1、 IN-WHEEL ELECTRIC MOTORS The Packaging and Integration Challenges Alexander Fraser Senior Mechanical Systems Engineer Protean Electric Ltd, UK ABSTRACT One of the fundamental advantages of in-wheel motors is that they free up packaging space on the vehicle platform. From retrofits to new vehicle de
2、signs, the freedom to add batteries, range extenders, or simply preserve cargo and passenger space is a major competitive advantage for all styles of in-wheel motor over more conventional electrical machines. This advantage needs to be preserved by not having to re-locate the friction brake to a pos
3、ition on the sprung mass of the vehicle, so both motors and brakes must be accommodated inside the wheels of a vehicle. To preserve as much of the retrofit market as possible, it must be ensured that the packaging solution inside the wheel does not require significant platform tear-up. In addition,
4、the whole system must be safe, reliable, and provide credible vehicle performance. This has led to some radically different in-wheel motor concepts being developed by various organisations around the world. Proteans unique solution to this packaging problem was, on the face of it, a bold one for a s
5、tart-up company; displace the brake to provide space for a motor, and develop a new packaging philosophy for the unsprung assembly incorporating both brake and motor. By working with an expert partner on the braking side, this, we believe, results in a much more harmonious use of the valuable space
6、inside the wheel than other in-wheel solutions, and meets or exceeds all the requirements outlined above. This presentation will provide details of the process that resulted in the Protean design, as well as highlight the engineering of the solution fitted to the front axle of the high performance P
7、rotean/Brabus E-Class vehicle. Figure 1: Unsprung assembly, Protean/Brabus E-class front-left corner. TARGET MARKETS AND REQUIREMENTS OVERVIEW Protean Electric is targeting both passenger vehicles and light commercials with a blank sheet electric motor design wholly dedicated to in-wheel application
8、s. Principally, the motor must be capable of being the sole source of tractive power for an electric vehicle or series hybrid, but must also work alongside conventional internal combustion powertrains to enable parallel hybrid architectures. To keep costs down, 2WD is an absolute must on smaller veh
9、icles and hybrids, and 4WD should lend credible performance to larger, heavier executive saloons and SUVs. The performance of these vehicles should be able to squarely address the market segment the vehicle is in with the performance, range and interior space that a demanding customer requires. A ke
10、y requirement that drives the Protean motor design is retro-fit ability. Clearly an in-wheel motor equipped powertrain opens up so many possibilities for sprung mass packaging and design changes that any vehicles that are designed from the ground-up around in-wheel motors will have their own protrac
11、ted development cycles measured on the scale of several years. Balanced against this, OEMs need to remove significant amounts of carbon emissions from their entire fleet right now, and a handful of compromised niche vehicles will not achieve this. The ability to engineer a hybrid or full EV on an ex
12、isting platform, with minimal tear-up of existing hardware or intrusion into passenger, storage or accident-vulnerable space cannot be underestimated and is one of the main advantages of choosing an in-wheel motor to propel a vehicle. It is therefore highly desirable for an in-wheel motor to be able
13、 to be fitted to a car without requiring any special wheel design or requiring risky suspension modifications, and this played a large part in the development of the requirements for the Protean motor. Regardless of the general architecture and form of an in-wheel motor, the unsprung environment is
14、a harsh one compared with the on-body requirements for a motor. The wading requirements of the vehicle mean that the motor needs to operate whilst essentially submerged, and the wheelhouse is subject to great extremes of temperature thanks to the friction brake and rejected heat from any on-body ICE
15、 powertrain. For a road vehicle the motor must survive repeated shock loads of up to 50g with continuous vibration spectra of up to 20g RMS for severe rough roads. Stones are expected to impact any exposed face of the device and the cabling to the motor must be able to survive repeated bending and a
16、rticulation with suspension and steering movement at temperatures well below zero.1 The main challenge however will always be package space, especially with a retro-fittable design. In appreciating this, it is clear that Nm/litre and kW/litre have to be maximised. It is already well known that the b
17、est motor technology for volume specific performance is liquid-cooled brushless permanent magnet, and so early-on in the programme the decision was made to focus on this technology for the first high volume production motor that Protean is developing. TORQUE AND POWER REQUIREMENTS The principal purp
18、ose of the drivetrain is to deliver tractive motoring effort to the tyre-road interface. As a free bonus, an electric drivetrain has the potential to apply similar tractive effort in the braking domain and hand a sizeable chunk of the vehicles kinetic energy back into the battery. Safety and/or regu
19、latory concerns aside, on first consideration it is tempting to believe that friction brakes can be replaced entirely with the electric drivetrain but a few simple calculations show the cost of such a move. Take for example a relatively light 2WD passenger vehicle at 1500kg. Assuming a certain few k
20、ey parameters it is quickly shown that the torque and power requirements are much greater in the braking domain than they are in motoring for a given wheel: Torque o Motoring - Pull-away torque for 30% grade = 650Nm (per motor) o Motoring - Maintaining 100mph at 6% grade =300Nm (per motor) o Braking
21、 - 1g braking = 1800Nm (per front wheel) Power o Motoring - Maintaining 100mph at 6% grade = 40kW (per motor) o Braking - Deceleration at a rate of 1g at 100mph = 250kW (per front wheel) Fundamentally this is because the vehicle is always required to be able to perform an emergency stop from all spe
22、eds during normal operation, but forward tractive forces do not need to be anywhere near as large as braking forces in order to provide credible or even sporty forward vehicle performance. A universal truth of passenger vehicles is that the most powerful actuator on the vehicle is not the power-trai
23、n, it is the brake system. It is important to note that the majority of this braking effort is carried out by the front wheels. In targeting 4WD applications, and not discerning (at the requirements level) between RWD or FWD in the case of a 2WD drivetrain, Protean clearly has to cater for, or not p
24、revent, the provision of these levels of braking power and torque in its motor design If the motor were designed to be able to create the levels of retarding torque required in braking, it is shown by the example above that the motor would need to produce around three to four times more wheel torque
25、 and over five times more wheel power than if it were sized for even sporty forward performance. By inspecting Figure 2 below it can also be seen that, further to the above design difficulties, an emergency stop is very rarely conducted and so motors which are sized for full braking performance are
26、very rarely of any use, and reclaim practically no extra regenerative energy than a drivetrain sized to regenerate up to 0.3g. The information displayed in Figure 2 was collated during initial developments on braking systems in conjunction with Alcon Components Ltd. Clearly, sizing the motor to meet
27、 emergency braking requirements will result in large implications for the cost, size and mass of the motor that are of little use. Figure 2: Brake utilisation curve for 2 million brake applications There are also some other show-stopping constraints you need somewhere to put this energy you are gene
28、rating with your motors. This will require either a much higher performance battery or other energy storage/dissipation devices which will add to the cost of the vehicle. Managing the failure modes of such a high performance motor will require some very restrictive safety requirements to be applied
29、at both motor and vehicle levels, further increasing cost. The conclusion to this section is that Protean believes in-wheel motor driven cars still need friction brakes and one of the major requirements on our motor is that it is sized for forward performance only, and a friction brake is used to “t
30、op up” the braking torque when under rare demanding braking circumstances, or when the battery has restricted charge current. By making the vehicle perform adequately in the motoring domain, almost all of the available regenerative energy is captured over the majority of real-world drive cycles anyw
31、ay, especially on front-drive applications where higher regenerative brake torque can be applied without stability concerns. There is optimism within Protean Electric that the present motor design will enable the safe removal of friction brakes from the rear axle in future applications, but this is
32、not currently permissible under present road vehicle legislation. So this leaves us to set torque and power requirements based on forward performance. When deriving torque requirements it is tempting to base the motor requirements from existing ICE-driven vehicles. This would go along the lines of m
33、atching the torque at-the-wheels of the target vehicles when in a low gear. However these low gear ratios (especially 1st) are sized more around the limitations of the powertrain and clutch life for example allowing low-speed crawling in traffic or around cark parks without having to “ride the clutc
34、h” constantly. Second or third gear would be a more realistic target, but then an electric motor torque curve is substantially different to an ICE, and the torque of an in-wheel motor can be changed on the scale of single milliseconds, as opposed to hundreds of milliseconds in the case of a suspende
35、d powertrain on driveshafts, giving a much greater “performance feel” 1 . When also considering automatic transmissions, the torque converter ratio introduces further uncertainty, and it rapidly becomes far easier and more consistent to size the torque requirement for grade-ability at GVW. Figure 3
36、shows the requirements for whole-vehicle torque requirements for 22% and 30% gradients over a large range of GVW values. Protean are aiming to make a single motor cover as many market segments as possible in 2WD and 4WD, so it can be seen, for example, that a motor that can propel a 1500kg car (torq
37、ue-wise) in 2WD requires 500Nm continuous and 650Nm peak torque capability respectively, and this motor will also serve up to a 2500kg car in 4WD. These drivetrains will of course also propel lighter vehicles with greater performance if so wished. Figure 3: Vehicle GVW vs. torque requirements for 22
38、% (continuous) and 30% (peak) grade Power requirements for motoring a vehicle however, can easily be derived directly from the target vehicle and its original powertrain. Although naturally Protean took great care to derive requirements from first principles also, it was quickly found that the same
39、power-at-the-wheels figure was broadly arrived at across a range of cases. The original powertrain must of course lend adequate top-speed and overtaking performance to the vehicle and these are power-limited driving scenarios. If an electric drivetrain can match this power then the performance will
40、generally be, both subjectively and qualitatively, much better, as the motor will have a much broader power peak and much quicker dynamic responses than the outgoing ICE powertrain. Figure 4 shows wheel power requirements for a range of vehicles up to 3500kg GVW. Using the same example of a 1500kg 2
41、WD vehicle, this would require each motor to give 30kW continuous power performance. Again this covers up to around 2500kg in 4WD configuration so complements the torque requirements well. Figure 4: Vehicle GVW vs Power requirements as derived from standard vehicle powertrain Now targets have been s
42、et for a range of performance requirements for the motor, the principles of the motor design and actual scalable performance characteristics of any particular motor layout must be well understood before settling on sizing for a single motor, which this paper shall now go on to describe in detail. TH
43、E SAFETY GOALS AND REQUIREMENTS FOR AN IN-WHEEL MOTOR Without question any vehicle drivetrain must be safe, and any drivetrain that has individually driven wheels presents a unique problem in this respect. A key safety goal for Proteans in-wheel motor design derived directly from ISO26262 3 - “no si
44、ngle fault shall prevent the driver from retaining control”. This safety goal, when applied to an individually electrically driven wheel, essentially boils down to a limit on the wheel torque error that the electric machine can cause through a single or random fault condition. This safety goal, alon
45、g with many others, was defined through following a process defined in ISO26262, which is the de facto standard for automotive functional safety. Any unintended single wheel torque will cause both an unintended acceleration/deceleration of the vehicle and also an unintended yaw moment and vehicle co
46、urse deviation. Whatever the response of the vehicle is to this torque, the driver must be able to respond with “reasonable” inputs to the controls, with a “reasonable” reaction time, and the car must not depart from its intended course more than a “reasonable” amount. The subject of driver modellin
47、g, the reasonable limits on control inputs and vehicle responses, and other specifics of Proteans safety concept formulation are well outside the scope of this paper but are covered extensively in 4 The single most effective way of addressing this issue for a broad range of motor faults is to use se
48、veral separate motors to drive a single wheel. This requirement applies directly to the control and power electronics, and the stator electromagnetics of the machine. Note the rotor assembly and stator mechanicals are comparatively benign objects safety-wise and all sub-motors can share these compon
49、ents without safety concerns. Following analyses covered in 4 Protean have settled on four sub-motors, each with of the full motor performance, as being required for safe vehicle operation in its present design, although the architecture is inherently scalable for different levels of subdivision. This means that many, normally potentially catastrophic failure modes can be mitigated locally inside the in-wheel motor, often with no net torque change at the wheel. Take, for example, a so-called “line-to-line” motor short, which would normally present a major safety conce
