Experimental study on automation and data-driven modeling in the punching-collar forming process [Dataset]

Summary The dataset originates from experimental investigations into a two-stage punching-collar forming process. Fluctuations in process and material parameters promote the formation of edge cracks along the collar, leading to undesired scrap. The dataset includes measurements that support the design of a fully automated progressive forming tool and enable detailed process analysis. More information can be found in the instruction: Instruction_for_the_application_of_the_dataset.pdf. Content This dataset (~455 MB) includes: force–displacement curves from the punching operation, force–displacement curves from the collar forming operation, labeled and unlabeled greyscale images of the collars for 216 experiments conducted under varied process parameters. The selected process and material parameters, along with the corresponding file naming convention, are documented in the file experiment_parameters.xlsx. The formed collars were evaluated based on the size and number of edge cracks and assigned to a quality scale. Published with the research presented in the paper: Korenek, J., Riemer, M., Kräusel, V., Langhammer, D. & Koschmider, A. (2025). Experimental study on automation and data-driven modeling in the punching-collar forming process. at - Automatisierungstechnik, 73(3), 185-197. https://doi.org/10.1515/auto-2024-0104