A probabilistic approach to the evaluation of the bending strength of timber beams with integration of data from on-site tests

摘要

A probabilistic method to evaluate the bending strength of maritime pine beams (Pinus pinaster Aiton) was explored. The beams were subjected to an experimental campaign that started by allocating them to a visual strength grade and segmenting them lengthwise into clear wood and weak zones. The weak zones are associated with a reduced bending strength that is caused by the presence of knots or grain deviation. Non-and semi-destructive techniques were used to gather information regarding the physical and mechanical properties of the clear wood. Bending tests were carried out until failure to determine the strength of the beams. The method to predict the structural behaviour of the elements was developed under a probabilistic framework based on Monte Carlo simulations in which the reference properties of the beams were assigned randomly based on their probability distributions. The results show that the characteristic strength values assigned to the timber pieces after strength grading only frequently underestimate their strength and that the probabilistic method applied could be used to safely explore the capacity reserve of the timber elements.