Materials and Structures/Mat6riaux et Constructions, Vol.

32, May 1999, pp 273-281

Assessment of washout resistance of underwater concrete:

a comparison between CRD C61 and new MC-1 tests.
M. Sonebi 1, P.J. M. Bartos I and K. H. Khayat 2
(1) The Advanced Concreteand Masonry Centre, Department of Civil, Structural and Environmental Engineering, University of Paisley,
Scotland, UK
(2) Department of Civil Engineering, Universit6de Sherbrooke, Sherbrooke, Canada

Paper received:May 26, i998; Paperaccepted:June3, 1998

A B S T R A C T R I~ S U M I~

An investigation was carried out to compare the per- Cette 6tude vise a comparer la r6pdtabilit~, la sensibilit6
formance and, in particular the repeatability and sensitiv- d'6valuation de l'ouvrabilit6 et la r6sistance au lessivage
ity of assessment of workability and washout resistance mesur& par deux m6tkodes bas&s sur diJff~rents principes,
of underwater concrete using two test methods based on du b6ton coul6 sous l'eau. La premi&e m6tkode est l'essai
different principles. The first method is the plunge test au plongeon, utilis8 en Am6rique du Nord, et la deuxi~me
(CRD C61), used in North America, and the second is est le nouvel essai de lessivage par arrosage d6velopp~
the new spray test (MC-1), developed at the University l' universitd de Paisley.
of Paisley. The sensitivity of the test results to variation L'6tude de la sensibilit6 porte sur l'effet de la variation
of dosage of the anti-washout admixture and of super- des dosages en agent colloidal et du superplast~'ant. Une
plasticizer, was investigated. corr61ation entre le lessivage mesur~ par l' essai au plongeon
Workability in terms of the slump and slump flow, and et celui par l'essai par lessivage a ~td &ablie.
the washout resistance measured by the two methods, were La fluidit~ mesur& par l'affaissement et le diam~tre
particularly sensitive to changes in dosages of the anti- d'6talement ainsi que le lessivage mesur~ par les deux
washout admixture and the superplasticizer. The assess- essais est particuli~rement sensible aux variations des
ment of the repeatability of the washout resistance indicated dosages de l'agent colloidal et du superplastifiant. L'6va-
that the coefficient of variation of the loss of mass after 3 luation de la r6p~tabilit~ de l'essai de lessivage a montr6
drops in the plunge test was lower than the coefficient of que le coefficient de variation apr& 3 plongeons est inf~-
variation of the spray test. The workability and washout rieur a celui de l'essai par arrosage. La fluidit~ et la perte
measurements by the two methods were sensitive to varia- de masse par lessivage obtenue par les deux essais sont sen-
tions of the dosage rate of the anti-washout admixture and sibles aux variations des doses d' agent colloidal et du super-
the dosage of the superplasticizer. In general the new spray plastifiant. Le lessivage obtenu par arrosage est g&&ale-
test was found to be more severe test which showed higher ment plus grand que celui de l'essai au plongeon. Cette
washout results than the plunge test. The investigation 6tude a aussi permis d'3tablir une correlation entre les deux
seeked to establish a relationship between both tests. essais de lessivage.

1. I N T R O D U C T I O N between washout loss obtained by the standard test

CRD C61 and the MC-1 test was established. Among
The objective of the research reported in this paper the investigated parameters were the concentrations of
was to compare the performance, namely the repeatabil- the AWA and dosage of the SP, and the W/C ratio.
ity and sensitivity, of two different methods for assess-
ment of washout resistance of fresh underwater con-
crete, namely the US standard test CRD C61 [1] and the 1.1 Review of washout tests
new test MC-1 [2, 3]. The effect of dosage of the anti-
washout admixture (AWA), the superplasticizer (SP), 1.1.1 Stream test
and of the workability measured by the slump and the This test was developed in Belgium [4]. It is a simple
slump flow tests was also investigated. The relationship test which gives a visual assessment of the degree of

E~t~iai~ie4~///)4)/ //) 5 /) /;i ) L / //)/ 7 ) / /

Ba~os i~a Sehio~Me~berand the Chairrndn~ the RILEM !,lSi WSM: Workabi!ity~speda! coniretemiXes:He is alsoa memberof
i 74~SCCi Se~compactingconCrete.

1359-5997/99 9 RILEM 273

Materials and Structures/Mat6riaux et Constructions,Vol. 32, May 1999

washout resistance. The test can be used in a laboratory

and on site.
The test procedure uses a 100-150 m m diameter gut-
tering channel 2 m long set at a slope of 15-20 ~ to the
horizontal (Fig. 1). A sample of test concrete is placed
300 m m from the raised end of the channel. Water is
then poured down the pipe and over the concrete sam-
ple. The amount of washout is determined visually.
Although a useful test, it is strongly operator-sensi-
tive when different concretes are compared. This is due
to the rate at which the water is poured over the sample. Fig. 1 - S t r e a m test.
Obviously, the faster the flow of water the greater is the
washout. An operator therefore must possess a substan-
tial previous experience in order to interpret correctly
the results of this test. It is very important that the same
volume of water be poured down the channel from the
same point in each of the tests.

1.1.2 Drop test

The basic equipment is a graduated cylinder filled
with water. A small scoop-full of concrete (300-500 g)
is dropped through the water. The resultant turbidity of
the water is to be visually assessed to determine the
degree of washout [4, 5]. It is a quick and simple quali-
tative test applicable in the laboratory. Instead of the
visual assessment, the degree of turbidity can be mea-
sured by the use of a spectrophotometer or similar
equipment.

1.1.3 pH factor test

This method was proposed in the Recommendation
for washout-resistance of underwater concrete in Japan
[6]. The procedure begins with a beaker filled with
water. A fresh concrete sample of an appropriate size is
divided into several parts and then dropped into the
beaker. Three minutes after the addition of the mixture,
a unit volume of the supernatant solution is decanted
into another beaker. The pH factor of this solution,
rounded to one digit below the decimal point, is then
determined and recorded. The higher the pH the higher
is the washout loss.

1.1.4 Plunge test Fig. 2 - P l u n g e test (CRD C61).

The test method probably originated in Belgium
where it had been used by the University of Ghent. The
original test used a small basket with relatively small when using a highly flowable concrete similar to mixes
diameter holes (3 mm), and the basket full of concrete used for underwater placement on site. Such fluid con-
was only immersed in water three times. The plunge crete inherently flowed out through the holes in the bas-
test is the only one of the three existing direct tests ket before any immersions [7].
which provides a quantitative value of the test result.
The test was adopted by the US Corps of Engineers into 1.1.5 Spray test
its standard by CRD C61 [1] (Fig. 2). The new test apparatus MC-1 developed by Ceza
Experiments carried out at Fosroc Technology and and Bartos [3] is based on the principle of a direct deter-
work at the University of Paisley showed that it was pos- mination of the resistance to washout of a concrete mix
sible to get apparently good washout resistance simply which produces quantitative results. The mass of a sam-
due to the aggregates in the mix blocking the holes of pie of concrete is determined before it is put into the test
the basket. The test was therefore modified to use a apparatus in which it is washed out by water.
larger size of basket with a larger diameter of the holes The test is simple. A mould is placed on a baseplate
(20 mm). The number of immersions was increased and filled with a sample of approximately 1 kg of fresh
from three to five. However, another problem occurred concrete (Fig. 3). The mould is removed and the base-

274
Sonebi, Bartos, Khayat

are plotted as curves of the change of the mass of the

sample during the duration of the test, and an expression
of the washout is provided as a percentage of the mass
lost compared to that of the original mass [3].

2. E X P E R I M E N T A L PROGRAM

The experimental program consisted of two phases.

The first phase concentrated on the investigation of the
repeatability and sensitivity of the CRD C61 "plunge"
test and MC-1 "spray" test. Properties of fresh concrete
made with W/C = 0.43 and three dosages of AWA vary-
ing between 0.30 and 0.60% by mass of cement were
evaluated. Five tests were used for the assessment of the
repeatability with W/C = 0.43 and 0.50% of AWA. For
Fig. 3 - M C - 1 apparatus. each AWA, three concretes were prepared with three
dosages of a melamine-based SP. Each mix was
repeated for two W/C = 0.43 and 0.47. In the sec-
Table 1 - Experimental program ond phase, the effect of the dimensions of the tube
and the basket of the plunge test on mass loss was
Mix W/C AWA (%) SP (I/100 kg of cement) Test type
investigated. A summary of the experimental pro-
Control* 0.50 0.0 1.0 A-B-C gram is given in Table 1.
1- 2 - 3 - 4 - 5 0.43 0.50 1.21 A- B-C

6- 7- 8 0.30 1.17-1.40-1.76 A-B-C

2.1 Materials
9 - 10 - 11 0.43 0.55 1.17 - 1.40 - 1.76 A-B-C

12 - 13 - 14 0.60 1.17 - 1.40 - 1.76 A- B - C A Portland cement, similar to that of ASTM

15 0.50 0.50 1.17 A-B-C
C150 Type I cement, with a Blaine fineness of
355 kg/m 2 was used conforming to Standard
16 - 17 0.43 0.50 1.67 - 2 . 4 9 A-B-C
BS.12:1996 [8]. A siliceous, riverbed fine aggre-
18 - 19 - 2 0 0.47 0.50 1.17-1.40-1.62 A-B-C gate was used. Its fineness modulus was 2.50, and
21 0.43 0.50 1.3 A-B-C-D its relative density value was 2.65. A coarse aggre-
gate consisting of round natural quartz and sand-
: 390 kg/m 3
A : Slump and slumpflow - B : Loss of mass using plunge test ( C R D C61) -
stone particles with a nominal aggregate size of 14
C : Loss of mass using " M C - 1 " test - D : Loss of mass usingplunge test with big tube. mm was used. The relative density value of the
coarse aggregate was 2.50. The grading of the fine
and coarse aggregate are shown in Fig. 4.
plate with the sample is placed on a frame suspended One type of a commonly used AWA was adopted for
from an electronic balance. A computer is connected to the trials, based on a hydroxypropyl methyl of cellulose.
the balance, and is set to a mode in which an increase of It was used at dosages varying from 0.30% to 0.60%, by
the pressure on the baseplate by the water spray will mass of cement. The powder-based admixture was
automatically switch on recording of the washout added directly into the mixer. A melamine-based SP
process. The recording begins automatically when the
tap on the pipe connecting the water tank with the
spray-head is opened. Water from the spray-head
100
washes out the sample for 4 minutes. A constant head of
water is maintained in the water tank (Fig. 3) in order to
8O
keep the same pressure throughout the test.
The measurement recorded directly from the balance .~ 6o
every two seconds is the mass of the sample resting on
the base-plate and "mass" due to pressure and volume of 8. =
40 i
the water poured onto the sample. The net amount of
the material lost is obtained automatically as the effect of 20
the pressure and mass of the poured water is subtracted
from the direct measurement from the balance (Fig. 3). 0 ~
0.01 0.1 1 10 100
A calibration procedure, in which a mock-up of the
hardened concrete sample is used instead of the fresh Sieve size (ram)

concrete, provides the combined values of the pressure

and mass of the water on the sample. The final results Fig. 4 - Grading o f aggregate used.

275
Materials and Structures/Mat~riaux et Constructions,Vol.32, May1999

was also used. The superplasticizer was compatible with

the AWA selected.

2.2 Mixing
All concretes were prepared in 15 kg batches and
were mixed in an open pan mixer. The mixing
sequence consisted of mixing the fine and coarse aggre-
gate first, then adding 50% of the mixing water. The
cement was then added along with the remaining mix
water. The AWA was added last together with the SP.
The concrete was then mixed for an additional 3 min-
utes. After a period of one minute of rest, the concrete
was remixed for an additional 2 minutes.

2.3 Test methods

Two tests were used in this investigation to deter-
mine the washout resistance of flesh concrete, the
plunge test (CRD C61) and the new spray test MC-1.
In a plunge test, the washout is measured by deter-
mining the relative amount of cement paste and fine
mortar lost when the concrete sample in a perforated
basket is passed through the water. This test consists of
placing a representative sample of concrete of a mass of
2.0 + 0.2 kg in a perforated basket which is then dropped
three times through 1.7 m of water. The change in mass
of the sample within the basket after each drop is mea- Fig. 5 - Plunge test (big tube).
sured. The results are expressed as a percentage the loss
of mass from the initial one (Fig. 2).
For the plunge test in a big tube (Fig. 5), a larger wire
basket with square holes of 50 mm was used and filled
with approximately 30 kg of concrete. It was then
plunged three times through the water in the Perspex
pipe. The diameter and height of pipe were 450 mm
and 3.6 m respectively. The tube was filled by water to a
level of 1.7 m.

2.4 Slump and slump flow tests

The workability of concrete was also evaluated using
the slump and slump flow tests [6, 9]. The latter consists Fig. 6 - Slump and slump flow.
of measuring the mean base diameter of the concrete
sample at the end of the slump test (Fig. 6). Because of
the viscous nature of the concrete containing an AWA, washout. In all cases, the volume of the coarse aggregate
the readings of the slump and slump flow measurements was limited to approximately 300 l/m 3 and the mass ratio
have been taken out after there was no discernible move- off'me aggregate to total aggregate was approximately 0.45.
ment of the concrete.

3. TESTS AND RESULTS

2.5 Mix design
3.1 Repeatability
The mix proportions of the investigated concretes are
sumrnarised in Table 2. For each concrete, the ranges of The repeatabilities were investigated for concrete
the concentrations of AWA and SP used are indicated. For made with W/C = 0.43, 0.5% of AWA and 1.21 1 of SP.
each series of AWA, a single batch of concrete was pre- The workability and washout resistance were deter-
pared for the determination of the workability and mined by measuring the slump and slump flow and the

276
Sonebi, Bartos, Khayat

Table 2 - Mix proportions tively). Generally, the standard devi-

Contml 1 - 2 - 3 6 - 7 - 8 9-10 12-13
ations of slump and slump flow were
Mix 15 16-17 18-19
-4-5 -11 -14 -21 -20 found to be lower than 10 and 25
mm, respectively, as indicated by the
Nature of test Rep. Sens. Sens. Sens.
JSCE recommendations for under-
w/c 0.50 0.43 0.43 0.43 0.43 0.50 0.43 0.47 water concrete [10].
Cement (kg/m 3) 390 450 450 450 450 390 450 450 The repeatability of measurements
Water (kg/m 3) 195 194 194 194 194 195 194 211
of washout resistance by the C R D
C61 and MC-1 methods is shown in
CoarseA. (kg/m 3) 971 971 971 971 971 971 971 920 Fig. 8. The dispersion of measure-
FineA. (kg/m 3) 794 794 794 794 794 794 794 745 ments of loss of mass using the MC-1
AWA (%) 0 0.50 0.30 0.55 0.60 0.50 0.50 0.50 was higher than that using the CRD
C61. The coefficient of variation of
SP (//100 kg 1.0 1.21 1.17 1.17 1.17 1.17 1.67 1,17
of cement) 1.40 1.40 1.40 2.49 1.40
the loss of mass measured by the MC-
1.76 1.76 1.76 1.3 1.62 1 was 12.7%, and was 5% for the
CRD C61 test. However, the coeffi-
Rep. : Repeatability Sens. : Sensitivity
cient of variation of the MC-1 test
results was considered acceptable
(approximately- 10%). The loss of
Table 3 - Results of repeatability tests
mass by washout using the MC-1 at
Mix W/C AWA (%) SP* Slump S. flow WSHcRD WSHMc.1 WSHMc.1/ the recommended time of 4 minutes
(mm) (mm) (%) (%) WSHcRD was higher than that obtained after
1 210 335 3.7 11.1 3.0 3 drops by the CRD C61 test (13.9%
2 220 335 3.8 16.0 4.2 vs. 3.6%). This reduction can be
3 0.43 0.5% 1.21 210 315 3.7 14.5 3.9 explained by the fact that the sample
4 215 335 3.3 15.1 4.6 of concrete placed on the baseplate in
5 210 325 3.5 12.7 3.6 the MC-1 apparatus was not retained
Average Av. 215 330 3.6 13.9 3.9 within the perforated basket as in the
Standard deviation s 4 8 0.18 1.8 0.5 CRD C61 test. The concrete sample
Coefficientof variation(%) v 1.9 2.4 5.0 12.7 12.8 in the MC-1 washout test was
Value r 11 22 0.50 5.0 exposed to the water from the spray
r = 2.8 s (the repeatability value r defines the level below which the absolute difference between two single test head without any restraint, and in
results isfound with a confidence level of 95% certainty and obtained under repeatability conditions [11] case of mixes less resistant to the
* : l/lOOkg of cement washout the loss of mass included
WSHcR D : Loss of mass by washout after 3 drops using CRD C61 test (%)
WSHMc_ 1 : Loss of mass by washout using "MC-1" apparatus (%)
particles of coarse aggregate. The
concrete sample placed in the perfo-
rated basket of the C R D C61 test
loss mass of the sample after 3 drops for the CRD C61 was restrained by the basket, and the loss of mass was lim-
test and after 4 minutes for the MC-1 test (Table 3), ited to "mortar" containing particles smaller than the 3
(Mixes 1-5). m m diameters perforations of the basket. On average, the
The results of workability tests are shown in Fig. 7. loss of mass obtained using the MC-1 test was approxi-
The coefficient of variation for the slump flow was mately 4 times higher than that of the CRD C61.
greater than that of the slump (2.4% vs. 1.9%) while the
standard deviation was smaller (8 m m and 4 m m respec-

17
,'jQ~

A A 15
E
g
o
Q,
"6 11
E
-l
~ 9
D.
= 3
7
E
5

3
1 2 3 4 5 1 2 3 4 5
Mix Mix
Fig. 7 - Repeatability test of slump and slump flow (W/C = 0.43, Fig. 8 - Repeatability test of washout made with plunge test and
AWA = 0.5%). M C - 1 ( W / C = 0.43 - A W A = 0 . 5 % ) .

277
Materials and Structures/Mat~riaux et Constructions,Vol.32, May 1999

Table 4 - Results of sensitivity tests

The effect of the variation of
dosage of AWA and SP on slump
Mix W/C AWA(%) SP* Slump S. flow WSHcRD WSHMc.1 WSHMc-1/
(mm) (mm) (%) (%) WSHcRD
and slump flow is shown in Fig. 9
and Fig. 10. A reduction of the
6 0.30 1.17 195 320 7.1 40.0 5.6
7 0.43 0.30 1.40 220 350 7.6 50.6 6.7
AWA from 0.55% to 0.30% or
8 0.30 1.76 235 410 8.0 57.0 7.1 increase of the AWA from 0.55% to
0.60%, for a fixed dosage of SP,
9 0.55 1.17 180 295 3.0 11.9 4.0
10 0.43 0.55 1.40 210 340 4.0 23.8 6.0
affected the slump and slump flow.
11 0.55 1.76 250 360 4.4 31.5 7.16 For example, increasing the AWA
from 0.55% to 0.60% (mix 13 vs.
12 0.60 1.17 160 275 2.3 12.6 5.4
13 0.43 0.60 1.40 200 320 3.2 19.9 6.0 mix 10) reduced the slump flow
14 0.60 1.76 220 340 3.9 24.4 6.3 and the slump by 6 and 5%, respec-
tively. In turn, the reduction of the
WSHcR D : Loss of mass by washout after3 drops using CRD C6I test (%)
AWA from 0.55% to 0.30% (mix 7
WSHMc t :Loss of mass by washout using MC-I (%)
* : I/1-O0-~gof cement vs. mix 10) increased the slump
flow and the slump by 3% and 5%,
respectively.
3.2 Sensitivity Reducing the dosage of AWA from 0.55% to 0.30%
and increasing the dosage of SP from 1.40 l to 1.76 l (mix
The objective of the sensitMty investigationwas to study 8 vs. mix 10) increased the slump and slump flow by
the effect of small variations in dosage of the AWA and the 12% and 20%, respectively. The increase in AWA from
SP on the workability and washout resistance. The 0.55% to 0.60% and the reduction of dosage of SP from
water/cement ratio was constant at 0.43. The dosages of 1.40 1 to 1.17 1 (mix 12 vs. mix 10) decreased the slump
AWA were 0.30%, 0.55% and 0.60% by mass of cement. and the slump flow 24% and 19%, respectively (Figs. 9
The dosage of the SP were 1.17 1, 1.40 1and 1.76 1per 100 and 10).
kg of cement (Table 4). Workability of the mix was assessed The effect of the AWA and SP variations on the loss
by the slump and the slump flow tests, and the washout was of mass by washout is shown in Figs. 11 and 12. The
assessedby the MC-1 and CRD C61 tests. reduction the AWA dosage from 0.55% to 0.30% (mix 7
Details of the compositions of mixes used in the eval- vs. mix 10) increased the loss of mass by washout by
uation of the sensitivity (Mixes 6-14) are given in Table 90% and 112% for CRD C61 and MC-1 tests, respec-
2. The results of the workability and washout tests are
given in Table 4.

Fig. 11 - Effect o f A W A and SP on loss o f mass using C R D C61

test ( W / C = 0.43).
Fig. 9 - Effect o f A W A and SP on slump ( W / C = 0.43).

Fig. 12 - Effect o f A W A and SP on loss o f mass using MC-1 test

Fig. 10 - Effect o f A W A and SP on slump flow ( W / C = 0.43). ( W / C = 0.43).

278
Sonebi,Bartos,Khayat

Table 5 - Results of effect of AWA and type of washout test tion of the loss of mass by washout,
of 43% and 47% for the CRD C61
Mix W/C AWA(%) SP(1) Slump [Link] WSHcRD WSHMc.t WSHMc-1/
(mm) (mm) (%) (%) WSHcRD
and M C - 1 tests, respectively.
Otherwise, in case of a simultaneous
Control 0.50 0.0 1.0 160 265 16.3 52.2 3.3
increase of AWA and SP (mix 14 vs.
15 0.50* 0.5 1.17 230 380 7.5 46.6 6.2 mix 10), the loss of mass by washout
18 0.47 0.5 1.17 240 435 6.2 37.4 6.0 remained generally the same when
19 1.40 255 490 10.2 51.5 5.0 both CRD C61 and the MC-1 tests
20 1.62 265 540 11.6 73.2 6.3 were used. The results were 4% vs.
22 0.43 0.0 1.17 225 380 35.7 56.2 1.6 3.9% and 23.8% vs. 24.4%, respec-
16 0.43 0.5 1.67 245 425 10.0 44.3 4.43
tively. These mixes also showed the
17 2.49 255 445 11.0 56.5 5.14 same slump and slump flow results
(210 m m and 200 mm, and 340 mm,
21 0.43 0.5 1.30 220 335 5.5 36.4 6.6
215 5.2 (2) respectively).
10.9 (3) The results indicated the com-
pensation between the opposing
* : 390 kg/m 3 of cement (1) : 1/100 kg of cement
(2) : Loss of mass by washout using C R D C61 basket and big tube effects of AWA and SP. The AWA
(3) : Loss of mass by washout using a big basket and big tube appeared to increase viscosity and
reduce washout, while the SP
enhanced the workability. For a
tively. However, the increase of the dosage of the SP constant dosage of the AWA, the increase in the dosage
and keeping the dosage of AWA to 0.30% (mix 8 vs. mix of SP from 1.40 I to 1.76 1 increased the workability as
10), the loss of mass by washout was increased by 100% well as the loss of mass by washout; however, the reduc-
for CRD C61 and 140% for the MC-1 tests. This sharp tion in the SP alone caused a decrease in values of these
increase of the loss of mass by washout was less affected properties. In the case of the AWA content of 0.55%,
by the increase in dosage of the SP compared to the the comparison between mixes 11 and 10, the slump
reduction of AWA dosage. flow and the loss of mass by washout were increased (by
The increase of AWA dosage and the reduction in the 6% for slump flow, 10% and 32% for CP,,D C61 and
dosage of SP (mix 12 vs. mix 10), caused the same reduc- MC-1 tests, respectively).

3.3 Effect of AWA dosage on workability and

400
O Slump (ram) loss of mass by washout
9 S. flow (ram)
E" 35O
g SP = 1.21 I The influence of the AWA dosage on the slump flow
and loss of mass by washout for a concrete mix made
3OO
n with the same content of cement (450 kg/m 3) and W/C
E (0.43), and a constant dosage of the SP (1.17 1), was
--=
r 250
o:
determined by varying the AWA dosage from 0% to
0.6% (Tables 3-5) as shown by Fig. 13 and Fig. 14.
The increase in dosage of AWA reduced slump and
150 slump flow as shown in Fig. 13. However, the increase
0 0.30% 0.50% 0.55% 0.60% in AWA from 0.3% to 0.5% slightly increased the work-
AWA (%) ability, because the dosage of the SP increased by 0.045 I.
Fig. 13 - Effect o f A W A o n w o r k a b i l i t y ( W / C = 0.43 - SP = 1.17 l). The introduction of AWA, for the given dosage of
SP, enhanced the washout resistance (Fig. 14). Similarly,
the increase in AWA content reduced the loss of mass by
washout measured with both the C R D C61 and MC-1
60 ] Slump = 160 mm H Plunge test (%)
tests. For example, adding 0.3% AWA dosage showed a
reduction in loss of mass by washout measured with
g 50 /l I Slump= 195 mm ==MC-1 a t 4 rain I
CRD C61 and MC-1 tests 80% and 30%, respectively.
When the AWA dosage increased from 0.3% to 0.5%,
3O the loss of mass by washout of CRD C61 and MC-1 was
=~ 2 o / I I Slump= Slump= Slump= I
reduced by 50 and 65%, respectively.
10

0
0 0.30 0.50 0.5 0.60
A W A (%)

Fig. 14 - E f f e c t o f A W A on loss o f mass by washout

( W / C = 0.43 - SP = 1.17 l).

279
Materials and Structures/Mat6riaux et Constructions,Vol.32, May1999

o4'~ iiiii!iiii!!!!!!!!!iii
80,
Y = - 0.35x2 + 10.7 x - 15.3
;5

="
70-1 ~ = 0.90

g ~0

== ~5

~0
"6 Big 4O4
== 5
basket
o, CRD C61
0 30 q

5 20 q

0 10q
CRD C61
450 mm tube 450 turn MC-1
200 mm tube OI
0 2 4 6 8 10 12 14
F i g . 15 - Comparison of loss of mass measured by four washout Loss of mass (CP.D CSl) (%)
tests ( W / C = 0 . 4 3 - A W A = 0 . 5 % - S P = 1.3 1).
F i g . 16 - R e l a t i o n s h i p b e t w e e n loss of mass using C R D C 6 1 and
M C - 1 tests.

3.4 Comparison of washout measured

by different plunge tests and the spray test loss obtained with the MC-1 test was about 66%. For
apparatus example, for an acceptable loss of mass by washout of
5% to 10% [10], the corresponding loss of mass by
The loss of mass by washout was measured using a ref- washout was estimated at 30% to 57% when the MC-1
erence concrete mix made with 450 kg/m3 of cement and apparatus was used.
W/C of 0.43, and 0.5% of AWA and 1.30 1of SP (mix 21). Following this investigation, the comparison
The losses of mass by washout measured using the CRD between the washout tests CRD C61 and MC-1 can be
C61 basket and the big basket and tube were compared to summarised as follows:
the loss of mass by washout obtained with the MC-1 test. - The test procedure of the MC-1 is simple, as is the
The results are shown in Table 5 and in Fig. 15. CRD C61.
The loss of mass by washout using the perforated - The mass of the test samples of both the MC-1 and
basket from the C1KD C61 test did not appear to be CRD C61 tests are acceptable (1 kg and 2 kg).
affected by the increased diameter of the tube used. The - The final test results of the MC-1 and the CRD C61
results of the loss of mass by washout measured by the are quantitative; however, an additional visual assessment
standard plunge test CRD C61 using a tube of 200 mm of the washout is also available when using the MC-1
diameter and the big tube of 450 mm diameter were test.
almost identical, namely 5.5% and 5.2% respectively. - Using the MC-1 test, the potential operator error was
However, the loss of mass by washout obtained using the practically eliminated compared to that of the C R D
big tube and the big basket was 2 times higher than that C61.
of the standard plunge test CRD C61, namely 10.9% - The loss of mass using the MC-1 test was higher than
and 5.5% respectively. This difference can be explained that of the CRD C61 test for the same concrete. The
by the fact that the big basket was made of mesh with MC-1 is a more severe test.
50 mm square openings, and the surface of concrete in
contact with the water was much higher than that of the
standard CRD basket with round 3 mm holes. The 4. CONCLUSIONS
large openings also allowed coarse aggregate to be
washout out. The loss of mass using the MC-1 test was Based on the comparative study between the loss of
higher than all other tests (36.4%). mass by washout of the underwater concrete measured
with the standard plunge test CRD C61 and the MC-1
test, the following conclusions have been reached:
3.5 Relationship between CRD C61 and MC-1 9 The coefficients of variation from the repeatability
washout tests tests of workability of underwater concrete determined
by the slump and the slump flow are small (1.9% and
A good relationship was established between the loss 2.4%).
of mass obtained by the standard plunge test CRD C61 9 The coefficient of variation of the loss of mass by
and the MC-1 test, as shown in Fig. 16. The coefficient washout obtained using the MC-1 test is greater than
of correlation between the loss of mass by washout after that of the CRD C61 (12.7% vs. 5%), but the variation
3 drops and that of the MC-1 at 4 min was 0.90. This of results obtained using the MC-1 apparatus is consid-
relationship is valid where the loss of mass by washout is ered acceptable.
more than 2% as measured by the CRD C61 test. The 9 The workability and loss of mass by washout using
loss of mass by washout using the MC-1 test began to the CRD C61 and the MC-1 tests are both sensitive to
level-offwhere the washout measurements by the CRD variations of the dosage of AWA and dosage of the SP.
C61 were about 15 to 16%. In this case, the washout 9 The loss of mass by washout and the workability

2 8 0
Sonebi,Bartos,Khayat

are influenced by the concentrations of the AWA and REFERENCES

the dosage of SP. This result is in line with results of
previous research [10]. [1] CRD C61, 'Test Method for Determining the Resistance of
9 The loss of mass by washout using the CRD C61 Freshly-Mixed Concrete to Washing out in Water', US Army
appeared not to be affected by the diameter of the tube Experiment Station, Handbook for Concrete, Vicksburg,
Mississippi, Dec. 1989.
(200 vs. 450 mm). The loss of mass increased when a [2] Ceza, M., 'Fresh non-dispersive underwater concrete : Properties
big basket with 50 mm square wire mesh and a big tube and tests', Ph. D Thesis, University of Paisley, United Kingdom,
of 450 mm diameter were used. Sept. 1996.
9 The relationship between the loss of mass by [3] Ceza, M. and Bartos, P.J.M., 'Development a apparatus for test-
washout obtained with C R D C61 and MC-1 tests ing the washout resistance of underwater concrete mixtures',
ACI Concrete in 'Marine Environment', Proceedings Third
appeared to be of a polynomial type (1K2 of 0.90). CANMET/ACI International Conference, SP-163, (V.M.
9 The MC-1 test proved to be more severe than the Malhotra, Canada, 1996) 111-126.
C1KD C61, and the washout test results were always [4] Davies, B. A., 'Laboratory methods of testing concrete for place-
higher9 ment underwater', Proceeding, Marine concrete 86-
International Conference on Concrete in 'the Marine
Environment', London, Sept. 1986 (Concrete Society, London,
1986) 279-286.
SYMBOLS [5] Sogo, S. and Haga, T., 'Underwater Concrete Containing
Segregation Controlling Polymers', Proceeding 5th International
W/C : Water/cement ratio Congress on Polymers in 'Concrete', Brighton, England, Sept.
SP : Superplasticizer (I per 100 kg of cement) 1987, 22-24.
[6] Japan Society of Civil Engineers, JSCE, 'Recommendations for
AWA : Anti-washout admixture (% of mass of cement) design and construction in antiwashout underwater concrete',
Av. 9 (~) Concrete library ofJSCE (67) (May 1991) 89 p.
v(x-x0
n

s" Standard deviation s = ~i--~ n - 1

-- 2 [7] Annett, M. F., 'Underwater concreting using admixtures',
C o n c r e t e 21 (9) (1987).
[8] BS 12, 'Specification for Portland Cement', British Standards
Institute, London, (1983).
v 9 Coefficient of variation V----'Z s [9] BS 1881: Part 102, 'Method for Determination of Slump', British
X Standards Institute, London, (1983).
~ ' Mean result [10] Khayat, K. H., Sonebi, M., Yahia, A. and Skaggs, C. B.,
x i 9 Individual result 'Statistical models to predict flowability, washout resistance and
n" Number of result strength of underwater concrete', in 'Production Methods and
Workability of Concrete', Glasgow, June 1996 (E & FN Spon,
WSHca D 9 Loss of mass by washout after 3 drops using Scotland, 1996) 463-481.
CRD C61 test (%) [11] ISO 5725, 'Accuracy (trueness and precision) of measurement
WSHMc_ 1 9 Loss of mass by washout using MC-1 test at methods and results', (1990).
4 min (%)

ACKNOWLEDGEMENTS

The Advanced Concrete and Masonry Centre

(Scotland) and the Fund of Formation of Research in
Quebec (Canada) are gratefully acknowledged for their
support9

IIIl~lii~

281