A Randomization Test for Comparing 1/4-Scale Kinetic Energy Penetrators
ARMY BALLISTIC RESEARCH LAB ABERDEEN PROVING GROUND MD
Pagination or Media Count:
Material properties of kinetic energy penetrators are compared at the Ballistic Research Laboratory in a 14-scale test environment. Metallurgists fire penetrators of various material compositions into semi-infinite steel blocks and record depths of penetration. Depth of penetration behaves approximately as a linear function of velocity, dv, over the range of the four-velocity design routinely employed. Under a common slopes assumption, a difference in performance between penetrators k and l is computed as d sub kv - d sub lv. This difference is determined graphically, occasionally with the benefit of a least-squares fit to each performance. Statements of significance are not made at present The approach presented is an attempt at a unifying structure within which inferences in this environment can be made both quantifiable and consistent. The recommended procedure combines existing techniques such as least squares with a new application of a randomization test in determining the significance of observed material differences. With this test supporting practitioners can make definitive statements as to the statistical significance of material differences observed.
- Statistics and Probability