A Comparison Between the EN 383 and ASTM D5764 Test Methods for Dowel-Bearing Strength Assessment of Wood: Experimental and Numerical Investigations C. L. Santos*, A. M. P. de Jesus † , J. J. L. Morais* and J. L. P. C. Lousada* *CITAB – Centre for the Research and Technology of Agro-Environment and Biological Sciences, University of Tra ´s-os-Montes e Alto Douro, 5001-801, Vila Real, Portugal † Engineering Department, University of Tra ´s-os-Montes e Alto Douro, 5001-801, Vila Real, Portugal ABSTRACT: The purpose of this paper is to provide a comparison between embedding tests covered by the EN383 and ASTM D5764 standards, highlighting some difficulties regarding the application of the referred standards. These embedding tests are essential to evaluate the embedding strength of wood, which is required for the assessment of joints strength. The proposed comparison is based either on experimental data and numerical simulations through the finite element (FE) method. Tests were performed on maritime pine wood (Pinus pinaster Ait. species) according to the longitudinal and radial directions, allowing the comparison of the embedding strength and elastic foundation modulus. Three dimensional FE models of the tests were built using contact elements technology and assuming the steel dowel and wood as linear elastic isotropic and orthotropic materials, respectively. The contact modelling is a challenging topic for which this paper also proposes some guidance. The test configuration proposed in the EN383 standard for assess- ment of the embedment strength in compression is more susceptible to the dowel bending than the half-hole test configuration proposed in the ASTM D5764 standard. The numerical simulation of the EN383 embedding test raises some additional difficulties regarding the dowel boundary conditions. KEY WORDS: dowel type connections, embedding tests, finite element method, pine wood Introduction Joints are often the weakest points in timber struc- tures. The loss of perfect continuity in the structure, which is caused by the presence of joints, will result in a reduction of the global strength. This implies an increase in dimensions of the assembled elements. About 80% of structural failures have their origin on connections [1]. The dowel-type connections are the main fastening technique used worldwide in timber structures. The singularity of wood joints is not only attributed to a combination of different materials such as wood and steel, but also due to the highly anisotropic behaviour of wood. Fundamental to an efficient utilisation of dowel-type joints is the understanding of their mechanical behaviour under load (e.g. load-slip behaviour, stress distributions, ultimate strength and failure modes). The mechani- cal behaviour of wood joints is a complex problem governed by a number of geometric, material and loading parameters (e.g. wood species, fastener diameter, end distances, edge distances, spacing, number of fasteners, fastener/hole clearances, fric- tion and loading configuration). Traditionally, dowel-type wood connections have been modelled based on Johansen’s model [2], pro- posed in 1945, which has been incorporated in design codes [3, 4]. Johansen proposed a beam (dowel) on foundation (wood) model based on maintaining equilibrium normal to the dowel and balancing internal and external moments. The foundation is assumed discontinuous, i.e. Winkler type as commonly assumed in design of foundations. Also, Johansen assumed perfect rigid-plastic response of both the dowel and wood. This approach considers the embedding strength as a material property, although it is a combination of many geometric and material factors. In order to account for some of these variables, many embedding tests are required. These tests have been normalised, with EN 383 [5] and ASTM D5764 [6] being the most relevant standards. The minimisation of the dowel bending in the Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Publishing Ltd j Strain (2010) 46, 159–174 159