Project Details
Projekt
Data-based identification and prediction of the die surface condition and interactions in sheet bulk metal forming processes from coil
Subject Area
Primary Shaping and Reshaping Technology, Additive Manufacturing
Automation, Mechatronics, Control Systems, Intelligent Technical Systems, Robotics
Automation, Mechatronics, Control Systems, Intelligent Technical Systems, Robotics
Term
since 2023
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 520194633
Manufacturing by forming offers the advantage of a material utilization of up to 100 %. Sheet bulk metal forming processes from coil can overcome current ecological and economic challenges through the efficient production of functionally integrated components. The production of defective components due to tool wear counteracts the high material utilization and thus reduces the energy efficiency. This results in increased product costs and environmental impact, as scrap costs have to be allocated to the good parts and defective components have to be disposed of and recycled. Online detection of signs of wear offers the potential to intervene in the process within a batch’s production and thus keep scrap levels low while positively influencing the environmental balance of the overall process. However, this requires a system that uses machine data in a knowledge model as part of an information model to predict tool wear and, thus, deviating component properties at an early stage. The aim of this research project is therefore to develop a tool wear model based on a full back extrusion process from the strip. Moreover, the correlation strengths between process parameters are to be determined from the automatic monitoring of the process by sensors, optical analysis and the microscopic examination of tool and component samples. These, in combination with a causal model constructed by experts, allow the generation of a quantified cause-effect graph. Additionally, experts may uncover previously unknown causal relations based on strongly correlated parameters that are validated experimentally and simulatively. The resulting quantified cause-effect graph is part of an information model formalized as an ontology, for product, process and the machine resource, including varying tools. With the help of an assistance system to be developed in the project, the information model can be interacted with in order to explain "backwards" observed deviations at product or tool by process variables acting on them and to predict "forwards" the effects of changes on dependent variables. Based on the extended parameter understanding, an intervention in the forming process within running batches and, thus, an increased resource efficiency shall be enabled.
DFG Programme
Priority Programmes
