Carbon foam characterization: sandwich flexure, tensile and shear response

The focus of this research is characterizing a new material system composed of carbon and graphite foams, which has potential in a wide variety of applications encompassing aerospace, military, offshore, power production and other commercial industries. The benefits of this new material include low cost, light weight, fire-resistance, good energy absorption, and thermal insulation or conduction as desired. The objective of this research is to explore the bulk material properties and failure modes of the carbon foam through experimental and computational analysis in order to provide a better understanding and assessment of the material for successful design in future applications. Experiments are conducted according to ASTM standards to determine the mechanical properties and failure modes of the carbon foam. Sandwich beams composed of open cell carbon foam cores and carbon-epoxy laminate face sheets are tested in the flexure condition using a four point setup. The primary failure mode is shear cracks developing in the carbon foam core at a critical axial strain value of 2,262 ??. In addition to flexure, the carbon foam is loaded under tensile and shear loads to determine the respective material moduli. Computational analysis is undertaken to further investigate the carbon foam's failure modes and material characteristics in the sandwich beam configuration. Initial estimates are found using classical laminated plate theory and a linear finite element model. Poor results were obtained due to violation of assumptions used in both cases. Thus, an additional computational analysis incorporating three dimensional strain-displacement relationships into the finite element analysis is used. Also, a failure behavior pattern for the carbon foam core is included to simulate the unique failure progression of the carbon foam on a microstructure level. Results indicate that displacements, strains and stresses from the flexure experiments are closely predicted by this two parameter progressive damage model. The final computational model consisted of a bond line (interface) study to determine the source of the damage initiation, and it is concluded that damage initiates in the carbon foam, not at the bond line.