Abstract
Some FRP confinement models available in the literature are based on standard cylindrical specimens and others are based on mixed sizes of cylindrical specimens. The accuracy of the latter models is questionable, as it depends on the percentage of increase in strength between unconfined and FRP-confined specimens and on the ratio of strength increase among the different sizes of specimens. The question which can be raised here is: is there a need to introduce a size factor for the test results which are based on non-standard sizes of cylindrical specimens before using them in developing analytical models for FRP-confined concrete? The output of this study answers this important question. Thirty-seven concrete cylinders with three different sizes were experimentally tested. Of these, 13 cylinders were control specimens, to be used as baseline for comparison, whereas the remaining 24 cylinders were wrapped with layers of CFRP jacket. Studied parameters were specimen size and confinement stress ratio. In addition to the experimental investigation, non-linear finite element analysis was also carried out using LS-DYNA software. The predictions of the numerical finite element models were found to agree well with the experimental results of the specimens tested in this study in addition to others selected from the literature. Based on this validation, the effect of specimen size on FRP-confined concrete cylinders was further investigated numerically taking into consideration various confinement ratios and cylinder sizes. The results show that the effect of specimen size on FRP-confined concrete is insignificant. (C) 2011 Elsevier Ltd. All rights reserved.