Numerical Simulation of Loading-Configuration Effects on the Identified Moment–Rotation Response of a Slotted-in Steel-Plate Glulam–Steel Hub Joint

Abstract

Moment–rotation (M–θ) relationships obtained from subassembly tests are often transferred directly to semi-rigid models of glued laminated timber (glulam) structures, although the identified response may depend on the loading arrangement. This study quantified that influence for a slotted-in steel-plate glulam–steel hub joint using refined three-dimensional finite-element models in strong-axis (MN) and weak-axis (MY) bending. Zero-preload three-point (3P) and four-point (4P) configurations were compared on a common joint-level basis, three 4P loading positions were assessed, and compression analyses were used to evaluate preload-induced eccentricity. After reconstruction of interface moments and local rotations, the initial-stiffness difference between 3P and 4P was 8.4% for MN and 4.0% for MY. Four-point loading reduced the shear-to-moment index from 0.317 to 0.023 in MN and from 0.146 to 0.006 in MY. Moving the loading points from 400 to 450 mm changed fitted stiffness by less than 0.1% while increasing actuator demand. At 100 kN axial compression, hub eccentricity reached 1.08 mm in MN and 15.77 mm in MY. The results demonstrate that loading configuration and preload-induced geometry should be considered when identifying rotational response for subsequent full-scale testing and semi-rigid modeling.