Retrofitting Reinforced Concrete One–Way Damaged Slabs Exposed to High Temperature

: Exposure of reinforced concrete buildings to an accidental fire may result in cracking and loss in the bearing capacity of their major components, columns, beams, and slabs. It is a challenge for structural engineers to develop efficient retrofitting techniques that enable RC slabs to restore their structural integrity, after being exposed to intense fires for a long period of time. Experimental investigation was carried out on twenty one slab specimens made of self compacting concrete, eighteen of them are retrofitted with CFRP sheets after burning and loading till failure while three of them (which represent control specimens) are retrofitted with CFRP sheet after loading till failure without burning. All slabs had been tested in a simply supported span and subjected to two-point loading. The main variables were the effect of different temperature levels (300ºC, 500ºC and 700ºC), different concrete compressive strength (20MPa, 30MPa and 40MPa) and cooling rate (gradually and sudden cooling conditions) on the behavior of retrofitted one way slabs .The structural response of each slab specimen was investigated in terms of load-deflection behavior, ultimate load carrying capacity and mode of failure. The experimental results, generally, indicate that slabs retrofitted using CFRP sheets restored flexural strength values nearly equal to or lower than those of the reference slabs, the retrofitted slabs exhibited larger deflection than the control slabs at ultimate loads. Retrofitted control slabs after loading regained about 93.95% to 97.92% of their original load capacity (before retrofitting) while the other slabs regained from 42.% to 84% of the load capacity of the original control specimens. Most of the tested slabs failed by concrete crushing at mid span and partial debonding of certain retrofitting systems was also observed for a few cases.

which was carried out to study the effect of fire flame on the behavior and strength of one way self compacted RC slabs specimens by burning eighteen of them in fire flame at different temperatures then loaded them until failure, while three specimens are loaded till failure without burning, they represent the reference specimens as indicated in Table 1 . In this study the same damaged specimens are retrofitted by CFRP and loaded till failure to investigate the efficiency of retrofitting system to restore the load capacity of burneddamaged slabs.

Material properties:
Type I Portland cement, fine aggregate with (4.75mm) maximum size, coarse aggregate with (10mm) maximum size, water , superplasticizer , silica fume and reinforcing bars are used in casting slabs.
Smooth reinforcing steel bars were of diameter 3mm (2.93mm). The reinforcing ratio was constant, details of reinforcement as shown in Fig 1 and 2. The results of yield stress test of steel bars show that the yield stress value of bars was (800 MPa) with modulus of elasticity (195.9 GPa). The unidirectional SikaWrap Hex-230C is an externally applied retrofitting system for RC slabs. Carbon fiber fabric SikaWrap Hex-230C and epoxy based impregnating resin Sikadur-330 properties are shown in Tables 2 and 3 as reported by the manufacturer.

Concrete mix proportions:
In this research, three groups of compressive strength of self-compacting concrete has been used. The ratios of mixing are resulted by casting trial mix cubes and testing in (7 days) age. Every trial mix has six cubes of (100 × 100 × 100mm). three cubes are tested to find out the compression strength of the concrete before burning and the other three are tested after burning. Table 4, shows the details of mixes, while Table 5 shows the compression strength before and after exposure to fire flame.

Experimental procedure
Twenty one reinforced concrete slabs were tested. All of specimens have the same dimensions, length is 500mm, width is 250mm and their thickness was 40mm. The reinforcement for all specimens are The slabs in this research are divided in to three groups; R, G and S. Group R contain three specimens which are not exposed to temperature representing the reference specimens; each specimen has a different strength (20, 30 and 40 MPa). Each of groups G and S has nine specimens; the difference between them is in the method of cooling after exposing to high temperature. Group G is gradually cooled while group S is suddenly cooled by water. Each groups G and S is divided into three subgroups; each of them has a different strength (20,30 and 40 MPa). The three specimens of each subgroup are exposed to a different temperature (300, 500 and 700 C 0 ). All slabs were tested under two point loads until failure and then were retrofitted with two layer of CFRP sheets and renamed (FR, FG and FS) for (group R, groups G and groups S) respectively. The detailed classification of groups is shown in Table 6.

Retrofitting process
After burning slab at high temperature and loading until failure, the slab was overturned (upside down) so that retrofitting should be done on its bottom surface. In order to ensure correct application of the external strengthening materials and to remove the deformation (curved shape) due to the loading at the first stage (after burning the specimens and before retrofitting them), it was considered necessary to improve the concrete surface characteristics on the contact areas to be bonded . It included removing the cement paste, The spalled areas at corner of slabs were patched conventional concrete repair methodologies, grinding the surface by using an electrical hand grinder, and removing the dust generated by surface grinding using an air blower. After that the primer was applied to repair surfaces with non-shedding brush. The adhesive was applied to the tension face of the slab Sikadur ® -330 (two-part epoxy impregnation resin) was used in this work for the bonding of CFRP sheet. The twopart epoxy was mixed according to the manufacturer's specification: 4 parts resin to 1 part hardener by weight. The epoxy system was thoroughly hand mixed for at least 3 minutes at room temperature. The CFRP sheets placed in the bottom of slab after applying thin layer of epoxy. The time gap between the CFRP sheets bonding and the slab test was at least 7days. The procedure of applying of CFRP is shown in Figs. 3 to 6.

Test setup
All specimens were loaded by two point loads until failure using hydraulic testing machine in the Structural Laboratory at the College of Engineering of Baghdad University as shown in Fig. 7. The distance between the specimen supports was 400mm and the distance between each point load was100mm. Deflection of the slab specimens was measured at mid-span using a dial gauge with travel distance of 50 mm and accuracy of 0.01 mm.

EXPERIMENTAL RESULTS AND DISCUSSIONS
Results of ultimate load, maximum deflection and percentages of residual ultimate load of specimens exposed to high temperature and retrofitted specimens are summarized in Table 7. The behavior of the retrofitted slabs was investigated in terms of load-deflection , ultimate load carrying capacity and mode of failure to make comparisons with the performance of the original unburned slabs (reference specimens). The role of different parameters affecting the strength and deflection of retrofitted specimens is discussed in the following paragraphs in some detail.

Load-mid span deflection relationship
The retrofitted slabs exhibited larger deflection than the damaged reference slabs at ultimate load. This observation deviates from test results reported by [Chan and Niall,2001], who showed that the deflections were higher in the specimens with preload prior to applying FRP laminates and the load-deflection behavior becomes brittle. All retrofitted slabs behaved similarly, it can be observed that the load versus midspan deflection response can be divided into two stages of behavior. The first limited stage was characterized by an approximately linear relationship between the load and the mid-span deflection. During this stage of behavior, the section was uncracked (no new cracks were observed) and both the concrete and steel, in addition to the CFRP sheet, behave essentially elastic. The second stage represents the behavior post initial cracking of the composite section (exhibit inelastic behavior) where the stiffness of the slab was decreased as indicated by the reduced slope of the load versus mid-span deflection curve. The stiffness of specimens R and specimens FR were nearly identical as shown in Fig. 8. Similar remarks have been made by [Obaidat et al, 2010]. It should be noted that repair of reference slabs after loaded allowed recovering the original stiffness. Slab specimens FG and FS for all compressive strength at all exposure temperature showed decreases in stiffness when compared with the of reference slab specimen R. The stiffness of specimens (FG and FS) is higher than (G and S) respectively at the earlier loading stage. After initiating new cracks or growing the old cracks due to loading, the stiffness of (FG and FS) is decreases than (G and S) respectively as shown in Figs. 9 to 11 . It should be noted that repair groups (S) and (G) could not restore the stiffness to the level of the reference slabs.

Ultimate load
From .42 at 300, 500 and 700 °C respectively . So, the average percentages of residual ultimate load carrying capacity for gradual cooling was larger than for sudden cooling by 3.26%, 8.34% and 9.95% for 300, 500 and 700 °C respectively. Clearly, the retrofitting system adopted here could not return the slabs (FG and FS) to their original strength prior to burning, this because, the retrofitting process including two zones: First, retrofit the tension cord (the damaged steel reinforcement bars due to burning and preloading) by using laminate CFRP sheets. Second, the compression cord, which is the default part of retrofitting, because there is no epoxy with low viscosity be ensure that inter and close all cracks .However , repairing with CFRP provided reasonable enhancement in restoring the original load carrying capacity. These results were due to the serious damages in slabs associated with exposure to high temperature and loading until failure .It is seen that the failure mode of debonding of FRP for slabs subjected to (700)

Mode of failure
In general, most of the tested slabs failed by concrete crushing at mid span and partial debonding of certain retrofitting systems was also observed for a few cases. For all slabs subjected to 300 0 C temperature and slabs FR group, few flexural cracks started first at the constant moment region, also few diagonal cracks near the supports were observed, with increasing applied load, one of these cracks developed diagonally to the nearest loading point. The slabs failed by crushing of the concrete in high moment region on the top surface and debonding was not observed with high elongate as shown in Fig. 12-A similar failure pattern was observed by [Ramanathan,2008] who noted that failure was characterized by compression failure of the concrete in the constant moment region on the top surface of slabs which should be expected for a section having a short effective depth. For all slabs subjected to (500) 0 C temperature ,cracking started in concrete layer between the CFRP and the embedded longitudinal steel reinforcement. The slab failed by crushing of the concrete on the top surface and CFRP debonding was not observed as shown in Fig. 12-B for slab FG20-500. For slabs FG20-700 , FG40-700 and FS40-700, the failure was by critical diagonal crack with debonding of CFRP sheet starting at the critical diagonal crack under point load and proceeded toward the support. The failure was extremely brittle and occurred at one end of the slab emanating from the support as shown in Fig. 12-C for slab FS40-700 and FG40-700. This type of collapse has also been reported as a failure mechanism for the externally bonded FRP on RC one-way For slabs FS20-700 , FG30-700 and FS30-700, diagonal crack originated under an external load point in the region of combined highest moment and shear and propagated toward the support .This crack generated high stresses in the CFRP sheet. Since the concrete could not maintain the interface shear and normal stresses, the CFRP separation was initiated at the location where the diagonal crack contacted the CFRP and propagated in the direction of decreasing moment (toward the support). Failure of the test specimen was by CFRP debonding at the end span with sudden failure by complete separation of the concrete as shown in Fig. 12 Fig. 13 it can be concluded that the specimens with high compressive strength show more stiff behavior than slabs with low compressive strength at all exposure temperature . The effect of exposure temperature was very clear on both slabs FG and FS, Fig. 14 show the comparison between specimens with different exposure temperature at the same compressive strength, the stiffness increased when the temperature decreased. When comparison are made between (FG) and (FS) for all compressive strength at all exposure temperature, it was shown that the stiffness of specimens cooled suddenly are lower than that specimens cooled gradually. This is because of the formation and propagating of the cracks due to the burning process, which had a great effect on the rigidity of the slab specimens, causing a high mid-span deflection. Also, the rate of cooling affects this propagation of cracks, where it increased as the rate of cooling increased, resulting in decrease in the rigidity and increase in deflection. In addition to the more cracks which allowed to higher debonding between the concrete and the steel bars. The ultimate loads of the specimens cooled suddenly were less than the specimens cooled gradually for the same compressive strength concrete because of the more minor cracks in the specimens cooled suddenly by water resulted in reducing in the bonding between the steel bars and concrete

Conclusions
Based on the experimental results and test observations the following conclusions can be drawn: -The test results indicated that a significant gain in flexural strength can be achieved by bonding CFRP laminates to the tension face of RC slabs which were burned and loaded till failure.
-Slabs FR repair with CFRP regained 97.92%, 93.95% and 97.17% of the original load capacity for the specimens FR20, FR30, FR40 respectively while for group FG and FS ,the average percentage of residual ultimate load carrying capacity for specimens (FG) is 78.74, 74.16 and 55.37 at 300, 500 and 700 °C respectively and the average percentage of residual ultimate load carrying capacity for specimens (FS) is 75.48, 65.82 and 45.42 at 300, 500 and 700 °C respectively .So, the average percentage of residual ultimate load carrying capacity for gradual cooling was larger than for sudden cooling by 3.26%, 8.34% and 9.95% at (300, 500 and 700 °C) respectively.
-The ultimate load of the retrofitted slabs (FG and FS) increased with an increase in the compressive strength of concrete and decrease due to exposure temperatures prior to applying FRP sheets.
-The retrofitted slabs exhibited larger deflection than the reference slabs at ultimate loads.
-Slabs with high compressive strength show more stiff behavior than slabs with low compressive strength at all exposure temperatures . The effect of exposure temperature was very clear on both slabs FG and FS, the stiffness increased when the temperature decreased.
-Most of the tested slabs failed by concrete crushing at mid span and partial debonding of certain retrofitting systems was also observed for a few cases.