The 7th
International Conference on FRP Composites in Civil Engineering
RETROFITTING AND FIRE TESTING OF RC SLABS USING CARBON-
FIBRE REINFORCED CEMENTITIOUS MATRIX (C-FRCM)
地球赤道
Jof SCHERER
Structural Engineer, R&D Manager, S&P Clever Reinforcement Company AG, Switzerland
jof.scherer@sp-reinforcement.ch
Brad ERICKSON
猪和羊合不合Senior Engineer – Composite Strengthening Systems, Simpson Strong-Tie Company Inc., USA
Michael McCULLAGH
Engineer – Composite Strengthening Systems, Simpson Strong-Tie Company Inc., USA
ABSTRACT: This paper prents a C-FRCM system. The reinforcing mesh is made of carbon fibre in the main direction and carbon or glass in the cross direction. The C-mesh has a special coating of amorphous silica to increa the bond between mesh and cementitious matrix. In addition, the paper prents experimental results of the tensile strength decrea of coated carbon fibre rovings after high temperature exposure up to 1000 ºC. In a fire, it is mandatory to retain a certain amount of residual strength after high temperature exposure in order to guarantee structural safety. Additionally, RC slabs were strengthened with the C-FRCM system and subjected to fire under rvice load. Temperature was controlled according to the European ETK-temperature curve. The measured temperature evolution on the carbon fibre rovings together with the previously performed tensile tests lead to the conclusion that the steel reinforcement is partially protected by the C-FRCM layer.
1. Introduction
Fiber Reinforced Cementitious Matrix (FRCM), also known under the designation ‘Textile Reinforced Concrete’ (TRC) (e Curbach and Jes 2011) is known to be an efficient strengthening method for existing reinforced concrete structures. Schladitz et al. (2012) performed static loading tests on RC beams strengthened with an C-FRCM layer and demonstrated a uniform enhancement in the ultimate bearing capacity when increasing the number of reinforcement layers. Ehlig et al. (2010) prent textile reinforced concrete plates tested after fire exposure under sustained load. It is shown that residual bearing capacity is considerable after one hour under high temperature. This paper shows the application of a C-FRCM system applied on a thin RC slab. One and two layers of the carbon mesh have been compared under flexural loading with a control slab. The result showed that ultimate load of the reinforced slab was incread about 70% thanks to the C-FRCM. It also quantifies the reduction in tensile strength capacity of a specific composite reinforcement for C-FRCM application after high temperature exposure. In the end, a high temperature test with a RC beam element retrofitted with a C-FRCM layer aims at quantifying the temperature development in the steel reinforcement, in the carbon fiber composite as well as on the interface concrete/shotcrete.
2. Experimental investigation
2.1. Flexural loading
At the Technical University of Fribourg, Switzerland basic experiments were performed on the C-FRCM-System (S&P ARMO-System). Figure 1 show the test t up.
Fig. 1 - Test arrangement at Technical University Fribourg, Switzerland
Interior r einforcement: longitudinal 6 Ø 12 (S 500) lateral Ø 8 s = 150 (S 500)
A reference slab (Test D0) was compared to a single layer (Test D1) and double layer (Test D2) C-FRCM strengthening using S&P ARMO-mesh L500. As spray mortar the cementitious wet spray mortar S&P ARMO-crete w was ud (Table 1).
Fig. 2 - Deflection curves under total load at mid-span
夏俊艾
for all test specimens
Table 1 - Application of the S&P C-FRCM System
英子的故事As the load-deflection curves (Fig. 2) show, the test specimens show only small differences in structural behavior at the un-cracked state.. The structural behavior of the test slabs change dramatically when the reinforcing steel begins to yield. When the reinforcing steel strain surpass the yield strain then the additional internal tensile forces must be carried primarily by the carbon mesh. The resulting forces and deflections at mid-span can be estimated from the load-deflection curves. The comparisons listed in Table 1 show that the yield loads and the respective deflections at mid-span increa only to limited degree with larger carbon mesh cross-ctional area.
Bad on the increa in deflection w u – w y from the point of the reinforcing steels’ yielding at Q y up the ultimate load Q u reduction factor can be determined for the flexural stiffness EI y at the plastic state of the reinforcing steel with respect to the stiffness at the cracked but still elastic state EI II . This can be compared to the deflection increa for lower working loads.
This clearly shows the influence of the carbon mesh strain stiffness. In contrast to the expected reduction of the flexural stiffness given the prent degree of reinforcement by a further factor of roughly 8 as compared to an unreinforced concrete girder, the reduction for Test Slab D1 is roughly below 2. For Test Slab D2 the stiffness is reduced by only about 25%. This very favourable influence of the carbon mesh is confirmed in the incread ratios Q u / Q y of roughly 25% per layer of carbon mesh.
2.2. Degrees of strengthening
Bad on the ratios of the ultimate loads Qu of the reinforced slabs the degree of strengthening by carbon mesh can be determined in comparison to the reference slab (Table 1). The values show that a doubling of the carbon mesh’s cross-ction area also results in a doubling of the additional load that can be borne until failure. For the tested structural system this results in a load increa of about 16.5 kN or 35% per layer of carbon mesh.
Table 2 - Characteristic values of the ultimate state and the yield state of the reinforcing steels
Q y [kN] w y [mm] EI y /EI II Q u [kN] w u [mm] Q u /Q y Q rest [kN] w remainder [mm] D s = w u /w y D0
外星传奇38.7 100% 61 7.98 47.2 100% 176.5 122% -- -- -- 2.89 100% D1
42.1 109% 62 1.95 63.8 135% 141.7 152% 52.4 82% 162.8 2.29 79% D2 44.4 115% 58 1.34 80.0 170% 132.7 180% 56.2 70% 155.1 2.29 79%
2.3. Crack patterns and strains in the tensile zone
Fig. 3 shows the crack patterns of all test girders at failure as well as the maximum load and the measured strain along the tensile edge.
The crack patters show a sufficiently fine distribution of the cracks for all experiments. For Reference Test D0, the crack spacing corresponds roughly to the distance of the lateral reinforcing as is to be expected.
Technical d ata:
S&P A RMO-‐mesh L 500 Elastic m odulus
≥ 240 k N/mm 2 Tensile s trength
≥ 4’300 N /mm 2 Weight o f C -‐fiber i n m ain d irection
187 g /m 2 Density
1.79 g /cm 3 Elongation a t r upture
京杭大运河始建于1.75 %
Technical d ata:
S&P A RMO-‐crete w Compressive s trength E N 12504-‐1
> 45 N /mm 2 Bond s trength E N 1542
≥ 1.5 N /mm 2 Elastic m odulus S IA 262/1-‐G
≤ 40’000 N /mm 2 Shrinking S IA 262/1-‐F ᵋcs(28) ≤ 0.80 ‰
For Test D1 and D2 with spray mortar strengthening, the tendency is for the crack spacing to be somewhat reduced. This is also to be expected due to the slight stiffening effect of the strengthening.
2.4. Failure behavior and residual resistance
Definitive failure principally occurs for all test slabs as compressive failure in the concrete compression zone. None of the slabs fail becau the reinforcing steel is pulled apart. Both the bond between spray mortar layer and concrete surface as well as the carbon mesh cross-ction area remained intact for both of the strengthened slabs when failure occurred in the bending compression zone.
In the load/elongation diagram (Fig. 4), the measured strain on the concrete tension side is shown. The elongation in the C-FRCM was not measured. The maximum elongation on the concrete of 0.8% was ud to calculate the theoretical elongation and stress in the C-FRCM mesh. The results showed clearly that there is a certain elongation required to activate the mesh. Therefore, in the design concept the theoretical modulus of elasticity is reduced by the factor 1.5 on one side. On the other side, the ultimate limit strain in the C-FRCM mesh in flexural enhancement is determined at 0.5
Fig. 3 -‐ C rack p atterns
Fig. 4 -‐ L oad/elongation d iagram 3. High temperature exposure at EMPA, Switzerland
The composite mesh type, namely C-FRCM mesh was tested at EMPA Dübendorf, Switzerland and at VSH Versuchsstollen Hagerbach, Switzerland in the current investigation. In longitudinal direction, 58.5 carbon strings per meter are ud for tensile force transfer. Since one string is compod of two roving. In theory, the unidirectional tensile resistance is in the range of 500 kN/m, which gives an average value of
4.27 kN per roving. The roving specimens were cut form the original mesh structure, as shown in Fig.
5. In transver direction, E-glass fibers are ud as a fixing element, but are not taken into account for any structural purpos.
Fig. 5 – C-FRCM mesh
寓言小故事
Fig. 6 - a) Tube furnace b) end closing of the tube
3.1. Tube furnace
For high temperature exposure, the specimens were installed in a tube furnace (e Fig. 6a) for 30 minutes or more. It is emphasized that the specimens had already entered the furnace during the heating-up period. Hence, a higher final temperature also involves a longer total temperature exposure and thus more conrvative results. Only a central gment with an approximate length of
40 cm was submitted to the high temperatures, whereas the remaining roving ends were kept at room temperature (e Fig. 6b). 3.2. Clamping configurations and test tup
An initial specimen length of 1600 mm was considered. The rovings were revolved three times around the rolls at both ends and eventually fixed with a clamping device. This procedure was performed according to the ISO 3341 (2000) testing recommendation. As a comparison, the reference tests were repeated with an epoxy coating and a special metallic clamping. Extensometers with a ba length of 150 mm were ud to record the vertical displacement subquently transformed into a tensile strain. A load cell measured the applied loading force. All the tests were conducted at room temperature (approx. 21°C) at a displacement velocity of 5 mm/min. An example of the test tup with roll clamping is given in Fig. 7a.
Fig. 7 - a) Test tup with roll clamps and - b) metallic clamps and epoxy-coated specimen
4. Results and discussions
4.1. Tensile tests
The results in terms of force-strain behavior are prented in Fig. 8. Ultimate tensile forces of the different configurations are summarized in Table 3. All specimens revealed a linear tensile behavior to failure, with the difference that the uncoated specimens need an initial time span before the real stiffness is developed. It can be obrved that an epoxy coating together with a metallic clamping configuration delivers higher tensile resistances and higher strains at failure than the roll clamping method, a relative increa of 85% in force. This clear enhancement is due to a more equal stress transfer in the rovings when an epoxy coating and metallic clamping are ud.
Exposure to an incread temperature up to 1000°C has a decreasing effect on both ultimate tensile force and strain at failure. A temperature of 300°C does not em to have any effect on the bearing capacity, the measured average ultimate force being even higher than the reference one (e Table 3). A further increa to 500°C of exposure during 30 minutes involved an average reduced strength of 1542 N. Both ries with 700°C and 1000°C heating temperature destroy the roving, leaving no no
挤公交车ticeable remaining tensile capacity left. Average forces in function of the exposing temperature prior to testing are given in Fig. 8.
Table 3 - Summary of the failure forces for all the tested configurations
Metallic c lamp Roll c lamps
Temperature e xposure [°C] RT RT 300 500 700 1000
Failure f orce [N] 4938 2664 2805 1542 60 19
Standard d eviation [N] 323 182 198 347 12 12
Fig. 8 - Force-strain curves for all the test specimens under different temperature configurations
Fig. 9 - Left: Ultimate tensile resistance for the C-FRCM mesh in function of temperature exposure during 30 minutes (roll
clamping)
Right: Temperature evolution in the steel, at the interface and in the mesh
4.2. Temperature tests of RC slabs with C-FRCM at VSH Hagerbach, Switzerland
In the testing gallery VSH Hagerbach, a RC slab (lower tensile reinforcement 5 ∅ 8 mm) with an C-F
RCM bi-mesh was submitted to a fire exposure for a determined time span. The C-FRCM bi-mesh is a bi-directional carbon-roving grid with a theoretical failure force of about 200 kN/m.
Fig. 10 prents the retrofitted concrete element. 20 mm of the initial cover of 30 mm is hydro-mechanically removed; the remaining reinforcement cover of 10 mm was subquently enhanced to 20 mm by applying a first layer of wet shotcrete of the type FRCM-crete. Afterwards, two layers of C-FRCM bi-mesh are introduced on the first shotcrete layer. Finally, a cond shotcrete layer with a thickness of 20 mm is put in place. Hence, the total concrete/shotcrete cover thickness with respect to the steel bars is 40 mm.
Fig. 10 - Retrofitting procedure for the C-FRCM bi-mesh (Upper picture: initial situation,
lower picture: first shotcrete layer + C-FRCM bi-mesh + cond shotcrete layer)
The strengthened RC slab is subquently submitted to a high temperature exposure from below as prented in Fig. 11. The goal of this investigation was to asss the temperature evolution in the steel
