Project Overview
Principle Investigators: |
Prof. Dr.-Ing. Bernd Engel Prof. Dr. Kristof Van Laerhoven |
Project Team: |
Jonas Knoche, M.Sc.Zeyneddin Öz,M.Sc. Alireza Yazdani, M.Sc. |
Research Institutions: |
Institute of Forming Technology (UTS), University of SiegenInstitute for Ubiquitous Computing (IUC), University of Siegen |
Semi-finished Material(s): |
Tube (1.4301, CW024A) |
Manufacturing Processes: |
Rotary draw bending |
Motivation
- Development of a digital representation for the two-stage rotary-draw bending of bend-in-bend geometries
- Evaluation of the influence of different semi-finished product properties and process parameters on the geometric characteristics of bent components
- Data-driven improvement of the active surface design of mould clamping jaws through a combination of empirical knowledge, simulation and process data
Aims
The overall aim of this project is to develop a data-driven design method for the effective surfaces of moulding tools in the rotary-draw bending of sheet-in-sheet geometries. For this purpose, sensitive process variables are determined that lead to fluctuating quality characteristics in bent components. The
automated and time-synchronised recording of these process variables in the process is carried out by sensors integrated into the forming system. The development of a method for parameterising the effective surfaces of mould clamping jaws takes into account the forming system properties and the real component geometry, and the process data is combined with simulation data and empirical knowledge, whereby data-driven and visualisable models are adapted and specific feature engineering and regression learning are applied. The aim is to
standardised collection and creation of a ‘benchmark’ reference data set with the data-based representation of the process sequence and initial adaptations of methods for explainability in combination with domain knowledge and the learned analysis results.
Working Program
Work Package |
Description |
WP1 |
FE process design, selection of measured variables, parameterisation of tool surfacessurfaces and design of the data platform |
WP2 |
Selection of sensors for the measured variables and development of the sensorconnections |
WP3 |
Automation and testing of data acquisition, development of interfaceinterface concepts and feature engineering |
WP4 |
Simulation-supported process modelling, selection of sensitive processoperation of the data acquisition infrastructure and feature generation |
WP5 |
Continuous operation with targeted variation of the process boundary conditions, analysisprocess data analysis and feature selection |
WP6 |
Development of a digital representation and redesign of moulding jawsclamping jaws |
Expected Results
The intended result of this project is a method for improving the effective surface design of mould clamping jaws in the two-stage rotary-draw bending of bend-in-bend geometries. For this purpose, synthetic data from simulations, process data from the forming process and empirical knowledge of the operators are combined knowledge of the operators are combined.
Contact
Univ.-Prof. Dr.-Ing. Bernd Engel
University of Siegen
Institute of Forming Technology
Breite Straße 11
57076 Siegen
E-mail: bernd.engel@uni-siegen.de
Website: https://protech.mb.uni-siegen.de/uts/en/
Prof. PhD Kristof Van Laerhoven
University of Siegen
Department of Electrical Engineering and Computer Science
Hölderlinstr. 3
57076 Siegen
E-mail: kvl@eti.uni-siegen.de
Website: https://www.eti.uni-siegen.de/ubicomp/home/index.html.en?lang=en
Jonas Knoche, M.Sc.
University of Siegen
Institute of Forming Technology
Breite Straße 11
57076 Siegen
E-mail: jonas.knoche@uni-siegen.de
Website: https://protech.mb.uni-siegen.de/uts/en/
Zeyneddin Öz, M.Sc.
Alireza Yazdani, M.Sc