73 same specifications, and the resulting clearance is not consistent with typical manufacturing practice for critical structures although it is not unusually high for non-critical factory or field assembly. The tightening torque of 35 N-m was applied by a digital torque wrench. After assembly, the samples that were selected for the hybrid joint were injected with SC-15 resin through the channels in the bolts so as to fill the clearance space between the threaded bolt shank and the work pieces. The injected inserts were cured by a convection oven at 60 °C for 2 h followed by post curing at 94 °C for 4 h. Figure 8.1b shows a cross-section of the bolted joint, wherein the channels are exaggerated for better visibility. As can be seen, the resin fills all the intricate areas including the threads of the bolts and, upon curing, creates a monolithic structure. 8.4 Experimental Setup All quasi-static testing was performed using a 100 kN load cell servo-hydraulic testing machine (MTS 810), and a typical test setup is shown in Fig. 8.2. The machine is equipped with a standard load cell and an internal crosshead displacement measuring extensometer. An external laser extensometer was also used to measure the joint longitudinal displacement, as shown in the figure. The tests were performed according to ASTM STP1455-EB. In this work, five tests were carried out for statistical significance. The average shear strength of the joint was calculated by dividing the maximum load reached in each test divided by the cross-sectional area of the bolt. 8.5 FEM Modeling 8.5.1 Modeling and Meshing In this study, the commercially available finite element software ABAQUS® was used to obtain the stress distribution and damage initiation near the holes in the bolted composite coupons. The overall dimensions for modeling of the GFRP with a drilled hole were obtained from the specimens used for the experimental tests. Eight-node quadrilateral in-plane general-purpose continuum shell elements with reduced integration and hourglass control were used to model the 16 ply plain-weave composite plates, [0]s8. A minimum of three integration points using the Simpson integration rule were assigned at each layer of continuum shell elements. As is shown in Fig. 8.3a, the specimen was partitioned at specific sections to apply boundary conditions and localized meshes, especially around the hole, where they need to be refined to capture more accurately the steep stress gradients there. During the meshing process, the bias method was implemented to allow the element size to be varied with respect to the areas near the holes: the nearest element to the hole has the smallest element size, as shown in Fig. 8.3a. Figure 8.3b shows the completely assembled lap joint as modeled. Fig. 8.2 Experimental setup 8 Numerical and Experimental Characterization of Hybrid Fastening System in Composite Joints
RkJQdWJsaXNoZXIy MTMzNzEzMQ==