Size Effect of Large Glued Laminated Timber Beams - Contribution to the Ongoing Discussion

Abstract

The effort to experimentally test glued laminated timber (GLT) beams is tremendous, especially for large beams. Such beams are increasingly used to realise wide-span timber structures, but experimental investigations of large beams are missing. In the literature, numerical simulation studies can be found instead to estimate the influence of the beam size. However, these studies come to partly different conclusions regarding the size effect, which confirms the high influence of the modelling strategy. Therefore, we conducted an extensive simulation program covering more than 8000 simulations to research the size effect, including large GLT beams of up to 3300 mm depth. The developed modelling concept considered the morphology of timber boards deterministically and accounts for discrete cracking to simulate the bending strengths of GLT beams subjected to four-point bending tests. We predicted the size effect on the characteristic bending strength for two commonly used strength classes. We found that the strength decreased with increasing beam size. The results showed that the strength decrease is not only caused by the beam depth but also the length. Furthermore, the influence of different global failure criteria was investigated, with which the results from the existing studies could successfully be reproduced.