Visualization of voids in actual C/C woven composite structure

Void statistic is an important C/C composite parameter since the voids presence decreases the predicted properties. On the other hand minimum number of voids is needed for medical application. Void study supposes finding its exact 3D shape. Since the void distribution is statistic, many samples must be processed. Therefore automated methods are necessary. In the paper the set of plan-parallel structure images were scanned by optical microscopy. The void contours on 2D images were found automatically by the use of MATLAB procedures with the same accuracy as hand marking. The 3D void surface was created by small triangles from its plan-parallel cuts. The 3D void was studied by virtual reality that makes possible almost all ways of inspection.