REDUCTION OF 40-YEAR-OLD REINFORCED CONCRETE BEAMS IN NATURAL CORROSION ENVIRONMENTS
3.4.3 Deterioration Progress of 40-year-old Specimens Under Combined Marine and Atmosphere Environments
a. Cracking Map
The actual appearance of RC-S-A is shown in Fig. 3.8 and the cracking map after 40 years exposure is shown in Fig. 3.9(a). Many longitudinal cracks appear in the tensile areas of the front and back surfaces. Cracks appear along nearly the entire length of the span. Longitudinal cracks are coincident with the positions of the tensile bars. The maximum longitudinal crack width in the tensile area is 10 mm. In some parts of the compressive bar areas, longitudinal cracks also appear, coinciding with the compressive bar positions. The maximum longitudinal crack width in the compressive bar areas is 0.6 mm. On the tensile surface of the beam, large longitudinal cracks appear throughout the beam span. The longitudinal crack widths are less varied with 5 or 10 mm. In the web of the beam, a few transversal cracks are observed. These coincide with the positions of the stirrups. Therefore, the transversal cracks in the middle span could initiate from pre-cracking or from the corrosion of stirrups. In addition, the corrosion cracks are more developed in the area of tensile bars than that with compressive bars, both in intensity and
Compressive Surface
Front Surface
Tensile Surface Back Surface
Max 0.4 mm Max 0.4 mm
Max 0.5 mm
Max 1.4 mm Max 0.25 mm Max 2.0 mm
Max 1.0 mm
Max 1.5 mm Max 2.0 mm
rust
rust rust
rust
rust
rust
III-12
width. More areas of concrete spalling are observed on the tensile surface.
Fig. 3.8 Crack opening on RC-S-A beam
(a) Appearance of RC-S-A
(b) Appearance of RC-S-B
Fig. 3.9 Cracking maps of RC-S-A and RC-S-B beams after 40 years
0.1 0.1 0.1 0.1 0.2
0.3 0.3 0.3
0.3 0.3
0.2 0.3
0.4 0.4 0.5 0.6 0.6 0.3
0.1 0.4 0.4
0.8 0.15 0.15 0.4
0.7 0.8
0.3 0.4 0.3 0.2
0.3 0.3 0.30.3 0.15
0.2 0.2
0.4 0.4 0.3 0.3 0.3
0.2 0.2 0.4
0.1 0.1 0.1
0.1 0.1 0.1
0.3 0.15
0.15 0.2
0.15 0.06
0.1 0.4 0.10.8 0.9 10
0.1 0.2 0.2 0.6 0.2
0.2 0.1 0.1
0.1 0.1
0.3 0.30.2 0.5 0.6 0.1
0.3 0.1
0.2 0.2
0.2
0.1 0.1 0.6 0.5 0.2
0.3 0.2
0.4 0.3 0.3
0.2
0.1
3 5 5
55 5 5 5
10
10 10 10
5 5 5 5 5 5
10 5 5 5
Compressive Surface
Front Surface
Tensile Surface Back Surface
0.3 0.8 0.3 0.15 0.3 0.3
0.5 0.6
0.3 0.2 0.4 0.5
0.1 0.15 0.5
0.2 0.2
0.06 0.08
0.1
0.08 0.08
0.7 0.6 0.7 0.8 0.8
0.1 0.15
0.7 0.6 0.4
0.4 0.4 0.5
0.08 0.1
0.4 0.3 0.3 0.06
0.2 0.1
5 0.15
0.2 0.15 0.2 0.15 0.15 0.25
0.15 0.3
0.4 0.4 0.5
0.08
0.5 0.15
0.08 0.08
0.03 0.6 0.6 0.25
0.20 0.15 0.08 0.08
0.6 0.5 0.5
0.5 0.6
0.15 0.08 0.1 0.08
0.6 0.6 0.6 0.6 0.6 0.6 0.10.5 0.2 0.08
0.6 0.5 0.5 0.4
0.4 0.5 0.7 0.3
0.3 0.3 0.25
0.8 0.35
0.6 0.2
0.4 0.20.25
0.5 0.4
0.5 0.4 0.7
0.35
0.25 0.5 0.40.6 0.35
0.2 0.6 0.60.3 0.4
0.50.4 0.40.2 0.3
0.15 0.20.3
0.50.6 0.5
0.2 0.15
0.4
0.15 0.3
0.40.3 0.3 0.5 Compressive Surface
Front Surface
Tensile Surface Back Surface
Steel bars Cracks Spalling
III-13
a. Front Side
b. Back Side
Fig. 3.10 Half-cell potential of RC-S-A and RC-S-B beams after 40 years
The cracking map of RC-S-B after 40 years is shown in Fig. 9(b).
Longitudinal cracks only appear in the compressive bar area on the front and back surfaces along the beam span. This areal restriction may be caused by the casting direction. Voids are present under the horizontal bars, particularly those located in the upper half of the beam, in a phenomenon known as the “top-bar effect” [3.28].
The maximum longitudinal crack width in the compressive area is 0.8 mm. On the compressive surface, similar to that of RC-S-A, only a few transversal cracks are observed. In addition, in the web of the RC-S-B beam, some transversal cracks appear at the positions of stirrups. Therefore, the transversal cracks are likely
-800 -700 -600 -500 -400 -300 -200 -100 0
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 Potential (mV:CSE/25oC)
Measurement point (mm)
RC-S-A Stirrup Fr RC-S-A Tensile Fr RC-S-A Comp. Fr RC-S-B Stirrup Fr RC-S-B Tensile Fr RC-S-B Comp. Fr
-800 -700 -600 -500 -400 -300 -200 -100 0
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 Potential (mV:CSE/25oC)
Measurement point (mm)
RC-S-A Stirrup Bc RC-S-A Tensile Bc RC-S-A Comp. Bc RC-S-B Stirrup Bc RC-S-B Tensile Bc RC-S-B Comp. Bc
III-14
caused by stirrup corrosion. Moreover, on the tensile surface of RC-S-B, many longitudinal cracks appear throughout the beam span. The widths of these cracks range from 0.15 to 0.8 mm. No concrete spalling is observed on the tensile surface.
b. Half-cell potential
The half-cell potential of RC-S-A and RC-S-B as shown in Fig. 3.10. Both pre-cracked and without pre-cracked beam categorize probability 90% of corrosion. Perhaps, after certain period of exposed, even without pre-cracked chloride ion reached surface of steel bar already. In addition, half-cell potential of steel bar of pre-cracked beam more negative than without pre-cracked.
a) RC-S-A
(b) RC-S-B
Fig. 3.11 Crack formation in beams at ultimate load
c. Mechanical behavior of the beams
The cracking pattern in the RC-S-A beam under the ultimate bending load is shown in Fig. 3.11(a). The cracking of the beam is gradually increased with increases in the load. During loading, several new cracks appear. From the bottom to the top of the beam, the existing cracks are slowly widened and extended.
Existing Crack Bending Crack
Back Surface
Front Surface Back Surface
Front Surface
III-15
Finally, at the ultimate load, in the middle of the beam span, large areas of spalled concrete appear suddenly on the tensile surface, as shown in Fig. 3.12(a).
The cracking pattern in the RC-S-B beam under the ultimate bending load is shown in Fig. 3.11(b). Similar to RC-S-A, the cracking of the beam gradually increases with increases in the load. Several new cracks appear in the middle of the span. These cracks propagate from the bottom to the top of the beam.
Furthermore, only a few existing cracks are extended; however, these cracks are widened. Crushing of the concrete at the top compressive surface of the beam indicates the ultimate load. Ultimate failure occurs by concrete crushing on the top compressive surface of the beam, as shown in Fig. 3.12(b).
(a) RC-S-A (b) RC-S-B Fig. 3.12 Failure condition at the ultimate load
The failure type of a normal reinforced concrete beam is designed to be ductile. Ductile failure is designed to occur by yielding in the tensile bars, followed by concrete crushing at the top compressive surface at the ultimate load.
The corrosion of the steel bars affects the failure type of RC, even at the same ratio of steel. The ultimate failure mode in the RC-S-A beam is indicated by the sudden large spalled areas in the middle span. Sudden failure by spalling is caused by bond failure between the concrete and the main tensile bar, as shown in Fig.
3.12(a). The ultimate failure of RC-S-A is categorized as brittle. In RC-S-B, the failure type at the ultimate load is signified by concrete crushing at the top of the compressive surface. This indicates that the tensile bar yields first in the tensile surface, and then suffers plastic deformation. Finally, the concrete at the top of the
CL
Large spalled area
Crushed concrete
CL
III-16
compressive surface of the beam is crushed, as shown in Fig. 3.12(b). Therefore, RC-S-B is categorized as experiencing ductile failure.
d. Chloride ion distribution
Chloride ion distributions of RC-S-A and RC-S-B as shown in Fig. 3.13. At 10 years of exposure chloride content around 8~12.5 kg/m3. While at 20 years of exposure chloride content around 9~15 kg/m3. Currently, the total of chloride content ranged from 5.68 kg/m3 to 6.75 kg/m3. This value is very larger compared 1.2 kg/m3 chloride content, which are considered in the concrete for corrosion initiation based on JSCE standard. The chloride content at 40 years’ exposure tend to decreased. This because the beams transported from marine environment to laboratory sheltered condition at 20~35 years. Further, from 35~40 years at Kyushu University site with un-sheltered condition. Perhaps, the chloride reduced due to undergone washing process from rain. One interesting point, after a long-lasting exposed, the difference between the amount of corrosion in cracked and un-cracked concrete will be minor.
Fig. 3.13 Chloride ion distributions
e. Corrosion Distribution of the Steel Bar
Corrosion distribution provides information on the corrosion behavior of steel bars, such as the type of corrosion regarding homogeneity. The corrosion behavior of steel is also characterized by the form of corrosion as either general or pit
III-17
corrosion. The type and form of corrosion are important parameters affecting the mechanical behavior of the corroded RC beam. Yuan et al. [3.9] reported that steel bars exposed to accelerated corrosion processes, such as electrochemical processing, show different types and forms than those in non-accelerated artificial and natural corrosion environments. These differences lead to different corrosion distributions on the surface of the steel bar in either the type or form of corrosion.
Therefore, the mechanical properties of an RC beam corroded by accelerated or artificial processes differ from those of a beam corroded by a natural process.
Corrosion maps of the tensile bars in RC-S-A and RC-S-B are plotted in both the downward and upward directions, as shown in Fig. 3.14. Both RC-S-A and RC-S-B beams are already corroded at the bar surfaces in both the downward and upward directions. Further, in the downward direction, the RC-S-A tensile bars show much more pitting corrosion than the RC-S-B bars do. The amount of pitting corrosion in RC-S-A is twice that in RC-S-B. In the upward direction, minor pitting corrosion appears in both beams. In the middle regions of the RC-S-A tensile bars, a failure point exists, while none exists in the RC-S-B tensile bars.
F : front; B : back
Fig. 3.14 Corrosion maps for the tensile bars in upward and downward directions
Contrary to the tensile bars, the compressive bars in RC-S-B show up to twice as much pitting corrosion than is seen in those of RC-S-A in the downward
F B
F B
F B F B Top Surface Direction
Top Surface Direction Bottom Surface Direction
Bottom Surface Direction RC-S-A
RC-S-B
General Pit Failure
III-18
direction, as shown in Fig. 3.15. However, RC-S-A has several failure points, whereas RC-S-B shows none. The small number of pitting points in RC-S-A is probably related to the change in pitting points to failure points, or to the interconnection of pits and increasing cross-sectional loss. In the upward direction of the compressive bars, RC-S-A has large non-corroded areas, while RC-S-B only shows small areas lacking corrosion.
F : front; B : back
Fig. 3.15 Corrosion maps for the compressive bars in upward and downward directions
Corrosion maps of the stirrups in the RC-S-A and RC-S-B beams are shown in Fig. 3.16(a) and Fig. 3.16(b), respectively. Several portions of the stirrups in both RC-S-A and RC-S-B are uncorroded, particularly in the top part of the beam.
However, both beams show significant pitting corrosion in the stirrups. Moreover, almost all stirrups in RC-S-A are deteriorated or disconnected, particularly in the bottom portion. In RC-S-B, the stirrups are not broken, as shown in Fig. 3.17. One factor in this concentration of corrosion is that the stirrups have the thinnest concrete covers because of the framing of the main tensile bars. Another contributing factor is the exposure of the beams to the tidal zone, consequently entailing wetting–drying cycles, particularly in the bottom half of the beam.
Therefore, the most deterioration occurs in the lower portions of the stirrups.
F B
F B
F B
F B Top Surface Direction
Bottom Surface Direction RC-S-A
Top Surface Direction
Bottom Surface Direction RC-S-B
General Pit Failure
III-19
The corrosion distributions shown in Fig. 3.14 and Fig. 3.15 occupy the full lengths of the steel bars. The corrosion distributions in the downward direction are more severe than those in the upward direction. These differences could be caused by the casting direction; over the length of the steel bar, particularly in the downward direction, gaps or voids are generated. The existing gaps or voids in the interfaces of the steel bar and concrete would promote corrosion, as oxygen or water can become entrapped in these features. The “top-bar” effect degrades the interfacial quality between the steel bar and the concrete under the horizontal parts of the steel bars.
(a) Stirrups of RC-S-A
(b) Stirrups of RC-S-B
Fig. 3.16 Corrosion maps for the stirrups
1 2 3 4 5 6 7 8
9 10 11 12 13 14 15 16
1 2 3 4 5 6 7 8
9 10 11 12 13 14 15 16
General Pit Deteriorated
III-20
Fig. 3.17 Stirrups
(a) Compressive bar
(b) Tensile bar
Fig. 3.18 Cross-sectional loss of compressive and tensile bars based on weight measurements
RC-S-A RC-S-B
disconnected
0 5 10 15 20
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
Loss of cross section (mm²)
Location (mm) RC-S-A Comp. Front
RC-S-A Comp. Back RC-S-B Comp. Front
RC-S-B Comp. Back stirrup
0 10 20 30 40 50 60 70 80 90 100
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
Loss of cross section (mm²)
Location (mm) RC-S-A Ten. Front
RC-S-A Ten. Mid RC-S-A Ten. Back RC-S-B Ten. Front RC-S-B Ten. Mid RC-S-B Ten. Back
stirrup
III-21
f. Cross-sectional loss of steel bar
Cross-sectional losses of the compressive and tensile steel bars in RC-S-A and RC-S-B are shown in Fig. 3.18(a) and Fig. 3.18(b), respectively. The appearance of an extracted tensile bar is shown in Fig. 3.19. Since the forms of corrosion in the steel bars are varied, determining the cross-sectional loss based on diameter loss causes an overestimate. Therefore, cross-sectional loss was determined by cutting the corroded steel bar into short pieces and then measuring the weight loss. The compressive bars in RC-S-A and RC-S-B show similar cross-sectional losses. The local cross-sectional losses of the compressive bars are 17.8% and 16.5% for RC-S-A and RC-S-B, respectively. For the tensile bars, the cross-sectional loss differs significantly between RC-S-A and RC-S-B, whereas those of the compressive bars are similar. The local cross-sectional losses of the tensile bars in RC-S-A and RC-S-B are 36.1% and 13.4%, respectively.
(a) RC-S-A tensile bars (b) RC-S-B tensile bars Fig. 3.19 Condition of tensile bars before cleaning
The cross-sectional losses and crack widths were then plotted as a relationship curve to understand the influence of crack width on the cross-sectional loss. The residual load-carrying ultimate flexural capacity of the beam is greatly affected by corrosion of the reinforcements. Reinforcement corrosion increases the volume of the steel bars, which generates corrosion cracking. At the end of the corrosion period, the loss of bonding between the steel bar and concrete is caused by cross-sectional loss. Therefore, cross-sectional loss greatly affects the load-carrying capacity of the RC beam. By estimating the
Deeper cross-section loss Lug and ribs still intact
III-22
cross-sectional loss of the tensile bars based on the crack width, the residual load-carrying capacity of the RC can be estimated.
Fig. 3.20 Relationship between crack width and cross-sectional loss in RC-S-B tensile bars
The crack width opening in the RC-S-A beam is too larger, and therefore not useful for estimating cross-sectional loss. Therefore, the data from RC-S-B was used to plot the relationship between crack width and cross-sectional loss. The RC-S-B beam lacks pre-cracks, so the cross-sectional loss is not overestimated.
The RC-S-B data is also more suitable because the crack width varies throughout the span. The relationship between crack width and cross-sectional loss shows a good correlation, as shown in Fig. 3.20. The R-squared value obtained from the linear regression line shows a relatively high value.
g. Tensile Test
Tensile testing was conducted in order to understand the effect of corrosion on the mechanical properties of the steel bars. Corroded steel bars, extracted from both the RC-S-A and RC-S-B beams, were cleaned using 10% diammonium hydrogen citrate (C6H14N2O7) in water. The positions of the steel bars extracted for the tensile test are shown in Fig. 3.4. The tensile test was conducted according to the JIS Z 2241-1998 standard at the tensile testing machines, as described in Fig. 3.21. The tensile bars are steel and 13 mm in diameter. The base length L was
y = 10.7 x + 12.4 R² = 0.8
0 5 10 15 20 25 30
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Loss of Cross Section (mm²)
Crack Width (mm)
RC-S-B front tensile bar RC-S-B middle tensile bar RC-S-B back tensile bar
III-23
determined as ten times the diameter of the bar. Further, the effective length Lo
was determined as eight times the diameter of the bar. The effective length was demarcated with two small notches in the steel bar surface, made using nails and a hammer. By using Vernier calipers, the length between the two points indicated by notches was measured. After the tensile test was conducted, the effective length Lo' was measured again to determine the elongation.
Fig. 3.21 Schematic of tensile test of steel bar
The elongations of the front and back RC-S-A tensile bars were 2.3% and 4.4%, respectively. Those in the front and back RC-S-B tensile bars were 12.1%
and 12.9%, respectively. The elongation of the control bar was 20.5%, as shown in Table 3.5. The results show that the elongation of the steel bar is significantly decreased with corrosion. However, in the base length of 600 mm in the steel bar extracted for tensile testing, the corrosion type was irregular along the length. The obtained cross-sectional loss was varied, and therefore the stress distribution was different along the bar. In particular, the elongation of the steel bar extracted from the RC-S-A beam was strictly limited. The control bar shows elongation during the tensile test by necking at the failure point, as shown in Fig. 3.22. However, the corroded steel bar extracted from RC-S-A shows almost no necking. In the corroded steel bar extracted from RC-S-B, some necking is observed. These results indicate that the existence of cross-sectional loss greatly affected the mechanical behaviors of the steel bars. The stresses are not distributed, but instead
III-24
concentrated in the areas where large cross-sectional losses exist. Then, the steel bar becomes more brittle.
Table 3.5 Summary of tensile test results
Item D13
Control
RC-S-A fr. ten.
RC-S-A bc. ten.
RC-S-B fr. ten.
RC-S-B bc. ten.
Loss of diameter (%) * - 11.1 15.5 3.1 2.9
Loss of cross-section (%) * - 31.4 37.4 17 15
Lo (mm) 100.5 101.4 99.9 101.0 100.3
L` (mm) 121.2 103.8 104.3 113.3 113.3
ΔL (mm) 20.7 2.4 4.3 12.3 13.0
Elongation (%) 20.5 2.3 4.4 12.1 12.9
Yield stress (N/mm2) 394.6 194.7 177.2 320.8 302.6 Ultimate stress (N/mm2) 540.7 260.1 300.6 489.3 482.2 Yield strain (×10-6) 1942 1289 1222 1741 1615 Ultimate strain (×10-6) 177553 7370 24678 38685 33384
Loss in yield stress (%) - 50.7 55.1 18.7 23.3
Loss in ultimate stress (%) - 51.9 44.4 9.8 10.88
* Based length 600 mm; fr. ten.: front tensile bar; bc. ten.: back tensile bar
Fig. 3.22 Failure points at ultimate load during tensile testing
III-25
The stress–strain curves obtained from the tensile test are plotted in Fig. 3.23.
The straight black line indicates the stress-strain curve of the control bar, which shows yielding and ultimate stresses of 394.6 N/mm2 and 540.7 N/mm2, respectively. The effect of corrosion on the mechanical properties of the steel bars is clear. The average losses in yield capacity for the tensile bars in RC-S-A and RC-S-B are approximately 50% and 20%, respectively. Loss in the ultimate capacity also occurs at the ultimate load. The average ultimate capacity losses of the two tensile bars in RC-S-A and RC-S-B are 48% and 10%, respectively. At the beginning of loading, the trends in the stress-strain curves of the extracted RC-S-A and RC-S-B tensile bars are similar to that of the control bar. With increases in the load, the strain behavior changes slowly. After reaching the yielding point, the strain behavior changes drastically. The tensile bars in RC-S-A reach the yielding point first, followed by the RC-S-B tensile bars, and then the control bar.
Fig. 3.23 Stress-strain curves for tensile tests of tensile bars and control
The stress is not distributed well in the steel bars because of the variations in the cross-sectional losses. Stress is not equally distributed along the steel bar, but instead is concentrated at specific locations. The stress becomes more seriously concentrated with higher cross-sectional losses. The concentrated stress causes
0 100 200 300 400 500 600
0 50000 100000 150000 200000
Stress (N/mm²)
Strain (x10-6)
D13 Control
RC-S-A front tensile bar RC-S-A back tensile bar RC-S-B front tensile bar RC-S-B back tensile bar
Loss of ductility
III-26
premature yielding of the steel during loading and reductions in the plasticity of the steel bar. Therefore, the RC-S-A tensile bar reaches the yield point first. The RC-S-B tensile bars have smaller cross-sectional losses, so the ultimate load approaches that of the control bar. However, the bar still experiences a greater loss of ductility than the control bar. For the control bar at the ultimate load, the strain is 177553 × 10-6, while in RC-S-B for the front and back tensile bars experience strains of 38685 × 10-6 and 33384 × 10-6, respectively. The mechanical behaviors of the steel bars are clearly affected by the corrosion, including the yield strength, ultimate strength, and ductility.
h. Bending Test
The load-displacement behavior of the RC-S-A and RC-S-B beams are analyzed and compared as shown in Fig. 3.24, while a summary of the bending test is presented in Table 3.6. The yield loading values in RC-S-A and RC-S-B are 42.8 kN and 91.9 kN, respectively. The ultimate loading values are 61.7 kN for RC-S-A and 97.4 kN for RC-S-B. The losses of ultimate capacity in RC-S-A and RC-S-B are 44.8% and 12.8% of the initial ultimate capacities Muo
(experimentally determined before exposure). Meanwhile, the loss of yielding capacity in RC-S-A is 53.4% of that in RC-S-B. However, according to the JSCE Standard Specification [23] where each safety factor is set to 1.0, both RC-S-A and RC-S-B capacities are below the calculated ultimate load of the RC beam, which is 99.4 kN. The displacement in RC-S-A at the ultimate load is higher than that in RC-S-B.
The higher displacement in RC-S-A may indicate poor bond strength in RC-S-A. Therefore, no further interlock force is applied in RC-S-A. In RC-S-B, the bond strength between the steel bar and concrete remains sufficient. This bond strength generates enough interlock force in the RC-S-B to counter deflection. In addition, two possible reasons may explain the rebar failure of RC-S-A at small elongation but large deflection: 1) The large deflection in RC-S-A might be caused by the loss of bond strength between steel bar and concrete. The loss of bond strength is generated from the large cross-sectional loss in the tensile bars.
As a result, the degradation in flexural stiffness may have caused excessive