Experimental and numerical analysis of structures with bolted joints subjected to impact load

The aim of this study is to analyze the transient behavior of structures with bolted joints subjected to impact or shock loads using experimental methods and Finite Element Analysis (FEA). Various factors that affect the response of the bolted joint structures for shock loading were studied, such as damping, preload, intensity of impact load and type of FE modeling. The objective of this work was to develop computational modeling procedures that provide structural analysts an improved physics-based shock model for combat vehicles focusing mainly on shock transmission across bolted joints. There is only a limited amount of published literature describing the proper method for analyzing the transient shock propagation across bolted connections for high impact loading. The initial case study focused on a simple cantilever beam with bolted lap joint subjected to relatively low levels of impact force. The second case study used a flat plate bolted to a hat-section and the third structure evaluated was two hat sections bolted together. These simple configurations are representative of structures found in many military ground vehicles that can be subjected to transient impact and blast loads. These structures were subjected to low impact loading (non destructive) using impact hammers and high impact loading (destructive) using an air gun and their responses were measured using accelerometers. LS-DYNA FE solver was used to simulate the shock propagation in bolted structures.;For all the bolted structures, the modal analysis was performed both experimentally and numerically. The results were in excellent agreement for lower modes and small deviation in higher modes. Secondly, the time history response of experimental and FE analysis are compared. Normalized Root Mean Square Deviation (NRMSD) criterion was used to compare the experimental and FE result. A full detailed FE model and a simplified FE model of the bolted structures were developed for impact analysis and their prediction were compared with the experimental results. In all the cases, the detailed FE model with 3-D solid elements showed good agreement with the experimental results. The simplified FE model with shell elements (bolts were not modeled) predicted higher magnitudes in the acceleration values. Addition of damping in the simplified FE model reduced the higher magnitudes in the predicted response and the results were in good agreement with the experiment. The simplified FE model developed for bolted joint structure in this report reduced the CPU time by one order (30 hours to 3.5 hours) and can be practically implemented in the full vehicle FE model for crash or blast analysis.