Advanced Axial Flux Motor Design
Calculate ideal torque from power and speed. Compare AFPM topology and cooling trade-offs before detailed design review.
For OEM product and engineering teams screening an operating point before detailed design validation.
·
Prepared by AFPM Motor; external benchmark sources are identified below.
Key Design Conclusions
Review these engineering facts before committing to an advanced axial flux topology.
Topology changes the axial force path and service access
A single-sided SSSR arrangement has a net axial attraction that the shaft, bearings, and frame must carry. SSDR and DSSR arrangements can oppose axial forces, but force balance depends on matched air gaps, magnetic loading, and structural deflection; topology alone does not guarantee cancellation.
The force path follows the arrangement of the active sides. Confirm net thrust and deflection with electromagnetic and structural analysis for the actual geometry.
Cooling limits depend on the complete operating point
Power and speed alone do not reveal winding loss, magnet temperature, or allowable continuous torque. Duty cycle, geometry, materials, coolant conditions, and flow paths determine whether a cooling concept can hold temperature limits.
A published 100 kW AFPM test motor demonstrated a direct-liquid cooling concept using oil; the result supports that specific prototype, not a universal cooling threshold.
Treat public power-density figures as configuration-specific benchmarks
Separate short-term peak output from continuous output and compare only figures with a stated mass basis and test conditions. Prototype records and production motor data describe different operating envelopes.
YASA labels its 2025 prototype result as short-term peak output and estimates a lower continuous output; its 750R datasheet separately publishes continuous and peak ratings with cooling conditions.
Torque Method and Design Validation
The configurator uses the shaft power and speed relationship T = P / ω, or approximately 9,550 × power (kW) / speed (RPM) for torque in newton-metres. It assumes the entered power is continuous mechanical output at the shaft and does not subtract motor, inverter, or drivetrain losses.
This operating-point calculation cannot estimate active mass, power density, air-gap shear stress, losses, or thermal limits. Those outputs require dimensions, electromagnetic loading, materials, duty cycle, and a defined cooling circuit.
Inputs needed for a feasibility decision
A design review needs a defined operating envelope and package, then checks the coupled electromagnetic, thermal, and mechanical constraints. At minimum, specify:
- Torque-speed duty cycle, overload duration, and lifetime
- Available diameter, axial length, shaft, and air-gap targets
- Voltage, current, inverter limits, and winding constraints
- Winding and magnet temperature limits plus ambient conditions
- Coolant type, inlet temperature, flow, pressure drop, and sealing
Advanced design review sequence
- Define the operating envelope. Record the torque-speed duty cycle, overload duration, voltage and current limits, package, environment, and lifetime.
- Select a topology and force path. Compare active sides, flux path, stator access, rotor support, and expected axial loads for the proposed geometry.
- Estimate electromagnetic losses and torque. Use a defined winding, magnetic circuit, and operating point; do not infer shear stress or power density from torque alone.
- Close the thermal and mechanical checks. Evaluate loss-dependent temperatures, coolant boundaries, air-gap deflection, bearings, shaft, and magnet retention.
- Validate the candidate design. Compare electromagnetic, thermal, and structural models with prototype measurements under a stated duty cycle before releasing a rating.
Carry assumptions and boundary conditions through each review gate. This sequence is a scoping framework, not a substitute for a geometry-specific design or test plan.
Peak output is not a continuous rating
Short-term peak power can be several times a motor’s continuous output because temperature rises over time. Compare ratings only when duration, coolant conditions, installation, and the mass definition are clear. The source-backed examples below are reference points for specific hardware, not scaling promises.
Axial Force Path and Cooling Trade-offs
| Topology | Force path | Cooling consideration | Key validation |
|---|---|---|---|
| SSSR single stator, single rotor | One active side; support the net axial attraction. | Stator access depends on housing and winding layout. | Bearing load, air-gap deflection, and rotor support. |
| SSDR single stator, dual rotor | Opposed rotor forces may reduce net thrust if symmetric. | Central-stator heat path and coolant access need a deliberate design. | Rotor retention, symmetry, and stator temperature. |
| DSSR dual stator, single rotor | Opposed stator forces may reduce net thrust if symmetric. | External stator surfaces can support jacket concepts; geometry still governs. | Stator support, rotor stiffness, both air gaps, and coolant flow. |
Public AFPM Benchmarks: Peak and Continuous Ratings
Manufacturer-published data, reviewed October 12, 2026. Ratings apply to the cited hardware and test conditions; they are not predictions for the configurator inputs.
| Hardware / date | Published output | Conditions and limits |
|---|---|---|
| YASA 750R datasheet rev. 11, June 2019 | 37 kg; up to 70 kW continuous at 3,000 rpm; 200 kW peak. | Continuous rating specifies 40 °C coolant; sheet notes results vary with cycle, cooling, and installation. |
| YASA prototype October 22, 2025 | 12.7 kg; 750 kW short-term peak; 59 kW/kg. YASA estimates 350–400 kW continuous. | The 59 kW/kg figure is a manufacturer-reported peak prototype result; continuous output is an estimate, not the same test rating. |
Design Risks and the Evidence Needed
| Risk | Why it matters | Next check |
|---|---|---|
| Axial force and air-gap closure | Force imbalance or deflection can reduce clearance or overload bearings. | Electromagnetic force calculation plus shaft, bearing, and structural deflection analysis. |
| Thermal overload | Copper and magnet temperatures depend on losses and heat paths, not the cooling label alone. | Loss map, duty cycle, thermal model, coolant boundary conditions, and prototype temperature test. |
| Cost and build variation | Precision gaps, magnets, winding process, and cooling hardware affect yield and supply cost. | Manufacturing review, tolerance stack, supplier quotes, and pilot-build yield plan. |
The benchmark values below come from the cited research and manufacturer publications. They are external reference data, not results produced by this configurator.
Sources & References
Axial Force Negative Stiffness in Axial-Flux Electric Machines (September 3, 2025)
Key Claim: Analyzes axial force variation with air gap in induction and PM axial-flux machines, validated against finite-element and experimental results. It examines double-sided machines, supporting the need to check residual force and force stiffness for the actual geometry.
Read IET research paperYASA next-generation axial-flux prototype announcement (October 22, 2025)
Key Claim: YASA reports 750 kW short-term peak output from a 12.7 kg prototype (59 kW/kg) and estimates 350–400 kW continuous output. The continuous figure is an estimate from the manufacturer, not the reported peak test.
Read YASA announcementYASA 750R electric motor datasheet, revision 11 (June 2019)
Key Claim: Lists 37 kg mass, up to 70 kW continuous at 3,000 rpm under its 40 °C coolant rating, and 200 kW peak at 700 V DC. The sheet also states dynamometer conditions and warns that performance varies with drive cycle, cooling, and installation.
Read 750R datasheetDirect Liquid Cooling in Low-Power Electrical Machines: Proof-of-Concept (December 2016)
Key Claim: Reports a 100 kW double-stator, single-rotor AFPM prototype using a closed oil coolant loop and evaluates the demonstrated cooling approach. This is a case study, not a general shear-stress rating.
Open IEEE paperFrequently Asked Questions
Design Feasibility
Which topology is best for my application?
There is no universal winner. Compare axial force, cooling access, rotor support, package diameter and length, speed, duty cycle, and manufacturing constraints. Use the tool for ideal torque demand, then validate a defined geometry.
How do you handle the axial magnetic pull?
Calculate net thrust over both air gaps and size the shaft, bearings, and frame for that load. Opposed active sides can reduce net axial force when the magnetic circuits and mechanical geometry are sufficiently symmetric; verify residual force and deflection.
Does the calculator tell me whether the design is feasible?
No. It calculates ideal torque from entered continuous shaft power and speed. It does not know motor dimensions, losses, materials, thermal limits, or structural loads, so it cannot validate feasibility.
Do SSDR and DSSR topologies automatically cancel axial force?
No. Opposed active sides can reduce net force, but air-gap mismatch, magnetic loading, tolerances, and structural deflection leave residual load. Calculate force and deflection for both sides in the actual assembly.
Manufacturing & Thermal Constraints
Can I choose cooling from power and speed alone?
No. Cooling selection needs loss estimates, winding and magnet limits, duty cycle, ambient conditions, coolant inlet temperature and flow, and the thermal path through the chosen geometry. Liquid and direct-oil concepts also add pump, sealing, pressure-drop, and integration work.
Is a coreless stator cheaper to manufacture?
Not necessarily. Removing a laminated core changes the electromagnetic design and manufacturing route; compare winding manufacture, support, magnet requirements, losses, and assembly yield against the saved core material and processing.
Why is peak power density different from continuous power density?
Peak output may be limited to a short duration before temperatures rise. Continuous output depends on steady-state heat rejection and specified coolant, ambient, and installation conditions. Always compare like-for-like ratings and mass definitions.
What information is needed for a real design review?
Provide the duty cycle, torque-speed envelope, voltage and current limits, package dimensions, target air gaps, thermal limits, coolant conditions, environment, lifetime, and manufacturing volume. Detailed electromagnetic, thermal, and structural analysis then checks the candidate design.
Does direct oil cooling guarantee higher continuous torque?
No. Direct oil can shorten the heat path in a particular design, but the result depends on flow distribution, losses, materials, seals, pump power, and temperature limits. Prototype thermal testing is needed to substantiate a continuous rating.
How should cooling options be compared on cost?
Include the motor cooling hardware plus pumps, plumbing, seals, controls, pressure drop, service needs, coolant compatibility, and production yield. Compare those system costs against the required duty cycle and thermal margin.
Ready to prototype your design?
Send us your calculated estimates and application constraints. Our engineering team will review the feasibility of your chosen topology and recommend an optimal manufacturing path.
Request Engineering Review