, ACI MATERIALS JOURNAL TECHNICAL PAPER

Title no. 85-M46

Response of Concrete to Sulfuric Acid Attack
by Emmanuel K. Attiogbe and Sami H. Rizkalla
The response oj four different concrete mixes to sulfuric acid attack
was evaluated in an accelerated laboratory test program. Small test
specimens cut from standard concrete cylinders and a 1 percent sul-
furic acid solution with a pH oj J were used in the test program.
Changes in weight and thickness of the test specimens were used as
physical indicators 0/ the degree of deterioration, while increase in
sulfur content of the test specimens was used as a chemical indicator
of the degree of deterioration.
The study shows that all three indicators of deterioration are effec-
tive measures oj concrete response to the acid attack. However, the
study suggests that the increase in thickness (expansion) of small
specimens (with large surface area-to-volume ratios) may be a more
consistent measure than the weight loss of larger specimens when
comparing the effects of different sulfuric acid concentrations on
concrete. Photomicrographs of the concrete microstructure show that
the concrete deterioration storts from the acid-exposed surface and
progresses inward. The degree of concrete deterioration is increased
by alternate wet-dry cycles of exposure to sulfuric acid. The rate of
concrete deterioration along the penetration depth of sulfuric acid
could be described by a variation in sulfur concentration with the
depth of acid penetration.
Keywords: acids; chemical attack; concretes; deterioration; electron micro-
scopes; expansion; measurement; microstructure; pH; sulfurj sulfuric acid;
thickness; weight (mass).
Concrete is susceptible to attack by sulfuric acid pro-
duced from either sewage or sulfur dioxide present in
the atmosphere of industrial cities. This attack is due to
the high alkalinity of portland cement concrete, which
can be attacked by other acids as well. Sulfuric acid is
particularly corrosive due to the sulfate ion participat-
ing in sulfate attack, in addition to the dissolution
caused by the hydrogen ion.' Since sulfur compounds
are formed as a result of the sulfuric acid-cement paste
reaction, the increase in sulfur content of concrete
specimens could be used as a measure of the chemical
manifestation of deterioration.
In previous studies, weight loss, reduction in com-
pressive strength, and change in dynamic modulus of
elasticity were used to evaluate the extent of concrete
deterioration due to sulfuric acid attack.
2
. ' These stud-
ies indicate that damage starts at the surface of the
concrete and progresses inward. However, the extent of
damage along the depth of penetration of acid is not
ACI Materials Journal I November-December 1988
clearly defined. This information is necessary to accu-
rately estimate the minimum thickness of the concrete
cover in reinforced concrete structures or to adequately
design for sacrificial layers in concrete structures ex-
posed to sulfuric acid solutions.
As expected, the previous studies have generally
shown that weight loss of the test specimens increases
with a decreasing pH level of the acid solutions. How-
ever, in a recent study,' a solution with a pH of 3 pro-
duced a greater weight loss than one with a pH of 2.
This apparent anomaly needs to be resolved for an ad-
equate comparison of the resistance of different con-
crete mixes to acid attack.
The present study is aimed at evaluating the response
of different concrete mixes to sulfuric acid attack, us-
ing both physical and chemical indicators of the degree
of deterioration. An accelerated laboratory test pro-
gram was conducted. The program involved alternate
acid immersion and drying of test specimens, as well as
continuous acid immersion of other test specimens.
Changes in weight and thickness of the test specimens
were used to evaluate the physical degree of deteriora-
tion of the concrete, while the increase in sulfur con-
tent of the test specimens, as measured with a scanning
electron microscope (SEM) equipped with an energy-
dispersive x-ray analyzer, was used to evaluate the
chemical change in the concrete. Photomicrographs of
the test specimens were used to study the extent of the
acid attack.
RESEARCH SIGNIFICANCE
This research sheds new light on the mechanism of
sulfuric acid attack on concrete. The experimental
techniques used in the study provide information on the
extent of damage that occurs in concrete as a result of
acid attack. Photomicrographs show the nature of the
Received Oct. 14, 1987, and reviewed under Institute publication policies.
Copyright 198B, American Concrete Institute. All rights reserved, including
the making of copies unless permission is obtained from the copyright propri-
etors. Pertinent discussion will bc published in the September-October 1989 ACI
Materials Journal if received by June I, 1989.
481
ACI member Emmanuel K. Attiogbe is a research associate at the University
0/ Manitoba, Canada. He received the BSc degree from the University of Sci
ence and Technology in Ghana and his MS and PhD degrees in civil engineer-
ing from the University 0/ Kansas. His research interests include durability and
micromechanics of concrete.
A CI member Sami H. Rizkalla is an associate professor and the head of the
Structures Division of the Civil Engineering Department at the University of
Manitoba, Canada. He received the BSc degree/rom Alexandria University in
Egypt and MS and PhD degrees from North Carolina State University. He is a
member of ASCE, PCI, and CSCE. He is also a member of the Executive
Committee of the Structures Division Of the Canadian Society of Civil Engi-
neers. His research activilies are in the areas of reinforced and prestressed con-
crete structures.
Table 1 - Characteristics and composition of
the concrete mixes
Mix numher
Concrete components M1 M2 M3 M4
Coarse aggregate, lh/ yd
l
1600 1600 1600 1600
Fine aggregate, lb/yd! 1400 1400 1400 1400
Water, tb/ yd
l
310 310 310 310
Type 10 cement, lh/yd! 540 540 540 -
(ASTM Type I)
Type 50 cement, lh/ yd!
- -
- 540
(ASTM Type V)
Waterreducing agent, oz - 22
- -
Superplasticizer, oz 65 65 65 65
Air-entraining agent, oz
-
- 7 -
Slump, in. 7.0 8.0 7.0 6.5
Air content, percent - - 6.0 -
Average 28 day compressive 5640 4670 5220 6230
strength, psi
Note: 1 Ib/yd! _ 0.593 kg/ m!; I oz = 29.6 mI; I pSI - 6.895 kPa; I m. -
25.4 mm.
acid damage to the cement paste matrix of concrete.
The study shows that the increase in thickness (expan-
sion) and the increase in sulfur content of concrete
specimens are good indicators of the degree of deterio-
ration. The effect of wet-dry cycles of concrete expo-
sure to sulfuric acid was investigated.
EXPERIMENTAL PROGRAM
Materials and test procedures
The experimental program utilized four different
concrete mixes. Three of the mixes were made with
CSA Type 10 (ASTM Type J) portland cement, while
one mix was made with CSA Type 50 (ASTM Type V)
portland cement. A water-reducing agent was used in
one mix and an air-entraining agent was used in an-
other mix. A superplasticizer was used in all mixes to
increase slump. Siliceous fine aggregate and a blend of
calcareous and siliceous coarse aggregate were used.
The characteristics and composition of the different
concrete mixes are shown in Table 1.
Standard 152 x 305 mm (6 x 12 in.) concrete cylin-
ders were made from each concrete mix. The cylinders
were moist-cured for 28 days. At the end of the curing
period, two cylinders were used to determine the com-
pressive strength of each concrete mix. To induce
stresses similar to those that exist in concrete under ser-
vice load conditions, a third concrete cylinder for each
mix was loaded to 40 percent of the compressive
482
strength. From this cylinder, small test specimens were
cut and used in the acid attack test program.
Test specimens-The test specimens for each con-
crete mix were obtained as follows. A disk approxi -
mately 6 mm (\I.i in.) thick was cut near the midheight
of each cylinder. Each disk was then cut into small 12
mm (y, in.) wide specimens. The final prismatic test
specimens, approximately 100 mm (4 in.) long x 12 mm
(V, in.) wide and 6 mm (\I.i in.) thick, were used to
evaluate the sulfuric acid resistance of the concrete
mixes. A total of six test specimens were used for each
concrete mix: two for periodic physical measurements,
two for continuous immersion in the acid solution, and
the remaining two for microscopic examination. The
test specimens were oven-dried to a constant weight for
24 hr at about 100 C, and their weights and thicknesses
were measured prior to the start of the acid attack test
program. The weight was measured using a balance
with a 0.01 g sensitivity; the thickness was measured
with a pair of calipers. The average thickness for three
locations along the length was recorded for each speci-
men. These locations were marked with waterproof ink
for subsequent thickness measurements following peri-
odic immersion of the specimens in the acid solutions.
The marks were retouched after taking measurements.
Additional test specimens for Mix MI were covered
with two coatings and used in the acid attack test pro-
gram. These coatings were applied to the test speci-
mens with a brush and allowed to air dry for 2 days be-
fore being immersed in the acid. The first coating, a
methyl methacrylate sealant, provided a visible layer
about I mm (0.04 in) thick on the surface of the test
specimens, while the second coating, an acrylic sealer,
penetrated without any visible layer on the surface of
the test specimens.
Sulfuric acid solutions-The test specimens were im-
mersed in jars filled with 'equal quantities of I percent
sulfuric acid solutions (pH = I). This concentration of
sulfuric acid is representative of that found in sewers
that are in the process of deterioration.'7 The pH levels
of the acid solutions were monitored at an average in-
terval of 2 days with a portable pH meter. All solutions
were replaced with fresh I percent solutions if the pH
of any solution exceeded a value of 1.1. All solutions
were also replaced prior to reimmersion of the test
specimens after taking measurements. The periodic use
of fresh acid solutions along with the use of small test
specimens provided an accelerated evaluation of the
acid resistance of the concrete mixes . The test program
was concluded after about 10 weeks when some of the
test specimens completely disintegrated in the acid so-
lutions.
Test measurements- Measurements of the test speci-
mens were performed after selected periods of immer-
sion in the acid. The test specimens for weight and
thickness measurements, as well as those for SEM ex-
amination, were removed from the acid, immersed in
fresh water several times , and oven-dried to a constant
weight for 24 hr. Upon removal from the oven, the two
test specimens of each mix that had been marked for
ACI Materials Journal I November-December 1988
Fig. 1 - SEM specimen as mounted on stud
measurements were lightly brushed, weighed, and their
thicknesses measured.
Scanning electron microscope (SEM) examination-
For the SEM examination of each concrete mix, small
fractured specimens, approximately 6 mm (\4 in.) thick
x 12 mm (\I, in.) long x 6 mm (\4 in.) high, were pre-
pared from one of the test specimens that had been
marked for microscopic examination. A typical speci-
men, as mounted on a stud, is shown in Fig. I. To ob-
tain clear photomicrographs, the top surface of the
fractured specimens was rendered conductive with a
layer of aluminum about 0.02 p.m thick. No conductive
coating was used on specimens for the energy-disper-
sive x-ray analysis.
Photomicrographs were taken of both unattacked
and acid-attacked specimens. Quantitative elemental
analysis of the specimens was performed at a magnifi-
cation of 500X, using computer software available on
the energy-dispersive system. Changes in.sulfur content
were of primary interest in this analysis. The elemental
analysis was performed at three locations along the
length of each specimen. At each location, the analysis
was performed on the cement paste matrix within an
area approximately 0.25 mm' (0.01 in.'), and three ad-
jacent areas along the thickness of the specimen were
analyzed.
EXPERIMENTAL RESULTS AND DISCUSSION
Expansion of test specimens
Percentage change in thickness of the test specimens
versus immersion time in the acid solutions are shown
in Fig. 2 through 5. Immersion time as defined in these
figures does not include the time used in drying the
specimens. Each data point is an average of two speci-
mens of the same mix. The data for the individual
specimens varied less than 5 percent from the average
values. The increases in thickness indicate that the
specimens undergo volume expansion or swelling as a
consequence of the sulfuric acid attack.
ACI Materials Journal I November-December 1988
~ 15
b. Mix M I : No Woter - Reducing Admixture
g 0 Mix M2 : Includes a Woter - Reducing
. ~ Admixture
o
Q. 10
"
!:!:!
J
~ 5
.50
g-
o
~
u 0 10 20 30 40 50 60
Immersion Time, days
70 80
Fig. 2 - Expansion oj specimens jor concrete mixes
with and without a water-reducing admixture
~ 15

l:l. Mix M I : No Air Entrainment
g * Mix M3 : Air Entrained
'in
"
8. 10
o!l
.E
CI)
'"
6
~
u
10 20 30 40 50 60 70
Immersion Time J days
80
Fig. 3 - Expansion oj specimens jar concrete mixes
with and without air entrainment
of!. 15
" .2
:g
8. 10
o!l
.E
CI)
'"
"
o
~
u
to. Mix M I : Type 10 (ASTM Type I) Cement
+ Mix M4 : Type 50 (ASTM Type,,) Gement
20 30 40 50 60
Immersion Time J days
70 80
Fig. 4 - Expansion oj specimens jor concrete mixes
made with Types 10 and 50 (ASTM Types I and V)
portland cements
Fig. 2 shows percentage change in thickness versus
immersion time for the concrete mixes without and with
a water-reducing admixture, Mixes M 1 and M2, re-
spectively. The test specimens for Mix M2 expand more
483
~ 15
o
.S;;
g-
o
.t::
U
a
~ Uncoated Specimen
x Cooted Specimen - No Visible Protective Surface Loyer
<> Coated Specimen - With Protective Surface Loyer
10 20 30 40 50 60 70 80
Immersion Time days
Fig. 5 - Expansion of uncoated and coated specimens
than the specimens for Mix Ml. This larger expansion
of Mix M2 is probably due to its lower compressive
strength compared to Mix MI (Table I). The lower
compressive strength is thought to be due to an exces-
sive amount of air entrainment caused by the addition
of the water-reducing admixture.
In Fig. 3, concrete mixes with and without air en-
trainment are compared. The air-entrained concrete has
a slightly smaller expansion. Air entrainment improves
the chemical resistance of concrete insofar as it helps to
produce a more uniform, well-compacted, and, hence,
denser concrete.
s
Fig. 4 shows that beyond about 20 days' immersion
in the sulfuric acid solution, the test specimens for the
concrete made with sulfate-resistant Type 50 (ASTM
Type V) cement expand more than those for the con-
crete made with Type 10 (ASTM Type I) cement. Thus,
in the long run, sulfate-resistant cement does not ap-
pear to provide concrete a better resistance to sulfuric
acid attack than that provided by normal portland ce-
ment. This observation is consistent with the findings
of previous studies'' and is explained by the fact that
sulfate attack is only one aspect of sulfuric acid attack
on concrete. I
As shown in Fig. 5 for test specimens of Mix Ml,
coated specimens expand less than uncoated specimens.
Also, the coated specimens with a protective surface
layer expand less than those without a visible protective
surface layer. These results support the well-known fact
that measures aimed at making concrete more im-
permeable to ingestion of chemical solutions are the
most efficient in providing resistance to chemical at-
tack.
The preceding discussions show that expansion of the
small test specimens used in this study provides a con-
sistent measure of the deterioration of concrete due to
sulfuric acid exposure.
Wet-dry cycles versus continuous immersion- Fig. 6
compares the expansion of test specimens that were al-
ternately immersed in the acid solutions and dried to
that of specimens that were continuously immersed un-
til the end of the test program. Specimens that under-
went alternate immersion and drying showed a greater
484
;fl 15
co
.11
I
.,
co
&. 10
"
W
.,
.,
0>
co
1
-'"
~
u
5
~
.S;;
t:; Wet - dry Cycles of Acid Exposure
0>
'"
o Continuous Acid Immersion co
0
.t:: a
u
a 2 3 4
Mix Number
Fig. 6 - Expansion of specimens that were subjected to
wet-dry cycles of acid exposure compared to expansion
of specimens that were continuously immersed in the
acid solutions
expansion than those that were continuously immersed.
This suggests that wet-dry cycles of exposure to sul-
furic acid solutions increase the degree of concrete de-
terioration. The increased permeability of concrete due
to drying cracks would lead to a greater volume of ma-
terial being attacked by the sulfuric acid.
Weight loss of test specimens
Percentage changes in weight of the test specimens
are shown in Fig. 7 through 10. Fig. 7 shows that
weight loss is greater for the test specimens of Mix M2,
which contained a water-reducing admixture, than for
the specimens of Mix Ml. Weight loss of the test spec-
imens for the air-entrained concrete, Mix M3, is smaller
than that of the specimens for the non-air-entrained
concrete, Mix M2 (Fig. 8). As shown in Fig. 9, the
weight loss is slightly greater for the test specimens of
Mix M4 with Type 50 (ASTM Type V) cement than for
the specimens of Mix MI with Type 10 (ASTM Type I)
cenient. The effect of using surface coats is shown in
Fig. 10 by the smaller weight loss of the coated speci-
mens in comparison to the uncoated specimens. The
weight loss is smaller for coated specimens with a pro-
tective surface layer than for those without a visible
protective surface layer. These results (Fig. 7 through
10) are consistent with those discussed in the previous
section based on the expansion of the test specimens.
It should be noted that the test specimens initially
gain weight followed by weight loss. The weight gain
cannot be attributed to saturation of the specimens
since the specimens were oven-dried prior to weighing.
The reaction between sulfuric acid and the cement con-
stituent of concrete results in the conversion of calcium
hydroxide to calcium sulfate (gypsum) which, in turn,
may be converted to calcium sulfoaluminate (ettrin-
gite). Each of these reactions involves an increase in
volume of the reacting solids by a factor of about twO.1O
The formation of calcium sulfate leads to softening
(decrease in density) of the concrete. Since weight de-
pends on both volume and density, the initial weight
ACI Materials Journal I November-December 1988
5
.E Immersion Time I days
'" - 5

.!: -10
.,
g
.s=
U -IS
-20
6. Mix M I : No 'Noter -Reducing Admixture
o Mix M2 : Includes 0 Woter - Reducing Admixture
Fig. 7 - Weight loss oj specimens jar concrete mixes
with and without a water-reducing admixture
-
.s=
5
'" -5

.5
-10
o
.s=
10 20 50
Immersion Time ,doys
U -15 l:l. Mix M I : No Air Entrainment
* Mix M3 : Air Entrained
,-20
80
Fig. 8 - Weight loss oj specimens jar concrete mixes
with and without air entrainment
gain of the test specimens is probably due to the rela-
tive increase in volume being greater than the relative
decrease in density.
Both the increase in volume and the decrease in den-
sity of the concrete due to the sulfuric acid-cement
paste reaction would be larger the higher the acidity
(the lower the pH) of the acid solution. This implies
that a stronger acid solution could produce a smaller
weight loss in a concrete specimen than that produced
by a weaker solution due to a significant increase in the
volume of the concrete in comparison to the reduction
in the density. This phenomenon explains the results of
the study reported in Reference 6, in which the weight
loss of concrete cylinders exposed to sulfuric/nitric acid
solutions with pH of 2 was smaller than the weight loss
of identical specimens exposed to the acid solutions
with pH of 3. In the same study, the weight loss of the
concrete cylinders produced by other acid concentra-
tions increased with growing acidity, as expected.
It seems, therefore, that weight loss may not be a
consistent measure when comparing the effects of dif-
ferent sulfuric acid concentrations on concrete. The ex-
pansion of concrete specimens may be a more consis-
tent measure than the weight loss of the specimens since
the sulfuric acid-cement paste reaction entails a volume
increase. In this regard, the use of test specimens with
large surface area-to-volnme ratios (such as those used
ACt Materials Journal I NovemberDecember 1988
5
:s.
V -5

.5: -10
g-
o
o -15
-20
60 70
t:. Mix M I : Type 10 (ASTM Type I) Cement
+ Mix M4 : Type 50 (ASTM Type 1[) Cement
Fig. 9 - Weight loss oj specimens jar concrete mixes
made with Types 10 and 50 (ASTM Types 1 and V)
portland cements

0
-
.s=
'"
...
;t
.5
.,
i?
0
.s=
U
5
0
-5
-10
-15
-20
Immersion Time , days
l:::.. Uncoated Specimen
x Cooted Specimen - No Visible Protective Surface Layer
<> Coaled Specimen - With Protective Surface Layer
Fig. 10 - Weight loss oj uncoated and coated speci-
mens
in this study) may be preferred since the acid attack is
a surface phenomenon, as indicated by previous
and supported by results discussed in a subse-
quent section of this paper. Further investigation is
needed to evaluate specimen expansion and weight loss
as measures of concrete deterioration due to acids with
different strengths and for specimens with different
surface area-to-volume ratios. In such an investigation,
more than two test specimens would need to be used for
the expansion and weight loss measurements to clearly
establish the statistical reliability of the tests.
ScannIng electron microscope (SEM) analyses
Micrographs - Photomicrographs of unattacked
and acid-attacked specimens are presented in Fig. II
and Fig. 12 through 15, respectively. Fig. 12 through 14
are micrographs of specimens that were not coated
prior to immersion in the acid solutions, while Fig. 15
is a micrograph of a coated specimen. The changes in
microstructure, as discussed later, are typical for all the
concrete mixes studied. However, as discussed in ear-
lier sections, the extent of deterioration as a function of
acid immersion time varies among the concrete mixes
due to differences in mix composition, such as the type
of cement and the admixtures used.
485
Fig. 11 - Microstructure of an unattacked specimen
Fig. 12 - Microstructure of an acid-attacked specimen
immersed for 32 days: region near acid-exposed surface
of the specimen
Fig. 13 - Microstructure of an acid-attacked specimen
immersed for 32 days: middle region of the specimen
A typical micrograph of the cement paste matrix on
a fractured surface of matured concrete is shown in
Fig. II. The predominant structure appears to be a
modified Type III calcium silicate hydrate (CSH) , one
of the prime hydration products of portland ce-
ment."!' A calcium hydroxide crystal is seen in the
lower right corner of the micrograph. Fig. 12 is a mi-
crograph of a specimen that had been immersed in the
acid for 32 days. A comparison of Fig. 12 to Fig. 11
486
Fig. 14 - Microstructure of an acid-attacked specimen
immersed for 71 days
Fig. 15 - Microstructure of a coated, acid-attacked
specimen immersed for 49 days; the coating is the dark
vertical band on the left
shows that the acid reacts with the cement paste matrix
reducing the matrix to a more porous material consist-
ing of smaller particles.
Fig. 12 shows a region near the acid-exposed surface
of the specimen, while Fig. 13 shows a region farther
from the acid-exposed surface of the same specimen.
These micrographs show that the deterioration is
smaller in the region farther from the acid-exposed sur-
face of the specimen than in the region near the acid-
exposed surface. This result supports previous studies
and clearly indicates that the acid attack is a surface
phenomenon: the deterioration starts at the surface of
the concrete and progresses inward.
A typical micrograph of a specimen that had been
immersed in the acid for 71 days is presented in Fig. 14.
This micrograph is representative of the total surface of
the specimen at this stage of the acid attack. The
spherical fibrous structures that are visible in several
locations on this micrograph are Type I CSH. These
structures are usually not visible on the fractured sur-
faces of matured cement paste and concrete. The frac-
ture path usually passes through Type III CSH (see Fig.
11 and References 13 through 15). Comparing Fig. 14
to Fig. 11 and 12, it appears that after 71 days immer-
sion, the acid has dissolved the Type III CSH to a large
extent exposing the Type I structures.
ACI Materials Journal I NovemberDecember 1988
Mg - Magnesium
At - Aluminum
Si - Silicon -
Ca
S - Sulphur
K - Potassium
-
Co - Calcium
-
Fe - Iron
- -
Si
-
I- Al -
-
S
A
Fig. 16 - Elemental energy-dispersive spectrum for an
unattacked specimen
A micrograph of a specimen that was coated before
being immersed in the acid for 49 days is presented in
Fig. 15. The protective surface layer of the coating is
shown as a dark vertical band on the left side of the
micrograph. A comparison of this micrograph to Fig.
12 shows that the extent of deterioration is less in the
coated specimen, even though it was immersed in the
acid for a longer duration.
Elemental analysis - Typical energy-dispersive spec-
tra obtained from the quantitative elemental analysis
are shown in Fig. 16 and 17 for unattacked and acid-
attacked specimens, respectively. These spectra were
obtained within three adjacent regions, each approxi-
mately 0.25 mm (0.01 in.) wide, along the thickness of
the specimens. Since sulfur compounds are formed as a
result of the reaction between sulfuric acid and cement
paste, sulfur components of the spectra are of primary
interest in these figures and are highlighted by the dark
shadings. The figures show that the sulfur content of
the acid-attacked specimen is higher than that of the
unattacked specimen.
Sulfur concentrations obtained from the elemental
spectra, such as those in Fig. 16 and 17, are shown in
Fig. 18 versus specimen immersion time in the acid so-
lution for the three adjacent regions analyzed along the
depth of acid penetration. Each data point is an aver-
age of the sulfur concentrations for three locations
along the length of the SEM specimen. Fig. 18 shows
that the sulfur concentration increases with immersion
time, the increase being larger in the region near the
acid-exposed surface during the initial period of im-
mersion than in the regions farther from the acid-ex-
posed surface. As the sulfur concentration in the re-
gion near the acid-exposed surface approaches an opti-
mum value, the concentration in the adjacent region
increases rapidly, while the concentration in the region
farthest from the acid-exposed surface increases less
ACI Materials Journal I November-December 1988
Mg - Magnesium
AI - Aluminum
l- Si - Silicon
-
S - Sulphur
Ca
-
K - Potassium
Co - Calcium
-
Fe - Iron
-
S
-
- -
Si
-
-
ifi
Ca
Mg '-I
~ ~ -
-
Fig. 17 - Elemental energy-dispersive spectrum for an
acid-attacked specimen
40
--0.00- 0.25 mm from Acid - Exposed Surface
----0.25-0.5Omm from Acid-Exposed Surface
;,l!
0 35
-"- 0.50-0.75 mm from Acid -Exposed Su1 e
"i
30
c
0
25
~
<:
20
~
0
u
~
"
or
.9-
"
C/)
////
/
d""'- ,.....0
,,' "" .. ..".. ..
.," ....... ~ .
.,. ... -
,," ..,.. ..
" , .,-"
_C1'-"-"-'-'-<Y'
10 20 30 40 SO 60 70 80
Immersion Time , days
Fig. 18 - Sulfur content of specimens versus immer-
sion time in the acid solution
rapidly. This result suggests that the concrete deterio-
ration starts at the surface and progresses inward, as
revealed by the photomicrographs.
Fig. 19 shows sulfur concentrations versus distance
from the acid-exposed surface of uncoated and coated
specimens. As in Fig. 18, each ordinate is an average of
the sulfur concentrations for three locations along the
length of the SEM specimen. For both the uncoated
and coated specimens , the sulfur concentration is
higher in the region near the acid-exposed surface than
in the regions farther from the acid-exposed surface.
The coated specimen has a smaller sulfur concentration
within each region than the uncoated specimen. The
smaller sulfur concentrations for the coated specimen
correspond to smaller expansion (Fig. 5) and weight
loss (Fig. 10) when compared to uncoated specimens.
To accurately estimate the minimum thickness of
concrete cover in reinforced concrete members exposed
487
40
~ 35r-__________ -,
~
_ 30-
. 25
e
C 20-
~
8 15
S 10-
--- Uncoated Specimen
.... - Coated Specimen
15.. ..................... _.M. __
Jl 5 ............ --........... """ __ ~ ~ _ ' "
O ~ ______ ~ ~ ______ ~ ~ ____ ~ ~ l
0.00 0.25 0.50 0.75
Distance from Acid - Exposed Surface, mm
Fig. 19 - Sulfur content of uncoated and coated spec-
imens versus distance from acid-exposed surface
to sulfuric acid, the rate of deterioration along the
depth of penetration of acid should be known. The
minimum cover thickness was estimated in Reference 5
by assuming the degree of deterioration to be linearly
related to the depth of penetration of acid, with full
sulfuric acid-cement reaction at the exposed surface and
zero reaction at the maximum depth of acid penetra-
tion. This relationship could be more accurately repre-
sented by the variation in sulfur concentration with the
depth of penetration of acid. Further investigation is
required to explore this issue.
CONCLUSIONS
The following conclusions are drawn from the test
results and analysis presented in this paper.
I . The increase in sulfur content of test specimens, as
measured with an SEM equipped with an energy-dis-
persive x-ray analyzer, is a good indicator of the extent
of damage in concrete due to exposure to sulfuric acid.
2. Photomicrographs, as well as the variation in sul-
fur concentration with distance from the acid-exposed
surface of test specimens, clearly show that deteriora-
tion of concrete due to sulfuric acid attack starts at the
surface and progresses inwards.
3. The increase in thickness (expansion) of small
specimens (with large surface area-to-volume ratios)
may be a more consistent measure than the weight loss
of larger specimens when comparing the effects of dif-
ferent sulfuric acid concentrations on concrete.
4. Wet-dry cycles of exposure to sulfuric acid in-
crease the degree of concrete deterioration.
5. The relationship between degree of concrete dete-
rioration and depth of penetration of sulfuric acid
488
could be represented by the variation in sulfur concen-
tration with the depth of acid penetration.
ACKNOWLEDGMENTS
This research was performed at the Structural Engineering and
Materials Laboratory of the University of Manitoba. Electron micro-
scope analysis was carried out on the JEOL JXA-840 Scanning Elec-
tron Microscope of the Department of Mechanical Engineering, Uni-
versity of Manitoba. The authors are grateful for the technical and
financial assistance provided by Barkman Concrete Ltd., Steinbach,
Manitoba.
REFERENCES
1. Mindess, Sydney, and Young, J. Francis, Concrete, Prentice-
Hall, Englewood Cliffs, 1981, pp. 548-555.
2. Meyer, Alvin H., and Ledbetter, William B., "Sulfuric Acid
Attack on Concrete Sewer Pipe," Proceedings, ASCE, V. 96, SA5,
Oct. 1970. pp. 1167-1182.
3. Hughes, B. P., and Guest, J. E., "Limestone and Siliceous Ag-
gregate Concretes Subjected to Sulphuric Acid Attack," Magazine of
Concrete Research (London), V. 30, No. 102, Mar. 1978, pp. 11-18.
4. Fattuhi, N. I., and Hughes, B. P., "Effect of Acid Attack on
Concrete with Different Admixtures or Protective Coatings," Ce-
ment and Concrete Research, V. 13, No.5, Sept. 1983, pp. 655-665.
5. Raju, P. S. N., and Dayaratnam, P., "Durability of Concrete
Exposed to Dilute Sulfuric Acid," Building and Environment, V. 19,
No.2, 1984, pp. 75-79.
6. Kong, Hendrick, L., and Orbison, James G., "Concrete Dete-
rioration Due to Acid Precipitation," ACI Materials Journal, V. 84,
No.2, Mar.-Apr. 1987, pp. 110-116.
7. Sand, W., and Bock, E., "Concrete Corrosion in the Hamburg
Sewer System," Science and Technology Letters, V. 5,1984, pp. 517-
528.
8. Wilder, Carl R., and Spears, Ralph E., "Concrete for Sanitary
Engineering Structures," Concrete International: Design & Con-
struction, V. 3, No.4, Apr. 1981, pp. 29-34.
9. Wenger, E. C., "Concrete for Sewage Work," ACI JOURNAL,
Proceedings V. 54, No.9, Mar. 1958. pp. 733-738.
10. Soroka, Itzhak, Portland Cem'ent Paste and Concrete, Mac-
millan Press, London, 1979, pp. 151-152.
11. Berger, R. L.; Young, J. F.; and Lawrence, F. V., Discussion
of "Morphology and Surface Properties of Hydrated Tricalcium Sil-
icate Paste" by M. Collepardi and B. Marchess. Cement and Con-
crete Research, V. 2, No.5, Sept. 1972, pp. 633-636.
12. Diamond, Sidney, "Cement Paste Microstructure-An Over-
view at Several Levels," Proceedings, Conference on Hydraulic Ce-
ment Pastes; Their Structure and Properties (Sheffield, Apr. 1976),
Cement and Concrete Association, Wexham Springs, 1976, pp. 2-30.
13. Lawrence, F. V., Jr., and Young, J. F., "Studies on the Hy-
dration of Tricalcium Silicate Pastes: Scanning Electron Microscopic
Examination of Microstructural Features," Cement and Concrete
Research, V. 3, No.2, Mar. 1973, pp. 149-161.
14. Attiogbe, Emmanuel K. , and Darwin, David, "Submicro-
cracking in Cement Paste and Mortar in Compression," 8M Report
No. 16, University of Kansas Center for Research, Lawrence, Nov.
1985,439 pp.
15. Attiogbe, Emmanuel K., and Darwin, David, "Submicro-
cracking in Cement Paste and Mortar," ACI Materials Journal, V.
84, No.6. Nov.-Dec. 1987, pp. 491-500.
ACI Materials Journal I NovemberDecember 1988