A Cement-Treated Base fr Rigid Pavement F.W. VAUGHAN, Sils Engineer, Prtland Cement Assciatin, and FRANK REDUS, Sils And Paving Engineer, Michael Baker, Jr., Inc. The new Jacksn (Mississippi) Municipal Airprt IS the first jet-age municipal airprt t use a cement-treated base under rigid pavement. Because f its gelgical setting m an area f few natural depsits f satisfactry base materials, the existence f a gd sandy sil n the site, and the need t pay special attentin t the highly expansive clay n the site, the use f the cement treatment was fund desirable. Results f labratry tests f varius mixes f the natural materials with cement are presented and discussed in describing the selectin f the design mix. Methds f wrking the brrw pit and cnstructing the cement-treated base are discussed in detail. THE COMING f the jet age in air transprtatin fund the airprt runways f the City f Jacksn, Miss, t shrt a length. A thrugh study was made t determine whether it was feasible t expand the existing field r whether an alternate site shuld be develped. The study shwed that the existing facilities culd be expanded nly at cnsiderable expense, and even then they wuld nt be entirely satisfactry. A grup f pssible new sites were studied minutely with respect t prximity t passenger rigin, availability f land, apprach clearances, tpgraphic features, finances required, and many ther details. One site rated higher than the thers with regard t mst f the criteria cnsidered and was chsen. This site was lcated in an adjining cunty just acrss the Pearl River Valley frm Jacksn. The selected site lies at the eastern edge f the alluvial valley f the Pearl River in the adjining lw, runded hills. The alluvium in this area is predminantly silt. The surface cvering f mst f the hill sectin is a thin layer f lean clay; hwever, remnants f sandy river terraces are fund n the tps f a few f the highest hills n the site. These surface layers are underlain by a thick, very expansive mntmrillnitic, marine depsit knwn as Yaz Clay. PAVEMENT DESIGN CONSIDERATIONS Brings shwed that Yaz Clay wuld be expsed in mst f the cut sectins, and labratry tests and past experiences indicated this t be the mst critical material frm the design standpint n the site. Results f the tests shwed the fllwing ranges f values: liquid hmit, 70 t ver 100; plasticity index, 40 t 70; vlume change frm LL, 80 t ver 200 percent; and swell (cmpacted samples saked 4 days), 10 t 20 percent. These values indicate this t be a treacherus material with which t deal. The sil IS stable in its natural state, but swells cnsiderably n being rewet after having dried. N instances f changed cnditins have been bserved where abut 7 ft f lean clay verburden exists. These factrs were studied and satisfactry means f keeping the Yaz Clay at its in situ cnditin during cnstructin and subsequent years f use were cnsidered. It was estimated that very little temperature change, if any, wuld ccur mre than 3 ft
belw the light-clred surface f a cncrete pavement; therefre, little r n misture change wuld ccur because f temperature variatin. This was cnsidered t be the nly type f prtectin against misture change in the cmpleted structure necessary as n grund water was encimtered n the site. Thus, a thickness f select subgrade, subbase, base curse, and pavement f 3. 5 ft was believed t prvide adequate thermal insulatin fr the Yaz Clay. Cut sectins had t be undercut 19 in. belw subgrade elevatin t prvide the 3. 5-ft thickness. T keep the expsed Yaz Clay frm drying during cnstructin, it was specified that the lean clay backfill shuld be placed immediately, r the Yaz Clay sprinkled sufficiently t prevent drying until the backfill culd be placed. The subgrade was verlaid with 12 in. f subbase and base and 11 in. f pavement, as shwn in Figure 1. All fills under areas t be paved were cnstructed f the lean clay; the Yaz Clay was depsited nly in median areas where it was very unlikely that any pavement wuld ever be cnstructed. A depsit f fine, clean sand cntaining thin clay and silt lenses was lcated near ne end f the runway, and ffered a surce f base curse material. Visual examinatin and labratry classificatin tests indicated that it was a material f adequate stability, but, n further cnsideratin, it was thught highly imprbable that sufficient mixing culd be effected t prduce a imifrm base n which t perfrm paving peratins. Small pckets f clean sand wuld tend t ravel under paving equipment. Als, the material was fine enugh t be easily transprted by pumping actin under mving aircraft if water entered the base. In its natural state, the sand was cnsidered unsuitable fr base curse immediately beneath rigid pavement. A study f the csts invlved shwed that because f haul distances, it wuld be cheaper t mix the sand with a mderate amunt f cement t prduce a satisfactry base than t imprt anther material. Alsait was believed that the cemented layer wuld prvide additinal prtectin and insulatin ver the Yaz Clay. These facts were the basis fr the decisin t make a design study f cement stabilizatin in the tp 6 in. CEMENT TREATMENT DESIGN Examinatin f samples and bring lgs frm the sand depsit prpsed fr use as base material shwed that the amunt f silt and clay in the final base material culd be cntrlled t a cnsiderable extent by the manner in which the material was taken frm the pit. It was prpsed that the material be remved by a shvel r dragline cutting up a vertical face. In s ding, the number f silt and clay layers cut culd be cntrlled, thereby cntrlling the amunt f material passing the N. 200 sieve. II" PAVEMENT 6" SAND SUBBASE 6"CEMENT TREATMENT T J SELECT SUB6RADE- -UNDERCUT YAZOO CLAY Figure 1. Typical cut sectin.
Labratry misture-density curves had been prduced fr untreated sil fr fur samples selected t cver the range f materials available in the brrw area in cnnectin with subbase design. The characteristics f the fur samples are given in Table 1. Because f time limitatins it was necessary t cmpact specimens cntaining cement nly at abut the ptimum misture cntents determined previusly fr the imtreated sils. The specimens were cmpacted in a standard 4-in. diameter mld, using mdified AASHO cmpactive effrt, and were brken by cmpressin at the end f 7 and 28 days f mist curing. Plts f cmpressive strength vs percentage f cement are shwn in Figure 2 fr the individual samples. These data indicate that the cmpressive strength btained varies with the amunt f cement used and als with the sample tested. TABLE 1 CHARACTERISTICS OF SAMPLES OF MATERIALS AVAILABLE IN BORROW AREA Sample N. Sample Percent Passing N. 40 N. 60 N. 200 LL PI Max. Dens. Opt. Mist. 1 Clayey sand 100 99 27 30 4 118.0 12.6 2 Clayey sand 97 52 18 25 3 122.7 10.6 3 Sandy silt 99 90 53 29 7 123.6 11.6 4 Silty sand 98 83 13 NP NP 116.7 11.2 10 = SAMPLE NUMBERS I- z Z 8 0 txj O / 200 400 600 COMPRESSIVE STRENGTH - PSI 800 Figure 2. Cmpressive strength vs cement cntent.
The cause f variatin in strength with sample tested was analyzed by pltting the percentage f material passing the N. 200 sieve against the cmpressive strength at 6 percent cement (Fig. 3). This curve indicates that maximum strength fr cement used is btained with abut 28 percent passing the N. 200 sieve. At this pint it was necessary t decide n the minimum acceptable cmpressive strength. Based n lng experience the Prtland Cement Assciatin recmmends a minimum f 300 psi at 7 days fr use with flexible pavements. It was felt that this figure culd be reduced slightly fr rigid pavements inasmuch as the design cnsideratins d nt envisin such strength. Examinatin f labratry specimens indicated that thse breaking at 200 psi r mre were well-cemented. This material s treated wuld be unlikely t pump, and wuld prvide a stable wrking surface fr paving peratins. Hwever, t add a little cnservatism t the requirements, a minimum strength f 250 psi at 7 days was selected. T arrive at the mst ecnmical and practical sil-cement mix which wuld prvide the selected minimum cmpressive strength a plt f cement cntent vs percentage f material passing the N. 200 sieve was made frm Figure 2 fr a cmpressive strength f 250 psi and is shwn in Figure 4. Examinatin f this plt shws that the cement required t prduce minimum strength varies inversely with the percentage passing the N. 200 sieve. This bservatin is based n samples 1, 2, and 4. Sample 3 was nt cnsidered in the analysis because the quantity f this material was limited and culd nly be used fr mixing with the sand in cases f deficiency in material passing the N. 200 sieve. A study f the bring lgs fr the pit indicated that a material with 20 t 30 percent passing the N. 200 sieve wuld be prduced in mst cases by pit peratin and that little mixing f additinal fines wuld be required. Figure 4 shws that such a material wuld require abut 4 t 5 percent cement t btain the desired strength. The labratry investigatins just described were all perfrmed n well-pulverized and blended samples. It was realized, hwever, that field cnditins wuld prbably be different, and it was specified that the cntractr shuld build a test sectin at the Tii 600 0. X h- z a: H CO 400 ^ > 200 CD CO Hi tr. NOTE : 1. = SAMPLE NUMBERS. 2. POINTS TAKEN FROM FIG. I FOR 6% CEMENT -4 ^ [ ^ \ L 10 20 30 40 50 60 % PASSING N. 200 Figure 3. Effect f fines n cmpressive strength.
z H Z LU NOTE : 1. = SAMPLE NUMBERS. 2. POINTS TAKEN FROM FIG. I FOR 250 PSI. / / y- 20 30 40 % PASSING N. 200 50 60 Figure It. Cement required fr 2^0 psi. beginning m which perating prcedures wuld be established. Cnstructin f the test sectin shwed that the pulverizatin requirements f 100 percent passing the 1-in. sieve and 75 percent passing the N. 4 sieve culd be met cnsistently, but that the field behavir f this material differed cnsiderably frm the we 11-pulverized labratry material. One f the first prblems encuntered during cnstructin was that f btaining sufficient blending f material n the fill t prduce a reasnably unifrm sil. This was shwn t be necessary by the discvery f a number f small spts (less than 100 sq ft) lacking sufficient fines and ther small spts having excessive fines. The lw amunt f cement (5 percent) did nt prvide sufficient cement t prduce minimum strength requirements in these areas. Misture requirements fr each f these spts differed cnsiderably frm that needed in mst f the base, and it was fund almst impssible t prduce satisfactry misture cnditins in the whle area at ne time. Befre begmning cnstructin f the untreated subbase, labratry misture-density relatinships had been determined n a typical sample meeting the specificatin requirements. It was bserved that the actin f the hammer n the small lifts in the labratry mld prduced further pulverizatin that was nt duplicated in the field by the actin f vibratry, pneumatic, and sheepsft rllers. A study f the effects f pulverizatin n density is shwn in Figure 5. (Xirve 1 represents a cmpletely pulverized sample. The sample represented by curve 2 received a mderate amunt f pulverizatin in the field during excavatin, and limited additinal pulverizatin in the labratry. CXirves 3, 4, and 5 represent samples with limited field and labratry pulverizatin. Sheepsft, pneumatic, and vibratry rllers were tried in varius weights and in varius cmbinatins, and the misture cntent f the material was varied cnsiderably during early stages f subbase cnstructin. Almst regardless f equipment, the density btained cmpared favrably with that f curves 3, 4, and 5 f Figure 5. The clay layers in the pit were generally in a very firm cnditin and at a misture cntent f abut ptimum r a little less. When cmpactin culd be cmpleted withut
120 -THOROUGHLY PULVERIZED - LAB (D-MODERATELY PULVERIZED - FIELD @ - SLIGHTLY PULVERIZED - FIELD Z >- O 105 10 15 MOISTURE CONTENT Figure 5. Misture-density curves, subbase material. g 100 >- 20 MOISTURE 25 CONTENT Figure 6. Misture-density curve, cement treatment (5.5 percent cement). 30
the clay becming wetter, the clay balls shwed little defrmatin under the rllers. Hwever, when the misture cntent f the balls increased appreciably, they flattened cnsiderably under the rllers and prduced a spngy cnditin. It then became necessary t dry the clay sufficiently t bring abut stable cnditins befre satisfactry cmpactin culd be btained. Based n this finding, a labratry curve fr cmpactin cntrl purpses was prduced by adding 5. 5 percent cement t a sample that had limited field and labratry pulverizatin (Fig. 6). Test sectin cnstructin indicated this t be abut the maximum density btainable m the field with reasnable cnstructin effrt. The cmpactin effrt cnsisted f 4 t 6 cverages f a sheepsft rller prducing a bearing f abut 300 psi, tw cverages f a vibratry rller, and 2 cverages f a light penumatic rller. Cres taken frm the test sectin shwed that cmpressive strengths were generally a little belw the 250 psi required, ranging frm abut 150 t 270 psi. Frm these tests it was apparent that additinal pulverizatin r additinal cement wuld be required t prduce a base curse with the desired cmpressive strength. It was decided that blending wuld be accmplished, the material pulverized mre, and 6. 5 percent cement added fr anther trial. At this pint in the histry f this prject (early Nvember 1961) a rainy seasn began and lasted well int December. It appears that weather cnditins will bviate further cement treatment wrk until the spring f 1962. In cnclusin, the design prcedures required fr the cement treatment fr this prject were f a rutine nature based n thrughly pulverized labratry samples. After the relatinships were established n ideal samples, hwever, it was necessary t perfrm labratry tests n samples which were cmparable t the mixes and blends prduced in the field t prvide cntrl standards. The critical item in this prblem has prved t be the degree f pulverizatin f the clay lumps, which must be taken int accunt during design testing t predict what can be btained in the field with satisfactin.