Experimental Investigation of Load Redistribution in a Historic-Type Timber Roof Segment Under Progressive Tie-Beam Weakening

This study examines how progressive loss of tie-beam stiffness redistributes deformation demand and internal force within an experimentally verified historic timber roof segment. A three-frame laboratory model was evaluated by close-range photogrammetry, inductive displacement sensing, post-test bending tests, and finite-element modelling. The measurements revealed clear frame-to-frame stiffness asymmetry, a distinct pre-cut settlement phase under sustained load, and accelerated deformation once the residual tie-beam section was substantially reduced. A refined model using member-specific elastic parameters reproduced the global loading response more accurately than a uniform-stiffness model. The damaged zone could also be represented by an equivalent rotational spring, although the fitted spring values remained response-dependent. The results indicate that lathing acted as a temporary redundancy path after severe weakening and that a nearly hinged residual section was not mechanically equivalent to a fully severed member within the tested configuration.

THE EXPERIMENTAL INVESTIGATION OF A MODEL OF A HISTORICAL-TYPE TIMBER ROOF SEGMENT UNDER PROGRESSIVE TIE-BEAM WEAKENING

This paper reports on an experimental study of a model of a historical timber roof segment subjected to progressive weakening of the central tie beam. A three-frame specimen was tested under controlled loading, short-term sustained full load, and sequential cutting of the middle tie beam. Close-range photogrammetry was combined with inductive displacement sensors and conventional control measurements. The study examined the agreement of optical and contact measurements, redistribution of displacement demand to the adjacent frames, and the temporary residual stability provided by lathing after complete failure of the middle tie beam. Across the tie-beam midspans, the mean absolute difference between inductive and paired optical measurements ranged from 0.07 to 0.40 mm, with a maximum point-wise deviation of 1.39 mm. After complete failure, additional settlement of the two intact frames reached 4.28 and 3.97 mm relative to the pre-cut full-load state, indicating longitudinal redistribution through the lathing.