Conc Mix Opt

7/28/2019 Conc Mix Opt

1/8

Concrete mix proportions for a given need can be optimized using

coarseness factor, mortar factor, and aggregate particle distribution

Concrete MixtureOptimization

Concrete mixture optimization

involves the adaptation of

available resources to meet

varying engineering criteria,

construction operations, and eco-

nomic needs. Economic considera-tions include materials, delivery,placement, and progress time re-

lated costs. Optimization is often

informally taken into consideration

before and during construction on a

non-quantitative basis by "adding

half a bag of cement," "cutting therock 100 pounds and replacing it

with sand," or adding a high-range

water-reducer. When mixtures are

optimized on a quantitative basis,

construction productivity will be

improved, durability increased and

both materials and constructioncosts reduced.

Methods for selecting cementi-

tious materials factors, entrained

air, and initial aggregate propor-

tions have been described in many

reports and will not be discussed

here. The purpose of this article is

to provide a quantitative method

for optimizing aggregate propor-tions and making adjustments dur-

ing progress of the work. The dis-

cussion is based on use of rounded

to cubical aggregates and mixtures

with ASTMC 494, type A or D ad-mixtures. Research leading to these

findings has spanned 15 years.

While only one research study is re

ported here (below), the findingshave been demonstrated many

times. From this study, we con

cl ude:

Keywords: aggregate size; concretes; costs; mix

proportioning; mortars (material); optimization;

particle size distribution; performance; quality

control; standards.

The accepted practice of estab-

lishing constant mixture propor-

tions by weight contributes to prob-

lems arising from variability in ag-

gregates and construction needs.

The method for selecting trialproportions is of minimal impor-

tance. Arbitrary means are as effi-

cient as complex procedures. Theonly meaningful factors are the

characteristics of the composite.

Once a composite is identified as

fulfilling a need, that combinationof materials and adjustment proce-

dures can be translated into a math-

ematical or graphical model as a

mixture design. This should includeprocedures for making adjustments

based upon statistical data and

variations in materials and con-struction needs. A mixture designmay be adaptable worldwide and

used indefinitely as long as aggre-

gate characteristics are similar ex-

cept for gradation and specific

gravity.

Mixtureproportions are the con-crete producer's solution to the de

sign, using those sound resourcesthat are available at the lowest

price.

Current ASTMand similar ag-

gregate grading limits do not con

tribute to mixture optimization, assuch standards do not address gra

dations of the blends.* Aggregates

that do not meet ASTM C 33 gra-

dation requirements, but are other

wise acceptable under a quality

standard, can be used to with equal

ease to produce high quality concrete

if they can be controlled to

produce a consistent, well-graded

composite.

`ASTM now recognizes this problem and a task

group has been set up to introduce a revision that

will provide for combined aggregate gradings in

addition to coarse and fine aggregate gradings.

Construction needs are becoming

increasingly complex and must be

considered second only to engineer-ing criteria when selecting mixture

design alternatives.

There are three principal factors

upon which mixture proportions

can be optimized for a given need

with a given combination of aggre-gate characteristics:

Fig. 1 Well graded mixture.

Fig. 2 Gap graded mixture.

June 1990 33


2/8

Table 1 Aggregate characteristics.

The relationship between thecoarseness of the two larger aggre-gate fractions and the fine fraction. Total amount of mortar. Aggregate particle distribution.

Mixture objectivesIt is difficult to picture the relation-ship of particles and their behaviorduring concrete mixing, deliveryand placement. Construction of alaid-up stone wall is, in part, com-parable to a concrete mixture. Themason selects the large stones andfills major voids with smaller stones

and bonds them together with mor-tar. The amount of mortar neededis a function of the relationship be-

tween the two stone sizes. If thelarge stones are rounded or squareand predominately one size, andthere are no smaller sizes, theamount of mortar needed to fill the

voids is increased.There is a major difference be-

tween the stone wall and the con-crete. The wall is static but the con-crete must have the rheological

properties necessary to providemixture stability, mobility, andcompactability.

Fig. 1 represents the profile of aconcrete composite with a good dis-tribution of large and smaller stoneparticles and the mortar to coat allsurfaces and fill the remainingvoids. Fig. 2 represents a conditionwhere there are no intermediateparticles. Since the volume filled bythe smaller stone particles cannot beoccupied by the large particles, that

volume must be provided by in-creasing the mortar. Increased mor-tar means increased sand, cement,and water. Such increases do notlead to the casting of high concretequality. However, it is not alwayspossible to place and finish mix-tures that are optimized for engi-neering needs alone. The mixturedesign must be compatible with theconstruction process to be used.

The concrete represented by Fig.1 is a uniformly-graded mixture andFig. 2 is a gap-graded mixture.While either concrete can beblended to produce almost anygiven strength, there is a vast dif-

ference in rheological properties.Normally, gap-graded or near gap-graded mixtures contain a greateramount of coarse particles thanshown here but that has an adverseeffect upon pumpability and finish-ability.

Coarseness factor chartThe Coarseness Factor Chart wasdeveloped during an investigationconducted under contract with theU. S. Army Corps of Engineers,

Mediterranean Division, for con-struction of the Saudi Arabian Na-tional Guard Headquarters, Ri-yadh, Saudi Arabia. The architec-tural design required constructionof white cement cast-in-place con-crete, sandblasted to produce a uni-form exposed aggregate surface.The investigation objective was toconfirm the potential for a contrac-tor to produce the specified finishand compressive strength.

Riyadh aggregates met the re-quirements of ASTM C 33 exceptfor gradation. Samples of local ma-

terials were shipped to MaterialsTesting Laboratory, Athens,Greece, for study. That laboratoryhad previously been the regionallaboratory for the MediterraneanDivision but was privately owned atthe time. The co-investigators werethis author and Wilhelm Voelker,Mediterranean Division.

The aggregate gradations andtechnical data, including the blendof the two sizes of wadi gravel, areshown in Table 1. Gravel B wasused for another series of mixtures

but the proportions were poorly se-lected. The results are used hereonly to indicate a range where theaggregate blend was too harsh toproduce placable concrete.

Five different mixture propor-tions were selected to determine themost suitable for use in casting thefinal samples. The proportions werebased upon the ACI 211 procedureexcept that the trial mixtures usedcontained 75, 81, 86, 90, and 93

percent of the coarse aggregate dryrodded unit weight. When ex-

pressed on the percent of aggregatebasis, the coarse aggregate variedfrom 63.3 to 76.6 percent of the to-tal aggregate absolute volume.

Since architectural samples wereto be consolidated by vibrator andthe slump was to be less than 3 in.(75 mm), the cylinders were consol-idated with the same equipment.That decision was important to

evaluate the response to equipmentused in the construction and to

3 4 Concrete International


3/8

Table 2 Trial mixture data, Athens tests.

Mix data Initial Tests Architectural

% of DRUW* 75 8 1 86 90 93 83 86

Coarse Aggregateol o 63.3 68.0 71.8 74.6 76.6 69.8 71.8

Cement lb 564 564564 564 564 564 564

Fine Aggregate lb 1175 1035 920 835 775 975 920

Coarse Aggregate lb 2075 2240 2380 2490 2575 2300 2380

Water lb 292 272 260 287 293 304

Admixture oz 23 23 23 23 23 23 23

w/c 0.52 0.48 0.46 0.51 0.52 0.54

Slump in. 2.25 3.0 2.0 2.0 2.25 2.5 2.0

Compressive 7-Day 2650 3080 3400 2870 3110 3600 3080

strength (psi) 28-Day 3588 4303 4393 4222 4233 4680 4647

*Dry rodded unit weight

properly consolidate the high coarseaggregate factor mixtures.

The water estimated for eachmixture was based upon the as-sumption that as fine aggregate wasdecreased, the water requirementwould decrease linearly. The firstbatch cast was the 75 percent mix-ture and the water was exactly asanticipated. From that point for-ward, the design water was eithertoo much or not enough.

Table 2 is a summary of the pro-portions and the test results with theactual water used. That series oftests were done using locally avail-

able grey cement. The second seriesfor architectural samples andstrength were cast using 83 percentand 86 percent coarse aggregatefactors and white cement. The wa-ter demand was higher with thewhite cement than with the local ce-ment due to differences in finenessof grind.

When the construction contractwas awarded, a training programfor Corps site personnel, archi-tect/engineers, and contractors washeld in Dallas, Texas. Aggregates

proposed for use in the project wereshipped to Dallas and used to pre-pare trial mixtures. The tests wererepeated with both Saudi and Dal-las aggregates and the results werecomparable. The optimum mixturewas used in construction with out-standing success.

Fig. 3 shows the results of testsgraphically. The X-axis is the per-cent of aggregate passing the No. 8(2.36 mm) sieve and the Y-axis rep-

resents both the compressive

strength and the water-cement ra-tio. From Fig. 3, the maximum

strength was produced at the lowestwater-cement ratio. The strengthcurve in Fig. 3 is comparable toProctor Curves used in soil testingto evaluate maximum density ofcompacted materials. The objectiveof materials blending for strength isto fill voids with sound, inert fillerto reduce the volume of binderneeded to produce a sound prod-uct. Portland cement concrete is nodifferent except for adjustments forconstruction needs. Fuller &Thompson' reported a means for

testing aggregate blends to provideminimum voids.

Each mixture was placed in apile, a vibrator inserted, and the re-

sponse observed. The 75 percentmixture responded sluggishly. The93 percent mixture responded rap-idly but voids were left betweensome of the coarse aggregate parti-cles after consolidation. The 86 per-cent mixture responded almost in-stantaneously and resembled a mix-ture with a 6 in. (150 mm) slump. Itwas apparent the particle distribu-tion of the mixture and response tovibrator are related.

Fig. 4 is a grading chart showingthe aggregate gradations and the

combined gradations of the coar-sest, finest, and optimum mixtures.

Fig. 4 Combined aggregate gradations.

June 1990 35


4/8

Fig. 5 Coarseness factor chart.

Optimizationcontinued

The chart used is divided into threesegments identified as Q, I, and W.

This was based on comments byother mix researchers about theamount and function of the "inter-

mediate aggregate" particles. Inter-mediate aggregate is defined asthose particles that pass the '/e in.

(9.5 mm) sieve but were retained onthe No. 8 (2.36 mm) sieve. The let-ter identifications were based on:Q The plus 3/8 in. (9.5 mm) sieveparticles are the high quality, inertfiller sizes. Generally, the more thebetter because they reduce the needfor mortar that shrinks and cracks.I The minus 3/8 in. (9.5 mm), plusNo. 8 (2.36 mm) sieve particles arethe intermediate particles that fillmajor voids and aid in mix mobilityor, if elongated and sharp, interfer-

ence particles that contribute to

mixture harshness.W The minus No. 8 (2.36 mm)sieve particles give the mixtureworkability, functioning as do ballbearings in machinery. Due to otherconnotations of this term, it is pos-sibly a poor choice but was selectedbecause workability at a given con-sistency is largely determined by thecharacter and amount of this por-tion of a mixture.

It was found that the aggregatesource was immaterial. What was

i mportant was the combined grad-ing curve. Of the three curvesshown, only one produced the opti-mum concrete to meet constructionneeds and produced the higheststrength. Based upon these and laterobservations, there are three impor-tant principles that can be stated:

For every combination of aggre-gates mixed with a given amount ofcementitious materials and cast at aconstant consistency, there is an

optimum combination which can becast at the lowest water-cement ra-

tio and produce the higheststrength. The optimum mixture has theleast particle interference and re-sponds best to a high frequency,

high amplitude vibrator. The optimum mixture cannot beused for all construction due tovariations in placing and finishingneeds.

A means to translate the findingsinto a usable reference was the nextstep. It is accepted concrete tech-nology that as the coarse aggregate

becomes finer, the need for sand tofill voids increases. As the sand be-comes finer, the amount should bereduced. These principles addresssand generically. As much as 20percent or as little as 0 percent ofthe sand can pass the 3/8 in. (9.5 mm)sieve and be retained on the No. 8(2.36 mm) sieve. They and the samesizes that occur in the coarse aggre-gate should be identified as inter-mediate particles. To properly ana-lyze a mixture, it is necessary toseparate the aggregates by sieve size

and study the distribution of thevarious sizes.

It was observed from the studiesand literature that a simple theorycould be stated: "The amount offine sand required to produce anoptimum mixture is a function ofthe relationship between the twolarger aggregate fractions." Laterthe following was added: "Theamount of fine sand needed to op-timize a mixture is a function of the

amount of cementitious materials inthe mixture."

This relationship was showngraphically in Fig. 5 with the plus

in. (9.5 mm) sieve as a percent of allplus No. 8 (2.36 mm) sieve particles(Q/[Q+I] ) shown on the X-axisand the percent of the total aggre-gate, with and without adjustmentfor cementitious materials factor,shown on the Y-axis. A trend ofneeds can be defined within thisframework.

The particle distribution of anymixture can be calculated and theresults plotted on the CoarsenessFactor Chart. The chart originallyconceived is identical to Fig. 5 ex-

cept for minor differences in style.The amount of the fine aggregate

in a mixture must be in balance withthe needs of the larger, inert parti-cles. If there is too much sand; themixture is "sticky," has a high wa-ter demand, requires more cementi-tious materials to produce a givenstrength, increases pump pressures,and creates finishing and crazing

problems. If there is not enoughsand, the mixture is "bony" and

36 Concrete International


5/8

creates a different set of placing andfinishing problems.

The results of the Athens testsand analysis of numerous other

mixtures was the basis for the trendbar location. Some of the pointsplotted for representative mixturesare shown on Fig. 5. The resultsproduced by these mixtures rangedfrom surprisingly good, highstrength, workable mixtures tothose which were a part of litiga-tion or claims. Data from earlyhigh-strength concrete was also in-cluded. The results of tests that

were not acceptable when the Ri-yadh 3/8 in. (9.5 mm) aggregate wasused as shown and helped define the

upper end of the trend bar.The trend bar is a reference only.If the aggregates are well- gradednatural sand and gravel or cubicalcrushed stone, the optimum mix-ture combined grading can plot inor near the trend bar. Such mix-tures generally must be placed bybottom drop buckets or by pavingmachines. The water demand forthese mixtures will probably be thelowest possible. The mixture willrespond very well to a large, highfrequency, high amplitude vibrator

even at a low slump. It cannot bepumped and can't be readily fin-ished in building slab construction.As the construction configurationand placement techniques vary fromthe ideal, the amount of fine parti-cles must be increased.

The amount of fine aggregateneeded is also influenced by theamount of cementitious materials.As cement content is varied, thesand content should be adjusted.One 94-lb (42.6-kg) US bag of ce-

ment is approximately equal to 2.5

percent of the combined aggregate.The adjusted amount of workabil-ity particles are identified by the

abbreviation W-Adj (Workability-

Adjusted). From the Riyadh tests,the base relationship for the minusNo. 8 (2.36 mm) particles W is thevolume equal to six US bags (564 lb[255.8 kg]) of cement. At that fac-tor W and W-Adj are identical. Ascement factors vary from 564 lb(255.8 kg), adjustments are madebased upon the absolute volume of

cement. If the cement factor ishigher than 6 bags, W-Adj will behigher than W and vice versa.

General-use mixtures are fre-

quently plotted 5 to 6 percentagepoints above the trend bar. For agiven combination of materials, it is

necessary to determine the opti-mum relationship for varying needs.When changes in materials grada-tions occur, adjustments can be cal-culated and the new aggregate pro-portions selected to closely approx-i mate the original mixture.

Mortar factor

The Mortar Factor is an extensionof the Coarseness Factor Chart.

The mortar consists of fine sand(minus No. 8 [2.36 mm] sieve) andthe paste. With reasonably soundaggregates properly distributed, it isthe fraction of the mixture that has

a major affect upon the engineer'sinterest in strength, drying shrink-age, durability, and creep. It is alsothe segment that provides the con-tractor's need for workability,pumpability, placeability, and fin-ishability. Thus, it is the amount ofmortar that is at the center of con-flict of interests. Neither should be

dominant. A mixture that is opti-mized for strength and shrinkagebut can

't be properly placed and

compacted will perform poorly re-gardless of the water-cement ratioshown on reports.

Most problems are caused bysand gradation variations that in-crease or decrease the amount ofminus No. 8 (2.36 mm) sieve parti-cles. Such variations affect the wa-ter needed to produce a given slumpand cement to maintain strength.The cement is seldom changed. The

need for mix adjustments was con-firmed and reported by Bloem. 2

It might seem that this is the mostreliable way to evaluate mixtureproportions, but there are pitfalls.Calculated mortar content is heav-ily influenced by water and en-trained air. An entrained air toler-ance of 1 percent of the volumeis the equivalent of allowing the

volume of water to vary slightlymore than 33 lb/yd' (20 kg/m').Such a variation can affect mortar

content by 0.02 percent and con-tribute to problems. In addition, asentrained air varies, water demandvaries so the effect of the combined

tolerances can be significant andcause construction problems.The mortar factor can be used to

judge the adequacy of the waterprovided in initial mixture propor-tions. ACI 211 provides guidancefor selection of water for each ag-gregate size. This is fairly accuratewhen adjustments are made for wa-ter-reducing admixtures. Generallywhen W-Adj is high but the mortaris relatively Iow, there is not enoughwater provided in the proportions.

The mortar factor needed for

various construction types varies. Amat foundation with the concreteplaced by chute requires less mortarthan the same strength concrete castin a thin slab to be trowel finished.Unless aggregate proportions areadjusted to compensate for differ-ing needs, changes in slump to in-crease mortar through the additionof water is the only option open tothe contractor.

Construction requirements thataffect mortar needs should be con-sidered when optimizing a mixture.

There are no fixed mortar factors,as they are influenced by particleshape, texture, and distribution.Approximate needs for ten con-struction classifications are shownbelow.Class 1: Placed by steep sided bot-tom-drop bucket, conveyor, or pav-ing machine. Approximate mortarrequired = 48 to 50 percent.Class 2: Placed by bottom-dropbucket or chute in open verticalconstruction. Approximate mortarrequired = 50 to 52 percent.

Class 3: Placed by chute, buggy, orconveyor in an 8 in. (200 mm) ordeeper slab. Approximate mortarrequired = 51 to 53 percent.Class 4: Placed by 5 in. (125 mm)or larger pump for use in verticalconstruction, thick flat slabs andlarger walls, beams, and similar ele-ments. Approximate mortar re-quired = 52 to 54 percent.Class 5: Placed by 5 in. (125 mm)

pump for pan joist slabs, thin orsmall castings, and high reinforcing

June 1990 37


6/8

Optimizationcontinued

steel density. Approximate mortarrequired = 53 to 55 percent.Class 6: Placed with a 4 in. (100mm) pump. Approximate mortarrequired = 55 to 57 percent.

Class 7: Long cast-in-place pilingshells. Approximate mortar re-quired = 56 to 58 percent.Class 8: Placed by pump smallerthan 4 in. (100 mm) Approximatemortar required = 58 to 60 per-cent.Class 9: Less than 3 in. (75 mm)thick toppings. Approximate mor-tar required = 60 to 62 percent.

Class 10: Flowing fill. Approxi-

mate mortar required = 63 to 66percent.

To maintain the water-cement ra-

tio constant, the total cementitiousmaterials factors will vary accord-ing to the mortar content. Thus,concrete with a higher mortar con-tent should cost more than thatplaced under restricted conditions.Decisions to use concrete with lowmortar and limited mobility shouldnot be made arbitrarily. In today'sconstruction economy, placementcost is important. A requirementfor a low mortar concrete can ef-fect not only direct placement costsbut also time of completion and

added project overhead.

Aggregate particle distributionPractically any sound aggregate canbe combined to produce a givenstrength concrete. However, when

particles are poorly distributed themixture can cause both construc-tion and performance problems. In

asphalt, poorly graded aggregatescan increase the need for filler andasphalt while too many fines cancontribute to excessive rutting andflow.

For the same reasons that parti-

cle distribution is important in geo-technical work and asphaltic con-crete, the aggregate fraction ofportland cement concrete must beadjusted to properly fill voids toprovide the mobility needed forplacement and finishing. Much hasbeen learned about the effects of

particle distribution from computersoftware that analyzes the particledistribution of every mixture forwhich gradations are provided.Both good and problem mixturesfollow consistent profiles.

A deficiency of particles passingthe 3/8 in. (9.5 mm) but retained onthe No. 8 (2.36 mm) sieve necessi-

tates use of more mortar. Generallysand and water are added, withoutthe addition of cement needed tomaintain a constant water-cementratio, to provide needed mobilityfor construction.

The typical single size stone andsingle sand mixes with gradationsmeeting ASTM C-33 size number 57(1 in. to No. 4 [25 to 4.75 mm])standards and concrete sand cancause both placing and finishingproblems. Fig. 6 is a computer-gen-erated particle distribution chart for

a mixture using aggregates comply-ing with the gradation acceptable byASTM C 33 size number 57 stoneand concrete sand. Such a mixture,if used at a reasonable mortar con-tent, will manifest finishing prob-lems. If the sand is increased to sat-isfy finishing needs, the strengthwill decrease because there will be ahigher water demand. The over-mortared mixture will cause thepumper problems because his linefriction will increase.

Fig. 7 describes an ideal solution

produced by the addition of a peagravel that is finer than that recom-mended by ASTMC 33 for sizenumber 8 (3/8 in. to No. 8 [9.5 to2.36 mm]). This ideal will seldom bepossible. Most local aggregates canbe blended in such a way as to pro-duce a uniform particle distributionwhen greater attention is paid to thecomposite than to individual stock-

piles.Fig. 8 reflects the particle distri-

bution produced using the 1923 ver-Fig. 6 Near gap graded mixture (1988 ASTM C 33).

3 8 Concrete International


7/8

sion of ASTM C 33 and recommen-dations of the first issue of thePortland Cement Association's De-

sign and Control of Concrete Mix-tures.'The effectiveness of this dis-tribution can be confirmed by ex-amining concrete that has been inservice over 50 years. When the sur-

face is abraded, there will be a highincidence of intermediate particlesexposed. This is in contrast to mod-ern mixtures that have a great dealof in. (12.5 mm) particles andlittle else between that size and the

mortar.Most industrial nations and some

sections of the U.S.A. use at least

three aggregate sizes (2 coarse and 1fine) to assure more consistent par-ticle distribution. The third aggre-gate is predominately intermediatesize (3/8 in. to No. 8 [9.5 to 2.36mm]) to provide a bridge betweenthe large particles and the mortar,fill major voids, and increases con-crete density. When size number467 (1 in. to No. 4 [37.5 to 4.75mm]) aggregate is specified, threecoarse aggregate sizes should be

used.Particle shape has a major effect

upon the influence of No. 4 and 8(4.75 and 2.36 mm) particles.Rounded pea gravel or cubicallycrushed stone are desirable. They

improve mixture workability,pumpability and finishability andproduce consistent, high strengthswith low shrinkage. When thesesizes are sharp and flat, they shouldbe limited because they cause mixmobility problems.

Similar information about parti-

cle distribution can be observed by

plotting the combined aggregategradation. However, this plot is notas definitive as the percent retained

on each sieve. When this guide isused, the trend line should be asnearly straight as possible. A goodreference is the 0.45 power chartused for asphaltic concrete. Forportland cement concrete, the gra-dations fall off the 0.45 power chartstraight line for the coarsest andfinest fractions.

SummaryMethods for selecting mixture pro-portions have been discussed formany years but there has been noconsensus of opinion about whatconstitutes a "good" mixture. Thatprocess addresses in-put and doesnot consider out-put. By use of thethree factors discussed here, it ispossible to stop the process of sub-mitting a new mix design for everyproject. All the producer need do isprequalify his mixtures, identifythose to be used, and provide statis-tical performance data.

The aggregates available in mostcommunities are fairly standard andused for many years. The only dif-ferences are in gradations and spe-cific gravities. The industry cancope with these variables when at-tention is given to the mixturerather than to stockpiles. Throughuse of the methods described hereinit will be possible to identify and useexactly the mixture needed for anyproject, and do it consistently.

This should not be interpreted asa suggestion that everything donetoday is done incorrectly. However,the current practices are wasteful

and contribute to many industryproblems such as unnecessarily highcosts, poor construction productiv-

ity, and reduced durability in theinfrastructure. It is an attempt todirect attention to performancepractices and concrete out-putrather than in-put. The entire con-struction industry is moving to per-formance specifications and con-crete should follow.

Each dawn is not a new day inthe life of concrete. The day hasbeen continuing for over 50 years

and we should know where we wantto go to achieve any result. Notevery mixture must be adjusted there is a great deal of flexibility inportland cement concrete. Were thisnot true, the quality work in placearound the world would not have

been possible. Needs, materials andconstruction are changing the accu-racy needed for concrete produc-tion to improve construction pro-

ductivity and in-place quality.

References

1. Fuller, William B., and Thompson,Sanford E., "The Laws of ProportioningConcrete," Transactions, ASCE, V. 59,1907, pp 67-143.

2. Bloem, D. L., "Effect of Sand Gradingon Mixing Water Requirement and Strengthof Concrete," Technical Information LetterNo. 106, National Ready Mixed ConcreteAssociation, Washington, D.C., July 1956,pp5-8.

3. Design and Control of Concrete Mix-tures, First Issue, Portland Cement Associa-tion, Chicago, (undated), approximately1925.

Selected for reader interest by the editors.

ACI fellow James M. Shilstone, Sr. ispresident of Shilstone & Associates,Inc., and Shilstone Software Co., Dal-las Texas. A graduate of the UnitedStates Military Academy, West Point,N.Y., he has been a member of the In-stitute for more than 20 years and hasserved on numerous ACI and ASTMtechnical committees. He was

awarded the Wason Medal in 1979and the Construction Practices Awardin 1987.

June 1990 39


8/8

OPTIMIZING BY THE 0.45 POWER CHART

The asphalt industry's 0.45 power chart isreference in the preceding article. Patrick Creegan

discussed the value of that chart in "Properly Coping

with the Low Water-Cement Ratios Required by ACI

350R-83" published in CONCRETEINTERNATIONAL, April 1990. ACI 350 covers

sanitary structures.The chart was developed by Nijboer as a means to

describe the ideal combined aggregate gradation for

asphalt. This could shown graphically as a the log of

the sieve opening in microns on the "X" axis and the

log of the percent passing on the "Y" axis. The

United States Bureau of Public Roads (BPR) verified

the work but used a non log relationship for the

percent passing. Since the BPR publication, the

asphalt industry has widely used the chart as their

standard for selecting combined aggregate gradations.

They do not control by control of stockpiles as isdone for concrete because two or more good

materials can be combined but produce a problem

combination.

Creegan states, "A straight line on this chart from

the origin to the point of 100 percent passing the

maximum size aggregate, represents the most densegradation of an aggregate having that maximum grain

size." Such an optimization reduces the need for

water to allow the production of concrete with a

water-cement ratio less than the specified 0.45. A

copy of Creegan's illustration is shown below with

optimum combined gradation lines plotted for 1/2",

3/4", 1", and 1-1/2" aggregates.

Asphalt makes use of fine mineral filler whileconcrete uses cementitious materials. We don't feel

inclusion of the cementitious material in the chart will

be beneficial unless the objective is the highest

strength with the lowest cement factor. In workingwith seeMAT-A to blend aggregates, on one occasion

we found the ideal cement factor was 410 lbs. per

cubic yard. As a result of not using the finest

particles, a concrete gradation falls off the 0.45power line at about the No. 16 sieve.

The 0.45 power chart and our concepts of mixand aggregate optimization are very similar. Whenthe combined gradation, expressed as percent passing,

for Figure 6 is plotted on the 0.45 power chart, there

is no correlation However, the gradation from Figure7 is very close.

We have found from mix analyses using the

seeMIX software program that the cementitiousmaterials factors affects the 0.45 power chart viability

and can contribute to using to much sand. The

Coarseness Factor Chart (Figure 5) is needed to make

the correction. The seeMIX program includes theanalytical methods and technology discussed thepreceding article. Call (800) 782-8649 for more

information on the software.

IN SUMMARY: When aggregates and

cementitious materials are optimized, concrete quality

will be improved, costs reduced by minimizing

wasteful use of materials, and long-term durability

will be extended.

Documents

Conc Mix Opt