“The Effect Of Particle Size Distribution (Psd) Concept Of CoarseAggregates On The Water Demand Of The Concrete Mix”

Dr. Khalid, A. A.Zakaria

Professor

Ghanim, H., KojaLecturer

Nadiya, S. I., Al SaffarAssistant lecturer

Civil Eng. Dept.

AbstractThe effect of particles size distribution of the coarse aggregates on the

water demand of the concrete mix was studied and analyzed using theconcept of particles size distribution in so far as they deviate from theavera size or fineness modulus as indicated by the standard deviation ofthe sample.

Six concrete mixes 1:2:4 by weight (320 kg/m3 cement content) weredesigned for a given slump of (30-60) mm. The main variable being thegradation of the coarse aggregates.

The results showed that an increase in the S.D of the coarse aggregateparticle distribution resulted in an increase in the water demand of themix, this was further substantiated by a decrease in the resultingcompressive strength. An extra 38 similar concrete mixes were selectedrandomly from the literature for further supporting evidence.

Keywords: Particle size distribution, Water demand, standard deviation,average size of aggregate

تأثیر توزیع مقاس جزیئات الركام الخشن على بعض خواص الخرسانة باستخدام مفھوم " "االنحراف المعیاري

خالد عبد العزیز زكریــــــا. دأستاذ

غانم حسین قوجــةمدرس

نادیة صدیق الصفـــارمدرس مساعد

یةقسم الھندسة المدن

الخالصة:

.النعومة وكما یفسر من خالل االنحراف المعیاري لتدرج نموذج الركام الخشنوبھطول ) kg/m3 (320ومحتوى أسمنت ) وزناً 4:2:1بنسبة (د ست خلطات خرسانیة تم اعتما

.حیث كان المتغیر الرئیسي في ھذه الخلطات تدرج الركام الخشن،(mm 60-30)یتراوح بین

.مقاومة االنضغاطالوقت نقصانطة للماء وبنفس زیادة في متطلبات الخل38)(وأخیراً تم اختیار

.إضافیةإثباتات وأدلةIntroduction:

For fully compacted concrete, w/c ratio is the main parametergoverning the strength of concrete, according to the law established byAbram [1], and his initial statement, “the strength of concrete is afunction of the ratio of cement to the free water in the plastic mixture“,agrees with the fact that the strength of concrete continues to increasewith the reduction of w/c ratio to a value of 0.2 or even lower.

At equal values of w/c ratios, the strength of workable concrete maybe influenced greatly by such factors as the grading, relative amount ofaggregates, the shape, surface texture, stiffness, and the maximum size ofaggregates.

A suitable gradation [2] of the combined aggregate in aconcrete mix is described in order to secure workability and tosecure economy in the use of cement. A well-graded mixtureproduces strong concrete than a harsh or poorly graded one. Singh[3] has proposed that, for constant mix proportions, the increase inspecific surface (index of grading) of the aggregate, causes adecrease in the amount of cement relative to the surface of theaggregate, thus causing more voids around the surface of theaggregate particles and decrease in strength.

Specific surface gives in somewhat misleading picture of theworkability to be expected and to overcome this difficulty,Murdock [4] has suggested the use of surface index, which is anempirical number, related to the specific surface of the particleswith more weightage assigned to the coarser material.

Received ٢٦ June 2005 Accepted 11 July 2006

The total surface index (fs) of a mixture of aggregates iscalculated by multiplying the percentage weight of materialretained on each sieve and the corresponding surface index and totheir sum is added a constant of 330 and the result is divided by1000 [5].

The specific surface varies with different types of aggregate dueto variations in the angularity. The angularity index (fa) dependsupon the grading of coarse and fine aggregates, angularity number[6] and the relative proportion of coarse and fine aggregates in themix.

The experimental evidence presented and mathematical analysis clearlyshows that the importance of fineness modulus is greatly underratednowadays. The limit of validity of this method is much wider and itsapplicability is much better than is generally believed [7].The grading is characterized numerically by the fineness modulus.This makes possible the development of formula to express theeffect of grading on the concrete properties [1]. The finenessmodulus represents an average particle size of the aggregate and,as such, it is a fundamental parameter of the particle sizedistribution.

Both the fineness modulus and the specific surface are measures ofaverage particle size. Experimental results, however, show bettercorrelation with fineness modulus [7].

All normal concretes containing the optimum coarse aggregatecontent (for workability) have fine aggregate to cement ratios wellin excess of (0.8). The more frequent cause for losses in strengthis at the other extreme: because the critical maximum value isexceeded, incomplete compaction occurs. However, for allpractical purposes, when the optimum coarse aggregate is beingused, these extreme conditions of either too little or too much fineaggregate are readily avoided [8].

Increasing the proportion of the rounded particles decreases the

percentage of voids. Since the cement paste required for concreteis proportional to the void content of the combined aggregates, itis desirable to keep the void content to a minimum [9,15].

Researchers [10, 11] have suggested that, an increase in themaximum aggregate size results in lower compressive strength inrich mixes and higher compressive strength in leaner mixes. At agiven w/c ratio and mix proportions, concrete with smallermaximum aggregate size develops greater strength than concretewith larger size [12]. The particle size distribution (as they deviatefrom the fineness modulus is indicated by the standard deviationof the aggregate particle) [13].

Statistical analysis indicates that the lower the value of thestandard deviation, the higher will be the percentage of particlesclose to the average size, conversely a higher standard deviationindicates a larger portion of particles in the coarse and finefractions.

If the coarse aggregate is regarded as standard, i.e., keptunchanged throughout a series of tests, it will be found that theuse of different sands in the mix results in different waterdemands. Consequently, it may be referred to the “water demandof the sand“. On the other hand, if the sand is regarded as standardthe use of different coarse aggregate in the mix also results indifferent water demand and it may then be referred to the “waterdemand of coarse aggregate”, [12].

Objective of the Research:

The object of the current work is to show that the particles sizedistribution of the coarse aggregate using the concept of standarddeviation has a definite effect on the mix water demand andultimately on the concrete compressive strength.

Experimental Program:

Materials used: -

Locally available materials were used. Their main properties are asindicated below:

Cement: the cement used was in accordance with Iraqi specification(IQS) No. 5 (1984) [16].

Fine aggregates: Fine aggregates used consisted of medium normalriver sand in accordance B.S. 882-(1992) [17].

Coarse aggregates: Coarse aggregates used were normal river gravel(irregular, almost rounded) in accordance with B.S. 882-(1992) [17].

Table (1) shows some relative properties with the sieve analysis of theused coarse aggregate.

Sieve

Index

Sieve

Size(mm)

(1)

Graded

20mm

(2)

ungraded

(3)

Graded

20mm

(4)

Graded

20mm

(5)

ungraded

(6)

UNESCO

4 40 100 100 100 100 100 100

8 20 95 90 98 100 95 90

7 10 30 20 45 60 70 50

6 5 0 0 5 10 65 10

–according to

B. S. 882-1992

5 2.36 0 0 0 0 0 0

4 1.18 0 0 0 0 0 0

3 600μm 0 0 0 0 0 0

2 300μm 0 0 0 0 0 0

1 150μm 0 0 0 0 0 0

F.M 6.75 6.9 6.52 6.3 5.7 6.5

S.D 0.536 0.538 0.545 0.64 1.004 0.82

Procedure

The above mentioned materials were used in preparing sixconcrete mixes 1:2:4 by weight, cement content (320) kg/m3,maximum aggregate size (20) mm and having a slump of (30-60)mm. The main variable was the gradation of the coarse aggregatesused as shown in Table (1).

For further evidence, number of mixes totaling 38 were chosenrandomly from the literature [12, 14] designated No. 7 to No. 44Table (3).

Investigated Parameters

The main investigated parameters of the present research workare as follows:

1. The standard deviation of the coarse aggregate particle sizedistribution was calculated, this is shown in col.3 of Table (3),and typical calculation is listed in Appendix (A).

2. The slump range for all the mixes considered with the w/cratios (water demand of the mix) are given in col. 4 and 5 ofTable (3) respectively. See also Figs (1, 2, and 3.)

3. The 28-day experimental cube compressive strength of100x100x100 mm cubes prepared, cured, and tested accordingto B.S 1881, 1983 parts 108, 111, and 116 respectively is listedin col. 6 of Table (3).

4. The calculated compressive strength using the followingequation which is predicted using the regression analysis on 44test result from the present study and the published data asshown in Table (1) is given in col. 7 of Table (3); this equationhas a correlation coefficient of (0.92).

D.S81.9S045.0C/W1.11269.96 ………..(1)

σ = Calculated Compressive strength of concrete (MPa).

W/C = Water cement ratio.

S = Slump (mm)

S.D = Standard Deviation of aggregates.

5. Col. (8) shows the Cal./Exp. values of the compressivestrength, see also Fig. (4).

6. Finally the A/C ratios, which are the richness of the mixesconsidered, are shown in col.2 of Table (3).

0

5

1 0

1 5

2 0

2 5

3 0

3 5

4 0

4 5

0 . 3 5 0 . 4 0 . 4 5 0 . 5 0 . 5 5 0 . 6 0 . 6 5 0 . 7

Water Cement Ratio(w/c)

Stren

gth (M

Pa)

ExperimentalCalculatedCalulatedExperimental

0

5

10

15

20

25

30

35

40

45

30 35 40 45 50 55 60

Slump (mm)

Stre

ngth

(MPa

)

A/C =7A/C =5.75A/C =4.8A/C =7A/C =5.75A/C =4.8

05

1015202530354045

0 10 20 30 40 50

Experimental Strength (MPa)

Cal

cula

ted

Stre

ngth

(MPa

)

Line ofEquality

Fig. (4) Experimental –Calculated Compressive Strength.

30

35

40

45

50

55

60

0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7

Water-Cement Ratio

Slum

p (m

m)

Fig. (3) Slump- Water Cement Ratio Relationship.

Type

of

agg.

28

days

MPa

7

days

MPa

WaterDemand

asrelatedto S.Dand

averageSievesize

w/cbased onS.S.D

for agive

nslum

p

S.D

ofagg.

Particles

%

passing

average

sievesize

Average

sieve

size(mm)

F.M

ofgravel

Mix

Graded

32.2

21.0S.D(Small)=higher

% ofparticlesclose toaverage

sievesize.

0.520.53630106.75

(1)

Ungraded

30.9

21.2S.D(Small)= higher

% ofparticlescloses toaverage

sievesize.

0.540.53820106.9(2)

Graded

30.5

19.4S.D(small)

butincrease

s inwater

demand

0.550.54545106.52

(3)

sent study.

due toincrease

in %passingaverage

sievesize.

Graded

28.6

17.9S.D(large)

indicating large

proportion of

particlesin thefine

fraction(60 %

passingsieve 10

mm)furtherincreasein waterdemand.

0.560.6410

60

4.75

10

6.3(4)

Ungraded

27.6

16.5S.D(large)furtherincreasein waterdemand.

0.591.00465

70

5

10

5.7(5)

Graded

(roun

25.8

16.0S.D(large)large

proporti

0.610.8250106.5(6)

ded)ons ofparticles

in thefine

fraction(50%

passingaverage

sievesize).

Cal/Exp.Strength

(cal.)

Strength

(exp.)

W/c

Ratio

Slump

(mm)

S.DA/CRatio

No.

0.973631.3532.20.52400.53661*

0.9414229.0930.90.54400.53862*

0.9072127.6730.50.55450.54563*

0.8797225.1628.60.56550.6464*

0.7934821.927.60.59401.00465*

0.7023318.1225.80.61500.8266*

0.9866725.1625.50.57450.572277

1.0669723.2621.80.585500.572278

1.055120.6819.60.61450.572279

0.9913233.1133.40.495550.57225.7510

1.0346431.6630.60.51500.57225.7511

1.0665529.6527.80.53450.57225.7512

1.073626.84250.555450.57225.7513

0.9491137.4939.50.46450.57224.814

0.9510536.14380.47500.57224.815

0.9921734.2334.50.485550.57224.816

1.0346431.6630.60.51500.57224.817

0.980722.3622.80.595450.5722718

1.0676821.1419.80.61350.5722719

1.0587218.2117.20.63500.5722720

0.9642914.8515.40.66500.5722721

0.9830730.7731.30.52450.57225.7522

1.0901529.8727.40.53400.57225.7523

1.22528.4223.20.545350.57225.7524

1.1130823.8221.40.58500.57225.7525

0.9437834.92370.485400.57224.826

0.9641232.78340.5500.57224.827

0.9970430.3130.40.52550.57224.828

1.037427.1826.20.55500.57224.829

1.0284627.4626.70.525500.829730

1.0241824.9924.40.545550.829731

1.0643823.3121.90.56550.829732

1.060121.5220.30.58450.829733

1.100618.4916.80.605500.829734

1.0275135.8634.90.45500.8295.7535

1.0120634.41340.465450.8295.7536

1.0468633.2931.80.475450.8295.7537

1.0412230.8229.60.495500.8295.7538

1.1579227.79240.52550.8295.7539

0.9740837.2138.20.44450.8294.840

0.9734235.5336.50.455450.8294.841

0.9519634.0835.80.47400.8294.842

0.990332.68330.48400.8294.843

1.0211230.9430.30.5350.8294.844

1.00679AverageCube Compressive Strength MPa

* Present Study

Discussion of Results

Table (2) shows the main investigated parameters of the sixmixes considered for the present work. It is clear from the Tablethat the water/cement ratios (i.e. water demand of the mix) isdependent on the standard deviation of the coarse aggregateparticles, this is based on the fact that a small standard deviationmeans a higher percentage of the particles close to the averagesize, that is a low water demand, hence higher compressivestrength both at 7& 28 days, this is true whether graded orungraded aggregates are used as the results of mixes 1,2 &3indicate.

A noticeable increase in the standard deviation (from 0.536 to0.64) indicates a decrease in the strength (as shown in Fig. (5)).This is clear for mixes 4,5, and 6 of Table (2), the reason for this

behavior being that an increase in the standard deviation indicatesthat a large proportion of the aggregates is in the fine fraction thisis further substantiated by the higher percentages passing theaverage sieve size (60, 70, and 50%) for mixes 4, 5, and 6respectively. This means an increase in the surface area hence anincrease in the water demand of the mix (w/c = 0.56, 0.59, and0.61) for mixes 4, 5, and 6 respectively.

This of course is followed by a decrease in the resultingcompressive strength both at 7 and 28 day.

ConclusionsBased on the finding of the present research the following mainconclusions may be drawn:

1. The concept of the particles size distribution of the coarseaggregates in so far as the deviate from the average size orfineness modulus as indicated by the standard deviation of thesample may be used in checking the water demand of the mix.

2. Generally speaking an increase in the standard deviationindicates a higher water demand and a strength reduction forconcrete mixes.

3. An empirical estimation based on the issue of standarddeviation was obtained to predict the compressive strength ofconcrete (σ) with a good degree of accuracy, as given by:

D.S81.9S045.0C/W1.11269.96

References:

1. Neville, A. M. “Properties of concrete”, 3rd Edition,Pitman – London 1983.

10

15

20

25

30

35

40

0.4 0.5 0.6 0.7 0.8 0.9 1 1.1

S.D

Cal

cula

ted

Stre

ngth

(MPa

)

Deviation.

2. Troxell G. L, Davis H. S, and Kelly J.W, “Compositionand properties of concrete”, 2nd Ed., Mc. Graw –HillBook Company (1968).

3. Singh B.G, “ Specific Surface of aggregate related tocompressive and flexural strength of concrete “, ACIJournal, (proceeding vol. 54), vol. 29, No. 10, April(1958), pp. 897-908.

4. Murdock, L. J., “The workability of concrete “,Magazine of concrete research, vol. 12, No. 36,November 1960, Cement and Concrete Association,London, pp. 135-44.

5. Raju, N. K., “Design of concrete mixes”, C.B.Spublishers and distributors, Delhi (India), Second Edition,1988.

6. B.S: 812-1960, “Methods of sampling and testing ofmineral aggregates, sand and fillers”, British StandardsInstitution, London, 1960.

7. Popvics S., ”The use of the fineness modulus for thegrading evaluation of aggregates for concrete “, vol. 18,No. 56, September 1966, pp. 131-142.

8. Hughes B.P., “Some factors effecting the compressivestrength of concrete”, Magazine of concrete research, vol.19, No. 60, September 1967, pp. 165-172.

9. Portland Cement Association, “Design and control ofconcrete mixture “. Part (3), General Information, 11thEd., July (1968).

10. AL-Rawi R.S, and AL-Murshidy K., ”Effect ofmaximum size and surface texture of aggregate inaccelerated testing of concrete“, cement and concreteresearch, vol. 8, No. 2 (1978), pp. 201-210.

11. Walker S., Bloam D.L., and Gaynor R. D.,“relationships of concrete strength to maximum size ofaggregate”, proceeding Highway research Board, vol. 38(1959), pp. 367-379.

12. Ali H. A., “Effect of aggregate characteristics on freshand hardened concrete using normal mixes”, M.sc.Thesis, Dept. of civil Eng., College of Eng., Univ. ofMosul, 1988.

13. Portland Cement Association (PCA), ConcreteInformation, Part1” Statistical Product Concrete”, June1970.

14. Othman, S.Y.,” Comparison between different mixdesign methods on some properties of concrete using

local aggregates”, M.sc. Thesis, University of Mosul,1986.

15. Zakaria, K. A. A., Kashmo;a, S. Y. and Jassim B. M.,“Study Effect of Total Voids Volume in the Concrete Mixon its Compressive Strength using the Accelerated TestMethod.” Tikrit Jour. Of Eng. Sci., Vol. 9, No. 3, Sept.2002, pp. 19-20.

16. IQS ع . ق. االسمنت البورتالندي م) ٥(المواصفات القیاسیة رقم .17. B.S 882-1992 “British Standard Specification for

Aggregate from Natural Sources for Concrete”, 1992.

Appendix (A)

Typical Calculations:

1. Average sieve size: grading 20mm aggregates

Col. (1) Table (1) = 75.6100

1001001001001001007050

Fineness modulus = 6.75 7.0

... Sieve index (7) is the average sieve size = sieve 10 (col. (3) Table (٢))

2. Standard deviation

(Col. 7) ungraded 20mm max. Agg. Size (F.M =5.7)

....).(2.%).(1.%101. 22 MFindexsievesievendonrtdMFindexsievesievestonrtdDS

2222 )7.55(65)7.56(5)7.58(5)7.59(0101. DS

=1.004