Material efficiency in timber high-rise buildings

Abstract

Timber high-rise buildings have been gaining in popularity. However, due to the low material stiffness of timber the stability systems of these buildings are often made with steel or concrete. To avoid this the material efficiency of the timber stability systems must be improved. The most material efficient stability systems for timber high-rise are the externally braced systems according to literature. The externally braced systems are the diagrid and external braced frame stability systems. In literature the diagrid is concluded to be more material efficient than the external braced frame. Nonetheless, in practice only external braced frame systems are used for timber high-rise buildings. This disconnect could be caused by the steel connections which are often simplified or overlooked in literature studies but have a large influence on the material usage of a timber stability system. The connections can increase the timber element size because the required connection does not fit, and the connections will decrease the global stiffness of the stability system. For this reason, the connections must be regarded when comparing stability systems.
One diagrid and two external braced frame designs will be compared with a parametric model. In this model other parameters will be studied as well. These parameters are: plot sizes of 27.2 x 27.2 m & 27.2 x 40.8 m, floor spans of 3.4 m & 6.8 m, permanent floor loads of 3.5 kN/m^2 & 5.3 kN/m^2 & 6.7 kN/m^2 and element widths ranging from 400 mm to 650 mm. The parametric model will first size the elements individually for the ULS checks and afterwards the elements will be sized per group for the SLS checks. The ULS checks start with a regular ULS member check, then a member check in the fire situation is performed and lastly the connection design component used. This connection design component will create slotted-in steel plate connections where the amount of steel per connection is calculated, and the timber element sizes are increased if needed. In the SLS checks the global lateral displacement is checked and the along-wind acceleration. The along-wind acceleration is often normative for the sizing of the elements in timber high-rise. With a higher global stiffness or higher building mass the acceleration decreases. To make a fair comparison in the results the amount of timber and steel of the internal structure is also added to the material required in the stability system in the façade.
From the results it is seen that all buildings are sized on the connection design or on the along-wind acceleration. The diagrid designs have a higher global stiffness so they are sized more often on the connection design whereas the braced frame designs are sized more often on the along-wind acceleration. The diagrid designs use 3x more steel than the braced designs. The results for the other parameters are:
Plot size: A smaller plot size requires a higher floor load to meet the acceleration requirement, therefore the large plot size is more material efficient.
Floor span: The floor span in the external braced frame has a large influence on the designs with a small plot size. That is because the facade with the smaller span becomes normative for the global displacement. When this happens a smaller floor span decreases the material efficiency significantly. In the larger plot size the influence of the floor span on the braced frame designs is insignificant. For the diagrid designs the larger floor span causes higher normal forces in the diagonals. This increases the material usage in the facade. However, considering the internal structure of the building the large floor span is still more efficient.
Floor weight: A higher floor weight is more timber efficient for a small plot size, and the lowest floor weight is the most material efficient for a large plot size. With a higher floor weight, the steel usage always increases since the connection designs are sized on the ULS checks only.
Element width: The steel efficiency increases when the element width increases. This is a result of the chosen connection design. The timber usage is similar for all the element widths.
Some of the designs with a small plot size and low floor weight are unfeasible since the elements in the facade are too large. This is a result of the choice to only increase the timber element heights and not the connection stiffness to improve the global stiffness. Therefore, it is recommended to study how the global stiffness can be improved more material efficiently. Other recommendations for further research are on the stability system design and the connection design. Since in this thesis many designs are fixed in simplified exploratory studies to decrease the design space of the parametric model.