Experimental study on mechanical joints with nail type fasteners

When timber is used as a building material a connection is required between the elements in order to transfer the load. A typical connection is the steel to timber joint with nail as connectors. These connections can fail in ductile manner, causing the nails to reach their plastic moment capacity, or in brittle manner so called plug shear failure, causing the timber to fail in combined tension and shear failure of the different faces of the plug. The Eurocode standard for calculating the ductile resistance for one shear plane steel- to timber connections with nail as fasteners is derived from Johansen’s theory. It divides five failure modes in two groups. Thick steel plate group and thin steel plate group. The criterion for the groups depends on the thickness of the steel plate used in relation to the fasteners diameter. If a plate is neither thin nor thick according to Eurocode linear interpolation of the resistance is allowed. The plug shear formulation in Eurocode separates the tension and shear resistance in such way that the higher of the two will decide the resistance of the joint. Seven different ductile nail patterns were designed in order to compare the test results with the Eurocode formulations and a simulation model which is based on the Johansen’s yield theory. The joints used both 2.5 and 5mm steel plates with 4mm in diameter nails making the patterns that use 2.5mm plates to count as joints that require interpolation of the resistance according to Eurocode. After testing all the ductile patterns it was shown that the 2.5mm plate joints had the same, and in some cases, higher failure load than their 5mm joint counterparts. The plastic hinges in the 2.5mm joints were formed at the same location as if a thick steel plate was used even though the 2.5mm plate was closer to the thin plate border of 2mm. Furthermore, nail spacing parallel to the grain did not seem to influence the resistance of the joint even though it should be reduced due to the risk of premature splitting along the line of the nails. In order to evaluate the current plug shear formulation in Eurocode and to develop an alternative formulation for plug shear failure, six plug shear patterns and three border patterns were designed. After testing all the patterns a nail density limit, where the patterns start to fail in plug shear, was discovered at around 600-700mm2/nail. The density of the timber seemed to influence the failure load in some patterns when plug shear failure occurred. With help of Matlab a formulation was designed with the data from this thesis test results combined with plug shear data from Johansson’s report. This new formulation includes the density of the timber and the different faces of the assumed timber plug with coefficients in front of them determined from the curve fitting solver function in Matlab. The stiffness theory which was proposed after observing the experimental test on plug shear joints seemed to capture important parameters in its formulation. It was considered to be a good candidate to predict the failure load when plug shear failure occurs