Strengthening of Reinforced Concrete Members Using BFRP Composites

Abstract

The fiber reinforced polymer (FRP) has emerged as an encouraging material to either rehabilitate the damaged/partially deteriorated/weakened structures or strengthen a sound structural member to fulfill the load requirements due to change in the mode of use of structures or meet the requirements of the updated design guidelines because of their various advantages like superior strength-to-weight ratio, lightweight, excellent durability etc. The available research on the rehabilitation of reinforced concrete (RC) flexural members using externally bonded FRP composites is primarily based on the application of synthetic fibers such as carbon fiber or glass fiber. It may not always be economically viable to use carbon fiber for the strengthening of RC members due to their very high cost. Although the strengthening schemes comprising of glass fiber sheets are economical than that of carbon fiber sheets, the increasing environmental awareness has encouraged engineers to explore more environment friendly materials such as natural fibers as an alternative to the synthetic fibers in the strengthening of RC members. On the other hand, basalt fiber, a mineral-based natural fiber, is getting a lot of attention from the industrial and academic communities. As basalt fibers are comparatively newer to civil engineering concerning the other synthetic fibers and only a few experimental investigations have been documented in the literature on the strengthening of RC beams with basalt FRP (BFRP) composites, an attempt has been made to explore the efficacy of the BFRP composites as an external shear strengthening material for RC flexural members. Total thirty-nine RC beams were examined under the four-point loading system. Thirty nine RC beams were primarily partitioned into two groups depending on the specimen’s cross section, i.e. R and T beams. Out of 39 RC beams, seven RC beams were tested in unstrengthened conditions and recognized as control beams, while the remaining thirty-two RC beams were strengthened with externally bonded fiber sheets in different configurations. The test parameters include shear span to effective depth ratio, type of strengthening scheme, width of BFRP strips, fiber types and orientations, effect of pre-damage level and the effect of end-anchorage schemes. In addition, the impact of six different types of end-anchorage systems, including one novel end anchorage system comprising of wooden plates, on the efficacy of U-jackets was investigated. The experimental outcomes have revealed that the failure load of strengthened beams improved by 5-89%, whereas the toughness enhanced up to 5 times in comparison to the reference specimen. It has been noticed that the degree of enhancement in shear capacity of strengthened RC beams significantly rests on the considered parameters. The overall effectiveness of basalt fiber sheets in increasing the shear capacity is 20% higher than that of the glass fiber sheets, while the pre-loading level has an adverse effect on the efficacy of external strengthening schemes. Although the endanchorage systems positively influenced the shear contribution by postponing or averting the debonding of fiber sheets from the web surface, their effectiveness varies with their location and type. The efficacy of end-anchorage system placed at the web-flange junction was highest among the considered anchorage schemes, followed by the mechanical anchorage system and the anchorage system comprising of wooden plates. The obtained shear contribution of FRP composites was compared with the predictions made by twelve widely used design guidelines. Besides this, a soft- computing tool, i.e. adaptive neurofuzzy inference system (ANFIS), has been explored to predict the shear contribution of FRP composites towards the shear resistance of RC beams. It is evident from the present study that there is a fine concurrence among the ANFIS estimations and the experimental shear contribution of FRP composites as the average ratio of the ANFIS estimations to the experimental outcomes and the R2 value were observed to be nearly equal to one.