Professor of Mechanical Engineering, The University of Michigan-Dearborn, Dearborn, MI
(Received 17 April 1989; accepted 7 September 1989)
The minimum variance strategy for specimen allocation in estimating median fatigue limits using small sample sizes is explained. Its implementation using microcomputer programs to select automatically the most effective stress amplitude for testing the next specimen is discussed. Applications involving modified up-and-down tests and estimation of strength percentiles are presented. An appendix illustrates strategies effective in preliminary tests.
Paper ID: JTE12461J