Intelligent Truss Design - Digitalization of the Building Process

Sammanfattning

Today, timber is widely used in construction in Sweden, but it is not used to its full potential and strength. Timber is a heterogenous material, but in the construction industry today, assumed to be homogenous. That timber is heterogenous is known, but the reason for it to be assumed homogenous is to simplify the processof grading it. The properties structural timber is assigned in the grading process are their characteristic values, also known as the fifth percentile value. This means that 95% of all timber in each strength class or grade has a value higher than the characteristic one. Another reason for doing this is to minimize the risk of failure. This may result in a potential under utilization if not other, more advanced meth ods are used. When designing trusses the characteristic value of the timber is used to limit the risk of failure, and not its individual value that can be gathered from the testing carried out during the grading process. If the truss designers can utilize the individual data instead of the characteristic value the trusses can be designed with higher utilization ratio. If the designers know which timber parts are weak and strong, the strong timber will be used where the stresses are the highest, and vice versa with the weaker timber. In this thesis, softwares such as Grasshopper, Matlab and MiTek Pamir are used to simulate and test whether the risk of failure can be decreased if the truss designers can use individual values instead of the characteristic value. The results show that if the individual values of each timber board instead of the characteristic value are used the risk of failure can be decreased.