PSI - Issue 54

Ahmed Selim et al. / Procedia Structural Integrity 54 (2024) 601–608 Selim et al./ Structural Integrity Procedia 00 (2019) 000 – 000

605

5

other types of FRP could have been employed interchangeably. The primary aim here is to compare the methods of fiber installation, rather than evaluating the inherent strength of the fibers themselves. The density of CFRP was kept at 1500 kg/m 3 which is within the typical range of CFRP densities from literature (Li et al., 2017). The elastic modulus and the used Poisson’s ratio were defined as 62 GPa and 0.22, respectively, which are typical values. 2.3. Constraints, boundary & loading conditions All the displacements at both supports were completely restrained to prevent any translational motion or rotation in all directions. The two rigid plates applying the two-point loads were restrained in all directions except in the y direction, U2. A static displacement-controlled loading condition was assigned. Reference points were assigned on the rigid plates as the point of loading. A linearly increasing displacement of 15 mm with 1 second period and 0.01 initial step increment were defined as the loading condition. A penalty formulation contact property with a friction coefficient of 0.15 was defined between surfaces of the support, loading plate and FRP for the first model. On the other hand, a higher coefficient of friction was used for the second model to simulate the enhanced bonding behavior of the prism. This value was equal to 0.7. Additionally, normal hard contact was defined between surfaces to prevent penetration and premature failure of prisms during loading application. 2.4. Mesh configuration & model verification An initial assumption of a mesh size of 10 to 20 mm was used based on the literature (Kyaure et al., (2021), Abed et al., (2021)). However, after carrying out more than 15 iterations, mesh sensitivity analysis revealed that a mesh size of 25 mm for concrete and 15 mm for FRP was found ideal for a reasonable computational time. The computational time for prisms was reasonable as expected because of the absence of any embedded reinforcements in the prisms and because the sizes of the prisms are relatively small. Fig. 3 and 4 shows the developed prism models.

Fig. 3. Prism externally bonded with FRP laminate model geometry: complete assembly with mesh configuration

Fig. 4. Prism externally bonded with near surface mounted FRP laminate model geometry: complete assembly with mesh configuration