.Lin. Ft. per Lin. t. T)erper 100100 Panels100 5226103646Rt.81781783410Lt.Mud J K DPanels per100 6340MlLt. 62417506753725044366550MlRt. laneALip CrackingTABLE 1T A B U L A T I O N OF S U R V E Y 01)74858782Lt.Percentof PanelsMap .702.191.541.791.501.34Lt.1.501.581 .111.701.41Rt.F.M 2 (Ml) .651.882.231.611.951.601.381through56-I10cement " X "15, 29, 43, 6016, 30, 4426, 33, 4747, 5913, 25, 39, 41, 48,557, 21, 35, 49,, 578, 14, 22, 27 36, 42,50, 569, 23, 37, 50, 583, 10, 17, 24, 31, 38,45, 5211. 19, 26A, 28, 34,53, 61, 624, 12, 18. 20, 32, 40,46, 541 through 1415 through 2829 through 4243 through 5657 through 61, nonAEA\[ odd numberedclassesAll even numberedclasses1,2,5,6,Total MapCrackingI'M 1 Sei tions lnclude{l2r.Ml) ,i(M2)cadOODHC3o

CLASSESCLASSESCLASSESCLASSESCLASSES40a10 DAYI a 2 - BASIC MIXED AGG.38 4 - MOWRY SHALE3 8 4 - FLY ASH3 8 4 - MONTEREY S H A L E5 8 6-LIMESTONE SWEETENED MIXED AGG.F i g u r e 5. F l e i u r a l streniith—classes of concrete.3 Mo 6 Mo 12 Mo l5Mo l8Mo 24Mo30Mo 36Mo 42Mo48Mo 54Mo 60MoAGEAIR ENTRANON AIR ENTRAINIILEGEND IIOCOtsHoOoU2ISOotsHWHiHwHo

308MATERIALSANDbeing common from one six months test periodto the next. A t 60 months test beams average760 psi, 100 psi more than at 10 days. Monterey sections contain more cracking of all kindsthan any of the other sections. Montereypanels have a large number of longitudinalciacks, and some sections are approaching acondition that may require some extensivemaintenance. The concrete adjacent to manyof the cracks is becoming laminated, in horizontal planes, and is in some cases beginningto spall out.CONSTRUCTIONBasic Aggregate with LimestoneSweeteningThe limestone sweetened sections on theproject are in generally excellent condition,with all brands of cement. Beam .strengthsaverage over 700 psi at 60 months and areslightly stronger than at ten days age. Beamstrengths were at their peak at 18 months andhave slowly declined since that time. Limestone sweetened sections have virtually nomap cracking and fewer cracks of other t.\-pesthan any of the other classes of concrete.E F F E C T OF AIR-ENTHAINMENTFlyashTest beams made f r o m concrete containingflyash show wide variations i n strength fromspring to fall, but still have much higheraverage strengths i n flexure than any of theother classes of concrete. A t 60 months flyashbeams average well over 800 psi, as comparedto less than 600 psi at ten days.The condition survey shows air-entrainedflyash-concrete sections to be less severely mapcracked than the basic aggregate sections, andmuch less than either of the shales. Air-entrained flyash sections were the only sectionshaving less map cracking than their non-airentrained counterparts.1400LEGEND 1300While the decrease in strength due to entrained air was expected, (see Fig. 6) someother results were not. The condition sui \'eyshave revealed that nearly every type of deterioration is present in greater degree in airentrained sections than in the non-air-entrained sections. M a p cracking, as well astransverse and longitudinal cracking, are fi-omslightly more, to two or three times as extensive in those sections where air-entrainmenthas been employed. Flyash sections, however,reverse this trend. Air-entrained flyash beamstrengths are lower, about 8 percent, thannon-air-entrained beams, but map cracking isless se\'ere and there are fewer longitudinalNON AIR ENTRAINEDAIR ENTRAINED;I200LilOO10009008007006005004001 10 Da3 Mo6 Mo12 Mo 15 Mo l8Mo24Mo 30Mo 36Mo 42Mo 48Mo 54Mo 60KAGEFigure 6. Fiexural strengtli—witli a n d w i t i i o u t air-entrainment.

PEYTON,STINGLEY,MEY'ER: POZZOLANS I N C O N C R E T Ecracks. The total for transverse cracks is stillhighest in the air-entrained sections. None ofthe sections are scaled whether air-entrainedor not.Result's of Nebraska DurabilityTestsDuring construction of the project tweh-e3- by 4- by 16-inch test specimens for eachsection were made and sent to the NebraskaTesting Laboratoi'v for directional freezingand thawing tests. The beams were frozenfrom one side only but were thawed uniformly.The degree of expansion and the loss in sonicmodulus were used as measures of the durability of the specimens.The durability of all classes of concreteexcept the flyash admixture was increasedgreatly by the addition of entrained air. Theflyashconcrete was, however, benefitedslightly. For most classes of concrete, thedurability did not vary greatly foi- the differentcements. S U R V E Y DATA B Y C E M E N T-concrete without air-entrainment, but no airentrained sections were built with this cementso i t has been omitted f r o m a large part of thetabulations. Flexural strengths of test beamsfollowed the same general seasonal trendsregardless of brand of cement. (See Figs. 7 & 8 )However, highest strengths were recorded forcement " R " at almost every test period. Thepeak strength for cement " R " occurred at 24months.The condition survey reveals that sectionsbuilt with cements " R " and " X " have theleast map cracking, and less transverse andlongitudinal cracking than any other brand.The other cements " C " , " H " , and " N " arenearly equal to each other in map cracking,transverse cracks, and longitudinal cracks,but in each case the values are about doublethose recorded for cements " R " and " X " .Observations of the project to date indicatethat some qualities inherent in cements " R "and " X " are in themselves inhibitors of mapcracking and other difficulties associated withsand-gi'avel concrete pavement.BRANDSFour brands, cements " C " , "H", " N " , and" R " were used in each class of concrete so areused in all comparisons of sections by class orbrand. Cement " X " was used i n all classes ofLEGEND. B309PAVEMENTCEMENT " C "Tendency Towards Declining Flexuralof Test BeamsStrengthsWhen all beam strengths were averaged(see Fig. 9), i t was evident that flexuralCEMENT - NCEMENT " H "gCEMENT "R120010 Da3 Mo6 Mo12 Mo15 Mol8Mo24Mo 30Mo36Mo42Mo 48MoAGEFigure 7. Flexural strength—four b r a n d s of cement.54Mo60Mo

310MATERIALS AND CONSTRUCTIONLEGENDCEMENT "X" NON AIR ENTRAINEDu 110052 80010 Da3 Mo6 Mo12 Mo 15 Mol8Mo24Mo 30Mo36Mo42Mo 48Mo 54Mo 60MoAGEFigure 8. F l e i u r a l atrengtli—cement " X " .1400LEGEND AVERAGE A L L BEAMSmm m1300- 1200UJ 1100a:t 1000u. 900wiIs 1 L1ti800700600500400fI 1 LLItlODo3 Mo6 Mol2Mo15 Mol8Mo24Mo 30Mo36Mo 11L11it42Mo 48Mo 54Mo 60MoAGEFigure 9. F l e i u r a l strength—all beams.strengths have declined since about the end ofthe second year. Each cycle is a little lowerthan the previous year's. A t the end of yearfive, the average strength of all beams is 695psi i n flexure, as compared to 605 psi at 10da3's age, and 900 psi at 24 months. Concretecontaining limestone sweetened aggregate,which is a standard aggregate for the state, isalso following this trend on this project, thoughat the end of year five i t has not decreasedas radicalh- as the basic aggregate sections.W E A T H E R E F F E C T ON BEAMSTRENGTHSWhen the project was planned i t was assumed that considerable variation in flexuralstrengths could be expected due to extremes i n

PEYTON,STINGLEY,MEYER:POZZ0L.\NSIi\ CONCRETEPAVEMENTOsa;LlJa.inhiXoz 3tr19491950ll il1951YEARUj1952il19541953F i g u r e 10. Monthly r a i n f a l l a t McPherson, K a n s a s .weather, both temperature and moi.stiire.These variations have, however, greatly exceeded that expected. For example, flyashl)eams, without aii-entr-ainment, and withcement " H " , dropped from 1215 psi i n the.spring of 1952 to 650 psi that fall, and recovered to 1300 psi the next spring. Beams inthis series, after a hot, dry summer, 1953, weredown to 750 psi and after a winter with moderate moisture, wer-e up to 1030 psi i n the springof 1954. When tested i n the f a l l of 1954, aftera hot, dry summer, they were at 946 psi.The above example is an extreme case butlar'ge variations occur i n all beams containingflyash, Monterey shale, or ;\Iowry shale, andsmall variations i n those beams made withbasic mixed, or limestone sweetened aggregate. (See Fig. 10)KFFHCT OF TRAFFICDuring the five year per'iod that this projecthas been under- observation north boundtraffic has ax-er-aged somewhat heavier- thanthat headed south. (See Table 2) This is dueto fuel tanker traffic or-iginating at McPher-son.i ; i Dorado, Wichita, and Augusta. Almost alltypes of damage are a little moi-e jironouncedin the north bound lane.TABLE 2LOAD CONDIDION S U R V E YNunibei' of .\xles bv Axle Load Gro ii)S—Directionally f ir1950 to 1954Coiintv: .McPherson; Routet U S 81 By-i ass; Station No.:059-081; Location: North of U S 50N.Average Daily Frequency.\xle Loadin PoundsXorthboundSouthbound1950 1951 I ),i2ll953 1954i966 1126' 497 561 659 77"o 77360 94 49 46 64 83 7527 20 34 48 4349 23 66 62 65168 202 242 235 314219 284 209 201 23435 66 40 27 42Under 0-1799918000-19999 2999930000-3399933Total axlescounted11044EFFECT31549 1044 1208 1314 1429 1.546OK S H O U L D E R.MA1NTEN.4.NCED u r i n g the wet periods of 1950 and 1951the shoulder-s of this project became badlyer-oded, and i n some places ruts reached thebottom of the slab. During 1952 a great num-

;!12MATERIALSANDher of slabs were pumping, blowing, andfaulting. This conditioti was corrected duringthe fall and winter of 1952 and shoulders havebeen well ni.'iintained since that time. Thismaiutcuanco, plus much di-ier weather, hasalmost eliminated pum])ing and blowing.M U D J A C K I N G (TWO METHODS)Along with the intensi\-e shoulder maintenance program instituted i n 1952, considerablemudjacking was done to correct faulted joints.Holes were drilled approximately 3 feet f r o mthe joint, and 3 feet in f r o m the edge, and acement-silt slurry was pumped in. Slabs wereraised about I 4 inch higher than the adjacentedge, and were found to subside very nearlyflush under traflic.A few slabs were raised by introducing aTABLE 3% ExpansionCIHSB1Chis.s(;i!issClasaClass23 FA.4 FA.3 MTY.030.020.004.004.006ClassClassClassClassClass4 MTY.3 MRY4 MRY56.001.010.010.001.010Cement " C " .Cement " N " .Cement "11".Cement " R "Cement " X " .0125.0059.0024.0125.0132TABLE 4MAP C R A C K I N GClass of Concrete orBrand of CementMap Cracking Factor(FM Ml X 2 M2 X lassClassClass.'Vir-entramed564 FA3 FA13 M R Y .4 M R Y .3 MTY24 MTY952.23C'ement " R " . . .Cement " X " . . .Cement " C " . . . .Cement " H " . .Cement " N " . . .1.831.651.86No air-ent. (uv.\ir ent. (av. all)CONSTRUCTION])ipe under the edge, and riiising the slab undertraffic, which eliminated the necessity forguessing the amount of subsidence whichwould occur when traffic was turned back on.While this method at first seemed to be favorably received, i t was later abandoned for themore generally used procedure. The principalobjection seemed to arise f r o m the edge preparation necessary to prevent slurry- seei)ing outwhile pumping.STRAIN GAGEREADINGSStrain gage points, two pairs i n each section,were placed i n the edge of the slab on 20-inchcenters. The points were protected w i t h greaseand screw caps and most are still in goodcondition. However, almost half of these gagepoints are no longer in use because of cracksfalling between points.Readings were taken, i n inches, spring andfall during the entire five year period.Table 3 presents a summary of gage readingsat 5 years compared to those at 6 months.I t will be noted that expansions as read inthe field are small as compared with laboratoryresults, and tliat cements fall i n somewhatdifferent order.MAPCRACKINGSince map cracking is held to be indicati\-eof abnormal exjxtnsion in concrete, this ])roject has been examined in detail twice eachyear for evidence of such cracking. A t the endof the fifth year particular care was taken toassure that all sections were examined undersimilar light conditions, all work being doneduring early- morning and late afternoon hours.M a p cracking was classified according toits severity as follows:Faint map is that which is just discernible.The cracks are very fine and tightly closed.Faint cracking is more easily found when theconcrete surface containing i t is damp, or])ainted. I t is the initial stage of maji cracking.M a p I , the second stage of develoi)ment, isthat which may- be easily- seen in concretesurfaces. The ci-acks are still tightly closed andthe "map" pattern may not be completelyoutlined, i.e., there may- be more cracking inone direction than i n another.Map I I . the third stage of development, isdiscernible in concrete surfaces from somedistance or f r o m a moving car. The f u l l "map"pattern is present, u

the five years which have elapsed since the construction of this project. # THE McPherson Test Road is a conci'ete pavement on Route US 81 in McPherson County, Kansas. The portion of this route comprising the test sections is a b\--pass around the cit\-of McPherson, Kansas. It was constructed in the summer of 1949 as