PhD on Sensor and actuator layout optimization for thermal control

The TU/e offers a PhD position within the 'ThermOpt' research project. This project is a collaboration between the TU/e and the company ASML, which develops lithography machines. The project involves two PhD positions, of which this vacancy is one.

The PhD position is embedded within the Dynamics and Control section of the Mechanical Engineering Department of TU/e. The Dynamics and Control section is home to 13 faculty members and approx. 40 researchers (PhD and postdoc) in the field of dynamical systems and control and offers a vibrant environment with excellent lab facilities on high-tech systems, robotics and automotive systems.

Project description

Background
ASML's lithography machines are used in semiconductor manufacturing for the production of computer chips (see figure below). In these complex high-tech systems, an increasing number of thermal control loops has been introduced to limit the undesired effects of thermal disturbances on (nano-meter) positioning performance. Advanced thermal control solutions will become even more important for future machines, as - on the one hand - the thermal disturbances in the machine will increase further (mainly due to the increasing power consumption in the machine) and - on the other hand - the performance objectives on positioning performance will become more stringent to enable the production of computer chips that are even faster, more powerful and energy efficient than the current generation.

Project objective

To support future system designs for lithography systems, including hardware design, sensor and actuator layout and control solutions, two central, underlying and inter-related challenges are addressed in the ThermOpt project:1) How to construct computationally tractable and accurate dynamic models of thermo-mechanical systems? and 2) How to optimize thermal sensor and actuator layout in support of thermal control?

The PhD project associated to this vacancy aims to address the second challenge (but will interface intensively with another researcher taking on the first challenge). Hence the main objective is to develop an optimization approach for thermal sensor and actuator layout in support of thermal control. Optimal actuator and sensor layout design should be developed in view of the achievable closed-loop control system objectives, such as minimizing thermally induced deformations of essential machine components, leading to higher positioning accuracy. This requires high-quality thermo-mechanical modelling of the hardware to-be-designed and accurate modelling and quantification of the disturbances and performance variables associated to the (thermal) control objectives.

The combination of the large design space on actuator and sensor location, the complexity of the physical model, model uncertainty and nonlinearities causes the decision space to be of large dimension and difficult to explore. The search for suitable techniques to intelligently explore this design space is at the core of this project. We envision to explore state-of-the-art optimization techniques in combination with AI techniques to develop computational tools to resolve this exciting problem.

People involved in supervision:

The starting dates are flexible but before January 2023.

Moreover, the project will offer to the students an extensive training program on Systems and Control in the scope of the Dutch Institute for Systems and Control (http://disc.tudelft.nl/ ). Moreover, a training program focusing on more generic and transferable skills required by professional researchers is offered (https://www.tue.nl/en/working-at-tue/phdpdeng/professional-development/) . This provides the students with a solid background for their research and future careers.