Lime and Fly Ash Proportions in Soil, Lime and Fly Ash Mixtures, and Some Aspects of Soil Lime Stabilization MANUEL MATEOS and DONALD T

Home , Lime (material)

Lime and Fly Ash Proportions in Soil, Lime and Fly Ash Mixtures, and Some Aspects of Soil Lime Stabilization MANUEL MATEOS and DONALD T. DAVIDSON, Respectively, Research Associate and Director, Soil Research Laboratory, Engineering Experiment Station, Iowa State University of Science and Technology

The main objects of the investigations were to find a possible best ratio of lime to fly ash and the optimum amount of the lime and fly ash admixture for stabUizing various textured soUs; to determine the effects of lime content and curing period on the strength of soU and lime mixtures; and to compare the strengths of soils treated with lime, lime and fly ash, or cement. Four soUs were used: dune sand, friable loess, aUuvial clay, and gumbotU. Four limes (a calcitic hydrated and three dolomitic monohydrated) and eight fly ashes were tried. Seven-day and 28-day curing periods were used with soil, lime, and fly ash mixtures, and strength contour lines for 28-day results are presented. Ninety-day curing was also used with soU-lime mixtures. Moisture-density and moisture-strength relationships of soU-lime mixtures are compared. The best ratio of lime to fly ash and the optimum amoimt of lime and fly ash were found to depend on the kind of lime, fly ash, and soU used. Recommended amounts of lime and fly ash for stabUizing the different soUs are given. Data are presented to show the influence of type of lime and kind of fly ash on the unconfined compressive strength of sta• bilized soils. A durabUity evaluation of selected soU-lime- fly ash and soU cement mixtures by the Iowa freeze-thaw test is presented. The optimum amount of lime for strength improvement increases with length of curing mixtures of soU and lime. Dolomitic monohydrate lime proved to be better than cal• citic hydrated lime, although the calcitic lime may be more beneficial in the additions of small amounts. Additions of lime to such clayey soils as gumbotU and alluvial clay modify the lubrication effect of water when the mixtures of soil and lime are compacted, and the mixtures may not have a well-defined optimum moisture content.

• SEVERAL SOIL admixtures are used to obtain a construction material with better engineering properties than those of the original soU. The most extensively used are cement and lime. Others, like lime with fly ash, appear to be satisfactory stabilizers but have not been much used because their characteristics and behavior when added to soUs are not well known. Since the early 1950's the Engineering Experiment Station SoU Research Labora• tory of Iowa State University, in cooperation with the Iowa State Highway Commission and the Iowa Highway Research Board, has been conductmg an extensive evaluation of 1*0 41 different methods of soil stabilization for road courses. Special attention has been given to the use of by-product or waste materials such as fly ash. Producers have the costly problem of disposing of over 10,000,000 tons of fly ash every year. Inasmuch as laboratory and field evaluations of soil stabilized with lime and fly ash have given promising results,- highway engineers and power industry management are interested in further improving this use of fly ash. The work done to evaluate lime plus fly ash as an admixture to soils has been re• stricted. General conclusions on the use of these materials have been based on results obtained with a limited variety of the component materials (soil, lime, and fly ash) or have been based on limited testing. An attempt has been made in this investigation to introduce a reasonable number of variables in the soil, lime, and fly ash. To have a comparative evaluation, the soils used were also stabilized with cement and with lime.

MATERIALS AND METHODS Materials Used Four natural soils (a dune sand, a friable loess, an alluvial clay, and a highly plastic gumbotil) were selected as representative of important Iowa soil types (Tables 1 and 2), Eight fly ashes were selected to represent variations in the properties of this by• product material (Table 3): Fly ash 1, collected by multiple cyclone and electrical precipitators, was from coal from districts 3 and 8 in Ohio and from northern West Virginia, and was processed through pulverizing mills so that 70 percent passed a No. 200 mesh. The sample was sent from the St. Glair (Mich.) Power Plant ol the Detroit Edison Company. Fly ash 2, collected by mechanical equipment, was from coal from northern niinois. The coal was burned in a B and W boiler. This sample was sent from the Sixth Street Power Station of the Iowa Electric Light and Power Company in Cedar Rapids. Fly ash 3 was collected by electrical precipitators from a dry bottom type of boiler using unwashed coal from western Kentucky. The sample was sent from the Paddy's Run Power Station of the Louisville Gas and Electric Company in Louisville, Ky.

TABLE 1 DESCRIPTION OF NATURAL SOILS Property Dune Sand Friable Loess Alluvial Clay Kansan Gumbotil Soil Res. Lab. No. (S-6-2) (20-2) (627-1) (528-8) Location Benton Co., Harrison Co., Harrison Co., Keokuk Co., Iowa Iowa Iowa Iowa Geological Wisconsin- Wisconsin-age Recent fill, allu• Kansan-age gum• description age eolian loess, friable, vial plastic, botil, highly sand fine• oxidized, cal• slightly calcar• weathered, grained, careous eous plastic, noncal- oxidized, careous leached Soil series Carrington Hamburg None Mahaska^ Horizon C C Undefined Fossil B Sampling depth (ft) 6-11 49-50 0-4 7.5-8.5

Underlies C-horizon loess of Mahaska series. 42

Fly ash 4, collected by mechanical precipitators, was from coal from northern Illinois burned in a Springfield boiler in the Sixth Street Power Station of the Iowa Electric Light and Power Company in Cedar Rapids. Fly ash 5 was collected by mechanical (centrifugal) precipitators. The coal from Illinois was pulverized in a ball mill before burning in the Riverside Station Power Plant of the Iowa-Illinois Gas and Electric Company at Davenport, Iowa. Fly ash 6 was collected by mechanical precipitators (multicone dust collector). The coal from Iowa (Monroe, Polk, Marion, and Mahaska Counties) was unwashed steam coal which was pulverized and tangential fired in the Des Moines Power Plant of the Iowa Power and Light Company. Fly ash 7 was collected by mechanical equipment (VGR multicone) in the Waterloo Power Plant of the Iowa Public Service Company. The coal from southern Illinois was washed, dried, and pulverized with Riley mills. Fly ash 8 was collected by mechanical precipitators (cyclone type). The coal from several Missouri and Kansas nunes was pulverized and burned in suspension in com• bustion engineering boilers in the Hawthorn Station Power Plant of the Kansas City Power and Light Company, Mo.

TABLE 2 PROPERTIES OF SOILS Property Dune Sand Friable Loess Alluvial Clay Kansan Gumbotll Textural composi- Uon^ (5f): Gravel (> 2 mm) 0.0 0.0 0.0 0.0 Sand (2-0.074 mm) 95.5 0.7 2.4 19.4 Silt (0.074-0.005 mm) 1.5 82.3 25.6 14.6 Clay (< 0.005 mm) 3.0 17.0 72.0 66.0 CoUoids (< 0.002 mm) 2.6 14.0 61.0 63.0 Atterberg Umits^: Liquid limit (?S) - 32 72 76 Plastic limit Cf-) - 25 26 26 Plasticity mdex Nonplastic 7 46 50 Classification: Textural^ Sand Silty loam Clay Clay fjnginecring (AASHO)a A-3 (0) A-4 (8) A-7-6 (20) A-7-6 (20) Chemical: Cat. exch. cap.® ' (me/100 g) 1.0 14.5 44.4 39.2 6.6 8.4 7,7 7.4 CarbonatesS 0.4 10.4 3.6 2.0 Organic matter'* (^ 0.1 0.1 1.6 0.1 Predominant clay Montmoril- Montmoril- Montmoril- Montmoril- mineral^ lonite (trace) lonite lomte lonite

*ASTM Method Dlj22-51»T. ''ASTM Method Dlt23-S1»T and Dl»2lt-51lT. "Triangular chart developed by U.S. Bureau of Public Roads. •ÂASHQ Method MII4S-U9. Ânmonium acetate (pH = 7) method on soil fraction 0.1*2 mm (No. 1*0 sieve). ^Glass electrode method using suspension of 1$ g soil in 30 cc distilled water, êrsenate method for total calcium. '^Potassium bichromate method. hi-raf diffraction analysis. 43

Most of this investigation was made using two commercial grade limes of the U. S. Gypsum Company. One is a hydrated calcitic lime, brand name "Kemikal"; and the other is a type N monShydrate dolomitic lime, brand name "Kemidol." Two dolomitic monohydrate limes, from Western Lime and Cement Company and from Rockwell Lime Company, were also used m a comparative study of some commercial dolomitic monohydrate limes with dune sand and fly ash 3 only. The analysis of all limes is given in Table 4. The Portland cement used was commercial type I of the Penn-Dixie Cement Corpora• tion of Des Moines, Iowa. The analysis of the cement can be found in O'Flaherty et al. (13). Distilled water was used throughout to eliminate the variable that might result from impurities added with ordinary tap water. Procedures The proportions of soil plus lime or lime-fly ash or cement are based on the dry weight of the soil and lime; soil, lime, and fly ash; or soil and cement mixtures. Mixing and Molding. —Mixmg of batches for preparing test specimens was done in a kitchen mixer at low speed. The dry ingredients were machine mixed for 30 sec. The mix water was added and machine mixed for 1 min. The mixture was hand mixed for about 30 sec to clean the sides and bottom of the mixing bowl. The mixture was then machine mixed for 1 min more. Molding of test specimens was started immediately after a batch was mixed, ex• cept where otherwise indicated. The Iowa State compaction apparatus was used to mold 2-in. diameter by 2 ± 0.05-in. high specimens. With this apparatus the equivalent of standard Proctor compactive energy was obtained when giving 5 blows on

TABLE 3

ANALYSIS OF FLY A8B

Loss Spec Percent Fly Source on Surface, Spec. Passing Asb Ignit Gravity No. 325 SlOi CaO AUOi Fe,0, SO, (*) (cmVll) Sieve (« T W) (*) (« (*) 1 St Qalr, 3 9 2,820 2 58 91 8 43 5 0 2 2 9 23 2 24 8 0 8 Mich 2 Cedar Rapids, 7 2 2,663 2.39 49 8 36 7 1 0 3 5 21 3 2.0 ' Iowa 3 Louisville, 2 6 3,226 2 60 86 1 42 9 0 8 5 7 23 4 20.0 2 3 Ky. 4 Cedar Rapids, 18 6 4,590 3 37 94 9 36 2 0 9 8 3 15 8 16 7 1 9 Iowa 5 Davenport, 0 7 578 3.43 22.6 11.3 0 3 12 3 0 9 88 4 3.2 Iowa 6 Des Bloines, 0.2 1,460 2.82 31 8 40 1 0 3 5 8 13 1 36.7 2 4 Iowa 7 Waterloo, 13 0 4,240 2 34 54 9 38 5 0 2 3.2 16 1 16 2 1 1 Iowa 8 Kansas City, 3 8 2,048 2 68 64 8 35 3 0 0 5 3 7 7 43 3 1 4 Mo

TABLE 4 ANALYSIS OF LIMES

FeaOs Loss Percent Lime and on Passing SiOa AlaOs CaO MgO SO, Ignit. No. 325 (1^ Sieve Calcitic hydrated 0.3 0.6 73.8 0.6 0.3 24.1 95.5 Dolomitic monohydrate 0.4 0.2 49.6 31.8 1,1 17.0 91.0 Special No. 1 0.6 1.1 48.3 33.2 - 16.8 99.2 Special No. 2 0.4 0.6 45.4 36.3 - 21.0 91.0 44 each side of the specimens using a 5-lb hammer dropping 12 m. with the molding apparatus fastened to a wooden table (8). After being molded, the specimen was ex• truded, weighed to the nearest 0.1 g and measured to the nearest 0.001 in. During molding, a wet cloth was kept over the bowl to prevent drying of the mixture. Curing. —Specimens of each batch were moist cured at 71 db 4 F, at a relative humid• ity of over 90 percent for the desired periods of time. To retain moisture better and to reduce absorption of carbon dioxide from the air, the specimens were wrapped in wax paper and were sealed with cellophane tape before being placed in the humid room. Strength Testing.—After each curing period, specimens were unwrapped and im• mersed in distilled water for 1 day. Then they were tested for unconfined compres• sive strength using a load travel rate of 0.1 in. per min. Tests were run in triplicate, and the average strengths are reported in pounds per square inch. This is in accord• ance with ASTM designation C-109-58 which requires a mimmum of three specimens for each set of curing conditions. Any readings that deviate excessively can be detected in three observations. Specimens that differed by more than 10 percent from the aver• age value of test specimens made from the same mix and tested at the same age were not considered in determining compressive strength. If two specimens were rejected, new specimens were prepared. Durability Tests. —The Iowa freeze-thaw test was used to evaluate the durability of selected mixtures J^9). Four 2- by 2-in. specimens from each mixture were cured 28 days in the moisture room. Two specimens, designated the control specimens, were then left immersed for 10 days; and the other two specimens, designated the freeze and thaw specimens, were exposed alternately to temperatures of 20 ± 2 F (16 hr) and 77 ± 4 F (8 hr) for ten cycles, each cycle lasting 24 hr. A vacuum flask specimen container was used to freeze from the top down and to supply unfrozen water, kept at 35 ± 2 F by a light bulb, to the bottom of the specimen throughout the test (Fig. 1). After these treatments, the unconfined compressive strength of the freeze-thaw specimens (pf) and of the control specimens (p^) were determined. These values were used to evaluate the durability of the stabilized soils. The index of resistance to the effect of freezing (Rf) was calculated

Stabilized Soil Specimen To Variac ac Synthetic Plastic Holder

Cork Gasket

W. Lr

Electric Bulb-

Vacuum Flask

Walter Flask Container-

Figure 1. Specimen container for Iowa freeze-and-thaw test. 45

100 p Rf= -^(^)

LIME AND FLY ASH PROPORTIONS AND CONTENT One of the questions in soil, lime, and fly ash stabilization is how much lime and fly ash are needed? The amount and proportions of the lime and fly ash admixture are governed by the desired strength in the stabilized soil and by economy. Laboratory-molded specimens having an unconfined compressive strength of at least 300 psi after 28 days curing followed by 24 hr immersion may be indicative of adequate stability for a base course mixture to withstand the imposed loads and the detrimental effects of freezing and thawing (2, 10). To be used, lime-fly ash stabilization has to compete economically with other soil stabilization methods that can produce equivalent strengths and durability. The price of lime ranges between $15 and $25 a ton, including transportation to the job site. Fly ash sells for about $1 a ton at the power plants, so even after transportation and other expenses are added the price of fly ash is much below that of lime. This econo• my favors the use of greater amounts of fly ash than of lime. Considerable work has been done to find the best proportions and amount of lime and fly ash, but the kinds of fly ashes have been limited. In this investigation, eight fly ashes were evaluated with the dune sand and three fly ashes with the other three soils. The fly ashes are from Iowa and other states of the Midwest. The reason for using eight fly ashes with the sand is that sandy and granular soils generally respond better to lime fly ash stabilization and are more sensitive to the amount and quality of fly ash than silty or clayey soils. Because the eight fly ashes represent a wide range in characteristics, sources, and pozzolanic activity, the strength results obtained with them may be indicative of the best ratio of lime to fly ash and amount of lime-fly ash used with sandy and granular soils. The number of fly ashes used with the loess and clayey soils was narrowed to three which represent the wide range m properties and composition obtainable in this ma• terial and should mdicate the effectiveness of fly ash addition with lime to silty and clayey soils and the best ratio and amount of lime-fly ash for stabilizing silty and clayey soils. Two types of commercial limes (a calcitic hydrated and a dolomitic monohydrate) were used with all the fly ashes and soils. Two other dolomitic monohydrate limes were used with fly ash 3 and dune sand to compare the effectiveness of these different commercial dolomitic monohydrate limes. The amounts of lime used were 3, 6, and 9 percent with all soils; with gumbotil, 12 percent lime was also tried. For each amount of lime four mixes were prepared—one without fly ash and three with 10, 17.5, or 25 percent fly ash. All the percentages were based on the dry weight of the total soil lime or soil lime fly ash mixture. These combinations of lime and fly ash gave sufficient data to plot strength contours, which was done for the 28-day cured, 1-day immersed strength results. After 7 days curing, the immersed strength developed was rather low, and contour graphs of immersed strength results did not show much and are not presented here. In preliminary work, not included here, moisture-density and moisture-strength relationships were determined to select the molding moisture content for every com• bination of soil, lime, and fly ash. At least four sets of tests were run for every com• bination of soil and fly ash. S^&ximum strengths for calcitic hydrated lime and for dolomitic monohydrate lime were obtained for practically the same optimum amount of water. The molding moisture content needed for maximum 28-day strengths was chosen.

Dune Sand Strength Contours. —The plotted strength contours (Figs. 2 to 10) indicate there is no optimum amount and ratio of lime and fly ash that might be used with any kind of 46

7 do. •trmgltl 28 do, •trtngth old contflun 7 40, •Innglh 26 do, •lioogm ond mntoon '67 '64 6(ptJI '37 '0 0 \200im

• • • — • • •2M 0 *3I 42 ir V^-7, .6, 0 24 0 0 V\ Caldtle CoWtic hydrat«d hydrated llin«,% llme,% 3 • , _ •o •o - D 0 26 46 •o ^ 'a 0 •21

1 1 1 1 1 1 ) 10 ir.8 IS J J- 10 C9 10 20 30 Ply oth no 1, % Fly oDi oa I, % Fly oih na2. % Fly ash no Z. %

T tor •traogtb 28 do, ilmglh ond oMoun 7 do> ttrmgt^ ZB doy fltrvngth OAd contours 274 2l(|ltll 'o 'o 'o •|J.7\\i6J t»2

• ^76" 6 "o •o •o - Ootamitlc Hiwtiytfra tfl IIIM, % 97 *J4~ b "b •o •o - 0 Vt

1 1 1 0 1 1 1 m IM 19 > to IM 25 10 » 90 Fir oM no. 1, % Fly nb no. I, % Fly oth no. 2, % Fly oMi no 2, %

Figure 2. Immersed unconfined compressive Figure 3. Immersed unconfined compressive strength values obtained for several com• strength values obtained for several com• binations of dune sand, lime, and fly ash binations of dime sand, lime, and fly ash 1 for 7- and 28-day curing periods, and 2 for 7- and 28-day curing periods, and strength contour lines for 28-day results. strength contour lines for 28-day results.

26 do, •trongfli and conloiin •a "n- aiptii 0 0 0 fei—-n—•s—

\50p«l •27 -62 IS-i' CadHc \ % •» y> CoklNc hydrotad 1 '73 lllM,% \ % 1>2 V- 1j ^) 1o i n S '52 V

•1n J179. J 1 L. 10 ITS 29 10 KO 90 Fly oih •IOL4, % Flyaihna4,% nyailin&9,% Flyo>hna9,%

26 dot ilronqth md 7 do, tfrangtt) 28 day strwiQih ond contoure ^ilol) 59 "76 64~ '2l(p«i) 'o ~*0 *0~ 'fa \ '65—'6a\^^

'O ^4 % '77-^' •6 '0 t) •4,>^^V Ooloinitic monohydrata Ilffl6,«

[0 'o 'o 'o -15 -JJ -44

_1 I L_ S 20 90 nyoih na4,% Flyoth na4,« n IT9 29 P1yoshn&9,% Fly am KL 9,%

Figure U. Immersed \mconflned compressive Figure 5- Immersed unconfined compressive strength values obtained for several com• strength values obtained for several com• binations of dune sand, lime, and fly ash binations of dune sand, lijiie, and fly ash k for 7- and 28-day curing periods, and 5 for 7- and 28-day curing periods, and strength contour lines for 28-day results. strength contour lines for 28-day resiilts. 47

Ze do, •trtngth and conlMri day ttranglh dnd Cdntduri eipti) 76 '69 •74 •fa t115-l " s ipal) 'n n '106

•57- 0 *ZT 76 *9a 6 \ Coleltle Colcltlc hrdrotad hydroted lllM.% llm<,% • lOOp.1 b V V •45 - V.' to 'V. 0 IT v *79~

1 1 1 1 1 1 0 to ir.B 29 iO K JO 0 » 179 a Fly osh na6, % Flyoilinae,% % Fly ofli no T, % Fly 01)1 no. r.

7 day ttrangth 2B do, ttrtngth ond conlourt r day •tTVigth doy Btrmglli and contouri

21 ((M) ~3T "57 1B4 ^9 \ 'ai \ 'im—^TT 21 (mil ia 13S

6 -SI •44 ^1 " 0 *60 *I2 *I7~ DoiofflWc Momltle moMhydrata llm>,%

6 •39- 0 'K» v»-

1 1 1 1 1 0 » na 2» 10 20 M 0 10 175 IB Fly oth no 7 Fly a«h no 7, « Fly osh na6,% ni all no 6,%

Figure 6. Immersed unconfined compressive Figure 7. Immersed unconfined compressive strength values obtained for several com• strength values obtained for several com• binations of dune sand, lime, and fly ash binations of dune sand, lime, and fly ash 6 for 7- and 28-day curing periods, and 7 for 7- and 28-day curing periods, and strength contour lines for 28-day results. strength contour lines for 28-day results.

7 day ttrongni 28 doy (Irength and eomoun 7 day Itrtngtli 26 day ftrtngtti ond eontourt

6 Ipiil 62 62 93 \06 6 Ipsll 97 158 265

• • — Colcllle Colcltlc b 34 64 75 ~ 91 \ I46\ 299 b '62 199 245 hydrotatf hydroted llma,% llme,% • • • — 24 97 260 b 'T 47 57 ~ b *3a 91 21T

1 1 1 J L 1 1 i 0 10 29 20 50 0 n irn 2S Fly osh na8.% Ftyashn&e,% Flyo«lino.3t%

28 doy •tr«nqth ond COntOuTi 7 day •tranglh 7 day •tringlh 26 day atrtngtli ond cantoart

21 Ipsll 90 50 69 29 y Ito ^io 4K 2(tpii) V 91 ISO

20 32 54 ~ Dolomltic • b 9 62 *4S T74 DotomlHc ntonohydrotQ llnw,% llmo.% • • — .1 b 16 24 b *90 93 160

1 1 1 J L 1 1 1 0 K> 17.9 a 20 ao to 179 29 Fly Oil Flyoslin&S,% kno.e,% Fly osh na 3,% Flyatfino.3,%

Figure 8. Immersed unconfined compressive Figure 9. Immersed unconfined compressive strength values obtained for several com• strength values obtained for several com• binations of dune sand, lime, and fly ash binations of dune sand, lime, and fly ash 8 for 7- and 28-day curing periods, and 3 for 7- and 28-day curing periods, and strength contour lines for 28-day results. strength contour lines for 28-day results. 48

lime and fly ash to stabUize dune sand. faw/v96 '^'2CM •«4~2M 1' •RTt(?5^^"*"' The contours obtained with the same fly ash but with different limes are quite 0 So 198 •»«- Sp>ci

Density. —The density varied with the 1 1 1 kind and amounts of lime and fly ash. nyiKliiw.3,% Flyothiia3,% There is no consistency in the densities attained with lime, calcitic hydrated or Figure 10. Immersed unconfined compressive dolomitic monohydrate. The densities strength values obtained for several com• apparently depend on the kind of fly ash binations of dune sand, dolomitic monohy- and the admixture proportions. drate limes, and fly ash 3 for 7- and 28- Lime. —It has been observed by other day curing periods, and strength contour investigators that, in lime-fly ash sta• lines for 28-day results. bilization, dolomitic monohydrate lime produces greater strength than calcitic hydrated lime (3, 7, 8, 11, 12, 14). The following analysis of the effectiveness of the limes is based on the varieties of lime and fly ash used m this investigation. In mixtures of dune sand, lime, and fly ash 1, 2, 4, or 7, dolomitic monohydrate lime was more effective than calcitic hydrated lime for both 7- and 28-day curmg periods. With fly ash 5, test results were erratic, and conclusions can not be made as to which lime was more effective. With fly ash 6, calcitic hydrated lime was more effective than dolomitic monohydrate lime. Seven-day strengths of mixtures of fly ash 8 with calcitic hydrated lime were greater than with dolomitic lime. But dolomitic monohydrate lime with fly ash 8 gave better 28-day strengths. Thus, no general con• clusion is possible as to which kind of lime (calcitic hydrated or dolomitic monohydrate) IS best in lime-fly ash stabilization of dune sand; the kind of lime to use depends on the properties of the fly ash. Nevertheless, it may be concluded on the basis of 28-day strengths only, that dolomitic monohydrate lime generally gives higher strengths than calcitic hydrated lime. The only exception to this was in mixtures containing fly ash 6. Tests with fly ash 3 deserve special discussion (Figs. 9 and 10). Three dolomitic monohydrate limes were used with this fly ash. Comparing the effectiveness of calcitic hydrated lime with the dolomitic monohydrate limes, it was observed that for 7-day strength the calcitic lime was more effective than the special dolomitic limes. All three dolomitic limes gave 28-day strengths much higher than the calcitic lime. Of the three dolomitic monohydrate limes tested the special fly ash 2 was more effective with the fly ash used. No explanation was found for the differences in strength produced by the different dolomitic limes. An investigation is presently being conducted to compare the effectiveness of various commercial dolomitic and calcitic limes (16). The effectiveness of dolomitic lime for soil stabilization seems to depend on the tempera• ture and period of burning, the amount of impurities, the gradation, and probably other factors (16). Fly Ash.—The mixtures made with fly ash 3 attained very high strengths. Mixtures made with two of the dolomitic monohydrate limes, showed a strength of 1,000 psi after 49

28 day •trength and contaura 28 days of curing (Figs. 9 and 10). This 7S(|W) 64 66 60 strength approaches that of a lean con• crete. Mixtures made with the other i9 •73 '65 ~ dolomitic monohydrate lime and fly ash Colcltic hydrated 3, showed a strength after 28 days curing lime.X of about 600 psi, also very good (Fig. 10). 72 '66 •72 V)- Strengths of about 500 psi for the same 1 t 1 curing period were obtained with calcitic 0 10 174 29 hydrated lime. Seven-day strengths of Ryoshr n l,% Flye>hml,% 200 or 300 psi, depending on the type of lime used, were obtained with this fly 7 day ctrongth 28 doy itrafiglh and coMaun ash. I74(pil) 166 tai "m |5oo "ss* "370—"as Fly ash 1 also gave good strengths.

After 28 days curing in mixes with dolo• isi *I48 *63 mitic monohydrate lime, 600 psi, were obtained. The 7-day strengths for the same mixes were close to 300 psi, but iiT 'oa *M0 •o»- the results obtained with this fly ash and 1 1 1 calcitic hydrated lime after 28 days cur• 0 10 r79 29 n 20 90 Fly oth ial,« ing were very poor, barely reaching 100 nyaslinal,% psi. Other fly ashes that gave strengths Figure 11. Inmiersed Tinconfined compressive strength values obtained for several com• over 300 psi after 28 days curing were binations of friable loess, lime, and fly- fly ash 6 in mixes with calcitic hydrated ash 1 for 7- and 28-day airing periods, lime and fly ash 7 with dolomitic mono• and strength contour lines for 28-day re• hydrate lime. Many fly ashes did not sults. produce the desired 300 psi after 28 days curing in mixes with either of the limes used.

7 doy •tranglh 26 doy ttrongtll ond conloun 7 doy strtnglh 28 doy ttrtnglh ond conteort 7e(ptii 62 112 '^50 174 I 216 76 Ipiil W 126 TS6 * is| lea ; '272

ISO 200 pil BO 200 pil 2S0

. oU >. 56 66 '94 Cokitle 105 'ue ^1*15 J 99 *IOt •|42 162- ColcIHe Jm 1161 hydralM liydrolad lta.t,% a>6,% 72 '»S no 166 173 221 • • — 72 *00 140 142 166 226 216

1 1 1 n ITS ts 10 20 30 1 1 1 0 10 174 29 10 to 90 Fly eih ns. Z, % Fly oth M e, % Fly Olll OOL 3, % Fly stk M 3, %

T doy ilronglh 26 day ttrtogth and contoort 7 doy ttrongth 28 doy ititnglh ond IT4|p«l) IB4 IS9 135 «0^ I74IPI1I 210 loe 'tis

• ISI US *i3e 354 3d I 366 356 171 DoloaHIc ISI *I6I %I2 mooeliydieto fS0Op,l ,110..,% . OT *I46 *i4e *I2I~ "F49^ '316 ViO '246 • — IIT BT 204 224 I249 394 413 460 1 1 1 0 10 17 9 29 10 20 90 1 1 1 \ I Ply ash W 2, % Fly otl> n 2, % 10 17 9 29 ny oth no S, * Fly oth 60. 3, %

Figure 12. Immersed unconfined compressive Figure 13. Immersed unconfined compressive strength values obtained for several com• strength values obtained for several com• binations of friable loess, lime, and fly binations of friable loess, lime, and fly ash 2 for 7- and 28-d^ curing periods, ash 3 for 7- and 28-day curing periods, and strength contour lines for 28-day re• and strength contour lines for 28-day re• sults. sults. 50

The results point out that the strengths obtamed depend greatly on the fly ash used. This indicates the great disparity of pozzolamc properties of fly ashes. Some of them with lime may give strengths comparable with those obtained with cement, whereas others produce scarcely any strength. Fly ash 3 was used with three different dolomitic monohydrate limes; the densities as well as strengths varied for mixtures with these three dolomitic limes, but the strengths were not in relationship to the density but to the admixture content and amount. Fly ashes of low specific gravity (Nos. 2, 4, and 7) imparted very low dry densities to the dime sand, lime, and fly ash mixtures. Friable Loess Strength Contours. —The strengths of the friable loess mixtures with lime were de• creased by the addition of fly ash 1. Additions of fly ash 2 did not increase the strength of the friable loess and lime mixtures to any great extent. Additions of fly ash 3 in• creased the strength some but not greatly. The strength contours with friable loess are therefore sparse and difficult to draw (Figs. 11 to 13). The only type of fly ash that may be recommended for use with lime to stabilize friable loess is a high quality fly ash like No. 3. The verticality of the contours with fly ash 3 suggests the use of mixtures contaimng small amounts of lime and large amounts of fly ash; for example, 3 percent dolomitic monohydrate lime, 25 percent fly ash 3, and 72 percent friable loess. If the price of the fly ash is prohibitive, this soil could be stabilized with lime alone. Density. —Calcitic hydrated lime gave lower density than equal amounts of dolomitic monohydrate lime. Fly ash 2, of low specific gravity, lowered the density in propor• tion to the amount of fly ash in the mixture. No correlation was found between density and strength. Lime. —Dolomitic monohydrate lime with or without fly ash always gave higher strengths than calcitic hydrated lime. Nine percent dolomitic monohydrate lime pro• duced an immersed strength of 400 psi, which is considered adequate for a road base or a subbase. Fly Ash.—Fly ashes 1 and 2 either did not greatly improve the strength of friable loess and lime mixtures or were deterimental to the point where they actually lowered the strength in some cases. This may be because friable loess has greater pozzolanic acticity with lime than fly ash 1 or 2. Fly ash 3 gave strength improvements to friable loess and lime mixtures, particularly for mixtures with low lime contents. This is the only fly ash tested that may be recommended for use with lime, preferably dolo• mitic monohydrate, in the stabilization of friable loess.

Gumbotil Strength Contours. —Strength contours tend to be horizontal for low amounts of lime and become vertical for high amounts (Figs. 14 to 16). This indicates that lime up to a certain amount mcreases strength, and then fly ash becomes important m the develop• ment of strength. No defimte ratio of lime to fly ash gives the highest strengths, recommendations on the amounts of lime and fly ash to use should be based on the need of a minimum amount of lime, which is about 5 percent. Low amounts of lime required high amounts of fly ash, and high amounts of lime required low amotmts of fly ash. Several combinations of lime and fly ash may be chosen depending on the desired strength. The amount of lime required will be between 5 and 9 percent, and that of fly ash between 10 and 25 percent. Density. —Density values did not correlate with strength, nor did they correlate with the kind of lime used. The fly ash of low specific gravity (No. 2) gave lower densities than the other two fly ashes. Lime. —The calcitic hydrated lime in small amounts gave greater strengths than small amounts of dolomitic monohydrate lime. Dolonutic monohydrate lime was more effective than calcitic in large amounts. This was observed for mixtures with and without fly ash. Large amounts of lime may stabilize gumbotil soil satisfactorily. For instance, 12 percent dolomitic monohydrate lime produced a 7-day strength of 190 psi and a 28-day strength of 298 psi. 51

T doy 91rrngt h 26 do, rtrtngth ond contaoro 26 day ttunqfti ond contowi 132 (Pti) 149 161 176 228 230 237 298 02(l»i| '206 199 209 229 / 313 313 323

• • • ^40 '237~ 129 *42 169 *62 219 207 241 248 129 \3 Im V6

Calcitic ^ Colelllc \ \ / • hydratad • • • lie *39 I4T *6I 199 Vs^22 8 238 242 il6 '212 •292 '235- lima, % ..200 p«J • 100 *oe 169 *64~ MS 109 180 206 loo *I24 *223 "266'' 149 141 269 396

1 I 1 1 1 1 1 1 0 10 r7s 2S n w JO 20 30 Fly oth na 1, % Fly otii no I, % Ryasfi na2,% Fly ash no 2.%

T doy ttranglh 28 do, ttransth ond conteurt 26 doy tUtntfh end contourt 190 (ptO 273 291 32t I90(psi) 273 "278 ~29S

191 *242 *2TI *S26 191 •260 •2S3 •305-

B9 •277- 89 209 *238 293~ liffle.%

• • — 3 93 90 93 3 u •at •93- W~-'20

1 1 1 1 1 1 -1- -i_ 0 N> 29 Flyaahno2,% Fly adi no I, % Fly a«h no I, % ny a>lina2,%

Figure lU. Immersed unconfined compressive Figure 1$. Immersed unconfined compressive strength values obtained for several com• strength values obtained for several com• binations of gumbotil, lime, and fly ash binations of gumbotil, lime, and fly ash 1 for 7- and 28-day curing periods, and 2 for 7- and I 28-day curing periods, and strength contour lines for 28-day results. strength contoiir lines for 28-day results.

Fly Ash. —AU three fly ashes tested were effective in improving the strength that may be obtained with gumbotil and lime alone. Immersed strengths, after 28 days curing, of from 400 to 500 psi were obtained. Consequently, the use of fly ash with lime may be recommended to stabilize gumbotil to meet the standards of a base course.

Alluvial Clay Strength Contours. —No definite optimum ratio of lime to fly ash was observed in the tests made with alluvial clay soil (Figs. 17 to 19). The dolomitic monohydrate lime content of mixtures was very critical for the development of strength. For large amounts of dolomitic lime, the fly ash content was more critical. With calcitic hy• drated lime, the fly ash content was almost the only component contributing to strength as seen by the verticality of the contours for mixtures with calcitic lime. The recommended amounts and kinds of lime and fly ash to stabilize alluvial clay are from 5 to 7 percent dolomitic monohydrate lime with from 10 to 25 percent of any fly ash used, or else 3 percent calcitic hydrated lime with 25 percent of fly ash 3. Fly ashes 1 and 2 are not recommended with calcitic hydrated lime because the same strengths may be obtained with dolomitic monohydrate lime only, in amounts from 6 to 9 percent. Density. —No relationship was found between density and strength. The statements made as to the relationship between specific gravity of fly ash and density of mixtures also apply here. Lime. —The calcitic hydrated lime gave better strengths than the dolomitic mono• hydrate for the lowest amount of lime (3 percent). The effectiveness is reversed for higher amounts. Without fly ash, 9 percent of dolomitic monohydrate lime may properly stabilize alluvial clay. Strengths of 173 psi after 7-day curing, and 345 psi after 28- days were obtained. 52.

26 day itrangtli ond eoitoun T dor alrtnglh 26 day tfrtngth and eontowt 132 (pil) 207 233 243 327 367 387 I09IPIII 148 «9 es |l66/ |09 '216 S-

• • Its 193 247 *266" 129 fas 162 l65- Coldllc hydrottd 200^ lteii.% * . — . — I1« 224 264 266 125 191 'lS4 153 132 'l92 216 *2I4

• • • — 1 1 1 ioo 191 162 273 0 n n3 29 Fly osb re 1 % Fly Oth no I, X 1 1 1 0 10 ITS t9 ttrtnglh Fly oth fio. 3, % Fly ah ne 3, % 7 day 26 day ttrtngth and conloiirt m (pall 217 233 264

7 doy cl 28 doy rirtngth ond contourt

190 (poo 240 298 290 173 'l9» *228 *2S9" Oolomlllc

llml.X • — • . • — ^100 p.1. 191 *233 *2S8 284 40 *76 93 99 60 119 131

Ootomltie • « — 6 inonofiydrQ 1 1 1 69 *89 304 29 O 10 20 30 2T8 lim*. % a 10 179 Fly otb no 1, % Fly otb no I, %

• • — s Figure 17. Immersed unconfined compressive 0 *S3 94 168 strength values obtained for several com• 1 1 1 binations of alluvial clay, lime, and fly 0 to M SB ash 1 for 7- and 28-day curing periods, Rf atti no 3, % 3, % Fly nth ne and strength contour lines for 28-day re• sults. Figure 16. Immersed unconfined compressive strength values obtained for several com• binations of gumbotil, lime, and fly ash 3 for 7- and 28-day cviring periods, and strength contour lines for 28-day results.

7 doy itrtngth 28 day itrtngth ond centeurt 28 day ttrtnglh ond caotgurt D6(pil) 64 175 183 "xij OT "27^ '306 )09lpti) 296 225 240 290 r302 347

200 Pil 300

129 153 193 I88" W *T4 219 '79~ Colclllo Colcltic i?a «' \>«> »« bydrottd li6'^'243 •28S>^ hydrottd r\ \ •llM,« Ilm6,% \200pil XsOO I2S 149 ae 29 *94 *203 236~ [132 239 294 343 ' '=^\ .... .1. ... 02 "204 *2e6 '229 J L 1 1 1 J L 0 10 179 29 Fly oth no 3, % Fly nb na 3, %

7 doy ilranglli 29 day atranglh ond cooloun mgth day ttrtnglh vtd conloura 173 (Pil) 234 213 214 ITS (pin *208 "2TI

173 'l92 *2IS 'l93~ T3 *2ie 2TI Oolomtilc w monohydrot*

• 3 "".« 3, 4B 130 137 *IS2~ 48 *9T :T8 •239-

1 1 1 1 1 1 29 0 10 29 0 10 to 30 0 D 179 Fly ath no 2, % Fly otb 00 2,X Fly oth na 9, % Fly oth no 3, %

Figure 18. Immersed unconfined compressive Figure 19. Immersed unconfined con^ressive strength values obtained for several com• strength values obtained for several com• binations of alluvial clay, lime, and fly binations of alluvial clay, lime, and fly ash 2 for 7- and 28-day curing periods, ash 3 for 7- and 28-day curing periods, and strength contour lines for 28-day re• and strength contour lines for 28-day re• sults. sults. > ' 53

Fly Ash. —The over-all effectiveness of fly ash 3 exceeded that of the other two fly ashes. Fly ash 1 was better than fly ash 2 with dolomitic monohydrate lime, but the effectiveness was reversed with calcitic hydrated lime (fly ash 2 was better than fly ash 1). Strengths from 400 to 500 psi may be obtained with dolomitic lime and fly ash. This IS an adequate strength level. Only fly ash 3 could be used with calcitic lime to stabilize alluvial clay. This is due to the small amount of calcitic hydrated lime required, although the strengths obtained, of the order of 350 psi, are rather low. Discussion Based on this study, no conclusions can be drawn as to the best ratio of lime to fly ash or as to the amount of lime and fly ash that could be used to stabilize any kind of soil. Based on the results obtained with dune sand the amount of lime recommended for sandy or granular soils is from 3 to 6 percent lime and that of fly ash from 10 to 25 percent. Unless fly ash is very high in pozzolanic quality, it should not be used with friable loess. If such a fly ash is available, 3 percent lime and 25 percent fly ash are recom• mended. The use of dolomitic monohydrate lime is favored. The amounts of lime and fly ash best for both alluvial clay and gumbotll soils vary. For gumbotll, between 5 and 9 percent lime and between 10 and 25 percent fly ash are recommended. For alluvial clay, between 5 and 7 percent dolomitic monohydrate lime and between 10 and 25 percent fly ash are recommended. Lower amounts of lime may be used if it is a calcitic hydrated lime. In general, dolomitic monohydrate limes give better strengths with fly ash than calcitic hydrated lime for the curing temperatures used (70F). It should be pointed out that with one fly ash (No. 6) calcitic hydrated lime was more effective than dolomitic monohydrate lime. For low amounts of lime, the calcitic hydrated is more effective than the dolomitic monohydrate in the stabilization of clayey soils with lime and fly ash; at higher lime contents, dolomitic monohydrate gives better strengths than cal• citic hydrated. Fly ash, unless of a high quality, is detrimental in the stabilization of friable loess; in all other soils it was beneficial, giving better strengths than mixtures of soil and lime without fly ash. A total of 22 fly ashes were studied as to their pozzolanic behavior, among them those used in these tests (15). No new information was found here as to the relation between pozzolanic activity of a fly ash and its physical or chemical cliaracteristics. The maximum dry density for the same compactive effort of soil, lime, and fly ash mixtures did not correlate with strength. Density varied with amounts and kinds of lime and fly ash. Dolonutic monohydrate lime gave consistently greater densities in friable loess, lime, and fly ash mixtures than calcitic hydrated lime. Fly ashes of low specific gravity produced lower densities than fly ashes of higher specific gravity. LIME STABILIZATION Lime stabilization of some soils may sometimes not be appreciably benefited by the addition of fly ash. To obtain data to evaluate the use of lime or lime and fly ash, an extensive study of lime stabilization was made. Maximum strengths up to 90 days were recorded, and up to 25 percent of lime was used (Table 5 and Figs. 20 to 22).

Results Dune Sand. -Though sand is not hardened by the addition of small amounts of lime, it was suspected that large amounts of lime might impart some cohesive strength. Therefore quantities of lime up to 25 percent were studied in mixtures with dune sand (Table 5). The large quantities of lime strengthened the dune sand; for instance, a mixture of 54

TABLE 5 STRENGTHS OF DUNE SAND STABILIZED WITH LIME

Dry Immersed Unconf med Compressive Lime Kind Percent (pcf) 7-Day 28-Day 90-Day'' Calcitic hydrated 3 110 0 11 11 6 113 0 11 12 9 117 8 25 31 12 119 19 30 42 15 120.5 30 51 ND 25 112 64 73 ND Dolomitic monohydrate 3 110 0 11 14 6 113 0 15 31 9 116.5 21 29 57 12 119 32 51 93 15 120.5 53 120 ND 25 120.0 112 215 ND

As molded. ''ND = not determined.

I I I I I I I I I I Immersed rnIIIIII1II Soil: alluvial clay Soil, friable loess compressive 90 day J zoo Immersed strength, psi I ^. compressive strength, psi 28 day Colcltlc hydrated lime. %

7 dflv

Colcltlc hydroted itme, I I I I I I I I I I I I Soil: olluviol clay Immersed 4oo 90 doy. ~i I I 1 I r compressive Soil, frioble li 2S doy 90 doy strength, psi zoo Immersed ^7 doy compressive ZBdoy strength, psi 3 6 9 12 Oolomltle monohydrate lime. %

oU»-r I I I I I Figure 21. Strengths obtained by addition 3 6 9 12 of different amounts and kinds of lime to Dolomitic monohydrate lime. % alluvial clay.

Figure 20. Strengths obtained by addition of different amounts of lime to friable loess. 25 percent dolonutic monohydrate lime and 75 percent sand had 7- and 28-day strengths after immersion, of 112 and 215 psi, respectively. But the use of so much lime is not economical. It was also observed that dolomitic monohydrate lime produced much higher strengths than cal• citic hydrated lime. The strengths obtained with lime may be greatly increased by the addition of a fly ash. 55

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Soil' gumbotil ^_^90 day Imirarsed "0°

compressive 28 day_ strength, psi zoo

~^ 7 doy 0 - 1 1 1 1 1 1 1 1 1 II \ 1 1 1 1 1 1 6 9 12 19 18 Calcitic hydroted lime. % I I I I I I I I I I I I I I I I I I I I I _| 90 day Soil gumbotil Immersed *<'<'

compressive 28 day strengtli, psi 2001 7 day

I I \^A'*T I I I I I I I I I I I I I I I I I I I e » 12 19 Dolomitic monohydrate lime, %

Figure 22. Strengths obtained by addition of different amounts and kinds of lime to gumbotil.

The added cohesive strength obtained by the addition of lime to clean sandy soils probably comes mainly from carbonation of the lime, but part of the strength may have been caused by the formation of calcium silicates, although this is not likely to have occurred at the curing temperatures used in this research. Friable Loess. —This soil shows a great pozzolanic activity with lime. It has been pointed out that based on 7- and 2 8-day curing periods the addition of some fly ashes dlmimshes the strength obtained with this loess and lime only, but the pozzolamc action between loess and lime continues and is important beyond 28 days (Fig. 20). Very small amounts of lime are needed to develop the full strength that may be obtained by addition of lime. Six percent of dolomitic monohydrate lime appears to be the best amount; use of greater percentages does not appear warranted. Friable loess should not be stabilized with lime and fly ash unless a very good quality fly ash is used. Six percent dolomitic monohydrate lime gave strengths of 150, 354, and 584 psi for 7, 28 and 90 days curing, respectively. These strengths were actually lowered by the addition of a medium or low quality fly ash. Gumbotil. —In the experiments with gumbotil, lime was added m amounts up to 25 percent (Fig. 22). For every curing period, a percentage of lime was found above which there was no appreciable increase m strength. This "breaking" percentage tends to be higher for the longer curing periods. (This was also observed in the re• sults with alluvial clay.) At least 9 percent of either the dolomitic or calcitic lime is recommended. With dolomitic lime, 200 and 300 psi may be obtained after 7 and 28 days curing, re• spectively. These figures are rather low and may be increased the addition of fly ash, or by substituting some lime for fly ash. Lime and fly ash may compete economically and in terms of strength with the minimum amount of lime required. Alluvial Clay. —The strengths obtained with alluvial clay stabilized with lime were relatively low (Fig. 21). The desirable value of 300 psi after 28 days curing may be obtained with 9 percent dolomitic lime, but for this amount the strength is not im• proved beyond 28 days. A recommended amount of lime is 12 percent, but the strengths obtained with this amount may also be obtained with an economically competitive lime and fly ash admixture.

Discussion This research indicates that a small amount of lime added to soil is not sufficient to obtain the maximum benefits of lime. Observation of 7- and 28-day strengths may lead to choose a smaU amount (Figs. 20, 21, and 22). But when curing periods were continued up to 90 days, the strength gain with time was found to be influenced by the amount of lime. The 7- and 28-day strengths of friable loess (Fig. 20) indicate that 56

3 percent lime is the best amount to stabilize this soil and that higher amounts do not particularly add to strength. But a study of 90-day strengths show that 6 percent of lime is the proper amount. Therefore, the amount of lime needed to stabilize a soil should be determined on the basis of long curing periods as well as short. If a high, long-term strength is desired, the highest economically possible amount of lime should be used. It was also found (Figs. 20, 21, and 22) that calcitic hydrated lime was more effect• ive than dolomitic monohydrate lime in small amounts of about 3 percent. Dolomitic monohydrate lime was more effective than calcitic hydrated in amounts of 6 percent or higher. Consequently, when small amounts of lime are used, the calcitic hydrated type should be favored. For high amounts, dolomitic monohydrate lime should be used. Nevertheless, it appears that for very long curing periods both limes will pro• duce the same strength. MOISTURE-DENSITY RELATIONSHIPS OF CLAYEY SOILS TREATED WITH LIME Some of the moisture-density curves for gumbotll and alluvial clay treated with lime and fly ash had a peculiar shape. There was no distinctive maximum density; it was often undefined. Fly ash was found not to be the cause of this. The shapes of the curves showing the moisture-density relationships of a friable loess-lime mixture follow the concept of a maximum density at an optimum moisture content (Fig. 23). This soil, friable loess, has a relatively low amount of clay, 17 percent. But for mixtures of gumbotll and lime or of alluvial clay and lime there is no defined maximum density for an optimum moisture content. Further, the drier the mixtures, the greater the dry density (Figs. 24, and 25). Both gumbotll and alluvial clay have a very high content of montmorillonltic clay, about 70 percent. It was sus• pected that high amounts of clay, at least of the montmorillonltic type, were the cause of the poorly defined shape of moisture-density curves.

1 1 1 1 1 1 1 1 1 1 Immersed 28 day ^ compressive' 100 strength, psi Iimnereed , coinpressive •--.^r day ~ strength, psi 14 18 22 IMoisture content, %

Mixture proportions 1 1 1 1 1 1 1 1 1 91% friable loess ) M 18 22 2S M 9% calcitic hydrated lime Moisture content. %

Miiture proportiom

89% gumbotil MX dolomitic monohydrate lime

density.

14 18 22 14 18 22 Moisture content, % Moisture content. %

Figure 23. Moisture-density and moisture- Figure 2k. Moistwe-density and moisture- strength relationships of mixture of 91 strength relationships of mixture of 85 percent friable loess and 9 percent cal• percent gumbotil and 15 percent dolomitic citic hydrated lime, compacted at standard monohydrate lime, compacted at standard Proctor compactive effort. Proctor compactive effort. 57

To find if the soil without lime had the same shape of moisture-density curves, some comparative tests were made. For instance moisture-density curves for alluvial clay with and without lime, compacted with the same compactive effort, are shown in Figure 26. A wide range of moisture contents was used in these tests. The curve for straight soil shows a continuous Increase in density as the water content increases up to a maximum density. Higher amounts of water decrease the density. The curve for the soil lime mixture shows a small increase in density with increase in water con• tent for very low amounts of moisture. From then on, the density decreases with the increase in water content. The addition cf lime to soils of high content of montmoril• lonltic clay distorted the shape of moisture-density curves. The moisture-density relationship of montmorillonitic clay soils stabilized with lime is probably affected by the flocculating effects of lime. The lime alters the character• istics of such clayey soils, converting them into a material with the workability of friable soils. At low moisture contents, the flocculating effects of lime impart to clayey soils a highly open structure. This facilitates the expulsion of air which be• comes more important to the increase of density than the lubricating effects of water. The free expulsion of air from a mass containing about one-third void space can in• fluence greatly the final compacted dry density at low moisture contents. The maximum strength does not occur at a point of maximum density (Figs. 24, 25, 26). A second point of maximum strength is initiated for high moisture contents. This is more clearly seen in the 28-day strength curve. Therefore, the concept that compaction should be at the optimum moisture content for maximum dry density should be reviewed. The strength gain or hardening of mixtures stabilized with lime and fly ash comes from the formation of cementitious products rather than from density. A high moisture content maintains a larger supply of water for the hydration process to proceed at a faster rate and/or for longer periods. It is therefore recommended that, in the stabilization of soils with lime or lime and fly ash, the molding or com• pacting moisture content be chosen on the basis of the maximum strength rather than on the maximum density of the mixture.

91Xoilii«oleloy,.^,_

coRipresslve strength, psi Imfflersad compressive «"> ze day lOOXoOwlol cloy no ilnngn^ strength, psi _l I I—I 1 1 J e kT a M n isiassMnce Moisture content, % Moisture content, %

Miiture proportions 91 % alluvial cloy "1 I I I r 1—I—r 9% calcitic hydrated llnis lOOXoSwIol

Dry density

(mod. Proctor

density, compaction), 9IX olluvlol cloy 9X cole. h. Dma pel

M B 22 26 30 6 S n 12 M W IS so 22 S4 2S 2S Moisture content, % Moisture content, %

Figure 25. Moisture-density and moisture- Figure 26. Moisture-density and moisture- strength relationships of mixture of 91 strength relationships of mixture of 91 percent alluvial clay and 9 percent cal• percent alluvial clay and 9 percent dolo• citic hydrated lime, compacted at standard mitic monohydrate lime, compacted at modi• Proctor compactive effort. fied Proctor compactive effoz>t. 58

PORTLAND CEMENT STABILIZATION An evaluation of lime fly ash stabilization is not complete without a comparison of its effectiveness with that of cement stabilization. Plastic soils to be stabilized with cement should be pretreated with lime to flocculate the soil particles and thus facilitate, the mixing process. Alluvial clay and gumbotil are montmorillonitic clay soils of high plasticity needing the lime pretreatment. Consequently alluvial clay was treated with 3 percent lime, and gumbotil with 4 percent in addition to cement. Both lime and cement were added together. Portland cement in the proper amounts stabilized any of the four soils tested (Figs. 27 to 30). Good strengths were obtained with at least 8 percent cement in dune sand, 6 percent in loess, 3 percent lime plus 6 percent cement in alluvial clay, and 4 percent lime plus 5 percent cement in gumbotil. These mixes gave 7-day strengths over 300 psi. Most of the final strength was develop• ed in the first 7 days. The rate of in• crease after 7 days was not very pro• nounced, except with the loess. In the length of time needed to develop strength lies an important difference between ce• ment and lime fly ash stabilization. The early strengths for lime fly ash were low, but the strength steadily increased with time at a fairly good rate. For long curing periods, the strengths with lime 600 and fly ash and with cement tended to InunAf S6d equalize. In many instances, strength eempranlve «oo (trangth, p

Figure 2?. Strength of dune sand cement mixtures. 600 Immaread compraaslva atrangth, pal 400

Soili frlabla loan

Curing parlod, doya Figure 28. Strength of friable loess ce• ment mixtures.

3\lima 4 9X ceraant Immaraad 3Xlima + 6X camant compraaalva Immaraad •"ol 4X luna-f 3X caiMiit atrangth. pal compraaalva 3X lima * 3X camant atrangthr-pal .

Soil, gumbotil Solli ollu«lol cloy

Curing parlod, doya Curing parlod, doya Figure 29, Strength of gumbotil cement Figure 30. Strength of alluvial clay ce- mixtures. ment mixtures. 59 monohydrate limes were used with gumbotil to compare their effectiveness in changing the plasticity in the lime treatement. The results were erratic and do not show con• sistently better improvements, based on strength, with one or the other lime. Further tests should be conducted to compare the effectiveness of both limes in treatments for soil cement stabilization. In the meantime, the cheapest one available is recommended.

DURABILITY EVALUATION A few mixes were selected with the proper amount of lime and fly ash for each soil, to compare them with mixes in which lime and/or cement was the stabilizer. The comparison mcluded freeze-thaw testing of selected mixes. The proportions used in the soil, lime, and fly ash mixtures and the lime and fly ash used were calculated to compete with the amount of cement and/or lime needed to stabilize the same soil. Use was made of the Iowa State equal-cost-lime method for soil, lime, and pozzolan mix design (5). The following were assumed:

1. Eight percent cement is required to stabilize dune sand. 2. Ten percent cement or 9 percent dolomitic monohydrate lime is required to stabilize friable loess. 3. Three percent lime and 9 percent cement are required to stabilize alluvial clay. 4. Four percent lime and 8 percent cement are required to stabilize gumbotil.

I I Soil dune sand S«il frloMa Ions

Lima or cement, % (4 rnidS • Lime or cement, ifc 4

19 20 29 Fly Qth, %

Fly oth. %

Soil gumbotil —I 1 r- i-(24 Soil: olluvlol cloy Lime or lime and c«ment,%

L me or lime and cement, % 4

20 30 40 Fly ash,% 20 30 40 90 Fly ash. %

Figure 31. Equal-cost-line charts for soil stabilized with selected admixtures of lime- fly ash compared with mixt\ires of soil-lime-cement or soil-cement. 60

5. The cost of lime or cement is the same, about $22 a ton. 6. The cost of fly ash is one-sixth that of lime or cement. 7. The cost of handling two materials (lime and fly ash: lime and cement), instead of one if stabilized with cement or with lime only, is equal to the cost of 1 percent of cement.

Dune Sand The sand, lime, and fly ash equal-cost-line graph for the selected mixtures is given in Figure 31. All the mixtures within the triangle ABC have the same cost or are cheaper than the required 8 percent of cement needed to stabilize dune sand. Based on 28-day strength requirements, lime and fly ash may be used economically to stabilize sandy soils (Table 6). Either lime and fly ash mixtures or lime and fly ash mixtures with chemical additives withstood the severity of freezing and thawmg tests and had enough residual strength to be considered adequately stable. A good quality fly ash (No. 3) was used in these tests. These results may not be reproduced with all kinds of fly ash. All five selected dune sand, lime, and fly ash mixtures gave 28-day strengths equal or greater than dune sand and cement for the same curmg period. It has been estimated that after freezing and thawing, the stabilized soil specimens should have a minimum strength of 250 psi (4, 6). This value was surpassed by all mixtures (p; col., Table 6). It is desirable that soil-stabilized specimens show an index of resistance to the effects of freezing (R*) of at least 80 percent to withstand Iowa climatic conditions satisfactorily (4, 6). Only mixes 4 and 6 gave indexes of resistance lower than 80 percent; however, they had Rf values of 78 percent, which should be adequate, inas• much as the values of pj and p^ are over 400 psi. Some mixtures contmued gaining strength during freezmg and thawing cycles and/or during immersion. None of the mixtures showed any visible damage from freeze-thaw, nor did they show any expansion. The as-molded dry density of the several mixtures changed by as much as 12 pcf, but there was no relationship between density and strength values.

Friable Loess Only one loess, lime, and fly ash mixture could compete economically and on a strength basis with loess and cement or with loess and lime mixtures. That one mix• ture was 72 percent loess, 3 percent dolomitic monohydrate lime, and 25 percent fly ash 3 (Table 7). It was compared with mixtures of the same soil stabilized with 9 percent dolomitic monohydrate lime or with 10 percent cement. The amount of 9 per• cent dolomitic lime was based on a previous evaluation using different amounts of lime (Fig. 20). Ten percent cement was chosen based on the ASTM requirements to stabilize this kind of soil (1).

TABLE 6 DURABILITY EVALUATION OF SELECTED ADMIXTURES TO STABIUZE DUKE SAND

Proportions Dry Unconfined Compressive Strength'' Mix No. Density^ Rfl (*) 28-DayC Pf^ p^^ (Pd) (*) 1 Sand, 92, p. cement, 8 112.6 474 507 517 98 2 Sand, 73, dol. lime, 3, fly ash 3, 24 124.3 792 821 966 85 3 Sand, 76, dol. Ume, 4; ny ash 3, 17.5 124.4 646 634 674 94 4 Sand, 82, dol Ume, 3; fly ash 3, 15 123.8 390 ND ND ND S Sand, 82, calc. Ume, 3; ny ash 3, 15 123.1 120 ND ND ND "As Doldsd. ' not dsteimlned. Âfter 28 days curing and 2ii far ijmssrsion in distilled water. Âfter 28 d^s caring, 2k hr Ijnmersion In distilled water, and 10 freeae-thaw cycles. °After 28 dsgrs curing and 11 dâ Ijmnersion in distilled water.' 61

Strengths of ^00 psi were obtained with all selected mixtures after a curing period of 28 days. The mixtures exposed to 10 cycles of freezing and thawing showed a strength either close to or well over 400 psi, which is adequate for a base course. The indexes of resistance were over the minimum of 80 percent desired. Friable loess can be stabilized with cement, lime, or lime and fly ash for use as a road base course material. The 10 percent cement mixture gives strengths that are much higher than those obtained with mixtures with lime or lime and fly ash. It appears that a lower amount of cement might also adequately stabilize friable loess. For in• stance, mixture 15 (Table 7) of 6 percent cement and 94 percent loess gave a strength of 495 psi after 28 days. This strength is comparable with that obtained with the se• lected mixtures of loess and lime and of loess, lime, and fly ash. Therefore, it is possible that 6 percent cement would be adequate for stabilizing this soil. Cement rather than lime or lime and fly ash, should be used to stabilize friable loess, unless the price of lime is much below that of cement, or a high quality fly ash is cheaply available.

Gumbotil Two fly ashes (Nos. 2 and 3) were used with dolomitic monohydrate lime to stabilize gumbotil and make an evaluation of the durability of these mixtures. The proportions

TABLE 7 DURABILITY EVALUATION OF SELECTED ADMIXTURES TO CTABILIZE FRIABLE LOESS

Dry Unconf ined Compressive Strength'' (psi) Mix No. Proporttons Density^ 28-Day'= p,* Pg* (*) (pel) a' 11 Loess, SO, cement, 10 103.5 645 567 682 83 12 Loess, 91; dol. mon. lime, 9 100.8 396 387 428 90 13 Loess, 72, dol mon. lime, 3, fly ash 3,25 99.1 462 441 521 85 14 Loess, 91; cement, 9 103.5 566 ND ND ND 15 tioess. 94; cement, 6 101.3 495 ND ND ND Âa molded. ^VD - not detemined. ^Âfter 28 dfiQrs caring and 21; hr Immnrsion In distilled water. *Âfter 28 daya caring, 2k hr Imiiarslon In distilled water, and 10 freeae-thaw ^les. Âfter 28 days caring and 11 d^ Imersion In distilled water.

TABLE 8 DURABILITY EVALUATION OF SELECTED ADMIXTURES TO STASUZE GUMBOTIL

Unconllned Compressive Strength** (psi) Mix No. Proportions Dry Density^ 28-Daye (pet) 21 GumbotU, 88, dol. mon Ume, 4, cement, 8 95.1 705 550 634 87 22 GumbotU, 69, dol. mon lime, 6, fly ash 2, 25 90.0 606 534 642 83 23 GumbotU, 69, dol. mon. lime, 6, fly ash 3, 25 94 1 682 529 780 68 24 GumbotU, 91, dol. mon Ume, 4, cement. 5 93.3 534 ND ND ND

HD - not detamlned. ^Ittar 28 d^s curing and 21; br immersion in distilled vat«r. ''ifter 28 d^ earing, 21i br iimsrslon in distilled water, and 10 freeze-thav cycles. 'After 28 d^s coring, and 11 dsgrs immersion in distilled water. 62 used, based on previous results, were 69 percent gumbotil, 6 percent lime, and 25 percent fly ash (Table 8). The strengths previously obtained with lime and gumbotil were rather low (Fig. 22) and do not recommend the use of straight lime stabilization for base course construction with gumbotil; therefore, the use of lime was not evaluated. The amount of cement to stabilize gumbotil, based on ASTM requirements, is about 12 percent W. This was the amount used in the durability studies. Without using lime, it would be impossible to field-mix gumbotil with cement, be• cause gumbotil is an extremely plastic clay soil. Hence, 4 percent of the required amount of cement was replaced by lime to decrease the plasticity of the soil. Both mixtures in which lime and fly ash were the stabilizing agents gave strengths comparable with that of the mixture of gumbotil stabilized with lime and cement. The strengths after 28 days curing were above 600 psi for both immersion periods and for all three mixes selected for the freeze-thaw studies. The strengths after freezing and thawing cycles were about 540 psi for the three mixes. These strengths are very good for this high-clay content soil and warrant the use of these mixtures as a base course material. The indexes of resistance are adequate for mixes 21 and 22 (Table 8). Mix 23 had a rather low index of resistance of 68 percent. This index value is due to a substantial gain of strength during the 11-day immersion period. If the strength after the Iowa freeze-thaw test is still 529 psi, gumbotil may be used in a base course when stabilized with the materials and proportions of mix 23; that is, 69 percent gum• botil, 6 percent dolomitic monohydrate lime, and 25 percent fly ash 3. The strength obtained with mix 24 shows that good strengths may be secured with lesser amounts of lime and cement. However, strengths equivalent to those of mix 24 may be also obtained with lesser amounts of lime and fly ash than used in mixes 21 and 22. It may therefore be concluded that gumbotil can be stabilized with lime and fly ash, competing economically with cement. The strengths obtained with the specimens prepared with gumbotil for the durability evaluation studies had greater strengths than specimens made with the same admix• tures in previous studies. This lack of reproducibility of strength was only found with gumbotil. It is possible that specimens prepared for the durability studies were benefited during curing by temperatures slightly higher than in the other studies, causing the strength differences noted.

Alluvial Clay About 12 percent cement is the smallest amount needed for stabilizing alluvial clay, according to ASTM tests (1). The lime and fly ash combinations that might give strengths comparable with those obtained with cement were those made with dolomitic monohydrate lime plus fly ash 3 (Table 9). Mixtures of alluvial clay and lime did not show high strength (Fig. 21), so they were not evaluated here.

TABLE 9 DUBABOJTY EVALUATION OF SELECTED ADMIXTURES TO STABILIZE ALLUVIAL CLAY Unconflned Compressive StrenKthO (psi) ProporUons Dry 28-Daye p,"* p/ R,' Mix No. (*) Density^ (pcf) (*) 31 Alluvial clay, 88, calc. hyd. Ume, 3, cement, 9 94.9 574 498 527 94 32 Alluvial clay, 69, dol. mon. lime, 6, fly ash 3, 25 93.6 513 475 563 84 33 Alluvial clay, 91, dol. mon. Ume, 3; cement. 6 94.0 470 ND ND ND "AS molded. \d - not determined. °After 28 dc^a curing and 21i br ijumerslon IA distilled water. '^After 28 days curing, 2U hr limnerslon in diatllled water, and 10 freeze-thaw eyclea. "After 28 d^ curing and U d^s iimeralon in dlatilled water, f 100 P, 63

Instead of using the full requirement of 12 percent cement, 3 percent lime and 9 percent cement were used. The lime was used primarily to give the soil friable characteristics that would allow better mixing with the cement. Lime also may counter• act any adverse effects from the somewhat high orgamc matter content of the alluvial clay. Both mixtures when tested by freezing and thawing gave strengths of around 500 psi for any of the three testing treatments tried. The indexes of resistance were also above the minimum desired. It appears that alluvial clay stabilized with the proper lime and fly ash admixture may have strengths and durability comparable with those of alluvial clay stabilized with cement, and may be economically competitive as well (Fig. 31). Mix 33, of 91 percent alluvial clay and 9 percent lime and cement, was not evaluated in freezing and thawing but gave seemmgly adequate strength. It is also possible that mixtures containing smaller amoimts of fly ash than mix 32 might give strengths equal to those of mix 33.

CONCLUSIONS On the basis of the investigation conducted, the following conclusions are made: 1. There is no optimum amount or ratio of lime and fly ash for stabilizing all soils. The amount and proportions of lime and fly ash to use depend greatly on the kinds of fly ash and soil, and somewhat on the kind of lime. For granular soils the amount of lime should be between 3 and 6 percent; the amount of fly ash, between 10 and 25 percent. For clayey soils, the amount of lime should be between 5 and 9 per• cent; the amount of fly ash, between 10 and 25 percent. 2. Dolomitic monohydrate lime generally gives better strengths in soil, lime, and fly ash mixtures than calcitic hydrated lime in normal amounts and when cured at ambient temperatures. 3. At low lime contents (about 3 percent) calcitic hydrated is more effective than dolomitic monohydrate for stabilizing clayey soils with or without fly ash; at higher lime contents, dolomitic monohydrate gives better strengths than calcitic hydrated. 4. The fly ashes used were beneficial to soil and lime mixtures for all soils ex• cept friable loess. With the friable loess, only a high quality fly ash was beneficial to loess and lime mixtures. 5. The moisture-density curves of montmorillonitic clay soils stabilized with lime are affected by the flocculating effects of lime. Sometimes the curves do not show a maximum density. 6. Portland cement is a very effective stabilizer for most soils. The strength gain of mixtures of soil and cement is rapid, and a large percentage of ultimate strength is developed in a relatively short time. But compacted soil, lime, and fly ash mixtures gain strength slowly. Full strength may not be developed for several years. The comparison of soil and cement and soil, lime and fly ash test specimens should be made on the basis of 28-day curing. After this period, soil cement should have de• veloped about 90 percent of the ultimate strength, and soil, lime, and fly ash only about 50 percent, depending on the soil, lime, and fly ash used. 7. Selected compositions of dune sand, lime, and fly ash can compete in strength, freeze-thaw resistance, and cost with mixtures of the same soil and cement. 8. Friable loess is most effectively stabilized with cement. If lime is cheap and a good quality fly ash is available, lime or lime and fly ash may compete with cement for stabilizing friable loess. 9. Additions of fly ash are beneficial to gumbotil and lime mixtures. Selected gumbotil, lime, and fly ash mixtures show good resistance to freezing and thawing, and may compete with gumbotil cement stabilization. 10. Additions of fly ash are beneficial to alluvial clay and lime mixtures. Lime- fly ash stabilization of alluvial clay may compete economically and strengthwise with cement stabilization. 64

ACKNOWLEDGMENTS The material for this paper was obtained as part of the research being done under Project 449-S of the Iowa Engineering Experiment Station, Iowa State University of Science and Technology. This project is under contract with the Iowa Highway Re• search Board of the Iowa State Highway Commission as their project HR-82. Sincere thanks are given to W. N. Handy of the Walter N. Handy, Inc., and J. R. Wilde of the Detroit Edison Company for furnishing and analyzing the fly ashes used, to U. S. Gypsum Company, Western Lime and Cement Company and Rockwell Lime Company for samples and analyses of the limes used, and to Penn-Dlxie Cement Corporation for the cement used in this investigation.

REFERENCES 1. "Procedures for Testing SoUs." ASTM Philadelphia (1958). 2. Barnes, H. F., "Effects of Sodium Carbonate on Lime StabiUzed Clay Soils." ^ M.S. thesis, Iowa State Univ., Ames (1959). 3. Crosby, R. L., "Lime-Fly Ash Ratio and Admixture Content Versus Strength of Stabilized Clayey Soil." M. S. thesis, Iowa State Univ., Ames (1957). 4. Davidson, D. T., and Bruns, B. W., "Comparison of Type I and Type m Portland Cements for Soil StabUization." HRB Bull. 267: 28-45 (1960). 5. Davidson, D. T., and Handy, R. L., "Section on SoU Stabilization." In "Highway Engineering Handbook." K. B. Woods (Ed.), McGraw-HUl (1960). 6. Davidson, D.T., Mateos, M., and Barnes, H. F., "Improvement of Lime Sta• bilization of Montmorillomtic Clay SoUs with Chemical Additives." HRB Bull. 262:33-50 (1960). 7. Davidson, D. T., Sheeler, J. B., and Delbridge, N. G., "Reactivity of Four Types of Fly Ash withXime." HRB Bull. 193:24-31 (1959). 8. Goecker, W.L., Moh, Z.C., Davidson, D.T., and Chu, T.Y., "StabUization of Fine and Coarse-Grained SoUs with Lime-Fly Ash Admixtures." HRB Bull. 129:63-82 (1956). 9. George, K. P., "Development of a Freeze-Thaw Test for Evaluating Stabilized Soils." M.S. thesis, Iowa State Univ., Ames (1961). 10. Hicks, L.D., "Design and Construction of Base Courses." HRB Bull. 129:1-9 (1956). 11. Lunger, R.R., "Use of Sodium Carbonate with Lime-Fly Ash for Stabilization of Sand." M.S. thesis, Iowa State Univ., Ames (1958). 12. O'Flaherty, C. A., "StabUization of GraveUy and Stony SoUs with Lime and Fly Ash." M.S. thesis, Iowa State Univ., Ames (1958). 13. O'Flaherty, C.A., Mateos, M., and Davidson, D.T., "Fly Ash and Sodium Carbonate as Additives to SoU-Cement Mixtures." Paper, 41st Annual Meeting, Highway Research Board (Jan. 1962). (To be published) 14. Segal, R., "Effect of Calcium-Magnesium Ratio on Lime Reactivity with Fly Ash." M.S. thesis, lowr. State Univ., Ames (1957). 15. Vincent, R.D., Mateos, M., and Davidson, D.T., "Variation in Pozzolanic Be• havior of Fly Ashes." Proc., ASTM (1961). 16. Wang, J.W.H., David ..on, D.T., Rosauer, E.A., and Mateos, M., "Com• parison of Various Commercial Limes for Soil Stabilization." HRB Bull 335:65- 79 (1962).