This is the new paper with Meryem Nurefşan Yabanigül, published at Automation in Construction. It is also Meryem’s master’s thesis. This study is about testing the production of curved surfaces with non-linear robotic hotwire cutting and shape memory alloys. Below is the abstract of it:

Robotic arms are being used by construction firms and schools of architecture around the world in design/build research and material studies. Some of these studies utilize robotic hot wire cutting (RHC) as an efficient production technique. However, this technique has limitations on the production of free-form objects. This paper addresses this limitation by proposing an alternative non-linear robotic hotwire cutting (NL-RHC) workflow. The original contribution of this workflow is the usage of a shape memory alloy (SMA) as a cutting wire, that enables the production of curved forms. We tested the workflow on the production of a Gyroid-like modular system to reveal its advantages and limits. We made several attempts to find the most accurate approximation of the Gyroid. Therefore, the results suggested that NL-RHC would be an efficient solution for the production of curved surfaces.

As a result of new studies, robot technologies can contribute to the increasing interest in complex mathematical objects in architecture. Currently, it is not possible to state that the method proposed in this research would present an ultimate solution, and always yield more effective and accurate results compared to its alternatives. However, we can foresee that non-linear robotic hotwire cutting (NLRHC), and the use of shape memory alloys (SMA’s) in RHC in general, would possibly provide effective alternative solutions if the geometric goal is suitable for it. Therefore, for the production of translational surfaces, as in the case of the Gyroid fundamental patch, NL-RHC can offer a significant advantage.