- Journal
- Année
- 2014
- Pages
- 1454-1456
Abstract
Southwest Research Institute (R) has been successful at performing finite element simulations of impact on fabrics by following a rigorous validation approach with feedback from tests performed in the laboratory. The validation procedure was presented in [ 1] and consisted of systematically testing and simulating different scale levels in the fabric, starting at the yarn level, and following with one layer and multi-layer targets. To ensure that the proper physics was being captured by the simulations the highly instrumented tests allowed measuring: 1) transverse wave velocity and critical velocity 1 for single yarn impact, 2) transverse wave velocity for single-layer impacts, 3) pyramid apex velocity and transverse wave velocity, as well as ballistic limit for multi-layered impacts. Additionally nickel-chromium wires embedded in the fabrics allowed following strain waves in the fabric. All these parameters were then compared to the results from the simulations. The logical next step was to add matrix to the fabric material to perform computations of impact on composite. During the project presented in this paper around 40 Kevlar/ resin composite panels were tested with the.30 cal FSP. The diagnostics used during the tests were: 1) Two high-speed video cameras to capture deflection and strain in the back of the target for digital image correlation and an additional top high(speed camera to measure residual velocities, 2) An ultra(high(speed Imacon camera to measure the transverse wave velocity on the panel, 3) nickel(chromium wires embedded in the targets. This paper focusses on the simulation results and how they compare to the deflections and ballistic limits measured during the lab tests. First the mesh geometry will be described and discussed, see unit cell in Figure 1a), as well as the material properties used which are all based on the literature or laboratory characterization tests. Then the simulation results for different impact velocities will be compared with the deflection vs. time history recorded in the tests. Various sensitivity studies were performed for the influence of the target size, and matrix and yarn properties. For example Figure 1b) shows the influence on the results for 5, 10, and 15(cm wide targets. The symbols are the data from the test obtained by measuring the pyramid width from images like the one in Figure 2a). The width of the pyramid is also measured in the computations, as in Figure 2b). Figure 1b) shows that to match the experimental data a minimum target size is required. This was expected as in small targets the release wave comes back from the boundary much earlier and the pyramid expansion stops. Ballistic limits for two different thicknesses were predicted by the model within a 10% error. The yarn material properties used were the same presented in [1] and previous conferences.