How To Evaluate The High-Speed Field-Weakening Capability of Axial Flux Motors?

In recent years, axial flux motors have gained increasing traction in the electric drive community. They are regarded by many as one of the representative directions for next-generation high-performance motors, especially in applications where volume and weight are critical—such as in-wheel motors, drones, electric sports cars, and hybrid systems. The axial flux structure does offer unique advantages in these scenarios.

However, to truly apply axial flux motors in high-speed electric drive systems, one technical challenge is unavoidable: Is the high-speed field-weakening capability sufficient? And how should it be evaluated?

I. First, understand: Why is "field weakening" always necessary at high speed?

A fundamental characteristic of permanent magnet synchronous motors (PMSMs) is that the back EMF increases with speed.

Back EMF can be simply understood as the voltage "generated" by the motor itself as it rotates. When the speed rises to a certain point, the back EMF approaches or even exceeds the maximum voltage that the controller can supply—i.e., the DC bus voltage. At that point, the motor can no longer inject current or increase speed; it is effectively blocked by a "voltage wall."

So what can be done?

The method is: actively reduce the motor's flux linkage, i.e., make the magnetic field "weaker", thereby lowering the back EMF and freeing up voltage headroom for higher speeds. This is what we commonly call "field-weakening control."

An analogy:
A standard bicycle needs a large gear ratio for starting and high torque; if you keep that large ratio at high speed, your legs cannot keep up with the cadence. Field-weakening control is like shifting to a smaller gear ratio at high speed, allowing the vehicle speed to continue increasing while your leg cadence stays the same.

II. The particularities of axial flux motors in field weakening

Axial flux motors are not a single fixed design, but rather a broad category of topologies. Their common feature is that the air-gap flux flows in the axial direction, rather than radially as in conventional radial flux motors.

This structure brings advantages such as high torque density, flat form factor, and short axial length, but also introduces some characteristics relevant to field-weakening capability.

One of the most influential factors is: many axial flux motors have relatively low inductance.

The field-weakening capability of a motor is closely related to the ratio of the direct-axis inductance Ld to the flux linkage Ψm. Simply put, the larger the direct-axis inductance and the smaller the flux linkage, the easier field weakening becomes in theory, and the stronger the high-speed extension capability.

Many axial flux motors, in pursuit of high torque density, have short magnetic paths and short end windings, resulting in naturally low inductance. Consequently, a larger field-weakening current is required to cancel the flux linkage, leading to a significant increase in copper loss at high speed, a drop in efficiency, and possibly even insufficient field-weakening capability.

But note: not all axial flux motors have poor field-weakening performance.
It depends on the specific design. For example, using multi-layer magnets, increasing the pole-pair number, optimizing winding configurations, or increasing the reluctance torque contribution can all improve field-weakening performance.

Therefore, when evaluating, one cannot simply say "axial flux motors always have weak field weakening"; it must be assessed on a case-by-case basis for each specific design.

III. What are the main indicators for evaluating field-weakening capability?

In engineering practice, evaluating the high-speed field-weakening capability of an axial flux motor typically involves the following core parameters and curves.

1. Characteristic current

Characteristic current is defined as:

 where Ψm is the permanent magnet flux linkage, and Ld is the direct-axis inductance.

If the characteristic current is less than the controller's maximum output current, then the motor can theoretically enter the "infinite field-weakening" region, allowing for very high speed extension.
If the characteristic current is greater than the maximum current, it means that even using all current for field weakening is insufficient to cancel the flux linkage, limiting high-speed extension capability.

Characteristic current is the most intuitive indicator for evaluating field-weakening potential.

2. Voltage limit circle and current limit circle

On the current vector plane, motor operation is constrained by two circles:

• Current limit circle: determined by the controller's maximum output current;

• Voltage limit circle: determined by the DC bus voltage and the current speed.

As speed increases, the voltage limit circle shrinks. The motor's operating point must lie within the intersection of these two circles.

A motor with strong field-weakening capability still has a sufficient intersection area at high speed to deliver useful torque.
A motor with poor field-weakening capability sees the voltage limit circle shrink too quickly, squeezing the operating point into a very small region, and output capability drops sharply.

3. Constant-power region width

In engineering, the "constant-power region" is often used to evaluate high-speed extension capability.
A wider constant-power region indicates that the motor can maintain relatively high power output at high speeds, rather than experiencing a cliff-like drop in torque.

If an axial flux motor is poorly designed for field weakening, its constant-power region will be noticeably narrow, and power at the high-speed end will drop quickly.

4. Ratio of maximum speed to rated speed

This ratio is also called the "field-weakening speed expansion ratio."
For typical passenger vehicle drive motors, an expansion ratio of 3 to 4 times or more is desired—for example, rated at 4000 rpm and maximum at 15000–18000 rpm.
Whether an axial flux motor can achieve this level depends on a combined evaluation of inductance, flux linkage, pole-pair number, and controller voltage capability.

5. High-speed efficiency and temperature rise

Field weakening is not just about "whether it can spin up," but also "how efficient it is when it does."
If the field-weakening current is excessively large, copper loss increases dramatically, leading to severe temperature rise at high speed, which may even cause irreversible demagnetization of the permanent magnets.

Therefore, when evaluating field-weakening capability, it is essential to also examine the high-speed efficiency MAP and thermal performance—not just the speed numbers.

IV. How to evaluate in practice? Three approaches

Approach 1: Electromagnetic simulation

During the design phase, engineers typically build an electromagnetic model of the axial flux motor using finite element simulation software.

By sweeping different speeds and current vectors, the following can be obtained:

• Back EMF waveform;

• Direct-axis and quadrature-axis inductances Ld, Lq;

• Flux linkage Ψm;

• Characteristic current Ich;

• Intersection of voltage limit ellipse and current limit circle;

• Theoretical maximum torque-speed curve;

• Efficiency MAP.

These data can comprehensively reflect the motor's field-weakening potential.
However, for accurate simulation results, the model must faithfully represent axial flux end leakage flux, 3D magnetic circuit paths, and saturation characteristics. As a result, simulation of axial flux motors relies more heavily on 3D finite element analysis compared to conventional radial motors.

Approach 2: Bench testing

Simulation is only theoretical; ultimately, bench measurements are required.

Bench tests typically include:

• Back EMF coefficient measurement;

•  Inductance measurement at different temperatures;

• Peak torque-speed external characteristic curve;

• Continuous rated power output capability;

• Sustained operation time at maximum speed;

• Efficiency and temperature rise in the high-speed field-weakening region;

• Performance degradation test after magnet demagnetization.

The most critical are temperature rise in the high-speed field-weakening region and continuous operation capability.
Many axial flux prototypes can reach very high speeds in short-duration peak tests, but if they overheat and derate after a few minutes of continuous operation, they cannot be considered as having truly qualified field-weakening capability.

Approach 3: Control strategy matching evaluation

Field-weakening capability depends not only on the motor itself, but also on the controller and control algorithms.

Common control strategies include:

• MTPA (Maximum Torque per Ampere);

• Field-weakening control;

• MTPV (Maximum Torque per Voltage);

• Voltage feedforward and current decoupling in deep field-weakening.

The evaluation should examine whether the control strategy can fully exploit the voltage and current limits of the axial flux motor.
If the motor body has good field-weakening potential but the control algorithm is inadequate, issues such as jitter, loss of control, or excessively low efficiency may occur at high speeds.

V. Which design improvements can enhance the high-speed field-weakening capability of axial flux motors?

If evaluation reveals insufficient field-weakening capability, common improvement directions include:

1. Optimize direct-axis inductance

Appropriately increasing the direct-axis inductance can lower the characteristic current and enhance field-weakening capability.
This can be achieved by increasing the pole-pair number, adjusting slot geometry, optimizing winding layout, increasing magnetic bridge width, etc.

However, increasing direct-axis inductance often conflicts with torque density, so multi-objective optimization is required.

2. Moderately reduce flux linkage

Using magnets with slightly lower remanence but higher temperature tolerance, or reducing magnet thickness, can lower back EMF and characteristic current.
The trade-off is a reduction in low-speed torque, which needs to be weighed against overall system requirements.

3. Increase reluctance torque contribution

By increasing the saliency ratio (difference between quadrature and direct-axis inductances), reluctance torque can assume a larger share of output, reducing reliance on permanent magnet flux linkage.
In this way, at high-speed field weakening, flux linkage occupies less voltage headroom, making speed extension easier.

4. Adopt hybrid excitation or variable flux

This is a more advanced approach.
Incorporating field windings or variable-flux magnets in the axial flux structure allows active reduction of air-gap flux at high speed, transforming "passive field weakening" into "active field weakening."

This design can significantly improve high-speed efficiency and constant-power region width, but increases structural complexity and cost.

5. Improve DC bus voltage utilization

If the motor itself has modest field-weakening capability, it is also possible to push back the "voltage wall" by raising the controller voltage level, using SiC inverters, optimizing modulation strategies, etc., to deliver more effective voltage from the same bus voltage.

This leaves more headroom for field weakening.

VI. Common misconceptions

When evaluating the high-speed field-weakening capability of axial flux motors, several misconceptions are worth noting.

Misconception 1: Only looking at the maximum speed.

A high maximum speed does not necessarily mean strong field-weakening capability—it may just be a short-term speed burst without sustained useful power output.

Misconception 2: Believing that axial flux motors always have poor field weakening.

As mentioned earlier, axial flux is a topological concept, and specific field-weakening performance varies widely. Some high-performance axial flux motors can even outperform comparable radial flux motors in field-weakening capability.

Misconception 3: Ignoring temperature effects.

Both permanent magnet flux linkage and winding resistance vary with temperature. At high temperatures, flux linkage decreases, altering the field-weakening current requirements. Evaluation should use hot-state parameters, not just cold-state data.

Misconception 4: Treating simulation maximum speed as actual usable speed.

Simulations are often run under ideal voltage, current, and temperature conditions. In actual bench testing, controller nonlinearities and thermal protection will limit the practical field-weakening region.

VII. Conclusion

The high-speed field-weakening capability of an axial flux motor is, in essence, a system-level engineering problem.

It depends not only on the electromagnetic design of the motor body, but also on the control strategy, inverter capability, and thermal management.
Evaluation should not rely on just one or two parameters; instead, it requires a comprehensive judgment based on characteristic current, voltage/current limit circles, constant-power region width, high-speed efficiency, temperature rise, and continuous operation capability.

For engineers aiming to truly push axial flux motors into high-speed electric drive applications, field-weakening capability is not a matter of "whether it exists," but rather "whether the motor can extend speed sustainably and efficiently."

In the future, with advances in new materials, new topologies, and multi-physics optimization techniques, the performance of axial flux motors in the high-speed field-weakening region will continue to improve.
And accurately and systematically evaluating their field-weakening capability will increasingly become a critical link in electric drive development.