Edinburgh Napier University

  An integrated programme of experimental and analytical work was carried out to evaluate the nonlinear
semi-rigid characteristics of timber connections using fully overlapping nails subjected to short
duration lateral loading and moment. The investigation is part of a continuing programme of research at
Napier University into the behaviour of timberjoints using fully overlapping nails as the connecting Z -1
mechanism.

The effects of the factors and material properties that influence the behaviour of nailed joints were
addressed in a structured and controlled way allowing semi-empirical models to be developed for the
lateral load behaviour of rnulti-nailed timber joints using steel and plywood gusset plates. A quality
control procedure was established for the testing programme and consistent standards were applied to
the preparation and testing work. The semi-empirical models that were developed included for the
effect of timber density: crussept late material effect- nail strength; number of nails: nail diameter; row
spacing and the effect of the moisture content in the timber. They covered joints assembled with and
without a gap between the timber and the gusset plates and for joints assembled with steel gusset plates,
the effect of the predrill size used in the gusset plate was also investigated. The model results compared
very well with the results from tests, accurately predicting the non-linear behaviour of the joints up to
failure.

An extensive analytical and experimental study was carried out to investigate the moment-rotation
behaviour of these types ofjoints. Two linear arýd four non-linear models were developed for each type
of joint and the efficiencies of the models were compared to detennine the one that best simulated the
joint behaviour. The linear models consistently underestimated the capacity of the joint, giving
conservative results. The best solutions were obtained by applying the torsion formula used for steel
connections and incorporating the nail behaviour models developed for the non-linear lateral load
joints. Account was taken of the non-linear behaviour of the connection and alternative models using
fixed and moveable centres of rotation were developed. Very good comparisons were achieved between
these models and the test results.

A detailed comparison was made between the behaviour of the joints using the lateral load displacement
models and Eurocode 5 (EC5) and it was concluded that EC5 rules did not accurately
simulate the behaviour of this type of joint. It was concluded that the nail spacing rules in the code did
not apply to fully overlapping nails. A limit state design method based on the principles used in EC5
has been developed from the models for the design of joints using fully overlapping nails and subjected
to lateral loading or moment.

The semi-rigid behaviour of the joints was also investigated and it was concluded that to safely predict
the response of structures assembled with fully, overlapping nails. the semi-rigid behaviour must be
included for in the analysis procedure. Rigidity factors, end fixing moment reduction factors and the
secant rotational stiffness coefficients for the joints were derived. It was also shown that where the
analysis was limited to the serviceability limit state. a modified elastic method of analysis could be used
and where it was beyond this state a non-linear method of analysis was required.