Field determination and modeling of load paths in wood light-frame structures

Low-rise buildings, constructed using wood, are vulnerable to extreme wind storms and earthquakes. While several experimental measurements of the environmental loads (mostly wind) on the building envelope have been made at full scale, none of these studies directly linked these external loads with the internal forces and displacements of the structure, as achieved in this research. The thesis presents the experimental and analytical work on two light-frame wooden structures, where one already existed (Forintek shed in Quebec City) and the other (UNB house) was built specifically for the research project on the University of New Brunswick campus in Fredericton. The research goal was to devise and demonstrate methods of identifying load paths in light-frame wood buildings subject to environmental loads. The objectives were also to improve the knowledge on the magnitude of the forces generated by environmental loads on typical low-rise buildings; to measure forces and deformations in test buildings and correlate them with the applied loads; and finally to develop accurate numerical whole-building structural models. These goals were achieved by carrying out experiments at the element level (studs, sheathings), subsystem level (shear walls) and on the whole-building level (finished and "realistic" light-frame timber buildings). The responses of these buildings to controlled static tests as well as natural environmental loads were observed and compared with a wind tunnel study and with detailed finite element models with good agreement. Shear walls were tested in isolation and as a part of the whole structure. The tests indicated that neither the strength nor the stiffness decreased by the same magnitude as the wall effective length is reduced. Therefore, the simple concept of effective length, being used presently, is invalid. For the Forintek shed, the structural monitoring was based on measurements of deformations within a representative segment of the wall and roof surfaces and a matching grid of wall and roof wind pressure taps supplemented with a wind tunnel study at Concordia University. In general, it was shown that the building surroundings had a great effect on the pressure distribution of the surface on the structure and that these effects are cannot always be determined intuitively. Both mean and peak pressure coefficient were measured and they compared well with corresponding values obtained in the wind tunnel tests. In general, the peak pressure coefficients from the full-scale tests were higher than those obtained from the wind-tunnel tests. The results from controlled static loads on the UNB house indicated that the load was distributed to all walls, and significant load sharing was observed. Mostly, this reflected not only the rigidity of the roof, but also the rigidity of transverse walls. The stiffness of the roof was sufficient to distribute load to walls farthest away from the load application point. Also, the expected vertical paths for load were not observed. It was also found that the internal forces are concentrated near the corners of the building. Under vertical loading on the roof, the load at the roof-to-wall interface was concentrated in a small region of the building plan around the application point. This was not the case at the superstructure-to-foundation interface. The test results also showed that the load was transferred to the transverse walls, even though there were only nominal connection between the wall and the roof trusses. The results from the analytical modeling showed good agreement with the full-scale test results for shear walls as well as for the whole building. The 3-D model was able to simulate the sharing of racking forces between shear walls, based on experiments reported in the literature. It was also able to reproduce static test results and predict the force measurements obtained from load cells underneath the house structure. In general, the errors in the numerical prediction were small. The model was able to predict the interaction between the roof system and the walls and the interactions amongst walls. The research relied on the collaboration of several researchers in industry and academia, and was funded by a CRD grant of the Natural Sciences and Engineering Research Council of Canada.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.