ABSTRACT
Like any other experimental technique, a successful physical model test should generate reliable (i.e., consistent and repeatable) quantitative data. Owing largely to cost and time constraints, model tests are rarely repeated to assess the reliability of the test results, although several researchers have discussed this issue (e.g. Corte et al. 1988; Arulanandan et al. 1994; Fragaszy et al. 1994).
