Effect of Fly Ash On Self-Healing of Cracks in Concrete: Conference Paper
Effect of Fly Ash On Self-Healing of Cracks in Concrete: Conference Paper
Effect of Fly Ash On Self-Healing of Cracks in Concrete: Conference Paper
net/publication/327226077
CITATION READS
1 420
2 authors:
Some of the authors of this publication are also working on these related projects:
EVALUATION OF CONCRETE DURABILITY INDICATORS WITH DIFFERENT SUPPLEMENTARY CEMENTITIOUS MATERIALS View project
High Volume Pozzolanic Green Engineered Cement Composites (ECC) or Strain-Hardening Cement Composites (SHCC) in India View project
All content following this page was uploaded by Shan Ratnayake on 25 August 2018.
Abstract—The design of water retaining structures is mainly methods of quantifying the ability to self-heal and do not
based on the serviceability limit state crack control. The allowable simulate the real situation of water retaining structures where
crack width depends on the self-healing ability of concrete and the water is flowing through a crack. The effect of pressure gradient
use of supplementary cementitious material like fly ash in concrete across a crack in concrete with ordinary Portland cement (OPC)
mixes might affect it. Therefore, an experimental investigation was has also been studied and an optimum pressure gradient for
carried out to find the influence of fly ash on self-healing process. sealing crack widths of different sizes was found [4]. Further,
Water is sent through artificially induced cracks in a specimen for smart cementitious materials which can self-heal with minimal
autogenous healing to take place at a constant pressure gradient external support have also been explored [5].
across the specimen. To determine the level of self-healing, the flow
through the crack was measured with time to obtain the sealing
time. Fly ash percentages of 20%, 30% and 40% were tested along C. Test methods used to study self-healing in previous studies
with a 0% fly ash mix. Insignificant variation was shown for initial • Tensile splitting test used to split cylindrical specimen and
drop in flow rate across fly ash percentages used in this study. water sent through the interconnected crack across the
Significant reduction in sealing time was observed for 20% and diameter and time taken to seal the crack was used to
30% fly ash mixes as compared to 0% fly ash whereas higher fly quantify performance [4]
ash percentages (40%) showed insignificant reduction. • Cylindrical specimen was loaded to 70% and 90% of
compressive strength to induce interconnected cracks.
Keywords— water retaining structures; self-healing; crack UPV, RCP test and sorptivity were measured to quantify
width; concrete; fly ash; sealing time
how much the specimen had sealed [3]
• Crushing of specimen to 75 µm fine particles and testing
I. INTRODUCTION for hydration degree to see if healing can be sufficiently
done by the unhydrated cement [6]
A. Introduction to self-healing of cracks in concrete
The phenomenon of self-healing of cracks in concrete has D. Trends shown with other parameters
been observed and studied for a long time. The process of self- Higher FA content showed positive trends in unhydrated
healing can be attributed to three major processes [1]: OPC, reduction of porosity [6], RCP test values, and an optimal
• Further hydration of unhydrated cement at the cracked FA content was recognized for each. FA mixes have shown
surface better performance in self-healing than ggbfs [5]
• Recrystallization of portlandite leached from the bulk Comparison of Portland limestone cement and OPC have
paste shown no appreciable difference in self-healing [7]. For a given
• Formation of calcite (CaCO3) hydraulic gradient across a crack, optimal crack width was
Design of water retaining structures (WRS) is based on the identified to seal crack in minimum time. Experiments
crack width limitation criteria. Previous studies have shown that conducted in the range 20 °C to 80 °C concluded that the rate of
supplementary cementitious material (SCM) used in the self-healing increased from 20 °C to 80 °C.
concrete mix have a large impact on the self-healing
performance. SCM includes pulverized fuel ash (fly ash) (FA), E. Importance and identification of knowledge gap
silica fume and ground granulated blast-furnace slag (ggbfs) However, the study of how supplementary cementitious
material such as fly ash, silica fume and ggbfs affect the process
B. Previous studies on self-healing of self-healing has not been looked into in depth. The most
Various parameters affecting this mechanism, such as crack commonly used SCM in the construction industry, especially in
width, water seepage rate through cracks and temperature have the construction of WRS is FA. With the introduction of FA, the
been studied in earlier researches [2]. Studies have been done on percentage of OPC in the mix reduces. Thereby, the amount of
strength, ultrasonic pulse velocity (UPV), rapid chloride Ca(OH)2 produced from the hydration of OPC varies and
permeability (RCP) and sorptivity of concrete, considering high consequently may affect the self-healing process. Hence the
percentages of FA in the mix [3]. Most of these are implicit
Since a major portion of the fly ash in the mix is expected to After the initial drop in flow rate, interval between readings
react within 28 days, it is acceptable to start conducting the were lengthened such that sufficient quantity of water was
testing self-healing of the specimens at 28 days. collected to take a reading (approximately 500 ml – 3000ml).
The measurements were taken until the cracks were sealed fully.
The cylindrical specimen was cured for 28 days before the The average flow rate in the time interval between time t1 and t2
tensile splitting test was done. was obtained from equation (1). For reasonable approximation,
the flow rate obtained was assumed to be the flow rate at time
(t1+t2)/2.
IV. RESULTS
The compressive strengths of cubes are given in Table III.
Average measured crack widths of cylindrical specimens
were in the range of 0.1 mm to 0.15 mm in most of the
specimens. Crack widths of some samples could not be lowered
beyond 0.2 mm even after maximum tightening of steel strap.
Fig 3. 0.1 mm crack through microscope
Since the two surface crack sizes measured on either end of
Fig 4. Fraction of reacted FA with time [8] Fig 5. Final arrangement of specimens and apparatus
TABLE III. FA% AND COMPRESSIVE STRENGTH
0.30
FA % Compressive strength (MPa) 4C
0 39.4 0.25
20 36.8 2A
All of the samples showed a rapid drop in flow rate at the 20C 0.125
start (See Fig 7) and then the flow rate reduction was gradual 0.100 30A 0.119
(See Fig 8). 40B 0.116
0.010
From the flow rate data, the initial reduction of flow rate to
0.01 ml/s was calculated by linear interpolation of the two
0.001
closest values. The time taken for that reduction for different
samples were plotted against the different crack widths as shown
in Fig 9. 0.000
0 240 480 720 960 1200 1440
Time (h)
TABLE IV. MEASURED AND CALCULATED CRACK WIDTH OF
SPECIMENS Fig 8. Flow rate with time (gradual drop)
Measured average Calculated crack width From Fig 9, it can be seen that between the crack sizes of
Sample name
crack width (mm) (mm) 0.094 to 0.15 mm, the time taken for flow rate to drop to 0.01
0A 0.12 0.094 ml/s was approximately the same for all specimens considered.
2A 0.20 0.199 Considering 20% FA specimens only, it can be seen that the time
2B 0.13 0.138 has not changed significantly from one specimen to the other.
2C 0.11 0.125
3A 0.10 0.119 The same can be observed for specimen with 30% fly ash
3B 0.13 0.145 (3B and 3C). 3A, however seems to be an outlier in this case.
3C 0.13 0.132 The non-uniform crack width across the sample may have
4A 0.098 0.154 affected the initial drop.
4B 0.11 0.116
4C 0.26 0.237
The same can be observed for specimen with 30% fly ash
(3B and 3C). 3A, however seems to be an outlier in this case.
900 2A
698 hours. Hence, all values will be normalized for 0.1 mm
crack width (CW) using a normalization factor in equation (3).
time taken to reduce flow to 0.01
800
700 The normalized sealing time (ST) is calculated by equation (4).
600
𝑆𝑇 (𝐶𝑊 0.1 𝑚𝑚)
500
ml/s (h)
4A
𝑁𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑓𝑎𝑐𝑡𝑜𝑟 = − (3)
400 𝑆𝑇 (𝐶𝑊 𝑥 𝑚𝑚)
300 3A
𝑆𝑇 (𝐶𝑊 0.1 𝑚𝑚) 698
200 0A 4B 2C 3C2B 3B =
𝑆𝑇 (𝐶𝑊 𝑥 𝑚𝑚) 342862.54𝑥 2 − 5580.88𝑥 + 2827.21
100
0 𝑆𝑇 = 𝑁𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑓𝑎𝑐𝑡𝑜𝑟 × 𝐴𝑐𝑡𝑢𝑎𝑙 𝑠𝑎𝑚𝑝𝑙𝑒 𝑆𝑇 − (4)
0.09 0.11 0.13 0.15 0.17 0.19 0.21
Based on the normalized sealing times given in Table V and
crack width (mm)
shown in Fig 12, the following observations can be made:
Fig 9. Time taken for specimens to reach 0.01 ml/s flow for different specimens
• FA mixes having 20% and 30% have shown
The non-uniform crack width across the sample may have approximately 25% reduced sealing time as compared to
affected the initial drop. the 0% FA mix
However, considering a much larger crack size of 0.20 mm, • FA mixes having 40% have shown only 6% reduction in
the time for this initial drop has been understandably delayed sealing time compared to 0% FA mix
due to the large crack size. Fig 10 shows the time taken to seal
the cracks completely. V. CONCLUSION
Some specimens were not used in this comparison since the The initial rapid drop in flow rate was common to all the
flow rate of those specimens showed a sudden increase after an mixes and insignificant variation was shown with respect to FA
accidental discontinuity in water supply. The sudden increase in percentages.
flowrate could be assumed to be due to the dislocation of the
particles inside the crack. As for the performance of full closure of the crack by self-
healing, there was a significant difference in the time taken with
For comparison purposes, the sealing time for specimens respect to FA percentages.
were normalized to that of a 0.10 mm crack width specimen
using the sealing time variation with crack width for pressure Since comparison between FA percentages and sealing time
gradient of 10 given in the reference [4]. could not be done with samples having different crack widths,
2500
Total time taken to seal 100% (h)
2000
2B
1500
4B 2C
1000 0A 3A
500
0
0.09 0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.20
crack width (mm)
Fig 10. Graph of time taken for specimens to seal Fig 11. Sealing time vs crack width [4]
900
ACKNOWLEDGMENT
0A
800 4B The authors would like to thank Siam City Cement (Lanka)
700
Ltd for providing the fly ash used in the experiment. The
2C
2B 3A assistance given by the technical staff of the Structural Testing
Sealing time (h)
600
and Material Testing Laboratories in the Department of Civil
500 Engineering is also acknowledged.
400
300 REFERENCES
200
[1] H. Huang, G. Ye, C. Qian and E. Schlangen, “Self-healing in cementitious
100 materials: Materials, methods and service conditions,” Materials and
0 Design, vol. 92, p. 499–511, 2016.
Sample [2] H.-W. Reinhardt and M. Jooss, “Permeability and self-healing of cracked
concrete as a function of temperature and crack width,” Cement and
Concrete Research, vol. 33, p. 981 – 985, 2003.
Fig 12. Normalized sealing time variation for different FA mixes
[3] M. Sahmaran, S. B. Keskin, G. Ozerkan and I. O. Yaman, “Self-healing
could not be done with samples having different crack widths, of mechanically-loaded self consolidating concretes with high volumes of
the sealing time of the specimens were normalized to 0.1 mm. fly ash,” Cement & Concrete Composites, vol. 30, p. 872–879, 2008.
[4] S. Nanayakkara, “Self-healing of cracks in Concrete under water
• FA mixes having 20% and 30% have shown pressure,” in New Technologies for Urban Safety of Mega Cities in Asia,
approximately 25% reduced sealing time as compared to Tokyo, 2003.
the 0% FA mix [5] M. Sahmaran, G. Yildirim and T. K. Erdem, “Self-healing capability of
cementitious composites incorporating different supplementary
• FA mixes having 40% have shown only 6% reduction in cementitious materials,” Cement & Concrete Composites, vol. 35, p. 89–
sealing time compared to 0% FA mix 101, 2013.
[6] P. Termkhajornkit, T. Nawa, Y. Yamashiro and T. Saito, “Self-healing
From these results, it can be seen that 20% and 30% FA ability of fly ash–cement systems,” Cement & Concrete Composites, vol.
mixes perform better than 0% FA in sealing cracks in the range 31, p. 195–203, 2009.
of 0.1 mm to 0.15 mm, which also matches with the literature. [7] S. M. A. Nanayakkara and T. Elakneshwaran, “Self-healing of cracks in
However, very high percentage of FA such as 40% does not concrete with Portland Limestone Cement,” in 6th International Congress
in Concrete Technology, Dundee, 2005.
show significant improvement in performance as much as the
20% and 30%. [8] Y. M. Zhang, W. Sun and H. D. Yan, “Hydration of high-volume fly ash
cement pastes,” Cement & Concrete Composites, vol. 22, pp. 445-452,
It is not recommended to use high volume FA mixes in the 2000.
concrete mix for WRS (more than 40% FA)