Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Based on test results from this research, and review of relevant literature, the following conclu- sions can be made for each of the studied topics: 1. Surface Roughness: a) Surface roughness did not significantly influence the performance of wet lay-up FRP sys- tems (with or without adequate anchorage) and whether FRP systems fail from debonding or rupture. b) The three groups of wet lay-up specimens exhibited different modes of failure. Specimens with no straps failed by debonding of FRP sheet, specimens with four and seven straps failed in the same way after the straps were ruptured, and specimens with 11 straps or full con- tinuous straps failed by rupture of FRP sheets. c) The higher level of anchorage did not affect the strength as much as it affected the ductility for the retrofitted beams. d) The smoothest concrete surface profile in this study (CSP 1) as defined by ICRI/ACI (1999) appears to provide an adequate surface roughness level for both wet lay-up and pre-cured FRP systems. However this study recommends CSP 2â3 to be conservative. 2. Surface Flatness: a) An out-of-flatness on the concrete surface in the form of 1â8 in. (3.2 mm) over a 12-in. (305-mm) length had a significant effect on the overall performance of the retrofitted beams. However, behavior of beams with 1â16-in. (1.6-mm) valleys was quite similar to that of control (level) specimens. Therefore, 1â16 in. (1.6 mm) over a 12 in. (305 mm) length seems to be an appropriate threshold for valleys and depressions. b) Peaks on concrete surface were found to be less critical than valleys of the same size. NCHRP Report 514 (Mirmiran et al. 2004) does not recommend threshold values for these peaks, and the results from this study do not support any changes. c) Bond reduction factor, which is a function of maximum strain in FRP at debonding, decreases with the increase of surface curvature or rise-to-chord ratio. d) All surface flatness specimens (wet lay-up or pre-cured) failed by debonding of FRP, regard- less of the type and extent of the out-of-flatness. FRP debonding in specimens with valleys was initiated at both ends of the valleys and propagated to both ends of the specimens. 3. Surface Voids: a) All specimens with surface voids and four straps failed by debonding of FRP that initiated at the mid-span and propagated toward the supports until one of the U-straps ruptured. Use of higher anchorage (i.e., 11 straps) increased load capacity and led to failure by FRP rupture. b) Load-deflection and load-strain responses of all specimens were quite similar, indicating that surface voids in the range studied do not significantly affect the overall behavior of FRP sys- tem, although there was a slight reduction in load capacity with an increased void diameter. 6 C H A P T E R 3 Conclusions
c) Using response surfaces with the design of experiment analysis, void depth appears to have no direct effect on the response of FRP system. The recommended 1â2 in. (12.7 mm) as a threshold for void diameter, as specified by NCHRP Report 514 (Mirmiran et al. 2004) appears conservative. 4. Surface Cuts: a) Cracking in concrete surface simulated as surface cuts did not appear to significantly affect the response of FRP retrofitted beams. Moreover, stress concentrations due to preload- ing cracks do not significantly affect the overall performance of FRP-retrofitted beams. Based on tests of 20 specimens, it appears that recommendation of NCHRP Report 514 (Mirmiran et al. 2004) for injection of cracks wider than 0.01 in. (0.25 mm) is conservative. A threshold value of 1â32 in. (0.8 mm) is recommended instead, primarily for durability con- cerns rather than structural reasons. The NCHRP Report 514 (Mirmiran et al. 2004) does not recommend a threshold for crack spacing. This study showed the crack spacing of 1.5 in. (38.1 mm) to be an appropriate threshold for crack injection, if crack width is between 0.01 in. (0.25 mm) and 1â32 in. (0.8 mm). b) FRP debonding was observed in all wet lay-up and pre-cured specimens except for when number of U-straps was significantly increased. Debonding first occurred at the mid-span and then propagated towards the supports. Failure occurred after the U-straps ruptured. An additional number of U-straps provided sufficient anchorage for the FRP system to prevent premature debonding. 5. NSM Groove Size a) A tolerance of ± 1â8 in. (3.2 mm) in the groove size of 9â16 in. (14 mm) seems to have no sig- nificant effect on the performance of FRP systems with NSM strips or bars. b) For both NSM strips and bars, the dominant failure mode changes from epoxy splitting to concrete splitting as the groove size is increased. 6. FRP Defects: a) The guidelines on the effects of FRP defects contained in the NCHRP Report 514 (Mirmiran et al. 2004) and the ACI 440.2R-02 (2002) are conservative. Disbonds and cracking influ- ence the bond performance only at the local level and do not significantly affect the overall structural response of the system. However, recommendations of NCHRP Report 514 (Mir- miran et al. 2004) are maintained for durability purposes. b) In general, disbonds located in the high moment regions are slightly more critical than those located within the shear span. c) Disbond locations, whether at the interface between the FRP layers, or within the adhesive, or at both locations should be treated equally, as they all result in very similar performance characteristics. 7. Environmental Conditions during Installation: a) Surface moisture, when measured by moisture-meter, should be limited to 4.3% to provide satisfactory bond performance. b) Relative humidity should be within 52% and 65% to provide satisfactory bond performance. c) Temperature should be within the range of 50° and 90°F (10° and 32°C) to avoid high viscosity of saturant and primer at low temperatures and rapid set of saturants at high temperatures. The above conclusions have been incorporated into the Recommended Construction Specifi- cations and Process Control Manual, as presented in Attachments A and B, respectively. Finally, it should be noted that the long-term behavior of FRP defects (due to improper surface preparation, FRP application, or environmental conditions during installation) under sustained loads, fatigue loads, aging, or severe environmental exposure has not been investigated in this study. Research in this area is needed to validate the findings and recommendations of this research. Conclusions 7
