Abstract
The inhomogeneous nature of polycrystalline graphite requires property measurements to be made over dimensions large enough to average the local variations in the structure. This is particularly true for mechanical integrity, and experimental data are presented which illustrate the importance of the real aggregate structure of graphite and the difficulties of interpreting strength data from different tests. The classical statistical treatments do not hold generally, and the problem of defining a failure criterion for graphite is discussed. It is suggested that the stress conditions in graphite components might be classified in terms of the dimensions and stress gradients related to the characteristic flaw size of the material as determined experimentally.
view the full text of this article (10 pages, format: PDF, size= 646kB)