Construction and Building Materials, Vol. 10, No. 2, pp.

123-129, 1996
Copyright 0 1996 Elsevier Science Ltd
Printed in Great Britain. All rights reserved
0950-0618/96 $15.00+0.00
ELSEVIER 0950-0618(95)00077-1

Effect of reinforcement corrosion on bond

strength

Abdullah A. Almusallam*, Ahmad S. Al-Gahtani, Abdur Rauf Aziz and

Rasheeduzzafart

King Fahd University of Petroleum & Minerals, Department of Civil Engineering,

Dhahran 3 1261, Saudi Arabia

Received 12 April 1995; revised 27 June 7995; accepted 6 October 1995

The effect of reinforcement corrosion on the bond strength between steel and concrete was inves-
tigated. The bond behaviour of reinforced concrete elements, including the ultimate bond strength,
free-end slip, and the modes of failure in precracking, cracking and postcracking stages was
studied. Also, the effect of different crack widths and the rib profile degradation for various degrees
of corrosion on the bond strength were evaluated. In order to establish different levels of corrosion,
a calibration curve establishing a relationship between the duration of the impressed current and
the corresponding degree of corrosion was prepared. The magnitude of corrosion was measured
as gravimetric loss in weight of the reinforcing bars. The results indicate that in the precracking
stage (O-4% corrosion) the ultimate bond strength increases, whereas the slip at the ultimate bond
strength decreases with an increase in the degree of corrosion. The degradation of bond results
from the crushing of concrete keys near the bar lugs. When reinforcement corrosion is in the range
of 4 to 6%, the bond failure occurs suddenly at a very low free-end slip. At this level of reinforce-
ment corrosion, a large slip was noted as the ultimate failure of the bond occurred due to the split-
ting of the specimens. Beyond 6% rebar corrosion, the bond failure resulted from a continuous
slippage of the rebars. The ultimate bond strength initially increased with an increase in the degree
of corrosion, until it attained a maximum value of 4% rebar corrosion after which there was a sharp
reduction in the ultimate bond strength up to 6% rebar corrosion. Beyond the 6% rebar corrosion
level the ultimate bond strength did not vary much even up to 80% corrosion. In terms of the effect
of rib profile, a sharp reduction in the bond strength was initiated when its degradation exceeded
25%. This decrease in bond strength continued up to 45%. Thereafter, there was no significant
effect of the rib profile degradation on the bond strength.

Keywords: bond (concrete to reinforcement); corrosion cracking (fracturing); slippage

The Arabian Gulf seaboard constitutes one of the most
aggressive environments for concrete durability in the
world. Concrete construction in the Gulf states is show-
ing an alarming degree of deterioration within a short
span of 10 to 15 years. The low durability performance
of concrete is due to several interactive factors. These
are characterized mainly by adverse climatic and geo-
morphic conditions in conjunction with inadequate
specifications and construction practices.
Research conductedI at King Fahd University of
Petroleum and Minerals, Dhahran, Saudi Arabia,
strongly suggests that reinforcement corrosion, with
cracking and spalling of concrete, is the main durability
problem of concrete and by far outweighs other forms
of deterioration. This is followed by deterioration due
to sulphate attack, salt weathering, and cracking result-
ing from other environmental effects. However, it
should be noted that corrosion of rebar is not only a
*Correspondence to Dr Abdullah Almwallam
t Deceased
predominant problem in this region, but of considerable
global concern. The deterioration of bridge-decks in the
United States, parking garages in cold countries, coastal
structures and many other structures, due to the use of
chloride-based accelerating agents, constitute probably
the largest group of reinforced concrete structures
undergoing deterioration due to a single causal factor,
namely reinforcement corrosion.
Although considerable research has been conducted
on the bond behaviour of rusted bars, little work has
been done on the bond behaviour in reinforced concrete
with significant damage due to reinforcement corrosion.
The objective of this study, therefore, was to assess the
effect of different degrees of reinforcement corrosion on
the bond degradation. Progressive degradation of bond
between concrete and reinforcing steel brings about a
reduction in the flexural strength of beams and slabs.
Data are required to understand the relationship
between the degree of reinforcement corrosion and the
bond strength. These data will provide an essential
input concerning the progressive loss of strength of

123
124 Effect of reinforcement corrosion: A. A. Almusallam
beams and slabs due to corrosion of reinforcement,
which is significant to develop models for predicting the
useful service-life of structures serving in aggressive
media.
Experimental programme
The experimental programme was designed to evaluate
the bond strength, including the ultimate bond strength,
free-end slip, and modes of failure of the test specimen,
in precracking, cracking, and postcracking stages, as
well as the effect of different crack widths and degrada-
tion of rib profile for various degrees of corrosion.
Materials and mix proportions
The concrete mixes were designed for a 28-day com-
pressive strength of 30 MPa. ASTM C 150 Type V
Portland cement, beach sand, and crushed limestone
were used to prepare the concrete mixtures. The fine
aggregate had a specific gravity of 2.64 and water
absorption of 0.56%. The coarse aggregate had a spe-
cific gravity of 2.64, water absorption of 2.3%, and the
maximum size was 19 mm. A coarse aggregate to fine
aggregate ratio of 2 and a water-cement ratio of 0.45
(by weight) were used.
Tensile reinforcement in the cantilever bond tests
consisted of 12 mm $J deformed bars meeting the
requirements of ASTM A615 for Grade 60 steel.
Test specimen
The selection of a proper type of specimen for bond
strength evaluation is of great importance as it signifi-
cantly influences the bond characteristics. The selection
of a proper specimen is still a matter of controversy and
until now no test specimen has been devised which fully
represents the actual bond behaviour.
The concentric pullout test (ASTM C234) is the most
widely used by researchers because of its simplicity. In
this test, a bar is embedded in a rectangular concrete
block and the force required to pull out the bar or to
make the bar slip excessively, is measured. However,
this test is used to compare bond strengths of different
concretes. In reinforced concrete beams or slabs, the
concrete surrounding the tensile reinforcement is in ten-
sion, whereas the concrete in this test is in compression,
which not only eliminates tension cracks in the speci-
men but also increases the bond strength. Moreover, in
this type of test, a small load causes a slip and the peak
bond stress moves in front of the slip6. Also some inves-
tigators have reported that in this concentric pullout
test, the bond stress is not influenced by the concrete
strength. In this type of set-up the specimen is not sub-
jected to external shear or bending moments, which are
present in the actual structures. For these reasons, the
failure patterns such as splitting or slipping in the pull-
out specimens are not realistic.
The tension pullout test6 is an improved form of the
concentric pullout test, in which compression on the
concrete specimens is eliminated. However, this method
et al.
introduces problems of spaced splices and crack pattern
is influenced by this interaction6.
The bond beam test6 is conducted on beams which
are significantly larger than those used in the pullout
tests. The beam tests truly represent the bond stress
conditions encountered in the actual flexural members.
Therefore, the results of this type of test are considered
more reliable. However, the major problem in the bond
beam test is the reaction restraint, which might increase
the splitting resistance because of the confining of the
concrete over the bar at the supports. In order to over-
come this problem, a number of modifications have
been made from time to time. Some of the modified
beam tests are the Bureau of Standards beam test6, the
University of Texas beam test6, and the West Virginia
test. Another problem associated with these beams is
that they are not only difficult to handle, but are costly
as well.
The cantilever bond test specimen was developed by
Kemp et ~1.~to overcome some of the objections raised
in connection with the ASTM C234 pullout test. Some
of the advantages of this bond test are:
(1) This test produces a bond stress situation similar
to that existing in the actual flexural members.
(2) Concrete and steel experience similar tensile
strains, and strain gradients produced in this spec-
imen are similar to those occurring in actual struc-
tures.
(3) Stresses generated in the cantilever bond test are
more representative of actual bond stresses in flex-
ural members because of the presence of both
external shear and bending moments in the test
specimen.
The relationship between bond, shear, and bending
(4)
moments can be varied easily to produce different
types of failures.
(5) This test specimen is smaller and cheaper than
beam test specimens.
(6) One or several bars with or without shear rein-
forcement can be used.
(7) The bar may be prevented from any of the confin-
ing action at the supports by the use of a shield or
soft covering.
A review of the merits and demerits of different bond
test specimens indicated that the cantilever bond test
has several advantages over the other test specimens.
This has prompted the investigators to select the can-
tilever bond test.
The size of the bond test specimen was selected in
order to fulfil the requirements of research with the
smallest size of the specimen. After making several
trials, a specimen of 152 X 254 X 279 mm with a 12 mm
@bar, was selected (Figure I). In order to avoid possi-
ble shear failure, inverted open stirrups were used in
such a way that they did not affect the bond resistance
of the specimen and did not interfere with the impressed
current. Further, in order to avoid compression failure
of concrete, compression reinforcement was also pro-
 Effect of reinforcement
pc t_ 89 _I:
Sleeves
Pull out -__T_ _ _
force - ---I=.
LF
Reaction
Nate: All dimensions
Figure 1 Design details of cantilever bond test specimen
vided. In order to avoid corrosion of shear and compres-
sion steel, the bars were isolated from the bond test bar.
The length of the tension bar was kept to 610 mm to
facilitate loading of the specimen in an Instron testing
machine and measurement of the free-end slip. The
length of the compression reinforcement was 254 mm.
Open legged, U-shaped stirrups were selected with a
spacing of 76 mm as shown in Figure 1.
The stirrups were tied with the compression rein-
forcement in the form of a cage. This assembly (com-
pression and shear reinforcement) was painted with
epoxy to avoid corrosion because of stray currents
during the corrosion process.
The weight, length, and diameters at different loca-
tions of the main reinforcement were measured and
recorded. An electric wire was soldered to the main
reinforcement for impressing electric current.
The embedment length (102 mm) was selected to
avoid yielding of the reinforcement under pullout force.
With this length, the bar tensile stresses at the point of
bar breakdown would be well below the yield strength
of the bar. In order to debond the remaining 178 mm
length of the reinforcement, 89 mm long PVC conduits,
slightly bigger in diameter than the main reinforcing
bar, were used on both ends of the specimen. Both ends
of the conduits were then sealed with silicon to prevent
the ingress of the concrete mortar inside the conduits.
This arrangement ensured a bar embedment length of
102 mm in contact with the concrete. The purpose of
using PVC conduits was to protect the reinforcement from
the confining pressure of concrete at the supports. Figure
I is a schematic representation of the test specimen.
Mixing, casting and curing
The concrete mixture constituents were mixed in an
electrically operated mixer. The cantilever bond test
corrosion: A. A. Almusallam et al. 125
102 _/_ ~pl,
4 #3 Bars JSI
Compression Reinforcement
Reinforcement @ %nm. c/c.
are in mm. unless othervise noted.
specimens were cast in 152 X 254 X 219 wooden
moulds. A groove with a depth of approximately 13
mm and a diameter slightly more than that of the bar
was made at the bottom of the mould so that some
length of the main reinforcement could protrude from
the specimen to be used for measuring the free-end slip.
The moulds with the reinforcements placed in their
proper positions were then placed on the vibrator, and
concrete was poured in three layers. Concrete cylinders,
75 mm $Jand 150 mm height were also cast to measure
the compressive strength.
Test techniques
To accelerate reinforcement corrosion, direct current
was impressed on the bar embedded in the pullout spec-
imen using an integrated system incorporating a small
rectifier with a built-in ammeter to monitor the current,
and a potentiometer to control the current intensity.
The specimens were partially immersed in water in a
glass fibre tank in such a manner that the reinforcement
was totally above the water resulting in the formation
of products due to reinforcement corrosion. The direc-
tion of the current was adjusted so that the reinforcing
steel served as the anode while a stainless steel plate
counter electrode was positioned in the tank to act as a
cathode. A schematic representation of the test set-up is
shown in Figure 2. A constant current of 0.4 A was
applied to the steel bar. This high current was impressed
to accelerate the reinforcement corrosion process and
meet the objectives of this study within a reasonable
time frame. In order to establish different levels of rein-
forcement corrosion, calibration curves establishing a
relationship between the duration of the impressed cur-
rent and the corresponding degree of corrosion were
prepared. The degree of corrosion was measured as
gravimetric loss in weight of the reinforcing bars. The
126 Effect of reinforcement corrosion: A. A. Almusallam et al.

machine was connected to a data logger and the

load and slip readings were recorded at preset intervals.
The Power Supply After conducting the bond test, the specimens were
broken and the weight loss due to corrosion of the
unsleeved portion of bar was determined by cleaning it
with Clarks solution (ASTM Gl-76). Similarly, the
degradation in the rib profile was determined by mea-
t
t suring the rib height before applying the impressed cur-
StaInless steel
Steel Bar Plate ICathode rent and after cleaning it with Clarks solution.
[Anode] /

Results and discussion

Ejjkct ojcorrosion on the slip and ultimate bond
strength
The relationship between the ultimate bond strength
Figure 2 Schematic representation of impressed current test setup
and the degree of corrosion for the cantilever bond test
is depicted in Figure 4. Figure 5 shows typical load
versus free-end slip relationships for different degrees of
current supplied to each specimen was checked on a
corrosion in the precracking stage (OWl/o). From the
regular basis and any drift was corrected by the adjust-
two figures it can be seen that, as the degree of corro-
ment of the potentiometer. The appearance of the first
sion increases from 0 to 4X, the ultimate load increases
crack was monitored carefully as this formed a refer-
from 6 1 to 7 1 kN, whereas the corresponding slip at the
ence time for selecting precracking and postcracking
ultimate load decreases from 0.68 to 0.238 mm. This
levels of corrosion.
can be attributed to an increase in the reactionary con-
The cantilever bond tests were conducted using a
finement of the bar in the concrete, as the amount of
specially designed loading frame which was fixed to
corrosion product increases and develops an expansive
the base of an Instron universal testing machine of
mechanical pressure on the surrounding concrete. It has
250 kN capacity under monotonic static load (Figure 3).
also been observed by some researchers- that for a
In order to measure the slip, one LVDT was used on the
small degree of corrosion of the reinforcing bar, signifi-
loaded end and the other on the free end of the
cant mechanical pressure is exerted on the surrounding
specimen. The output from the LVDTS and the testing
concrete before cracking occurs. This pressure is caused
by the development of expansive corrosion products
which increase the reactionary confinement and the
mechanical interlocking of concrete around the bar.
/
Secondly, in the initial stages of corrosion, the rough-
ness of the bar is also increased, thereby enhancing the
MS TRON
friction between the bar and the surrounding con-
crete . Both these effects increase the bond resistance
and decrease the slip of the bar. None of these condi-
tions is obtainable in the uncorroded bars. Also, from
Figure 5, it can be seen that the load-slip curve for the
_ 12mm.0 specimen with uncorroded bar shows a linear relation-
bar.

PVC
Conduit BO

--Specimr

Loading
Frame

- To oat;
= Logger.

ot
/
Figure 4 Relationship between the ultimate bond strength and dif-
Figure 3 Cantilever bond test loading arrangement ferent degrees of corrosion
 Effect of reinforcement corrosion: A. A. Almusallam et al. 127

men, the ribs of the bars were still effectively bearing

against the concrete keys. When pullout load is applied
to the specimen, the ribs bearing action against the con-
60 crete causes the horizontal bearing stresses as well as
hoop stresses. Compared to the reference specimen,
these specimens are weaker in tension due to the pres-
ence of the corrosion cracks. Therefore, the hoop tensile
stresses further widen the corrosion cracks thereby
reducing the confinement and facilitating the bar slip
which eventually causes the bond failure of the speci-
mens. Bond failure in this range occurred due to the
l- splitting of the test specimen along the corrosion cracks.
I
The reason for this type of failure is that the concrete is
I I t I 1 I I I I

.oo .50 1.00 1.50 2.00 2.50 3.00 3.50 4.00
Slip Imm)
Figure 5 Relationship between load and slip in precracking
ship between load and slip up to almost two-thirds (40
kN) of the ultimate load. Thereafter, the load-slip curve
exhibits a significantly increased slip with a relatively
smaller gain in the bond resistance. This curve, for spec-
imens with corroded bars, shows linear behaviour up to
almost 60% of the ultimate load for bars with 3.6% cor-
rosion and right up to the ultimate load for bars with
4% corrosion. Thereafter, the curve breaks off, indicat-
ing a rapid increase in slip without an increase in the
load. The maximum bond strength was obtained at 4%
corrosion which is almost 13% higher than the bond
strength of the specimen with uncorroded bar. In this
precracking stage, the bond failure of the specimen
resulted from the crushing of concrete keys adjacent to
the lugs without any splitting of the specimen.
The relationship between load and free-end slip for
specimens in the cracking stage (4-6%) is shown in
Figure 6. The specimens with a relatively low level of
corrosion, in the range of 4% to 5.09%, exhibited a
linear relationship between load and slip up to almost
ultimate bond strength; thereafter, a sudden drop
occurs which continues to show increasing slip at
decreasing loads. The reason for the sudden drop in
load-slip curve at ultimate load can be attributed to the
sudden splitting of the concrete cover over the bar.
Although corrosion had caused fine cracks in the speci-
k.50 5.0 0 already weak due to the presence of cracks due to rein-
forcement corrosion. When the pullout load is applied
stage to the specimen, the splitting failure occurs due to the
radial hoop component of the bearing pressure devel-
oped by the lugs on the surrounding concrete. This
hoop stress is operative in the tension zone of concrete
which further facilitates and somewhat advances the
splitting tension failure of the specimen.
The load-slip curves in the specimens with post-
cracking corrosion (8% or more) are shown in Figure 7.
Considerable slip occurred at ultimate load stage,
before failure was observed. Also, the value of the ulti-
mate bond strength remains more or less the same,
ranging from 13.17 kN to 8.46 kN for corrosion rang-
ing from 11.72% to 80%, respectively (Figure 4). Similar
findings were reported by Cabrera et al.. The reason
for this behaviour can be explained as follows:
(1) With an increased level of reinforcement corrosion,
a significant degradation of the lugs occurs due to
which the deformed steel bar starts acting more
and more like a plain bar. The horizontal compo-
nent of the bearing force produced as a result of
mechanical interlocking of the ribs is therefore sig-
nificantly reduced. The pullout action is therefore
facilitated and occurs at a lower load.
(2) Accumulation of a heavy layer of the flaky cor-
roded material around the bar acts as a lubricant
and reduces the friction component of the bond
strength.

80-

70
-I-
i
60. 60

1 I I I I I I I I -i
.dO .50 1.00 1.50 2.00 2.50 3.00 3.50 5.00 4.50 5.00
00 .50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 5.50 5.00
SLip lmml Slip Imml
Figure 6 Relationship between load and slip in the cracking stage Figure 7 Relationship between load and slip in the post-cracking stage
 128 Effect of reinforcement corrosion: A. A. Almusallam et al.

(3) The third important factor in the post-cracking a0 ,

7
stage is the width of the crack corresponding to a
particular level of corrosion. The formation of
cracks around the reinforcing bar reduces its con-
finement in the concrete to the extent that practi-
cally any cumulative negative effect of the above
mentioned two parameters (1) and (2) does not
affect the ultimate bond strength, which remains at
a low average value of about 10 kN.

Due to these factors, after 7.8% of corrosion, there is

not much reduction in the bond strength and it remains
almost constant even up to 80% of corrosion, as is clear
from Figure 4.
00 00 SO 1.00 1.50 2.00 2.50 3.00 3.50 1.00 4.50
Crack width [mm)

Effect of crack width on the bond strength
Figure 8 presents
the relationship between the crack
width and the degree of corrosion, whereas the effect of
crack width on the bond strength is depicted in Figure
9. From this figure, it is clear that the ultimate bond
load is reduced abruptly from 65 kN to 40 kN at the
initial stages of the formation of cracks and the associ-
ated resulting reduction in the confinement. After a cer-
tain degree of corrosion, the effect of crack width on the
bond strength decreases, and beyond a certain level, it
does not affect the bond strength significantly. The
reason for this is that, because of sufficiently wide
cracks, the reinforcing bar has already lost much of its
confinement, and a further increase in crack width does
not cause any further reduction in the confining force.
The interactive effect of reduction in the lug size and
confinement with increasing corrosion and cracking
width therefore facilitates the pullout action to such an
extent that the bar is easily pulled out. Any further
decrease in the lug height and widening of the crack
does not affect the bond strength to any great extent.
The results of this part of the study indicate that crack
width plays a critical role in the performance of the
structure with respect to bond strength. The critical
crack width was noted to be 0.3 mm for a c/d value
of 5.
Figure 9 Effect of corrosion crack width on ultimate bond strength
Effect of rib profile degradation on the ultimate bond
strength
Figure 10 shows the relationship between the loss in the
rib profile and the ultimate bond strength. Initially an
increase in the bond strength was noted with a loss of
rib profile, up to 26%. However, sharp reduction in the
ultimate bond strength for rib degradation in the range
of 26 to 43% was recorded. After 43% loss in the rib
profile, the bond strength does not vary significantly
right up to 100% rib degradation. The cause for such a
change in the ultimate bond strength with rib profile
degradation can be explained as follows:
(1) Initially, in the precracking stage, the increase in
the bond strength, in spite of a 26% loss in the rib
profile, can be attributed to the increase both in
the reactionary confinement and the roughness of
the bar in contact with concrete. The increase in
bond strength, as a result of both these factors,
overshadows the reduction in the bond strength
due to rib degradation.
(2) A sharp reduction in the ultimate bond strength
for a reduction in the rib profile from 26 to 43%
may be attributed to the formation of small corro-
sion cracks around the reinforcing bars and a sig-
nificant reduction in the rib profile with resulting

270 -
Y
r604
Fz 50 -
IA

p 40-
0

D30-
5
2
Z 20 -
i
10 -

Oi I I I I , I I I I
0 10 20 30 50 SO 60 70 80 90
Oegree of corrosion
Figure 8 Relationship between the degree of corrosion and crack
Loss in ribs profile, percent
width Figure 10 Effect of loss of rib profile on ultimate bond strength
 Effect of reinforcement corrosion: A. A. Almusallam
reduction in the interlocking action between ribs
and concrete.
(3) The insignificant change in the bond strength due
to the reduction in the rib profile in the range of 43
to 100% indicates that the ribs are degraded to the
extent that their interlocking action with the con-
crete becomes negligible. The confinement of the
bar with the concrete is therefore significantly
reduced. Thus, beyond this level of corrosion, fur-
ther degradation of the ribs does not affect the
bond strength to any noticeable extent.
However, the individual effects of all the above-men-
tioned parameters on the bond strength can be isolated
from the data developed in this study.
The authors would also like to point out that in this
study the tension bars were isolated from the shear rein-
forcement, to evaluate the effect of steel corrosion on
the bond strength. However, in the real structures, the
tension bars are normally connected to the shear rein-
forcement. Therefore, the effect of rebar corrosion in
such situation needs to be evaluated.
Conclusions
(1) The ultimate bond strength increases by about 17%
and rebar slip decreases in the early stages of rein-
forcement corrosion. This behaviour is noticeable
up to 4% corrosion and is ascribable to an increased
rebar roughness and its confinement by concrete.
(2) With the appearance of the first corrosion crack at
about 5% gravimetric weight loss (c/d=5) the bond
strength decreases gradually for an additional 1%
corrosion, and thereafter decreases rapidly. At
additional 2% corrosion, after cracking, the bond
strength drops to a residual value of about 35%.
Thereafter, corrosion up to 80% causes only 20%
loss in bond strength.
(3) Corrosion cracking is of critical importance from
the standpoint of bond strength. Very little corro-
sion after cracking is required to reduce bond
strength to an unacceptable level.
(4) The bond failure mechanism indicates that imme-
diately after corrosion cracking, high values
(103%) of bond resistance are still obtainable
because the interstitial corrosion products are neg-
ligible and noticeable rib profile degradation has
not occurred. Hence, the specimen fails at these
high ultimate bond values because of the splitting
of concrete caused by the hoop tensions generated
by the rib action in concrete.
(5) At a corrosion level of about 12%, the failure
mode changes from splitting to continuous slip-
page of the bar. It is at this corrosion level that the
interactive effect of loss of rib profile, lubrication
by the interstitial corrosion layer, and loss of
confinement causes a sharp change in the mode of
failure.
et al. 129
(6) At about 7% rebar corrosion, the rib profile loss is
about 45%, whereas at 12% rebar corrosion, the
loss in rib profile is around 70%. This explains the
bar slippage mode of failure at about 12% corro-
sion level.
(7) Corrosion levels between 5 and 7% are observed to
cause significant increase in crack width as well as
loss of rib profile. This explains the sharp reduc-
tion in the bond strength from 30 to 70% in this
range of corrosion.
A crack width of 0.3 mm and rib profile loss of
(8)
25% determine the critical values beyond which a
sharp reduction in bond strength occurs.
Acknowledgement
The authors gratefully acknowledge the support of
King Fahd University of Petroleum and Minerals,
Dhahran, Saudi Arabia, in the conduct of this research,
and appreciate Mr Efren Superales in typing this paper.
The assistance of Mr Mohammed Mukarram Khan in
the conduct of the experiments is also greatly appreci-
ated.
References
1 Rasheeduzzafar and Dakhil, F. H. Field studies on the durability
of concrete construction in a high chloride-sulphate environment.
Int. J. Housing Sci. 1980, 4(3), 203-232
2 Rasheeduzzafar, Dakhil, F. H. and Gahtani, A. S. The deterior-
ation of concrete structures in the environment of the Middle
East. J. Amer. Concr. Inst. 1984, 81(l), 13-20
3 Rasheeduzzafar, Dakhil, F. H. and Gahtani, A. S. Corrosion of
reinforcement in concrete structures in the middle east. Concr.
Intern. Des. Constr. 1985, 7(9), 48-55
4 Al-Tayyib, A. J., Rasheeduzzafars and Al-Mana, A. I.
Deterioration of concrete structures in the Gulf States. In Proc.
First Internaiionul Conference on Deterioration and Repair of
Reinforced Concrete in the Arabian Gulf; Bahrain, 1985,~~. 10-27
5 Standard specifications for deformed and plain billet-steel bars
for concrete reinforcement. Annual Book of ASTM Standards,
ASTM A615-89, Vol. 1.04, American Society for Testing and
Materials, Philadelphia, PA, 1990, pp. 383-391
6 Ferguson, P. M. Reinforced Concrete Fundamentals. 4th edn,
John Wiley, New York, pp. 169-210
I Thompson, P. W. Bond stress distribution in concrete beams and
eccentric pullout specimens. M.S. Thesis, University of Texas,
May 1961
8 Kemp, E. L. Bond in reinforced concrete: behaviour and design
criteria. AC1 J. January-February 1986, 50-57
9 Kemp, E. L., Brezny, F. S. and Unterspan, J. A. Effect of rust
and scale on the bond characteristics of deformed reinforcing
bars. ACI J. 1968, 65(9) 1968, 743-756
Al-Sulaimani, G. J., Kaleemullah, M., Basunbul, I. A. and
Rasheeduuafar, Influence of corrosion and cracking on bond
behaviour and strength of reinforced concrete members. ACI
Struct. J. 1990, #(2), 220-231
Cabrera, J. G. and Ghoddoussi, P. The effect of reinforcement
corrosion on the strength of the steel/concrete bond.
International Conference, Bond in Concrete, from Research to
Practice, Vol. 3, Riga, Latvia, October 15-17 1992
Bamforth, P. B. and Pocock, D. C. Minimising the risk of chlo-
ride induced corrosion by selection of concrete materials.
Corrosion of Reinforcement in Concrete, eds, C. L. Page, K. W.
Treadaway and P. B. Barnforth, Society of Chemical Industry,
Elsevier Science, Amsterdam, 1990, pp 178-187