Metal Additive Manufacturing In Power Electronics, Machines And Drives: Opportunities and Challenges
Interview with Dr. Nick Simpson, University of Bristol
Could you start by discussing the research you are undertaking at the University of Bristol, the significance of computational design in optimizing metal additive manufacturing processes for electrical machines?
The world is in an electrification transition, in pursuit of Carbon Net Zero.
Electrical machines (motors and generators) are at the heart of this transition, with electrified ground and air transport being a significant focus. The automotive and aerospace sectors have identified the need for between 9 and 25 kW/kg by 2035 with >96% efficiency, in stark contrast to the 2-5 kW/kg available today. For a given power (kW), the physical size and mass (kg) of an electrical machine is ultimately limited by internally generated losses, manifesting as heat, which act to raise component temperatures above the insulation material rating.
As the electrical windings are a primary source of loss and engineers are hitting limits with conventional manufacturing, we adopt the unrivaled geometric and design freedom of metal Additive Manufacturing (AM) to simultaneously enable new conductor geometries and topology, form embedded thermal management features such as cooling fins and channels, and to integrate value-add features such as terminals and voids to accommodate sensors.
This approach is a radical departure from convention requiring parallel work streams to:
1) prove the electrical properties of the resulting AM components,
2) demonstrate the efficacy of the loss mitigation topologies and cooling schemes to the Power Electronics, Machines and Drives (PEMD) community to stimulate adoption, and
3) to develop parametric design tools capable of mixing boundary and mesh geometry representations with an awareness of Design for AM (DfAM) and facilitating interaction with electromagnetic and thermal simulation to drive design.
Design Innovations in Electrical Windings through Additive Manufacturing
With metal additive manufacturing offering unprecedented geometric freedom, yet facing crippling cost constraints, how do you balance these factors in designing electrical windings that are both highly performative and cost-efficient?
As an academic research group, The Electrical Machine Works enjoys the freedom to focus on the engineering, design and performance challenges, however, as we see adoption from our automotive and aerospace partners we cannot ignore the economics. Aerospace and high-end automotive can typically absorb the higher costs as we find more cost-effective manufacturing practices.
These typically comprise more dense bed packing, alterations to geometry (build orientation, simplification of parts) to improve build volume efficiency and exploring alternatives to our typical Laser Powder Bed Fusion (LPBF) approach. In addition, we develop in-house post-processes for surfaces and application of electrical insulation coatings to allow fine control and cost- and sustainability- conscious production.
Software Solutions and Design Challenges
Could you outline the software workflow employed in the design and manufacture of these components?
Conventionally, electrical machines are initially sized using analytical equations based on the desired rated torque and power. Due to the symmetry of the active region of electrical machines, the design is then iteratively refined using 2D numerical electromagnetic models coupled with thermal and mechanical analyses as necessary. The near final design is then validated using computationally expensive 3D numerical models or through physical manufacture and test.
Adopting AM from the outset significantly increases the design space and can make the problem inherently 3D. As such we develop computationally efficient modeling methods using Open Source and Commercial tools to drive design.
As far as possible, we aim for parameterisation and design automation so a large part of our work is designing new algorithms to generate geometry based on electromagnetic and thermal behavior.
How do these tools enhance or restrict your design process, and what advancements would you like software companies at CDFAM to focus on, to better facilitate bringing these solutions to market?
Unfortunately the Venn diagram of software engineers, computational designers and electrical machine designers seldom overlaps.
As such, while The Electrical Machine Works has a deep capability in the design, manufacture and test of electrical machines employing AM components, our in-house tools are not polished, optimised, or production ready, often requiring Open Source and Commercial codes to be glued together.
One of the motivations to speak to the CDFAM community is to learn how we could formalise our approaches and further stimulate adoption.
Collaborative Efforts in Advancing Metal Additive Manufacturing
Could you highlight any collaborative projects or cross-disciplinary efforts that have significantly advanced your research and how do these collaborations enhance the field?
The Electrical Machine Works is funded through a combination of a UK Research Innovation (UKRI) Future Leaders Fellowship, Engineering and Physical Sciences Research Council (EPSRC) and Innovate UK grants, and affiliation with the Future Electrical Machines Manufacturing (FEMM) Hub, and industrial funding totaling approximately €2.5M.
We work with AM machine builders, specialist AM suppliers, academic institutions, Catapult organisations and industrial end-users in automotive and aerospace.
Within our team we have electromagnetic, thermal, mechanical, material science and AM expertise. The cross-sector collaborations enable us to credibly showcase the significant performance benefits offered by metal AM in electrical machines and will, in time, drive adoption of the technology.
Future Research Directions and Applications
Looking ahead, what are the next steps in your research, and how do you hope to spread adoption of this approach beyond academia?
The Electrical Machine Works has worked on many strands of metal AM for electrical machines.
The near future will see us bring these strands together, along with our collaborators, to demonstrate the benefits of AM technology in achieving ultra-high power density electrical machines. This demonstration will naturally be disseminated to our industrial partners and is seen as a potential catalyst in driving adoption outside of academia while bolstering academic activity.
Finally, how can the CDFAM community contribute to overcoming the current computational design challenges faced in the field? Are there specific areas of collaboration or research that you believe are particularly promising?
Although the design methods of The Electrical Machine Works are likely considered Computational Design, we do not understand the depth and breadth of the subject, nor the tools available or best practices.
For us, it is important to start a dialogue with the CDFAM community to highlight our multi-physics needs (electromagnetic, thermal, mechanical) and to understand how we might mutually benefit in applying Computational Design to electrical machines that adopt AM.
Don’t miss out on the opportunity to learn from, and connect with experts in computational design and advanced manufacturing at CDFAM Berlin, May 7-8 2024
