The current experimental study presents the results of bond strength loss (steel bar concrete) due to the corrosion damage of steel bar specimens, semiembedded in concrete, at various times of exposure to corrosive environment. In this case, a correlation was made between the width of the surface cracks of concrete caused by reinforcing steel corrosion and bond strength for different distances between stirrups and different cover thickness of concrete. The study indicates close relationship between the width of surface cracking, the percentage mass loss of embedded reinforcing bar, the distance between stirrups, and the cover thickness. In addition, mathematical predictive models of bond strength loss of corroded specimens were proposed. The model outcomes showed that the cracking development on concrete surface up to a width of 1.6 mm is accompanied by an exponential reduction of bond strength loss between steel reinforcement and concrete. Furthermore, the investigation has shown that the increase of transverse reinforcement (stirrups) percentage and the cover thickness play a significant role in durability of reinforced concrete elements and in bond strength maintenance between rebar and concrete.
Corrosion of reinforcement steel consists of one of the main degradation problems in reinforced concrete structures. In many coastal regions, where the prevailing high (mean) temperature is combined with high concentration of chlorides, the corrosion phenomena are intense [
According to the existing knowledge, corrosion is an electrochemical process, in which steel tends to return to its original form (ore) by forming iron oxides on its surface (rust). Originally, the surrounding concrete protects the steel reinforcement acting as a physical barrier and forming a thin protective film of hydrated iron oxide at its interface with steel, because of the high alkalinity of concrete (pH∼12.5). Nevertheless, aggressive environmental factors gradually penetrate the concrete through its pores, and when chloride ions reach a critical concentration rate in conjunction with the pH drop, steel reinforcement depassivates, and corrosion initiates [
The consequences of corrosion damage on the mechanical properties of the steel reinforcement are significant and well-documented, especially when it comes from chlorides action [
The development of iron oxides at the surrounding concrete is accompanied by tensile stresses and surface cracks.
Generally, cracks are inherent in reinforced concrete structures and are caused by a number of different types of actions. One of the most severe forms of cracking in hardened reinforced concrete structures is the result of the corrosion of steel reinforcing bars. However, in accordance with Eurocode 2 [
Reasons of “health monitoring” of coastal structures, which suffer from the presence of chloride ions, often push inspectors to relate the service life of structures with that of chloride content in concrete, as it is a generally followed standard practice. Moreover, it is known that the predictive life-time models about the onset of corrosion of reinforced concrete elements, in their majority, require limit value (threshold) chloride content (CTV). Nevertheless, in the existing literature to date, neither has a model been prevailed nor been established (a specific chloride content in concrete) that has been widely verified by the researchers. This is largely related to the inability to accept a certain reliable method of calculating the chloride threshold value, resulting in misinterpretations and different results. The most important specifications related to chloride content in concrete and available in the literature, such as [
Technical assessments inspection of durability of the load-bearing structure is a very complex process. In reinforced concrete structures where the rebar is not optically accessible, chloride measurements and corrosion damage estimation are of low reliability. Contrary to the above, the occurrence of surface cracking in concrete due to corrosion is visible and measurable, and the crack width is easily quantified. Based on this fact, there is already a tendency of some researchers to link the surface cracks of the concrete, the corrosion damage of steel reinforcement, and the degradation of bonding locally to the surrounding concrete [
In the current study, the results of a broad experimental investigation on the correlation of the bond strength with the surface cracking width are discussed and analyzed. The main input parameters in this correlation are the cover thickness of concrete
Experimental procedure consisted of two phases. Initially, 48 reinforced concrete specimens were made, the classification of which was based on the cover thickness and the amount of stirrups (the distance between them or their absence). Subsequently, controlled corrosion damage to the steel reinforcement was carried out by accelerated electrocorrosion exposing the specimens to the corrosive environment for 5 different times. Thereafter, the percentage mass loss of the corroded reinforcement was estimated and correlated with the surface cracking width of concrete. Finally, the bond strength between steel reinforcement and concrete of corroded specimens was tested by pull-out tests, which were compared to the corresponding noncorroded reference specimens.
In this manuscript, a parallel aim is to study structural reinforced concrete elements of existing structures. Hence, concrete C20/25 is chosen, as this category constitutes a representative sample given the fact the majority of structures in many regions in the Mediterranean basin such as Greece, Turkey, and Italy were built up by ignoring or adopting late the incentives of EN206 on the use of concrete strength class C25/30 C30/37 in coastal environment and likewise with the rules about cover thickness. Therefore, the preparation of (prismatic) 240 × 200 × 310 (mm) reinforced concrete specimens, Figure
(a) Typical cross-sectional view (cover thickness,
Before mixing the concrete in molds, reinforcing steel bars were carefully cleaned, and their initial mass was documented. Thereafter, all steel bars and stirrups were aligned and fastened to the molds. Compaction of concrete was conducted with the help of table vibrator. The concrete specimens were kept for three days in a room at a temperature of 22°C and continuous wetting. Thereafter, the molds carefully removed, and the specimens were immersed in water for a period of 28 days (curing). For the concrete mixture, a Portland cement was used, with water-cement ratio of 0.55 and 20 mm maximum size coarse aggregate. The 28-day recorded compressive strength of the concrete was measured on 200 mm3 cubes and with average
Prior to initiating a series of specific tests, appropriate labeling was carried out which correlated with the specific characteristics of each specimen. The investigation of RC elements’ bond behavior was conducted by examining three main parameters, the cover thickness (
Tested parameters of RC elements’ bond behavior (cover thickness, transverse reinforcement (stirrups), and electrocorrosion time).
Tested parameters | ||
---|---|---|
Cover thickness, |
Transverse reinforcement (stirrups) | Exposure time |
|
No stirrups (N) |
Noncorroded 0 |
The occurrence of corrosion phenomena in the reinforced concrete elements in the real natural environment takes place slowly over a period of more than 20 years. In order to study the effect of different intensities of corrosion phenomena on bond strength, experimental tests of accelerated corrosion on reinforced concrete specimens (which had previously been properly maintained for 30 days after their concreting) were carried out by the anodic corrosion method. A power supply is used to induce the corrosion by applying direct electric current to semiembedded reinforcing steel bar, which acted as anode. The specimens were fully immersed in electrocorrosion cells filled with 5% sodium chloride (NaCl) solution by weight of water, in the presence of a stainless-steel bar (cathode of the circuit), which was positioned in the same direction as the semiembedded steel bar. Solution 5% NaCl was selected because it simulates a severe corrosive environment or a coastal environment [
(a) Simplified procedure of accelerated electrocorrosion. (b) Metal mold of steel reinforcement before casting.
The induced corrosion was limited to a 250 mm section of the semiembedded reinforcing steel bar, the rest of which was protected with a wax layer. Correspondingly, the stirrups were exposed to electrocorrosion along the upper part of the specimens on the side of the main reinforcing bar by 80 mm from the concrete surface. The remaining area of stirrups legs and the other 4 auxiliary corner bars Φ8 were suitably protected with a special anticorrosive coating (epoxy resin) (Figure
The cracks that appeared on the concrete surface, after specimen removal from the electrocorrosion cells, were recorded, mainly on the upper side of the specimens where the main reinforcing steel bar Φ16 and the unprotected legs of the stirrups were placed. The average width of surface concrete’s cracking was mapped, measured, and calculated for each specimen (Figure
Occurrence of surface concrete cracking (parallel to the axis of main steel bar) due to corrosion.
A series of mechanical pull-out tests, based on ASTM C234-91a [
For this purpose, an apparatus was designed (Figure
(a) Special apparatus which transfers the force from steel bar to concrete. (b) Experimental pull-out test.
The calculation of the bond strength of reference specimens and corroded specimens was recorded taking the nominal diameter of the reinforcement bar into account. The average bond strength between steel and concrete
Calculation of mean bond strength between steel and concrete (pull-out test).
The results of the experimental work are summarized in Table
Experimental results (mass loss of steel reinforcement, average crack width of concrete, and bond strength loss).
Specimen | Mass loss (%) | Average crack width (mm) | Bond loss ratio (MPa) |
---|---|---|---|
25-N-0 | 0 | 0 | 1 |
25-N-1 | 0.97 | 0.20 | 0.88 |
25-N-2 | 2.05 | 0.35 | 0.62 |
25-N-3 | 4.12 | 0.55 | 0.41 |
25-N-4 | 5.83 | 0.95 | 0.19 |
25-N-5 | 7.73 | 1.45 | 0.15 |
25-S240-0 | 0 | 0 | 1 |
25-S240-1 | 1.03 | 0.35 | 0.77 |
25-S240-2 | 2.22 | 0.55 | 0.60 |
25-S240-3 | 4.38 | 0.85 | 0.51 |
25-S240-4 | 5.74 | 1.05 | 0.48 |
25-S240-5 | 8.47 | 1.45 | 0.40 |
25-S120-0 | 0 | 0 | 1 |
25-S120-1 | 1.17 | 0.35 | 0.95 |
25-S10-2 | 3.22 | 0.70 | 0.82 |
25-S120-3 | 5.84 | 1.20 | 0.71 |
25-S120-4 | 7.06 | 1.10 | 0.50 |
25-S120-5 | 8.32 | 1.40 | 0.68 |
25-S60-0 | 0 | 0 | 1 |
25-S60-1 | 0.85 | 0.20 | 1.10 |
25-S60-2 | 1.72 | 0.55 | 1.02 |
25-S60-3 | 2.96 | 0.65 | 0.92 |
25-S60-4 | 5.82 | 0.90 | 0.90 |
25-S60-5 | 8.68 | 1.00 | 0.82 |
25-N-0 | 0 | 0 | 1 |
40-N-1 | 0.52 | 0.20 | 0.78 |
40-N-2 | 2.36 | 0.55 | 0.45 |
40-N-3 | 5.12 | 1.00 | 0.25 |
40-N-4 | 7.81 | 1.20 | 0.32 |
40-N-5 | 8.26 | 1.45 | 0.15 |
40-S240-0 | 0 | 0 | 1 |
40-S240-1 | 1.04 | 0.30 | 0.92 |
40-S240-2 | 2.24 | 0.90 | 0.53 |
40-S240-3 | 4.10 | 1.40 | 0.40 |
40-S240-4 | 5.77 | 1.50 | 0.27 |
40-S240-5 | 8.81 | 1.60 | 0.28 |
40-S120-0 | 0 | 0 | 1 |
40-S120-1 | 2.06 | 0.30 | 0.96 |
40-S240-2 | 3.14 | 0.65 | 0.83 |
40-S120-3 | 5.02 | 1.15 | 0.51 |
40-S120-4 | 6.44 | 1.45 | 0.43 |
40-S120-5 | 8.76 | 1.50 | 0.47 |
40-S60-0 | 0 | 0 | 1 |
40-S60-1 | 0.96 | 0.25 | 0.98 |
40-S60-2 | 2.86 | 0.60 | 0.87 |
40-S60-3 | 3.82 | 0.85 | 0.85 |
40-S60-4 | 6.46 | 0.90 | 0.82 |
40-S60-5 | 7.96 | 1.00 | 0.68 |
After the measurement of crack width at discrete points of each surface crack and the calculation of the mean cracking width of each specimen as well as the mass loss of steel reinforcement, Figures
Correlation of average crack width on concrete surface and of percentage mass loss of steel bar (cover thickness 25 mm).
Correlation of average crack width on concrete surface and of percentage mass loss of steel bar (cover thickness 40 mm).
Furthermore, average crack width and bond strength loss were correlated after the conduction of pull-out tests on corroded specimens. The results of this investigation are presented in Figures
Correlation of bond strength loss and average crack width (cover thickness 25 mm).
Correlation of bond strength loss and average crack width (cover thickness 40 mm).
In this study, two different predictive models of bond strength loss due to corrosion of steel reinforcement were presented, based on the correlation of bond strength loss of corroded specimens and surface cracking of concrete. An allometric equation is proposed for the first predictive model (Figure
Allometric predictive model of bond loss as a function of average crack width, cover 25 mm (a) and cover 40 mm (b).
Exponential predictive model of bond loss as a function of average crack width, cover 25 mm (a) and cover 40 mm (b).
The functions of both predictive models are as follows:
Parameters by regression analysis for the allometric predictive model.
Cover 25 mm | Cover 40 mm | |||||||
---|---|---|---|---|---|---|---|---|
No stirrups | Φ8/240 | Φ8/120 | Φ8/60 | No stirrups | Φ8/240 | Φ8/120 | Φ8/60 | |
|
0.233 | 0.475 | 0.743 | 0.870 | 0.282 | 0.435 | 0.576 | 0.773 |
|
−0.845 | −0.448 | −0.239 | −0.156 | −0.640 | −0.633 | −0.461 | −0.181 |
|
97.3 | 98.7 | 99.4 | 79.3 | 93.0 | 94.0 | 87.1 | 62.1 |
Parameters by regression analysis for the exponential predictive model.
Cover 25 mm | Cover 40 mm | |||||||
---|---|---|---|---|---|---|---|---|
No stirrups | Φ8/240 | Φ8/120 | Φ8/60 | No stirrups | Φ8/240 | Φ8/120 | Φ8/60 | |
|
1.435 | 0.736 | 0.274 | 0.117 | 1.257 | 0.724 | 0.499 | 0.260 |
|
96.2 | 96.5 | 97.7 | 45.9 | 97.5 | 96.0 | 91.9 | 80.1 |
All 48 specimens did not exhibit any surface cracking after their concreting and 28 days of maintenance. Thus, the recorded surface crack widths were solely due to corrosion of steel reinforcement. The loss of passive steel protection resulted in rapid corrosion and thereby in linear development of various range cracking of concrete along the steel bars. Diagrams of correlation between percentage mass loss of steel bar and surface cracking of concrete depicted the following.
The group of specimens with concrete cover thickness equal to 25 mm (Figure
In the group of specimens with concrete cover thickness equal to 40 mm (Figure
However, as corrosion increases, specimens with stirrups Φ8/60 mm recorded a remarkable curtailment of surface cracking development. The remarkable differentiation of specimens with stirrups Φ8/60 mm has been (about) equivalently observed for both groups, the one with concrete cover thickness of 25 mm and the one of 40 mm; since for mass loss percentage between 8.5% and 9%, an average range of cracking of only 1 mm has been recorded.
In addition, observing the bond loss drop of corroded specimens, Figures
In both groups of steel specimens, concrete cover thickness of 25 mm and 40 mm (Figures
In the same groups of specimens, it is obvious form Figures
Between the two groups of specimens regarding the concrete cover thickness, the change of bond strength recorded in the first group (cover thickness equal to 25 mm) was clearly foreseeable than the second one.
The rare fitting of stirrups by 240 mm (group of specimens with cover thickness of 25 mm) does not seem to affect bond strength performance at percentages less than 40%, contrary to the absence of stirrups (Figure
In the case of cover thickness of 40 mm (Figure
Between the two groups of specimens, those with stirrups Φ8/60 mm and Φ8/120 mm, a rapid degradation of bond strength occurs in specimens with cover thickness of 40 mm against compared to those with cover thickness of 25 mm. There is a significant reason for this behavior; given the same total dimensions of specimens, on one hand, the concrete cover of 40 mm is associated with smaller effective cross section and on the other hand, in case of thicker concrete cover, more extensive damage occurs due to the inner development of oxide in concrete on grounds of corrosion of steel reinforcement. Specimens with stirrups per 240 mm for both groups of cover thickness do not show signs of same behavior. A possible explanation of this differentiation could be synergy of various factors such as the scale of specimen and the incurred reduced corrosion damage of sparse stirrups.
The bond strength between concrete and steel shows obviously a significant decrease by increasing the range of surface cracking of concrete. Both Figures
The knowledge of bond strength between concrete and steel bar is a very important aspect in the mechanical behavior of RC structures, both in new structures’ design and in the assessment of existing ones. In the existing literature, there are various predictive models of bond strength loss of corroded RC specimens that provide linear or exponential bond loss [
Thus, the evaluation of corrosion level of existing structures is a really complicated issue. Steel reinforcement is rarely visible or accessible, and the corrosion damage is uneven along the steel bars. The evaluation of corrosion damage of steel bars in experimental studies is calculated through mass loss of random length’s part. Hence, the measured mass loss has no local characteristics at the position of maximum loss of the cross-sectional area, but an average uniform mass loss along the steel bar. Against this background, it becomes obvious that a more suitable technique of correlation between damage corrosion and degradation of bond strength concrete steel should be adopted. As it is well known, surface cracks indicate the presence of corrosion of steel reinforcement. Fischer and Ožbolt [
Referring to the literature, it emerged that studies associating bond strength loss with crack width are limited. In contrast, as correlation indexes are sometimes determined the corrosion level of steel bar and sometimes the depth of chlorides’ penetration and their percentage concentration. Tahershamsi et al. [
Given this tendency to approach the issue of bond strength between concrete and steel, the developing predictive models of the current study linked the bond strength loss of corroded specimens to the average width of concrete surface cracking. As it is demonstrated, the equation of prediction chosen was allometric in the first model and exponential in the second. Even though in literature are met linear predictive models of corrosion ongoing in relation to crack width, such as by Cabrera [
In the case of presence of stirrups Φ8/60 mm, in both predictive models, a decrease of factor
In this paper, through its efforts to validate and correlate the experimental results with those of the literature, appears that the experimental ones display large dispersion with each other and with those from the literature, due to various parameters such as concrete category, nominal diameter of steel reinforcement, concrete cover thickness, presence or absence of stirrups. As shown in Table
Review parameters of bond strength loss from existing studies and present experimental study.
Study | Test |
|
Steel bar diameter (mm) | Cover (mm) | Stirrups |
---|---|---|---|---|---|
Almusallam et al. [ |
Pull-out | 30 | 12 | 63.5 | No |
Rodriguez et al. [ |
Beam end | 40 | 16 | 24 | Φ8/100 |
Zandi and Coronelli [ |
Beam end | 34–38 | 20 | 30 | Φ8/48 |
Fischer and Ožbolt [ |
Beam end | 41 | 12, 1 6 | 20, 35 | No |
Zhao et al. [ |
Pull-out/Beam end | 41.9 | 18 | 66 | No/Φ8/100 |
Tahershamsi et al. [ |
Beam end | 37–49 | 2 × 16 (bundled) | 30–70 | Φ8/300 |
|
|
|
|
|
|
Even though mass loss of steel reinforcement due to corrosion about 8-9% and corresponding cracks up to 1.4 mm–1.6 mm displayed dramatic decrease of bond strength, often the encountered surface crack widths in existing structures present larger values. From this point of view, the present session will expand with further running experiments, so as to extend to wider cracks.
In the present study, an extensive experimental study was conducted on reinforced concrete elements, where the corrosion damage of the steel reinforcement was correlated with the average crack width on surface of concrete and the bond strength loss between steel and concrete. From this investigation, the following outcomes were obtained: The corrosion damage of steel reinforcement causes surface cracking width on concrete, which is closely related both to the existence and amount of stirrups and to the cover thickness of concrete. The use of dense stirrups Φ8/60 mm, in the case of cover thickness The development of surface cracking in concrete with an average width up to 1.60 mm is associated with an exponential reduction of bonding forces steel reinforcement concrete. In both cases of concrete cover (25 mm and 40 mm), the presence of dense connectors, Φ8/60 mm, was accompanied by a clear limitation of the surface cracking development up to 0.90 mm of width, corresponding to an average mass loss of about 8.5%–9%. The densification of stirrups (through the confinement) contributes positively to maintaining bond between steel reinforcement and concrete. Prediction of bond loss was made by an allometric and exponential function. Predictive results through the exponential function were found to be in good agreement with the approximation of contemporary corresponding studies in the literature.
No data were used to support this study.
The authors declare that they have no conflicts of interest.