Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/269152651

Evaluation of the Segregation Resistance of Fresh Self-Compacting Concrete


using different test methods

Conference Paper · January 2005


DOI: 10.1617/2912143624.031

CITATIONS READS

8 4,734

1 author:

M. Sonebi
Queen's University Belfast
290 PUBLICATIONS   2,566 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Use Compressive Membrane Action theory and optimised concrete mixes to reinvent box culvert design and construction View project

Organisation of third International Conference on Bio-Based Building materials View project

All content following this page was uploaded by M. Sonebi on 14 August 2015.

The user has requested enhancement of the downloaded file.


PROJECT TESTING SCC-SEGREGATION TEST METHODS
François CUSSIGH1, Mohammed SONEBI2, Geert De schutter3
1 GTM Construction, Paris, France
2 ACM Centre, University of Paisley, Paisley, Scotland (UK)
3 Department of Structural Engineering, Royal University of Ghent, Belgium

ABSTRACT : The study described is a part of a large European research project on self-
compacting concrete: “TESTING – SCC”. The University of Paisley, Ghent University and
GTM Construction are responsible for Task 3-3 on the investigation of resistance of
segregation. The objectives are: to identify suitable tests for assessment and measurement of
the segregation resistance of SCC, for use in the lab and on site; to identify characteristic test
values for each test; focusing on static segregation. Three test methods, described in RILEM
compendium of tests report TC 145 WSM, have been examined: settlement column test, sieve
stability test, and penetration test.
In order to assess sensitivity and variability of these test methods, the three partners
evaluated in parallel a common programme of tests on mixes designed to have different levels
of segregation resistance. From a satisfactorily stable control SCC mix the segregation was
increased by adding 10 and 20 L/m3 of water, or 1 and 2 kg/m3 of superplasticizer. The
segregation was also measured on hardened concrete and compared to the results obtained
from fresh concrete. Interesting results have been obtained showing acceptable repeatability,
reproducibility and sensitivity and the next step will be to compare segregation measurements
on fresh SCC and real in-situ segregation.

KEYWORDS: Self-compacting concrete, penetration test, segregation, settlement column test,


sieve stability test

THE TESTING SCC PROJECT


The aim of the project is to develop test methods for fresh-compacting concrete to
establish a basis for European standards, and to support and accelerate further introduction of
this technology in general construction.

THE OBJECTIVES OF THE PROJECT


- Principally to find tests which identify the three key properties of SCC, for mix design
purposes in the lab, and for compliance purposes on site.
- to find tests which will enable engineers and suppliers to clearly identify both ‘good’ and
‘bad’ SCC
- to be able to recommend a range of results for the chosen tests, which enable non-
specialists to identify suitable SCC, and be considered for compliance in specifications.
- to confirm the scientific basis of these tests by fundamental rheological measurements of
the concrete.
- to assess the use of these tests in real construction.
- to avoid duplication of work in different European countries and establish agreed
guidelines for a draft standard.
- to encourage the use of self-compacting concrete in general construction and promote the
potential economic and environmental benefits of this technology.
The three key properties of SCC are :
- Filling ability: the ability of the concrete to flow freely under its own weight, both
horizontally and vertically upwards if necessary, and to completely fill formwork of any
dimension and shape without leaving voids.
- Passing ability: the ability of the concrete to flow freely in and around dense
reinforcement without blocking.
- Resistance to segregation: during placement and while flowing, the concrete should retain
its homogeneity. There should be no separation of aggregate from paste or water from
solids, and no tendency for coarse aggregate to sink downwards through the fresh concrete
mass under gravity.
The structure of the work-plan is divided into seven work-packages and work-package
3, the laboratory work, is the heart of the project. It is made up of three tasks which are
strongly inter-related. The paper present the results obtained in task 3-3 which is concerned
with resistance to segregation. The three partners are GTM, University of Paisley (UoP) and
Royal University of Ghent (RUG).

THE OBJECTIVES OF TASK 3-3


Task 3-3 aims at identifying suitable tests for assessment and measurement of the
segregation resistance of SCC mixes, for use in the lab and on site. It will also try to identify
characteristics test values, for each test, for segregation resistance assessment.
After discussion among project partners, three test methods have been selected for initial
assessment:
- Settlement column segregation (1-4)
- Sieve stability (1, 5)
- Penetration apparatus (6-8)
In order to check the variation inherent in the test methods the three partners [GTM,
UoP and RUG] have performed in parallel a common lab trials programme with different
mixes.

MATERIALS AND MIX PROPORTIONS


The concrete mixes investigated in this study were prepared with CEM II cement with
interground limestone - Skövde Byggcement produced by Cementa in Sweden. Limestone
powder - Betocarb 80 - from Omya (UK), produced from very high purity carboniferous
limestone was used. . The relative density of the limestone powder was 2.68 and the material
was finer than cement.
Continuously graded Glensanda (Scotland) crushed granite aggregates 4/8 mm and
8/16 mm were used. A marine sand from Denmark was employed. The relative density
values of the coarse aggregates and sand were 2.68 and 2.65, and their absorption rates were
0.87%, 0.72 and 0.30%, respectively. Glenium 51 superplasticizer was used which had solid
content and specific gravity of 30% and 1.05, respectively.

Four mix proportions have been developed from a control Mix C :


- Mix SW1 Slight segregation with more water compared to control mix
- Mix SW2 Strong segregation with more water compared to control mix
- Mix SA1 Slight segregation with more SP compared to control mix
- Mix SA2 Strong segregation with more SP compared to control mix

The mix proportions of all SCC mixes are given in Table 1.


Table 1 – Mix proportions of all SCC mixes
C SW 1 SW 2 SA1 SA2
M aterial/T ype kg/m³ kg/m³ kg/m³ kg/m³ kg/m³
Skovdö, Bygg Cem ent 470.0 470.0 470.0 470.0 470.0
Lim e ston e powder, Betocarb 80 133.0 133.0 133.0 133.0 133.0
W ater 187.0 197.0 207.0 187.0 187.0
Glen ium 51 5.13 5.13 5.13 6.13 7.13
RN class A 790.0 790.0 790.0 790.0 790.0
Glen sanda 4/8 m m 234.0 234.0 234.0 234.0 234.0
Glen sanda 8/16 m m 543.0 543.0 543.0 543.0 543.0
T otal 2372 2382 2392 2373 2374
T otal vol. [l] 999 1009 1019 1000 1001

Powder con tent [kg] 603 603 603 603 603


Powder con tent [l] 199 199 199 199 199
W /P ratio (weight) 0.33 0.34 0.36 0.33 0.33
W /P ratio (volum e) 0.99 1.04 1.09 0.99 0.99
Paste con ten t (vol.% ) 41.1 41.7 42.3 41.2 41.2

EXPERIMENTAL PROGRAMME AND TEST METHODS


Three (or in some cases four) successive and identical batches of mix C have been
produced. On every batch, slump-flow and T500 (the ‘basic’ tests for fresh SCC) were
measured after 30 minutes. The target slump flow of mix C was 705 ± 25 mm. The 3
segregation tests were performed and a concrete specimen cast to check segregation on
hardened concrete. The same tests were performed with the four other mixes: SW1, SW2,
SA1, SA2 which had been deliberately varied to induce segregation. The influence of the
speed of pouring of the sieve stability test was also investigated. The segregation of hardened
concrete was also measured for all mixes.
The sieve stability test is shown in Fig. 1. The test aims to determine how likely a SCC
mix is to segregate by allowing a 10 L concrete sample to undergo static segregation for 15
minutes (in a bucket). Then the top layer of the sample (4.8 kg ± 0.2) is poured onto a 5 mm
sieve and some mortar passes through the sieve. The potential segregation is expressed as the
ratio between the mass of mortar collected through the 5 mm sieve and the initial mass of the
top layer(1, 5). More mortar passing through the sieve indicates a greater liability to
segregation.

Figure 1. Sieve stability segregation test (1, 5)


Fig. 2 shows the settlement column test (1-4). The dimensions used for the settlement
column test used at UoP, GTM and RUG are: 500 x 150 x 100 mm, 400 x 150 x 100 mm, and
400 x 100 x 100 mm, respectively. The test involves samples of concrete being taken from
the top part and bottom part of a column shaped apparatus via doors after a controlled jolting
cycle and standard settlement period, except at RUG laboratory, where the sample of concrete
was rodded with a tamping bar similar to that used in the standard slump test. Those samples
were analysed to determine proportion of coarse aggregate. Segregation resistance is
expressed as the ratio between coarse aggregate mass in the top part and coarse aggregate
mass in the bottom part. A lower ratio indicates more coarse aggregate in the bottom layer,
therefore an increased liability to segregation.

Figure 2. Settlement column segregation test (1-4)

The penetration test is shown in Fig. 3 (6-8). The test consists of a frame, a slot, a
reading scale and a penetration head (an aluminium cylinder). The head is placed on the
concrete surface and falls freely under its own weight (54 g) into the concrete. Segregation is
expressed as the fall depth (mean value of three measurements corresponding to three
different points on the surface of the concrete). If the upper layers of concrete have been
subject to the settlement of coarse aggregate, and consist largely of mortar, the penetration
depth will be greater.
.

Figure 3. Penetration test (6-8)


RESULTS AND DISCUSSION
Results obtained are presented through graphs showing variations of slump-flow and
segregation behaviour in the different mixes. The points on the graphs represent the range
and average results

Segregation of fresh concrete


Fig. 4 presents the variations of slump-flow versus the segregation obtained for the
three segregation tests for all mixes. It can be seen that the increase in slump flow led to an
increase of the penetration and sieve stability values and a decrease of the settlement column
ratio. The increase of slump flow by adding more SP or water resulted in a reduction of the
segregation resistance measured by the three apparatuses of segregation.
44

39
SIEVE STABILITY TEST (%)

34

29

24

Mix C
19 Mix SW1
Mix SW2
14 Mix SA1
Mix SA2
9
690 710 730 750 770 790 810 830 850
SLUMP FLOW (mm)

45

40

35
PENETRATION TEST (mm)

30

25

20

Mix C
15
Mix SW1
10 Mix SW2

Mix SA1
5
Mix SA2
0
690 710 730 750 770 790 810 830 850
SLUMP FLOW (mm)

Figure 4. Graphs sieve stability vs. slump flow, penetration vs. slump flow, settlement column
vs. slump flow for all SCC mixes
80

70

SETTLEMENT COLUMN TEST (%)


60

50

40

Mix C
Mix SW1
30 Mix SW2
Mix SA1
Mix SA2
20
650 670 690 710 730 750 770 790 810 830 850
SLUMP FLOW (mm)

Figure 4. Graphs sieve stability vs. slump flow, penetration vs. slump flow, settlement column
vs. slump flow for all SCC mixes (continued)
The objective of examining sensitivity of the three test methods is to study the effect of
variations in dosages of superplasticizer and water on segregation resistance. The sensitivity
of the tests can be defined as the ability to detect and measure the differences of segregation
resistance. Table 2 summarises the differences between all segregation tests for the three
laboratories; the percentage figures in bold represent the degree of change from the control
mix results. It can be seen that the sensitivity is good (in general) for the three apparatuses.
Table 2. Sensitivity of the tests
GTM C SW1 SW2 SA1 SA2
12.3 17.4 37.6 17.5 25.1
Sieve Stability
0% +41.5% +206% +42% +104%
3.7 7.2 42.1 9.3 26.8
Penetration test
0% +95% +1040% +151% +624%
Settlement 90.3% 95.6 76.6 96.5 92.1
column 0 +5.7% -15.2% +6.9% +2%

RUG C SW1 SW2 SA1 SA2


17.6 30.2 45.2 23.2 35.5
Sieve Stability
0% +71.5% +157% +32% +102%
11.7 31.8 42.3 25.4 40.2
Penetration test
0% +172% +261% +117% +244%
Settlement 72 76.1 32 74.4 69
column 0% +6% -56% +4% -4%
UoP C SW1 SW2 SA1 SA2
23.9 34.9 58.1 38.5 52.3
Sieve Stability
0% +46% +143% +61% +119%
19.2 30.3 44.3 40.8 43.9
Penetration test
0% +58% +131% +113% +129%
Settlement 75.8 56.5 41.5 57.8 53.9
column 0% -26% -45% -24% -29%
Comparison between labs
It is important to note that although materials and mix proportions were the same, the
results varied between the three labs. Visually, even the reference mixes differed in terms of
cohesion and viscosity, and were, in fact, observed to be different concretes. Slightly
different water contents were used to achieve the target slump flow, but more importantly,
this can be attributed to differences in mixers and in concrete temperature. It has therefore
been difficult to assess variation of the segregation test methods because the concretes
produced in the 3 labs differed.

An assessment of the reliability of the test methods can be obtained by analysing correlation
between test methods.
Fig. 5 represents the values obtained from the penetration test and sieve stability for
each mix and laboratory. There is a good correlation between these two tests because the
results of both are dependent on the depth of mortar in the top layers of the sample – this
depth is greater when the concrete has segregated..
50

sw2
sa2 sa2 sw2
40 sa1 sw2
PENTRATION (mm)

sw1

30
sa2 sw1

sa1

20
C

sa1 GTM
10 C
RUG

sw1 PAISLEY

0
10 15 20 25 30 35 40 45 50 55 60
SIEVE STABILITY (%)

Figure 5. Graph penetration/sieve stability

Criteria
For the three tests, criteria have been proposed to qualify the resistance to segregation of
the mixes: good or poor resistance (1).

Penetration test : - Pd < 8 mm concrete has good resistance to segregation


- Pd > 8 mm concrete has poor resistance to segregation
where Pd is the measured penetration depth

Settlement column test : - segregation ratio lower than 0.9: the mix is liable to segregation.
- ratio greater than 0.95: good resistance to segregation.
Sieve stability test : - less than 5% of the sample passing the sieve: could be too
cohesive/viscous.
- 5-15% of the sample passing the sieve: optimum amount of
resistance to segregation.
- 15-30% of the sample passing the sieve: likely to be susceptible to
segregation.
- more than 30% of the sample passing the sieve: likely to be
susceptible to severe segregation.

In later work, the existing criteria will be further evaluated by means of the obtained
test results. A more fundamental investigation on alternative criteria is still going on.

Segregation of hardened concrete


The method developed by T. Sedran was used to assess the segregation on hardened
concrete (9). This method consists of measuring the depth of the first two coarse aggregate
particles (> 8 mm) from the top of 160 x 320 mm cylinders. Segregation is considered to
exist if the average depth of the two particles is greater than 10mm.

Table 3. Average depth from the top (mm) of the first two coarse aggregate particles

Mix Mix C Mix SW1 Mix SW2 Mix SA1 Mix SA2
6 4 17 1 4.5
RUG 0.5 1 10 0.5 5.5
1.5 11.5 8.5 10.5 7
Average 2.6 5.5 11.8 4 5.6
6.5 5.5 29 4.5 11.5
GTM 7 10 21 10.5 70
10 8.5 22 16 70
Average 7.8 8 24 10.3 50.5
1.5 3 10 4.5 11
1 2 18.5 10.5 10.5
UoP
1 14.5 29 7 10
3.5 9.5 22.5 8 29.5
Average 1.8 7.3 20 7.5 20.3

For the three laboratories, the increase of SP or water dosages did lead to an increase
of the depth of the first two coarse particles in the top of the cylinder. In these mixes,
segregation is considered to be present if any one of the three average depths is greater than
10 mm. Mixes SW2 and SA1 exhibited segregation in all three labs. Mixes SW1 and SA2
also exhibited segregation with only one exception each.
Comparisons based on the suggested criteria for segregation resistance of fresh
concrete and that assessed from hardened concrete are given in Table 4.
On the basis of Table 4, we can assess the validity of the criteria for each test. Even if
accuracy of segregation measurement on hardened concrete is poor (with 160*320 cylinders
cast in lab), it is clear that further testing is necessary and comparison of test results with real
segregation will be done on full scale trials (task 4 of the project).
Table 4. Comparison of segregation on cylinder and with tests

Sieve stability test Penetration test Settlement column


Evidence of Indication Is Indication of Is Indication Is
Mix segregation of criterion segregation criterion of criterion
in cylinder? segregation? OK? OK? segregation OK?

‘C’
RUG no no YES no YES yes not defined
GTM no no YES no YES no YES
UoP no yes NO yes NO yes NO

‘SW1’
RUG yes yes YES yes YES yes not defined
GTM no yes NO no YES no YES
UoP yes yes YES yes YES yes YES

‘SW2’
RUG yes yes YES yes YES yes not defined
GTM yes yes YES yes YES yes YES
UoP yes yes YES yes YES yes YES

‘SA1’
RUG yes yes YES yes YES yes not defined
GTM yes yes YES yes YES no NO
UoP yes yes YES yes YES yes YES

‘SA2’
RUG no yes NO yes NO yes not defined
GTM yes yes YES yes YES no NO
UoP yes yes YES yes YES yes YES

Effect of the speed of pouring / manual pouring in sieve stability test


In this part of the programme, an attempt has been made to examine the influence of the
speed and the method of pouring using mix SW1. Automatic pouring was performed as
normal with the tilting machine. Manual pouring was performed with the same bucket from a
height of 50 cm. Three speeds of pouring were used: 8 seconds for the rapid one; 35 seconds
for the average one and 1 minute for the slow one. The results are given in Table 5.
The results indicate that the sieve stability test was totally independent of both kinds of
pouring (automatic/manual) and speeds of pouring.
Table 5. Results on sieve stability tests with different speeds of pouring and manual
pouring
SIEVE STABILITY TEST

SW1
Batch 1 2 3
22,5 - - rapid (8sec.)
MANUAL POURING - 22,5 - average (35 sec.)

- - 23,1 slow (1min.)

Slump flow 735 750 760

22,4 - - rapid (8sec.)


AUTOMATIC
- 21,6 - average (35 sec.)
POURING
- - 24,1 slow (1min.)

Assessment of variability
It is important to emphasise that each trial, including the repeats at each laboratory,
was conducted on different batches of material. Proper calculations for repeatability and
reproducibility cannot therefore be made, since they have to be based on results of tests
carried out on identical materials. The partners used the same sources of all materials (and of
course the same proportions) but, given the innate variation of raw materials for concrete,
they cannot be considered identical.
In Fig. 6 the standard deviations (sd) of each segregation test on fresh concrete are
plotted against the average results.
For the sieve stability test, at levels of stability lower than 30% (the absolute limit for
non-segregating mixes) the average standard deviation was 1.8 %.
For the penetration test, at levels of penetration below 10mm (near the limit for non-
segregating mixes), the average standard deviation was 1.7%.
For the settlement column test, at segregation ratios above 85% (near the proposed
limit for non-segregating mixes), the average standard deviation was 2.9%.
In general for all three tests, SDs were higher in the concrete where segregation was
present.
ALL LABS / SIEVE TEST :STANDARD DEVIATION / MEAN VALUE
14

12

STANDARD DEVIATION (%)


10

6
SW2

4
GTM

2 RUG
SA2
UOP

0
0 10 20 30 40 50 60
SIEVE AVERAGE (%)

ALL LABS / PENETRATION TEST :STANDARD DEVIATION/ MEAN VALUE


14

12
STANDARD DEVIATION (mm)

10

6
SW2
4
3 GTM
SA2 RUG
2
UOP

0
0 10 20 30 40 50 60
PENETRATION AVERAGE (mm)

ALL LABS/SETTLE. COL. TEST:STANDARD DEVIATION/MEAN VALUE

14

12
STANDARD DEVIATION (%)

10

8
SW2

6
SA2

4
GTM
RUG
2
UOP

0
20 30 40 50 60 70 80 90 100
SETTLEMENT COLUMN AVERAGE (%)

Figure 6. Standard deviation vs. average value


CONCLUSIONS
Based on the preliminary results presented in this paper, the following conclusions can be
drawn:
• in assessing the segregation of fresh concrete, the sensitivity of all three tests to the
variation of the dosages of superplasticizer and water was considered good.
• generally, variability of results was high in the highly segregating mixes. This is
considered to be due in particular to sampling problems. Acceptable levels of
variability will depend on final criteria chosen, but it is already clear that values of
standard deviation can be low in concrete mixes with good resistance to segregation.
• choice of suitable tests for identification of resistance to segregation can only be made
after full scale trials in task 4 when verification with in-situ segregation can be done.

ACKNOWLEDGEMENTS

The study was carried out at University of Paisley, University of Ghent and GTM and is part
of the Growth project “Testing SCC : Measurement of properties of fresh self-compacting
concrete” contract GRD2-2000-30024. The European Commission funds this project.

REFERENCES

1. Bartos, P. J. M., Sonebi, M., Tamimi, A.K., “Workability and Rheology of Fresh
Concrete: Compendium of Tests,” Report of RILEM Technical Committee TC 145-
WSM: Workability of Special Concrete Mixes, RILEM Publications S.A.R.L, Paris,
2002, 127 p.
2. Rooney, M. J., Bartos, P.J.M., “Development of the Settlement Column Segregation
Test for Fresh Self-Compacting Concrete,” Proceedings of the 2nd International
Symposium on Self-Compacting Concrete, K. Ozawa, Ouchi, M., Tokyo, Japan, pp.
109-116, 2001.
3. Sonebi, M., Rooney, M., Bartos, P.J.M., “Assessment of the Segregation Resistance of
Fresh Self-Compacting Concrete,” to appear in 6th CANMET/ACI International
Conference on recent Advances on Concrete Technology, Bucharest, June 2003, 12 p.
4. Sonebi, M., Rooney, M., Bartos, P.J.M., “New Test Method to Evaluate the
Segregation Resistance of Fresh Self-Compacting Concrete using the Settlement
Column Test,” submitted to Cement and Concrete Research, 2003, 26 p.
5 Recommandations provisoires sur les bétons auto-plaçants, Documents Scientifiques
et Techniques, Association Française de Génie Civil, July 2000, 62 p.
6 Skarendahl, Å., Petersson, Ö., “State-of-the-Art Report of RILEM Technical
Committee 174-SCC, Self-Compacting Concrete,” Paris, RILEM Publications
S.A.R.L, 2000, 154 p.
7. Bui, V. K., Montgomery, D., Hinczak, I., Turner, K., “Rapid Testing Method for
Segregation Resistance of Self-Compacting Concrete,” Cement & Concrete Research,
Vol. 32, pp. 1489-1496, 2002.
8. Bui, V. K., Akkaya, Y., Shah, S.P. “Rheological Model for Self-Compacting
Concrete”, ACI Materials Journal, Vol. 99, No. 6, 2002, pp. 549-559.
9. Sedran, T., “Rhéologie et rhéometrie des bétons – Application aux bétons
autonivelants”, Ph.D thesis, Ecole Nationale des Ponts et Chaussées, Paris, 1999,
219 p.

View publication stats

You might also like