Over the last two decades many constitutive models with different degrees of accuracy have been developed for analysis of sawn timber and engineered wood products. However, most of the existing models for analysis of timber members are not particularly practical to implement, owing to the large number of material properties (and associated testing) required for calibration of the constitutive law. In order to overcome this limitation, this paper presents details of 1D, 2D and 3D non-linear fi nite element (FE) models that take advantage of a quasi-brittle material model, requiring a minimum number of material properties to capture the load-defl ection response and failure load of timber beams under 4-point bending. In order to validate the model, four tapered timber piles with circular cross-section (two plains and two retrofi tted with steel jacket) were tested and analysed with the proposed 3D FE modelling technique; and a good correlation between experimentally observed and numerically captured ultimate load was observed. Consequently, it was concluded that the developed FE models used in conjunction with the quasi-brittle constitutive law were able to adequately capture the failure load and load-defl ection response of the fl exural timber elements.
In this study, flexuralbehaviors of glue laminated timber beams manufactured from Pinussylvestristree were investigated by comparing the results with those of massive timber beams. The main variables considered in the study were number of laminations, types of adhesive materials and reinforcement nets used in the lamination surfaces. In scope of the experimental study, glue laminated beams with 5 and 3 lamination layers were manufactured with 90 x 90 mm beam sections. In the lamination process epoxy and polyurethane glue were used. Morever, in order to improve the bond strength at the lamination surface, aluminium, fiberglass and steel wire nets were used at the lamination surfaces. Load–displacement responses, ultimate capacities, ductility ratios, initial stiffness, energy dissipation capacities and failure mechanisms of glue laminated beams were compared with those of massive beams. It was observed that the general bending responses of glue laminated beams were better than those of massive beams. In addition to that the use of reinforcement nets at the lamination surfaces increased the ultimate load capacities of the tested beams. The highest ultimate load capacities were oberved from the tests of glue laminated beams manufactured using five laminated layers and retrofitted with polyurethane glue using steel wire reinforcement nets, in the direction normal to the lamination surface. Finally, the finite element simulations of some test specimens were performed to observe the accuracy of finite element technology in the estimation of ultimate capacities of glue laminated timber beams.
The determination procedure of the failure mechanism of CLT shear walls due to the failure of joints was presented in the 45th CIB-W18 meeting in Vaxjo1. It showed that the reliability based analysis based on the ultimate capacity of fasteners predicted quite well the failure process of shear walls when a rigid loading beam was applied. However, the failure process due to the failure of hold-down connectors was not very clear when the flexible loading beam was used. Therefore additional lateral loading tests were conducted by using flexible loading beam as shown in Fig.1 with different procedures to determine the failure mode. This new procedure based on the yield strength of shear plates and the ultimate capacity of hold-down connectors showed better determination of the failure mechanism of CLT shear walls without conspicuous slips between CLT panels.
This paper shows the racking test results of CLT shear walls with different failure modes. The failure modes of shear walls were designed by using reliability analysis considering the failure of the hold down connections at the bottom end of shear wall and that of the joints connecting two CLT panels at the centre of the wall. It was shown that the design of joints with the yield capacity Py for the central joints SP and the ultimate capacity Pu for the hold down connection HD (Mode III) determined well the precedence of HD failure without slips in SP and showed high capacity, while Modes I and II failure showed higher ductility than Mode III failure.