Thermally Informed Bending: Relating Curvature to Heat Generation Through Infrared Sensing

Abstract

The process of bending flat ductile surfaces such as metal sheets into curved surfaces that match a precise digitally predefined geometry so far has required either using a mold or a predetermined gauge and alloy-based predictive model to determine the necessary bending arc. Instead, our approach relies on the material itself to report its bending state during a simple actuation movement enabled by a robotic arm. We detect spontaneous heat generation triggered by the transition from elastic to plastic deformation using an infrared camera in real time. This information is used to determine the motion path of the robotic arm in order to reach the desired final geometry in one movement rather than incrementally. The bending process is generated through a RAPID code with a trap loop to guide the robot and a Processing script to control the thermal sensing and trap loop activation. Results for an IR sensor-guided geometric trap loop and the determination of a meaningful threshold for the onset of plastic deformation are discussed. This method’s efficiency and applicability to large scales of production and to a wide array of ductile materials with varying elastic modulus, suggest a potentially significant contribution to the fabrication of curved architectural skins.