The second part of our PhD interview series with colleagues from RWTH Aachen, which started with a compelling interview with Christoph Wellmann, continues with an exchange about the development of thermal systems for electrified vehicles. This is the research topic of Kai Franke, research associate at the Chair of Thermodynamics of Mobile Energy Conversion Systems.
Please describe your research to a wider stakeholder audience and highlight innovative aspects.
My research focuses on the development of thermal systems for electrified vehicles, which are essential for fulfilling various thermal requirements. These include climate control in the passenger cabin, cooling the electric drivetrain, and the regulation of thermal conditions of batteries
My work involves designing these thermal systems, including selecting suitable actuators such as coolant pumps, valves, and heat exchangers. Beyond the physical system design, my primary focus is on the control strategies for these systems. Due to increasing complexity driven by integrating advanced components like heat pump systems, the associated control strategies have also become more intricate. However, there is a persistent need for greater efficiency in these systems.
Therefore, the primary innovation in my research lies in developing efficiency-oriented operational strategies for thermal systems in electrified vehicles. A key methodology utilised is machine learning, particularly reinforcement learning algorithms, which enable the system to autonomously adapt and optimise operational strategies against multiple objectives. For instance, reinforcement learning can maintain optimal operating temperatures in the system while simultaneously minimising energy consumption.
Moreover, employing machine learning allows efficiency-focused control strategies to be established in early development phases, significantly reducing the effort required in calibrating conventional control methods.
What is the connection with the ESCALATE project? How do you benefit from the project, and what synergies exist?
ESCALATE offers an essential framework for conceptualising, simulating, and evaluating the thermal system and its control strategies, specifically for a battery-electric truck of ELECTRA. My responsibilities within ESCALATE include defining application scenarios and deriving critical test cases relevant to the thermal system design. Initially, various system topologies suitable for the vehicle concept are identified and analysed through simulation, creating a fundamental basis for subsequent design and optimization activities.
A particularly valuable aspect of the ESCALATE project is the availability of detailed operational data and specifications from project partners like ELECTRA. This includes existing system topologies, operational boundary conditions, and specific component parameters, all crucial for developing accurate simulation models. This collaboration effectively addresses the common challenge of insufficient component-level information, thereby greatly enhancing the realism and applicability of my research outcomes.
Additionally, project partners are committed to implementing various operational strategies or elements of the developed control concepts onto their vehicle control units. This real-world application provides an opportunity to validate theoretical assumptions and comprehensively assess the potential benefits of optimisation-based operational strategies.
Moreover, ESCALATE allows for validation of the developed simulation models at the component level and supports the exploration of advanced operational strategies. A unique feature explored within ESCALATE is the “hotel function,” typically exclusive to trucks, where both reinforcement learning and model predictive control methodologies are employed. This enables a direct comparison of these methodologies in terms of their effectiveness in cabin climate management.
Thus, ESCALATE significantly enhances my research capabilities by offering realistic scenarios, collaborative expertise, and practical validation opportunities, thereby highlighting strong synergies between theoretical innovation and industrial application.
With your expertise and the perspective of your research, where do you currently see the biggest challenge in a large-scale deployment of zero-emission trucks?
The primary challenge for the large-scale deployment of zero-emission trucks lies in the limitations associated with battery systems, particularly concerning their weight and achievable energy density. This limitation is reflected in regulations, such as those in Germany, where the permissible total weight for battery-electric trucks has been increased to 44 tonnes.
Due to the high mass of batteries, new challenges arise, including notably sluggish thermal behaviour, which can necessitate significant energy for thermal conditioning of battery packs. This aspect underscores why examining heat pump systems for the “hotel function” in trucks becomes particularly interesting, as heat pumps offer substantially higher efficiency compared to conventional heating systems, presenting an opportunity for potential energy savings.
Achieving desired target ranges of 800 to 1000 kilometres under realistic operational conditions remains a substantial challenge. Contrary to some perceptions, I don’t consider charging infrastructure a critical barrier, as trucks are often charged overnight or at fleet manager facilities. In my assessment, charging capacities of approximately 350 kW are initially sufficient for existing use cases. However, the necessity for widespread availability of 1 MW charging capacities to enable broad adoption remains uncertain.
Another significant challenge involves ensuring the safety and regulatory compliance of high-voltage batteries in trucks. Early and reliable detection of thermal runaway events is crucial, coupled with advanced safety systems capable of alerting or even waking the driver if asleep within the vehicle cabin.
Lastly, warranty and durability considerations pose substantial challenges compared to conventional diesel vehicles. Electric trucks must compete with diesel trucks, routinely achieving mileage exceeding one million kilometres. The economic viability of current battery systems under such demanding operational conditions remains questionable.
What do you think could be the impact (or lack thereof) of projects like ESCALATE on the promotion of zHGV technology/solutions?
The project promotes collaboration between industry partners and research institutions, especially crucial in the commercial vehicle sector, where research opportunities are frequently constrained by budget limitations. ESCALATE and similar initiatives substantially contribute to elevating industry-wide knowledge and methodological expertise. By engaging partners across the complete development process, these projects build a strong foundation for future innovations and collaborations.
How will your solution enhance the efficiency of the thermal control of a truck? Could these strategies also be applied to fuel cell vehicles?
My research focuses on enhancing the efficiency of thermal control systems in electrified trucks by investigating improvements on two key levels: hardware modifications and optimised operational strategies.
On the hardware side, I explore the potential of altered system topologies and the integration of heat pump systems through preliminary simulations. Although implementing a real heat pump system in our test vehicle is economically impractical, these simulations aim to identify potential energy-saving benefits.
Regarding operational strategies, I am integrating a reinforcement learning-based strategy to complement this approach. Comparing these two advanced control strategies is crucial for identifying the ideal method and exploring synergies between them. Industry partners frequently express interest in this comparative analysis since selecting the optimal optimisation-based operational strategy is often non-trivial.
Beyond energy efficiency, my research aims to optimise cabin comfort comprehensively. Rather than simply regulating cabin temperature, the control system will focus explicitly on occupant comfort by managing temperature, relative humidity, and CO2 concentration within the cabin. Such comfort-focused regulation is important for truck drivers, as maintaining optimal driver alertness significantly enhances road safety, critical due to the high mass and consequential severity of truck-related accidents.
These advanced operational strategies are also applicable to fuel cell vehicles, as they share similar cabin climate control requirements with battery-electric vehicles. However, implementing heat pumps in fuel cell vehicles may be less beneficial since the fuel cell provides an effective onboard heat source, capable of sufficiently meeting heating demands.
Moreover, the methodologies and particularly reinforcement learning strategies developed in my research are fully transferable to the thermal management of fuel cell vehicles, thereby extending their applicability across the broader commercial vehicle sector.
In conclusion, my research contributes significantly to advancing thermal system efficiency and cabin comfort in electrified commercial vehicles, offering broadly applicable strategies across both battery-electric and fuel cell-powered platforms.