Influence of proof loading on the reliability of members

Proof loading concept is a recognized quality control technique to improve the characteristics of the lower tail of strength distributions of structural timber products. Very few comprehensive studies are available to quantify the effectiveness on the use of proof loading relating to the choice of proof load level, the potential damage on the members resulting from the application of proof load, and the improvement of performance in the context of reliability based design methods. One of the difficulties is the need of a rather large sample size for an experimental-based study to develop statistically meaningful solutions. This paper illustrates the use of damage accumulation model and reliability based design analyses to quantify the effectiveness of proof loading. The performance of No. 2 and better Western Hemlock 38×140 mm dimension lumber in bending is considered. Damage accumulation laws are established to consider the residual strength of members that survived a proof load. Reliability analyses are conducted to compare the performance of proof-loaded and non-proof loaded members subject to snow load conditions in two locations in Canada. For a given reliability index (β), the improvement of performance can be quantified as characteristic strength adjustment factors for proof loading in terms of the ratio of the performance factors (φ) between the original and proof load material. Conversely the gain in reliability, β, for a given φ value is also apparent. The adjustment factors depend on the proof load level, the β level, and the distribution selected for fitting the strength data after proof loading.