Balancing design and circularity

Abstract

The steel sector is responsible for 4-5% of the total greenhouse gas emissions in the world. Reusing structural steel elements can potentially decrease the need for the production of new steel. However, designing a load-bearing structure with reusable elements poses challenges to the design process. The starting point of the process is different due to a limited availability of structural elements. The design of the load-bearing structure should be based on the measurements of the available reusable elements, while traditionally the amount of elements and their measurements were based on the design. However, the freedom in a design is often limited by certain architectural and structural constraints. Therefore, the focus of this thesis is on how to organize an optimization process in which the aim is to design a load-bearing structure containing the least amount of new steel by means of implementing reusable elements, while taking into account the architectural and structural constraints.

In the optimization method developed in this study, a primary design for a load-bearing structure functions as an input. The method consists of four big steps: the definition of the geometry, the assignment of (reusable) elements, structural calculations and the formulation of results. The goal is to assign reusable elements which are respecting the given constraints: the minimum and maximum UC-value allowed and the maximum deviation in length. The output is a modified design in which reusable elements are implemented, in a way that lowers the amount of steel required to realize the design.
By performing a case study, the model is tested and the influence of the constraints and the implemented stock can be determined. Several analyses are conducted, using three different stocks. Stock 1 and stock 2 can be considered more diverse than stock 3. 
The characteristics of the resulting designs are influenced by the relation between the stock and the original design. For the more diverse stock, there are only results whenever the minimum allowed UC-value is set to 0.01. While the required amount of steel lowers (up to 79% for stock 1 and 74% for stock 2), the actual UC-values of the implemented elements can be considered low and the change in design can be considered big (up to 19 out of 24 angles in the beam-configuration changing more than 10%). For the less diverse stock, all analyses give results, independent of the values for the constraints. In this case, the results UC-value are high compared to the results related to stock 1 and 2, and the changes in design are less (0 changing angles). Besides, all elements can be reused, which lowers the amount of required new steel to zero. 
Therefore, applying the optimization method developed in this study to a design for a load-bearing structure results in a modified design in which the amount of new steel required to realize the design is minimized. However, the actual UC-values corresponding to the different elements and the changes in the design are dependent on the available stock and the constraints.