How MBSE Digital Twins Expand Electric Aviation

Utilizing CAE solutions to develop reliable, power-dense ePropulsion systems for the future of aviation.

MSC Hexagon has submitted this post.

Written by: Andrew Johnston, Head of Electrical & Electronics Engineering, Hexagon

Several key goals influence all aerospace system development programs: achieving high efficiency, low mass, high reliability, high levels of safety and robust ability to withstand environmental factors. In addition, there is an increasing motivation toward the development of sustainable solutions. The additional complexity for electric propulsion systems is that the technology employed is constantly evolving, with little in-service experience to support the demands of the safety case.

While the automotive sector has done much to progress the general technology for electric powertrains, such experience and the typical level of rigor applied in this domain are not nearly enough to satisfy the safety integrity and reliability demands of aerospace. This does not prohibit the use of modern or leading-edge technology in aerospace, but success will only be achieved through the application of robust Systems Engineering, complemented by integrated Model-based Systems Engineering (MBSE) workflows.

The application of robust MBSE is paramount to achieving aerospace goals, including the ability to better manage system-wide interfaces to support “propulsion as a service.” The main problem with electric powertrains is that they are already very efficient and power-dense. How then do we improve this to deliver increasingly demanding airframe requirements?

While there is no single one-size-fits-all design or technology solution, certain trends can be identified depending on the airframe. These trends are important to understand in the context of the requirements which apply, as it is impossible to consider the infinite number of possibilities when developing any ePropulsion system.

For example, given a typical propeller speed of circa 2,500 rpm and assuming a direct-drive architecture, an outer rotor solution can provide higher power and torque density benefits over inner rotor solutions purely from an electromagnetic perspective. However, depending on the requirements, the scalability of outer rotor solutions is somewhat limited and can become difficult to integrate into the wider system. To achieve higher power density targets, a way to reduce the mass of the electrical machine is sought after.

There are several ways to achieve this. Perhaps the most obvious one is to use a gearbox coupled to the electrical machine, enabling a highly efficient higher-speed and more compact machine at the detriment of system-level propulsion efficiency and additional potential failure modes associated with the gearbox. Nonetheless, the efficiency shortfall of a geared solution when compared to a direct-drive solution is likely to be better overall due to the total mass reduction achieved at the airframe level.

Figure 1. Example cutaway of direct-drive outer rotor eVTOL Propulsion motor (left) and a higher power density inner rotor motor with integrated gearbox eVTOL Propulsion system (right). When designed against the same requirements, a substantial reduction in propulsion system mass can be achieved.

Figure 1. Example cutaway of direct-drive outer rotor eVTOL Propulsion motor (left) and a higher power density inner rotor motor with integrated gearbox eVTOL Propulsion system (right). When designed against the same requirements, a substantial reduction in propulsion system mass can be achieved.

A key challenge associated with light-weighting is sustaining both performance, safety and reliability requirements. For Power Electronics, Machines and Drive (PEMD) technologies, this can typically result in the need to manage thermal performance with a demonstration of robustness against the external environment which can vary dramatically for an aerospace application.

To achieve higher power densities from the motor perspective, the increased current density is desired. This is possible so long as the waste heat can be extracted from the machine effectively. To improve thermal management, a method of directly cooling the motor windings is sought. Hexagon, combined with its capability in high fidelity metrology measurement and inspection, is refining a novel winding concept that permits cooling through the bulk of the motor slot, using oil, as a ‘direct stator cooling’ solution.

Figure 2. Novel motor winding solutions permit direct cooling which can be optimized through the use of Cradle CFD.

Figure 2. Novel motor winding solutions permit direct cooling which can be optimized through the use of Cradle CFD.

This concept ensures that there are no hot spots in the machine which could otherwise lead to long-term reliability issues and fundamentally permits higher current to pass through the windings, which vastly increases the power and torque density potential. Such a cooling system can be virtually assessed, optimized and validated through the use of CFD analysis.

Another benefit of this novel motor concept is that it vastly improves manufacturing repeatability due to the deterministic nature of the assembly process. This results in overall system reliability gains as the windings do not rest on one another as they would in a traditional machine, therefore preventing any thermal or vibration-induced fatigue on the winding insulation, mitigating the risk of insulation breakdown and motor failure while retaining a respectable copper fill factor.

In addition to thermal management, a demonstration against external shock, vibration and fatigue is also required to ensure reliability and safety integrity are maintained through life. Aligned to the objective of reaching higher power and torque density solutions, enabled by higher speed electrical machines, this also increases the demand for power electronics. In particular, the devices used require lower switching losses and the ability to operate at higher temperatures.

Wide Band Gap devices, such as Silicon Carbide (SiC) and Gallium Nitride (GaN) offer attractive performance gains compared to traditional Silicon (Si) devices, owing to the underlying physics of higher electron mobility, saturation velocity and higher thermal conductivity, which effectively results in the potential to operate at higher frequencies at lower losses. Being able to operate at higher switching frequencies means that higher fundamental output frequencies can be realized, which permits higher speed motor control. SiC devices are currently the go-to device for high performance, higher power applications.

Figure 3. Example of a close-up view of a typical Silicon-based power switching device which makes electrification possible, alongside an FE-based bond wire fatigue and CFD-based thermal analysis to aid design reliability and robustness, using MSC Apex, Marc and Cradle.

Figure 3. Example of a close-up view of a typical Silicon-based power switching device which makes electrification possible, alongside an FE-based bond wire fatigue and CFD-based thermal analysis to aid design reliability and robustness, using MSC Apex, Marc and Cradle.

Other benefits of SiC devices, when compared to traditional Silicon (Si) based devices, include the removal of diode reverse recovery from the switching waveform, which directly removes a significant switching loss component, allowing greater efficiency to be achieved. For a given device size, the on-state resistance is also lower for SiC and GaN devices compared to Si, which allows further efficiency gains to be realized.

With Si-based technology, a maximum fundamental output frequency of around 1 kHz is commonly observed. It is possible to create a highspeed machine based on this output frequency, but this can only be achieved through the use of a low pole number machine. As an example, if a machine design required >30,000 rpm with a maximum inverter output frequency of 1 kHz, the only credible choice would be a 2-pole machine.

Wide Band Gap devices support the generation and control of higher frequency fundamental waveforms from the inverter. Maximum inverter output frequencies of around 2.5 kHz can typically be reached with current SiC technology, which in the aforementioned example would mean speeds of up to around 75,000 rpm could be reached before being forced into a 2-pole design solution. This means that higher speed electrical machine designs with a higher pole number can be employed. The benefit of using a higher pole number in the machine is that the thickness and therefore the mass of the stator and rotor yoke can be reduced, allowing a more power-dense design.

This is just one example of the intricacies of electrical machines and power electronics development, which cannot be developed in isolation if the best system-level results are to be achieved. Other important considerations include but are not limited to: skin effect, proximity effect, core loss, partial discharge, passive filter components, total harmonic distortion, track impedance, resonance, phase delay and of course electromagnetic compatibility (EMC). There are also other architectural considerations of paramount importance, for example, to ensure single fault tolerance as well as mitigating the effects of soft errors through Single Events Effects (SEE) caused by ionizing radiation.

The solution requires a robust architecture which would typically include replicated functionality and elements of diversity in the design.

The perhaps obvious and most significant benefit of SiC is its ability to operate at higher temperatures. This means that the cooling requirement can be reduced, which can lead to a physically smaller cooling system for the inverter, therefore contributing to an increase in the inverter, and system, power density.

Nonetheless, optimizing the cooling system remains of paramount importance, as this can help reduce the electromechanical fatigue on the components and thus ensure and even improve reliability.

Figure 4. Examples of FEA-based stress analysis using MSC Apex/MSC Nastran to achieve reliable, robust, yet optimized light-weight solutions for inner and outer rotor machine and gearbox configurations.

Figure 4. Examples of FEA-based stress analysis using MSC Apex/MSC Nastran to achieve reliable, robust, yet optimized light-weight solutions for inner and outer rotor machine and gearbox configurations.

Hexagon has a high-fidelity co-simulation modelling solution, whereby thermal analysis can be linked with shock or vibration analysis, forming the foundations of a digital twin, the models of which can be validated through test and in-service data sets to enable a true digital twin. This analytical evidence can be used to support the development program at reduced technical risk, with a more accurate and robust safety case, as well as exploring the operational space of the product when it is in-service at no risk. This can be used to consider life extension or performance optimization options through the ability to run the digital asset ‘ahead of time’ and under different environmental conditions, to help mitigate any hazards and risks associated with the physical asset.

In conclusion, the integration of MBSE and appropriate CAE workflows enables the design space to be rapidly yet robustly explored, which helps ensure requirements are complete, correct and validated, the design intent verified, and risk is reduced overall. The application of MBSE is a cost-effective way to deliver complex systems and fundamentally underpins the ability to develop digital twin assets that can be used to support physical in-service assets to meet the emerging needs of the market.

To learn more, visit Hexagon.

This article was originally published by Hexagon in Engineering Reality Magazine, Summer 2022.