Systematic design of glulam trusses

Sammanfattning: The design of a structure should be regarded as the design of a system. In this licentiate thesis a method (concurrent engineering) for the design of systems is studied parallel with the development of the system itself. The purpose of the work is twofold: To study the application of concurrent engineering (CE) as a method to manage the design work for the design of a glulam truss. To develop a glulam truss and investigate if it can be made competitive on the Swedish market. The qualitative study started with the formation of a CE team consisting of an architect, a contractor and a manufacturer of glulam with the researcher as the project manager. The idea in CE is to let members from different parts of the design work (design, production and manufacturing) solve the problem together. Through the concurrent work of these members a proposal of a glulam truss was posed. The proposal was then verified and refined by the CE team through several iterations between resistance considerations and production issues. The qualitative study showed that the members of the CE team are equally important for the success of the development project. The project manager needs to be independent and have a wide knowledge base. Furthermore, the production issues were investigated early in the design process, which is an advantage for the manufacturer of the truss. The quantitative study focused heavily on the solution for the joint in the glulam truss. A connection type new to the Swedish market was proposed; nails with slotted-in steel plates. The joint consists of steel plates placed in internal slots in the glulam member which is assembled by shooting the nails through the glulam and the steel plates with a nailer. Medium-sized joints were tested in tension parallel to the grain and the resistance was found to be on average 8.35 kN/nail and two steel plates. Production considerations for the joint type revealed that sawing the slots to accommodate the steel plates could be a problem and this was further investigated in laboratory tests on full-scale joints. The full-scale tests were performed using the theory of statistical experimental design aiming to show tendencies in the behaviour rather than performing a parameter study. The specimens were designed without eccentricities in the joint. The results showed that buckling tendencies of the steel plates must be suppressed and therefore the design of the full-scale joint was adjusted to counteract this. Tension perpendicular to the grain in the joint was qualitatively studied. The results showed that the occurrence of a compressive force close to the tensile force does not affect the resistance in tension perpendicular to the grain for this joint design and this failure mode will not set any demands on the resistance of the truss. The glulam truss was optimised with respect to material cost to investigate the competitiveness on the Swedish market. First order theory was used assuming linearly elastic material behaviour. The rotational capacity of the joints was included in the analysis. Together with estimates for the production cost of a joint, the results showed that a glulam truss using this connection type would have a production cost of approximately 19,200 SEK/truss. The competitiveness of this cost must be judged by the market, but steel trusses have a production cost of about 15,600 SEK/truss while other glulam structural elements range from 18,300-31,000 SEK/element.