ABSTRACT The European Yield Model (EYM) is a commonly accepted procedure to calculate the capacity of laterally loaded connections with dowel type fasteners. The standard procedure is specified in EC5 (EN 1995-1-1 2004) and uses characteristic values as input parameters for the calculation of characteristic values of the resistance of the connection. These characteristic values are the basis for an engineering design of structures. Nevertheless, there are situations, e.g. the design of a test set-up or a pushover analysis in earthquake engineering, when mean values of the resistance of a timber connection are needed. In this context, analytical equations and mechanical models were assessed on two levels by comparing theoretical values to experimental results. The first part of the research considered all the input parameters. The mechanical definition of the plastic moment proved to be more appropriate than the definition of a modified plastic moment according to EC5 for the determination of the yield moment of nails. On the other hand, it was confirmed that the embedment strength of Norway spruce softwood can be predicted reliably and slightly conservatively according to code regulations given by EC5. The definition for the withdrawal strength according to EC5 also seems to define a conservative lower boundary. The situation is not so clear for the embedment strength of oriented strand board (OSB). Test results are very scarce and do not account for the inhomogeneity regarding the thickness of the boards with higher density at the surfaces, which is the section where contact forces are transferred. In the second part, over 110 specimens considering eight different types of nailed connections were tested to assess different calculation procedures. The study focused on nailed connections with OSB and solid timber. This combination together with smooth nails as fasteners is very common in Europe for light-frame structures, especially in areas of low and medium seismicity because they exhibit large deformation capacity without losing strength. On the level of mean values, relationships are found between 1.00 and 1.47 for experimental results over calculated resistance, and it was found that the EYM is more conservative when the thickness of the sheathing is 10 mm instead of 18 mm, which means, that the expected failure mechanism changes from one to two plastic hinges. The findings are approximately the same if the relationship is studied on the level of characteristic values. Here, the relationship of values from experimental investigations over analytical calculation lays between 1.27 and 1.58 for a mechanism with one plastic hinge and between 0.90 and 1.12 for a mechanism with two plastic hinges. Based on these results, the question arises, if the EYM is the appropriate tool to predict the capacity of laterally loaded connections if comparatively large deformations occur between the connected parts. This question should be studied more detailed in the future.
Read full abstract