FAQ on axial flux motors: part 2

Radial flux AC motors offer some performance and packaging benefits but also bring some thermal and manufacturability issues.

This part of the FAQ continues the exploration of AFMs and RFMs. The first part examined the axial-flux motor (AFM) in more detail and compared it to the very widely used radial-flux motor (RFM). The final part of this FAQ looks at some real-world issues associated with the adoption of AFMs.

Q: Can you briefly explain the operation of the RFM?
A: Traditional radial-flux brushless DC (BLDC) motors consist of a rotor made of permanent magnets located inside a stator. The copper windings are wrapped around slots. The resultant flux is generated perpendicular to the axis of rotation. In this case:

A stator contains support known as a yoke, which is outfitted with “teeth” containing electromagnetic coils.
The teeth function as alternating magnetic poles.
The rotor’s magnetic poles interact with the alternating magnetic flux of the wound stator teeth, resulting in the motor’s torque.

Q: What about the same considerations for the AFM?
A: The flux is generated parallel to the axis of rotation because of the way the coils are wound. This has the advantage of simplifying the motor’s fabrication in principle (but not in practice). There is no need for a yoke. However, this “simplification” brings new production problems that may outweigh the benefits, as discussed further below.

Q: Is the axial-flux approaching a new development?
A: Yes and no. The world’s first motor, invented by Michael Faraday in 1821, was the axial-flux type. However, its further development and use was limited by the materials, understanding, and various realities of that period.

What we know as the RFM was devised in the 1830s by Thomas Davenport, a blacksmith from Vermont, who was able to eventually patent it. This motor is known as the simple DC motor (alternating current had not yet been “discovered” at the time – electricity came from crude batteries). The irony is that Davenport died in financial ruin as a result of his relentless, dogged pursuit of the commercial success of this now very widely used motor topology.

Q: Why are AFMs now getting all this attention?
A: Practical AFMs have only become viable from a performance, cost, and manufacturing perspective in the last 40 to 50 years. This is due to advances in materials and production as well as computer-aided design (CAD) and finite-element analysis (FEA) for electromechanical, magnetic, mechanical, and thermal analysis. The CAD-based design supports the optimization of the performance across many related variables. It also helps assess the losses and temperature rises in motors, which are often limiting factors in the performance or practicality.

Q: What are some of the details of the AFM construction?
A: There are three primary aspects:

Disk-shaped stator and rotor: Axial flux motors feature both the stator and rotor designed as flat, circular disks. This design allows for a compact and space-efficient motor, making it particularly well-suited for applications with limited installation space.
Radial magnets: The stator and rotor in an axial flux motor incorporate magnets arranged radially, enabling an efficient and direct interaction between the magnetic fields. This radial magnet arrangement enhances the motor’s torque production, making it suitable for various industrial and automotive applications.
Winding coils: Copper wire coils are wound on the stator of the axial flux motor. These winding coils are crucial in generating electromagnetic forces that drive the motor’s rotation. The precise arrangement and control of these coils are essential for the motor’s performance.

Q: Are there variations in the basic AFM design?
A: Of course, nothing in motors is ever simple, as was seen in the “family tree” figure. For axial-flux design, there are two basic approaches:

1) AFM magnet with skewed fan shape and slotted motor: here, the “cogging” torque cannot be avoided due to the interaction between the permanent magnet and the stator tooth, leading to torque ripple, vibration, and noise (Figure 1 left). (“Cogging” is the intermittent jerking motion that occurs when the shaft of a conventional brushless motor rotates and is due to unevenness of the magnetic field.)

2) Laminated-type AFM magnet: AFMs also have issues with eddy-current losses, which increase self-heating (think of the flat-top induction stove which uses these eddy-current losses to advantage!), and their efficiency will also be affected by the high-temperature demagnetization. To overcome this, “magnet segmentation” is used to decrease the AFM eddy-current losses (Figure 1 right).

Figure 1. There are two widely used internal arrangements of the AFM structure: (right) skewed fan shape and slotted motor and (left) laminated-type magnet (Image: Stanford Magnets/Oceania International LLC).

Q: What are the relative attributes of AFMs compared to RFMs?
A: This is not an easy question to answer. Proponents of AFMs tend to “play up” the virtues and downplay their weaknesses. At the same time, RFMs have a long history in both manufacturing processes and in-the-field performance, and those are significant benefits in most motor applications. Finally, the relative attributes do have many exceptions based on the specifics of their design and construction, as well as the application details.

Q: There must be some general characterizations, right?
A: Yes, there are some generalizations — and they are generalizations — detailing what AFMs offer compared to RFMs:

Compact structure, especially short axial sizes.
Small volume.
Low weight.
High torque density (Note that this is not the same as torque!)
High power density.
Small end-winding.
Highly efficient and thus reduced cooling requirements.

Q: It seems like the AFM benefits are clear, but are they?
A: Not really. While they can reduce cooling needs, they also have less surface area to evacuate heat than radial-flux motors.

Q: What are some of the issues associated with using AFMs rather than RFMs?
A: From an electromagnetic perspective, it’s been known for a long time in the academic world and benchtop that the AFM topology is more effective with higher torque density, which is a key figure of merit. Nonetheless, nearly all traction motors in railcars and EVs — applications where torque density is critical — still use the RFM.

Q: Can you give some specifics?
A: The challenges of building axial flux motors range from the design phase to large-scale manufacturing.

Consider CAD tools and design software. The electromagnetic design of radial flux machines can largely be done with two-dimensional (2D) calculations, while the magnetic flux of axial flux machines flows in three dimensions and requires 3D simulation software. This level of tool only became available recently and requires powerful hardware. Note that very effective RFMs were designed and used prior to the existence of any CAD at all.

Q: But once a design is done, what’s the impediment to using AFMs?
A: In the single stator — dual rotor topology of a practical yokeless AFM, more on that later), there are two air gaps: one between the first rotor disk and the stator and another between the second rotor disk and the stator. A uniform air gap is essential between the rotor and stator to minimize acoustical noise, vibration, and harshness (rough running). Since RFMs have only one air gap, this uniformity is easier to establish and maintain compared to doing so in an AFM, both in the manufacturing phase as well as field deployment and use.

Q: Is there more?
A: Absolutely. Recall the perspectives on the thermal performance of the two motor types. The specifics are subtle: In the case of the RFM, the stator is typically cooled indirectly from the outside; in extreme cases, this is done via a water jacket in the housing. More recent RFMs are cooled via direct oil cooling, which is more effective than indirect cooling since it removes the heat directly from where it is generated.

For AFMs, the cooling thermal-flow path is more complex, so getting the heat out and away from the motor is more challenging, even if the amount of heat to be removed is less. The windings of AFMs are located deep within the stator and between the two rotor discs, making it difficult to dissipate the heat.

Q: What about production issues?
A: The manufacture of RFMs is a very mature process and highly automated. There are many steps, such as material sourcing and preparation, punching parts, welding, coil winding, and final test, which are characterized and optimized. In contrast, practical volume production of AFMs is more challenging and less standardized, for now.

The final part of this FAQ looks at some real-world issues associated with the adoption of AFMs.