Reproducibility of liquid oxygen impact test results

Results for 12,000 impacts on a wide range of materials were studied to determine the reproducibility of the liquid oxygen impact test method. Standard deviations representing the overall variability of results were in close agreement with the expected values for a binomial process. This indicates that the major source of variability is due to the go - no go nature of the test method and that variations due to sampling and testing operations were not significant.