The Challenge

Virtual (model-based) techniques are nowadays common tools to design manufacturing processes with optimal use of material and process capabilities. However, model-based techniques are subjected to different types of uncertainties which can hamper their reliability. The uncertainties can originate from variations in process inputs such as material properties, process settings (machine dynamics, lubrication) but also from insufficient accuracy of the models themselves e.g. as a result of coarse approximations or unrealistic assumptions. If these uncertainties are neglected, very often an optimal deterministic process design is found which is at the limits of the material, process capabilities and the accuracy of the models. In real-life, where variations are always present, this leads to unreliable production processes.

The main goal of the current project is to develop deep learning strategies based on physics informed neural network (PINN) methods for building digital twin of a multi-stage sheet metal forming process. The digital twin will utilize virtual (e.g. finite element) models as well as measured input and output data from the multi-stage sheet metal forming process in order to identically represent the system behavior. Therefore, the digital twin of the process will be able to cope with different sources of uncertainties in the process, reduce the discrepancies between predicted results and the measured output data. Such digital twins will be used to improve process settings, control and design, ultimately leading to first-time-right and zero-defect manufacturing.

Job Description

For this project, we are looking for a PhD candidate with relevant expertise. The candidate will report his/her research during bi-weekly meetings of our research group and quarterly meetings with the project partners. The candidate is expected to present his/her results at international scientific conferences and publish them in academic literature. Furthermore, the researcher will be encouraged to tutor MSc students who do their final assignment on sub-projects pertaining to the current research project.

The PhD research will be performed in the chair of Nonlinear Solid Mechanics and co-supervised from the chair of Precision Engineering.

• MSc. degree in Mechanical Engineering or a related field with excellent results and special interest in modelling, data analysis and production processes.
• Prior knowledge of nonlinear solid mechanics, machine learning, computational methods and optimization.
• Curious and highly motivated to do interdisciplinary scientific research with a clear industrial application.
• Working independently and taking ownership for your own research.
• Strong written and verbal communication skills. Proficiency in English is required, both spoken and written. You need to provide IELTS test results (minimum score 6.5) or TOEFL-iBT (minimum score 90). The latter is not necessary for native speakers of English or whose MSc program was in English.

• The UT provides a dynamic and international environment, combining the benefits of academic research with a topic of high industrial relevance, excellent working conditions, an exciting scientific environment, and a green and lively campus. We offer:
• A full-time 4-year PhD position;
• We provide excellent mentorship and a stimulating international research environment with excellent facilities;
• You are offered a professional and personal development program within Graduate School Twente;
• A starting salary of € 2.443 gross per month in the first year and a salary of € 3.122 in the fourth year gross per month;
• A holiday allowance of 8% of the gross annual salary and a year-end bonus of 8.3%;
• A solid pension scheme;
• Minimum of 29 holidays per year in case of fulltime employment;