Towards optimisation of a double-sided punched metal nail plate connector

Abstract

Timber joints fastened with prototype double-sided nail plates were fabricated from European whitewood of densities generally in the range 350 to 600 kg/m3 . Joint behaviour was characterised from experimental results of 150 joint tests conducted to current European standard methods. Three characteristics, namely tooth load-carrying capacity, slip modulus and ultimate modulus, were required to fully describe the joint behaviour envelope up to maximum load. Joint performance was found to be well correlated to timber density and empirical models relating each characteristic to density were derived.

For analytical modelling purposes, each nail plate tooth was assumed to act as a dowel. Theoretical capacity based Johansen's pin joint theory and the Eurocode 5 embedding strength prediction for driven nails over-predicted the empirical model by 15% at 350 kg/m3 density, improving with increasing density to 8% at 600 kg/m3 . Also, effective embedding strengths derived from Johansen's theory were perfectly correlated to those
predicted from an empirical model based on tests conducted using a rigid embedment tool on fine-sanded wood surfaces. On the assumption of no joint separation at failure, the effective embedding strength was half the rigid embedment value. The proportion decreased for assumed larger gaps.

Modelling of nail plate fastened timber joints is complex and often relies on empirical calibration. In contrast to the conventional approach of using non-linear springs to lump tooth-wood interaction at the tooth root, modelling in this thesis treated the double-sided nail plate as separate from the timber. Tooth-wood interaction was modelled using contact mechanics and effective material properties determined from inverse modelling.
FE modelling was done using ABAQUS 6.3 software. Analysis using a two-tooth joint model showed that tooth capacity per unit length increased with decreasing tooth length but other factors outside the scope of the study precluded the determination of optimal tooth length. Parametric studies based on the nail plate tooth width were undertaken using a ½ joint model. At optimum, the tooth width increment was 25% and the capacity gain was 12%. Base plate buckling, which was absent in the original configuration, was well pronounced at optimum.

Date of Award	Jun 2005
Original language	English
Awarding Institution	University of Brighton
Supervisor	P.D. Rodd (Supervisor), Z. Guan (Supervisor) & David Pope (Supervisor)