Validity of Distortion‐Energy‐Based Strength Criterion for Timber Members

Abstract

The validity of an interaction formula, based on distortion energy principles, in the design of nonprismatic glulam members is examined. The distortion energy distributions in various nonprismatic members tested to date are determined by utilizing an orthotropic trapezoidal finite element method of analysis. The distortion energy at the location of failure initiation in all the test specimens was found to be significantly lower than that at the critical location on the tapered edge. In full-scale members failing in radial tension, the distortion energy at these two locations was found to differ by a factor ranging from 12 to 23, and in those members failing in bending, the factor ranged from 1.5 to 7. In double-tapered members, the redistribution of stresses caused by the apex was found to result in low distortion energy levels throughout the apex region, and failures triggering in this region cannot be attributed to distortion energy reaching a threshold value. The use of a distortion-energy-based strength criterion for nonprismatic glulam members is not supported by the results of this study, nor by the observed ultimate behavior of the members tested to date.