# Technology

Technology

Technology

Description

**4.Electric Motor Topology Investigations**

Using a bespoke motor design tool, candidate electrical motor topologies including the Interior Permanent Magnet Motor (IPM), Permanent Magnet assisted Synchronous Reluctance Motor (PMaSynRel), and Induction Motor (IM) are quantitatively and holistically compared.

**4.1.Motor Design Tool**

A multi-domain tool had been developed within UoN for the sizing of electrical machines for traction, which is ideal in this kind of scenario where the machine doesn’t have tight volumetric/space constraints, and hence an optimal design can be found over a large search space. The layout of the sizing tool is shown in Fig. 4.1, and described in detail in [13]. The input variables, top level (shaded in blue) are the variables which can describe any arbitrary geometry of the machine. These include amongst others the stator/rotor diameters, slot/teeth geometries, magnet dimensions, bridge/rib dimensions etc. The middle level shown in Fig. 4.1 (shaded in yellow) describes the calculation/optimization core. The ‘Matlab’ blockset within this level contains the multi-domain calculation scripts which for any arbitrary geometry based on the input variables calculates the performance against set requirements. The DOE is the design of experiment which contains the initial parameter population generated using a pseudo-random sequence, while the MOGA block defines the parameters for the Multi-Objective Genetic Algorithm (GA) include the mutations and convergence criteria. Given the specific loss of a lamination material say at a frequency of 1000 [Hz] and at an induction of 1 [T] - W_{Fe_1000,1}, for any stator fundamental frequency - *fs*, and iron fluxdensity level *Bs*, the specific iron losses are approximated from :

where *Bs* is the on-load flux density.

The bottom level shown in Fig. 4.1 (shaded in green) defines the output constraints (which make a design as feasible/unfeasible) and optimisation goals. In this case, the optimization has two goals, to minimize the size (i.e. maximize the kW/L) and the cost ($/kW).

**4.1.Comparison of IPM and PMaSynRel Topologies**

Multiple V-type IPM and a PMaSynRel machines are designed using the intelligent multidomain GA-based design tool. Constraints are set on the maximum temperature (<200 oC), current density (<25 A/mm2), efficiency (>91%), total volt-ampere rating (<250 kVA) and maximum flux density (<1.6 T).

Fig. 4.2 shows the optimization results for the V-shaped IPM and PMaSynRel machines. The IPM power density trend increases with pole number, peaking at around 10-poles, at which pole number the best cost-performance ($/kW) is obtained for this topology. This is because the machine cost for the IPM is dominated by the magnet quantity, and a more power dense machine naturally utilizes less magnets. However at the higher pole numbers, the efficiency of the IPM topology is not too high, primarily due to the increase in iron losses. On the other hand, for the PMaSynRel, best power densities are observed for low pole numbers. This is due to the fact that for the PMaSynRel as the pole number increases, there is less drawing-space available for tailoring the barrier geometries, resulting in the rapid loss of reluctance torque. For the PMaSynRel the cost-performance ($/kW) and efficiency are also better for the lower pole numbers (6, 8 poles). For the lower pole numbers, where the PMaSynRel has a very strong saliency, the kVA requirement is also markedly lower with respect to the IPM topology.

The efficiency is looked at into more detail through the generation and comparison of efficiency maps for both topologies as shown in Fig. 4.3. Results show distinct features between the two machines. The IPM machine has better efficiency at the base speed and rated load, meanwhile the PMaSynRel machine has wider high efficiency region especially at lower loads. This is particularly important as the drive cycle for full electric vehicle indicates lower load operation more frequently.

In order to further investigate two machines in terms of efficiency, another optimization is performed, this time with the optimization goal changed to maximizing efficiency instead of power density. Summary of two optimizations is shown in Table 4.1. It is observed that optimizing for increasing the efficiency, results in 20% and 25% less power density for IPM and PMaSynRel machines respectively. In terms of cost, IPM machine is less affected with 4% increase whereas the cost of PMaSynRel increases significantly by 18%. Importantly the PMaSynRel shows a better efficiency improvement with 1% increase whereas the IPM exhibits 0.4% efficiency increase. This reinforces the point that the PMaSynRel is in general more suitable for typical EV duty cycles.

**4.1.Rare-Earth vs. Ferrite Magnets**

Considering the low cost of ferrite magnets when compared to NdFeB, a comparative study is done replacing the NdFeB magnets of PMaSynRel machine with Ferrite magnets. Results are summarized in Table 4.2, showing a very significant decrease in power density by about 33%. Another detrimental effect of using ferrite magnets is the increase of volt-ampere requirement by 25%.

For additional comparison, converter VA is kept constant as the original rare-earth based machine and the motor design tool is re-run with ferrite magnet properties. Results show that in order to use the same converter for ferrite assisted machine, the active mass of the motor more than doubles.

**4.4. Performance with Induction Machines**

In-house design tools are used to design IM with 4 and 6 poles. Water jacket cooling is used in conjunction with spray cooling although spray cooling is less effective with longer machines, as in the case of IM. Summary of the design study is given in Table 4.3, with the induction machine heavier/larger due to the lower pole number, as well as due the rotor losses which are fundamental to the machine operation.

Cost is estimated based on active mass usage and shown in Table 4.4 for the different machine types investigated within this report. The electrical steel and armature winding (copper) cost for the rare-earth based IPM and PMaSynRel are similar since they have similar power densities, while the cost increases for the ferrite-based machine and IM, since these machines are bigger. In terms of the rotor material (magnets or copper alloy), the advantage of the PMaSynRel is clear in that the magnet cost is reduced by 36%. Overall the PMaSynRel with rare earth magnets is the most cost effective solution followed by PMaSynRel with ferrite magnets, with the IPM and IM having similar cost performance.

**4.5. Preliminary studies on AC effects**

In the analysis presented thus far, for the armature winding a slot fill of 40% using finelystranded wire without additional AC losses is considered. This is akin to using a Litz-Wire winding. While armature windings using Litz Wire are possible, for automotive applications it is best if they are avoided due to the cost associated with using Litz Wire. To that end, in this section, the PMaSynRel machine is analyzed considering the AC losses. For the estimation of AC losses with random windings, the bundles are grouped with realistic strand positioning as shown in Fig. 4.4, and the proximity and circulating current effects calculated using Finite Element Analysis.

Fig. 4.5 shows the results of this analysis considering the main performance metrics of relevance to this development. In terms of power density, random windings are better for 6 poles. For 10 and 12 poles, using Litz wires can be advantageous, although their cost is markedly higher, leading to a significant drop in cost-performance. Efficiency is evaluated at two operating points (base speed and maximum speed) for different pole numbers corresponding to a frequency range between 450 Hz and 2.1 kHz. At base speed conditions using random wires provides better efficiency until 10 poles (750 Hz) while Litz wire becomes better for 12 poles. From the preceding analysis, with PMaSynRel machines, there is no specific benefit of going to high pole number designs, and hence random windings can be considered as overall advantageous with respect to Litz-wire windings.

**5.Summary, Conclusions, and Next Steps**

Taking into consideration the end-application’s various importance criteria, all considered machine types are scored in Table 5.1. Upon discussions with NEC, highest customer importance is given to the efficiency over the duty cycle, followed by the power density and cost-performance. All these aspects have been quantified within this phase of FAST project for the various motor topologies, and based on the presented analysis it can be concluded that the PMaSynRel machine using NdFeB magnets shows best compromise, since it achieves comparable power density to the more traditional IPM with less magnet translating into superior cost-performance. Additionally, the topology exhibits better efficiency at intermediate power nodes over a wide speed range, where the machine is operating for most of the time, and its suitability with the lower pole numbers means that it does not necessitate the use of costly transposed wire for high speed operation. Based on the aforementioned considerations it is the selected topology for the subsequent detailed design stage.

With the topology identified and foundations set, in the next stage more detailed design aspects will be investigated including:

Detailed quantitative comparison between traditional random windings and advanced higher-frequency hairpin windings. This will include welded hairpin layouts as well as the more recent continuous hairpin winding layouts. The comparison will include also advanced features for minimizing the AC losses within hairpin windings

Modelling and design of the spray-cooling system

Tailoring of the rotor design for the maximisation of reluctance torque, minimisation of magnets, whilst ensuring a high degree of mechanical factor of safety

The above technical electromagnetic, mechanical and thermal aspects will be performed with strong manufacturing input. The discussions with machinery companies will intensify over the next period to ascertain that the proposed solution is one suitable for volume manufacturing.

**6.References**

[1]"Advanced Propulsion Centre UK." <https://www.apcuk.co.uk/app/uploads/2020/11/Technology-Roadmap-Electric-Machines.pdf> >(accessed 17 July, 2021).

[2]C. Jung, "Power Up with 800-V Systems: The benefits of upgrading voltage power for battery-electric

passenger vehicles," IEEE Electrification Magazine, vol. 5, no. 1, pp. 53-58, 2017.

[3]H. Sobue et al., "Analysis and Experimental Comparison of Acoustic Noise of Three Switched Reluctance Motors Made of Conventional Steel, High Silicon Steel, and Amorphous Iron," IEEE Transactions on Industry Applications, vol. 57, no. 6, pp. 5907-5915, 2021.

[4]D. Golovanov et al., "4MW Class High Power Density Generator for Future Hybrid-Electric Aircraft," IEEE Transactions on Transportation Electrification, pp. 1-1, 2021, doi: 10.1109/TTE.2021.3068928.

[5]I. Tanaka and H. Yashiki, "Magnetic and mechanical properties of newly developed high-strength

nonoriented electrical steel," IEEE Transactions on magnetics, vol. 46, no. 2, pp. 290-293, 2010.

[6]F. Zhang et al., "Performance Entitlement by Using Novel High Strength Electrical Steels and Copper Alloys for High-Speed Laminated Rotor Induction Machines," Electronics, vol. Early Access, 2022.

[7]YAMATO. <https://www.yamatogokin.com/en/alloy-products/nc-alloy/><https://www.yamatogokin.com/en/alloy-products/nc-alloy/>products/nc <https://www.yamatogokin.com/en/alloy-products/nc-alloy/><https://www.yamatogokin.com/en/alloy-products/nc-alloy/>alloy/ <https://www.yamatogokin.com/en/alloy-products/nc-alloy/><https://www.yamatogokin.com/en/alloy-products/nc-alloy/>(accessed 4 July 2021.

[8]F. Zhang et al., "Back-Iron Extension Thermal Benefits for Electrical Machines With Concentrated Windings," IEEE Transactions on Industrial Electronics, vol. 67, no. 3, pp. 1728-1738, 2020, doi: 10.1109/TIE.2019.2903758.

[9]F. Zhang et al., "Electrical Machine Slot Thermal Condition Effects on Back Iron Extension Thermal Benefits," IEEE Transactions on Transportation Electrification, 2021.

[10]F. Zhang et al., "A Thermal Modelling Approach and Experimental Validation for an Oil Spray-Cooled Hairpin Winding Machine," IEEE Transactions on Transportation Electrification, pp. 1-1, 2021, doi: 10.1109/TTE.2021.3067601.

[11]C. Liu et al., "Estimation of Oil Spray Cooling Heat Transfer Coefficients on Hairpin Windings With Reduced-Parameter Models," IEEE Transactions on Transportation Electrification, vol. 7, no. 2, pp. 793803, 2021, doi: 10.1109/TTE.2020.3031373.

[12]C. Liu et al., "Experimental investigation on oil spray cooling with hairpin windings," IEEE Transactions

on Industrial Electronics, vol. 67, no. 9, pp. 7343-7353, 2019.

[13]D. Golovanov, L. Papini, D. Gerada, Z. Xu, and C. Gerada, "Multidomain Optimization of High-PowerDensity PM Electrical Machines for System Architecture Selection," IEEE Transactions on Industrial Electronics, vol. 65, no. 7, pp. 5302-5312, 2018.