Selection and Validation of Experimental Specimens for Determining Model Inputs for Matrix Compression Damage Propagation in Fiber Reinforced Composites

Abstract

The matrix compression parameters in continuum damage mechanics models are often calculated from assumptions requiring detailed knowledge of fracture and frictional behavior of the material. A direct method for experimentally determining the matrix compression damage parameters would simplify the process and potentially increase accuracy of parameters. In this study, specimen geometry is determined and validated to directly evaluate the energy dissipated by matrix compression damage experimentally. Initially, candidate specimens were determined from literature on fiber compression testing including compact compression (CC), center notched compression (CNC), and four-point bending (4PB) specimens. These specimens were modeled using the finite element program ABAQUS. Geometry and boundary conditions of the specimens were varied to test different specimen geometries and fixture types. The matrix compression damage isolation of the specimens was evaluated by measuring the region of matrix compression damage without other damage modes present in the material. 4PB specimens showed damage primarily at the loading points. Both CNC and CC specimens showed fairly good isolation of matrix compression damage with the former showing a tendency to split off-axis depending on the fixture used and the latter showing tensile splitting after significant compressive damage growth. CC specimens were selected because they require less complex loading fixtures and are less dependent on the fixtures for damage isolation than CNC specimens. The geometry was varied on the CC specimens to increase the compression damage isolation. Specimens were manufactured for experimental validation. A layup with the fiber direction of all plies parallel to the notch tip was used to isolate the loading to the matrix. It was determined that 20 plies was sufficiently thick to prevent bucking. The specimens showed good isolation of compression damage at the notch tip. This is due to the stress concentration at the notch tip that lowers the load required to cause compression damage to below the global buckling load. Ultimate failure was due to tensile splitting opposite of the notch, but only after sufficient compressive damage growth. The size of the damage zone was able to be tracked visually from video of the tests. Load-displacement data and the damage zone size were used to calculate the strain energy release rate using the basic compliance calibration method. This method is limited because it is based in fracture mechanics principals and may not be accurate for the damage modes present in the material, but is sufficient for initial validation of the CC specimen. The strain energy release rate for the material was measured to be 35 in-lbs./in2. CC specimens show promise for measuring the energy dissipation of matrix compression damage for use in continuum damage mechanics models due to their ability to isolate compressive damage modes without buckling. Refined data collection methods can be implemented to increase the accuracy and generality of the strain energy release rate measured.