ÄûÃʵ¼º½

The title of the thesis: Simulating Electromechanical Powertrains: System-Level Performance and Component-Level Condition Monitoring

Thesis defender: Milla Vehviläinen
Opponent: Prof. Aki Mikkola, LUT University, and Prof. HÃ¥kan Johansson, Chalmers University of Technology, Sweden
Custos: Prof. Anouar Belahcen, Aalto University School of Electrical Engineering

As transport electrification accelerates, understanding how electric vehicle powertrains operate and how their components wear in service becomes increasingly important. In electric vehicles, the powertrain is a key subsystem that, in addition to the electric motor, consists of several mechanical components whose task is to transmit the motor’s rotational energy to the wheels as forward motion. For efficient and safe powertrain research and testing, it is important to complement experimental measurements with computational simulation models. Simulations make it possible to run predictive tests repeatedly and to examine variables that are difficult to access experimentally.

In this doctoral dissertation, electromechanical powertrains are investigated using various simulation models from two complementary perspectives: system-level performance assessment and component-level condition monitoring. The condition of individual components affects the operation of the entire powertrain, while system-level optimisation decisions in turn influence component loads and wear.

System-level simulations were used to assess how the number of gears and different driving cycles affect the performance of a heavy-duty electric vehicle in various driving situations. At the component level, dynamic contact models were adapted to model fault geometries in critical components, such as bearings and gears, and to generate realistic condition-monitoring data, thereby reducing the need for extensive and resource-intensive experimental measurements. The results show that a multi-speed powertrain can improve the energy efficiency of heavy-duty electric vehicles, particularly in hilly driving conditions. The research also provides an efficient way to generate reliable simulated data that can be used to develop machine learning methods for predictive maintenance.

The thesis work lays the foundation for a model-based approach in which vehicle performance and component condition can be evaluated simultaneously. This enables smarter and more sustainable design solutions as well as more advanced predictive maintenance – a step towards more energy-efficient and reliable electric transport.

Key words: computational simulation, powertrain, performance, condition monitoring

Thesis available for public display 7 days prior to the defence at . 

Contact:
E-mail: milla.vehvilainen@vtt.fi  
LinkedIn: www.linkedin.com/in/milla-vehvilainen