Experimental and Design Evaluation of Multi-bolted GFRP Joints

Abstract

For infrastructure system implementation, high strength/stiffness to self-weight ratio, corrosion resistance and ease of fabrication are just a few of the many advantages of pultruded fiber reinforced polymer (PFRP) composites over traditional construction materials. However, challenges of engineered FRP materials are addressed via continuous updating of design codes. For example, the anisotropic nature of FRP composite materials includes different mechanical properties in different directions (pull or longitudinal direction vs. transverse to pull direction) of a structural member, resulting in design complexities over isotropic materials. Typically, infrastructure systems are built by bolting FRP structural members because of certain advantages of bolted joints including higher joint efficiency over adhesive joints. Advantages such as ease of field installation and inspection for structural integrity, and excellent clamping capability at a joint excel high stress concentrations around joint, differing thermal coefficients and susceptibility to corrosion with steel bolts. In general, connections of pultruded FRPs are designed and constructed as simple framing joints that are pin connected. Stress concentration is induced around a bolted location due to sudden changes in stiffnesses of gusset plate(s) connecting the FRP members and the loss of FRP material to create connector holes. Orthotropic nature of FRPs leads to different failure modes. This study is mainly intended to evaluate doubler and tube-in-tube concepts for their feasibility and effectiveness of multi-bolted joints in composite under different static loading to ensure load/moment transfer and minimize the stress concentration in joints. Experimental and theoretical evaluations of this study will provide fundamental understanding about failure modes and joint efficiency of bolted joints in FRPs. Moment transfer efficiency of splice joint specimens with tube-in-tube concept under bending was investigated. These specimens consisted of tight fit square tube sections in a way that a smaller profile with a specific length was connected to a telescopic two-equal-length-component beam member. A 2.5×2.5 in. section overlap was multi-bolted (two bolts at each side of the joint) at mid-span of the beam made up of a 3×3 in. profile. Joint efficiency of the specimens was evaluated under 41-in. span in three-point bending tests. Average joint efficiency in bearing failure mode was evaluated ~23%, over the 3×3 in. beam member with no discontinuity (i.e., no joint). Joint efficiency in bearing of the same configuration was investigated by two methods with (a) adding a bottom plate to, and (b) wrapping of the splice joint. Bottom plate redirected the joint to have a controlled