TECHNICAL NOTE: LOWER TOLERANCE LIMIT APPROACH TO EQUATION-BASED RATIONAL DESIGN VALUES FOR T-SHAPED MORTISE AND TENON JOINTS

Abstract

A nonlinear regression expression was fitted to the test data obtained from a study of the bending moment capacity of 320 rectangular T-shaped mortise and tenon furniture joints consisting of 64 configurations of five specimens each. A statistical lower tolerance limit approach was then used to explore the degree to which these values should be reduced when used for design purposes and the confidence that a user might have in these reductions. The procedure followed was to apply statistical lower tolerance limit techniques to the ratios obtained by dividing each test value by its corresponding estimated value. To gain insight into the relationship of a specific confidence proportion level and its corresponding reduction factor on the percentage of an estimated value that could be used for design purposes, lower tolerance limits were computed for four confidence proportion levels. The results illustrate a statistical technique that can be used to determine reduction factors and the impact of the selection of any of the given confidence proportion levels on design values.