Axial Flux Motor Magnets Checker for 10.5 kW / 45 Nm Designs
Default 10.5 kW / 45 Nm point
2,228 rpm
9.7 kPa estimated average shear; no pass/fail threshold is applied.
Screening calculations onlyFor a 10.5 kW, 45 Nm permanent-magnet axial-flux brushless motor design, this page screens speed and geometry assumptions but cannot confirm that a catalog unit exists. BIZ Karts lists those ratings for a permanent-magnet brushless motor, but its listing does not identify the topology. Request the supplier datasheet before selecting magnets; all outputs are preliminary screening calculations.
Published: April 4, 2026 · Evidence reviewed: October 3, 2026
Quick starts
5-phase dual-rotor and 100 rpm class points require high shaft torque. Use a preset to avoid unit mistakes, then tune one variable at a time.
This value is copied into the handoff summary only; no phase-current, DC-bus-current, copper-loss, or inverter-load calculation is performed.
Profile: Higher remanence at same SH temperature class; common for compact torque targets.
The temperature reference is an example grade profile, not a design limit. Verify the supplier’s exact grade, geometry, and operating conditions.
Report Summary: core conclusions and who this fits
Tool layer solves the immediate sizing question. This section turns the output into an actionable decision with suitability boundaries.
2,228 rpm
Calculated from P = T * w at 10.5 kW / 45 Nm.
9.7 kPa
Uniform tangential loading over the active annuli using an area-weighted mean radius; no universal pass threshold is implied.
371 Hz
Derived from speed and pole pairs; phase architecture and inverter strategy must still be validated separately.
30 C
A catalog temperature comparison only; it does not establish demagnetization or service-life margin.
Share duty-cycle, thermal limit, and preferred grades to get a practical recommendation.
| Audited gap | Decision risk | What was added |
|---|---|---|
| Core conclusions were not directly mapped to evidence IDs. | Readers could not quickly verify which source supports each action. | Mapped core conclusions to their supporting evidence and expanded the evidence table. |
| Material section lacked quantified thermal-vs-magnetic tradeoff. | Teams may over-upgrade to high-temperature grades and miss torque-density targets. | Added numeric Br/BHmax/max-temp comparison and explicit counterexample path. |
| Supply-risk layer lacked dated disruption timeline. | Procurement plans might assume static price/lead-time behavior. | Added 2024-2025 concentration + export-control shock facts with actions. |
| Lacked explicit operating point analysis for the 10.5 kW / 45 Nm alias. | Readers could confuse shaft speed, DC-bus current, and motor phase current. | Shows ~2,228 rpm at the stated shaft point and labels 218.75 A at nominal 48 V as ideal DC-side arithmetic before losses. |
| Earlier mechanical integration guidance exceeded its cited evidence. | Readers could mistake a topology choice or supplier claim for a universal design rule. | Removed unverified bearing-force and air-gap claims from the evidence-backed conclusions. |
| Lacked direct-drive generator start-up context for the 100 rpm alias. | Coreless topology evidence could be overgeneralized into a mandatory 100 rpm design rule. | Retained the cited coreless/cogging trade-off and marked 100 rpm start-up as application-specific. |
| Conclusion | Applies when | Counterexample / limit | Evidence tags |
|---|---|---|---|
| Use continuous (S1/30-minute equivalent) power as checker input, not peak-only S2 values. | Applies at architecture screening and supplier comparison stages. | If only burst/peak data are available, results are preliminary and can overstate feasibility. | S2-S3 |
| Thermal-grade upgrades (SH -> UH/SmCo) are not free; magnetic loading can drop and force size/current redesign. | Applies when hotspot uncertainty is high or expected operation exceeds SH margin. | Blindly shifting to higher-temp grade can reduce Br/BHmax and fail the same torque target. | S11-S12 |
| Single-country magnet sourcing should be modeled as schedule risk, not just unit-cost risk. | Applies to EU/US programs exposed to rare-earth magnet imports. | If dual-source and fallback-grade contracts already exist, 2025-style shocks are less disruptive. | S1, S9-S10 |
| Axial-flux brochure comparisons remain non-normalized unless duty cycle, cooling, and rating basis match. | Applies to cross-vendor benchmarking and make/buy decisions. | When an internal protocol harmonizes duty cycle and cooling, comparison confidence improves. | S6-S7, pending data note |
| At the stated shaft point, 10.5 kW / 45 Nm corresponds to about 2,228 rpm. At nominal 48 V, 10.5 kW also corresponds to 218.75 A ideal DC input before losses or bus-voltage sag. | The P/T speed calculation assumes the power and torque describe the same operating point. The P/V current is DC-side arithmetic, not phase current. | Actual speed, bus current, and current waveform depend on the duty point, voltage under load, inverter, and motor efficiency. | Calculated from P/T and P/V |
| BIZ Karts lists the 10.5 kW / 45 Nm permanent-magnet brushless motor on its product page and separately lists a 48 V / 350 A Curtis ECU part. | Neither listing identifies the motor as axial flux or states whether 350 A is a continuous, peak, DC, or phase-current rating. | Do not treat one product’s controller specification as a general design requirement or evidence of an axial-flux implementation. | S15-S16 |
| Coreless axial-flux topologies can avoid cogging torque, but that does not make them a universal requirement for 100 rpm generators. | The cited paper supports the coreless/cogging trade-off; 100 rpm start-up still depends on the prime mover, load, winding, and converter. | An iron-core topology may remain viable when its cogging torque is compatible with the available starting torque. | S18 |
| Area | New fact | Decision impact | Time scope | Sources |
|---|---|---|---|---|
| Supply volatility update | USGS 2026 shows apparent U.S. consumption jump (9,010 -> 27,000 t REO from 2024 to 2025) and NdPr oxide increase ($55/kg -> $69/kg). | BOM sensitivity checks should use updated 2025 price/disruption window, not older 2024 baseline. | 2024-2025 data, published Feb 2026 | S1 |
| Power rating boundary | UN/ECE R85 defines maximum 30-minute power as 30-minute average and requires >90% net-power speed band for the test. | If input power is short-term S2 peak, this checker can overstate feasibility. Convert to continuous basis first. | Regulation text in force; accessed 2026-04-04 | S2 |
| Peak vs continuous evidence | EMRAX 228 v1.6 publishes both S2 peak and S1 continuous ratings (124/75 kW, 220/130 Nm). | Use continuous (S1/30-minute equivalent) for architecture screening; keep peak for overload checks only. | Version 1.6 (Mar 2025) | S3 |
| EU compliance planning | CRMA sets 2030 10/40/25/65 benchmarks and permanent-magnet labeling/data-carrier + recycled-content disclosure timelines. | EU-bound product programs need magnet traceability and recycled-content roadmap in sourcing plans. | Regulation dated Apr 11, 2024 | S4-S5 |
| Concentration and disruption timeline | IEA reports around 90% magnet production concentration in China (2024), and a 2025 control wave linked with large regional price dislocation. | Single-country sourcing should be treated as a schedule risk, not only a cost variable. | 2024-2025 events, published Oct-Dec 2025 | S9-S10 |
| Thermal class tradeoff is quantifiable | Catalog data show moving from SH to UH can increase operating temperature class but reduce Br/BHmax significantly. | Do not upgrade temperature class without re-running torque and volume assumptions. | Catalog data accessed 2026-04-04 | S11-S12 |
| 48 V ideal DC current at 10.5 kW | At a nominal 48 V DC bus, 10.5 kW / 48 V = 218.75 A ideal DC input current before conversion losses or bus-voltage sag. This is not motor phase current. | Use actual minimum bus voltage, efficiency, transient duty, and inverter/battery ratings for system sizing; the checker does not infer those values. | Calculated from 10.5 kW / nominal 48 V | P / V |
| One commercial 10.5 kW / 45 Nm example | BIZ Karts lists the motor at 10.5 kW / 45 Nm; a separate BIZ Karts parts listing describes a 48 V / 350 A Curtis ECU. Neither listing establishes an axial-flux topology or motor-current basis. | Treat this as one product specification, not a market standard, proof of axial-flux feasibility, or a transferable controller-sizing rule. | Manufacturer product and parts listings accessed October 3, 2026 | S15-S16 |
| Coreless topology and cogging | A cited axial-flux machine-model paper identifies the absence of cogging torque as a coreless-topology advantage; it does not establish a universal 100 rpm design requirement. | Treat low-speed start-up as a system-specific torque and electrical design question, and compare the coreless topology tradeoffs. | Paper published 2020; accessed October 3, 2026 | S18 |
1. You need a fast go/no-go before detailed FEA cycles.
2. The product is in early architecture phase and magnet grade is still negotiable.
3. You can bound hotspot temperature with at least one realistic cooling scenario.
4. Your sourcing team needs a risk-aware fallback grade path.
1. You are already in final release and only tolerance confirmation remains.
2. Cogging, NVH, and harmonic torque ripple are dominant constraints.
3. Duty cycles include severe overload transients without thermal test evidence.
4. Contract terms require certified lot-by-lot demagnetization curves.
Deep Layer: method, evidence, and data caveats
We separate deterministic physics calculations from market evidence, and we explicitly label unknowns.
Formula block
base rpm = 9550 * P(kW) / T(Nm)
f_elec = rpm * polePairs / 60
average shear = T / (annulus area * mean radius), assuming uniform loading over entered active surfaces and an area-weighted annulus radius.
Boundary logic
The temperature comparison flags when the entered target reaches the selected profile reference; it is not a demagnetization limit.
Shear stress and electrical frequency are reported as calculated values without generic pass/fail thresholds.
Uncertainty policy
Unknown values are shown as N/A with reason.
Vendor benchmarks are marked as non-normalized claims.
| Decision gate | Boundary condition | Fallback when data is missing |
|---|---|---|
| Is your 10.5 kW value continuous (or 30-minute) power? | Use continuous/rated value for this checker. Peak-only values (S2, dyno burst) are not equivalent. | Mark result as preliminary and request S1 curve or 30-minute rating from supplier. |
| Do you have temperature-dependent B-H curves for the shortlisted lot? | Thermal margin alone is insufficient for irreversible demagnetization sign-off. | Treat demag risk as "pending confirmation" and block final design freeze. |
| Is the product sold into EU categories covered by CRMA Article 28/29? | Labeling, data-carrier traceability, and recycled-content disclosures may become mandatory by timeline. | Open a compliance workstream before procurement lock. |
| Are you comparing axial and radial options with harmonized duty cycles? | Public cross-vendor datasets are not normalized across cooling and duty definitions. | Use internal A/B test protocol instead of direct brochure comparison. |
| If you upgrade magnet grade for heat, did you recheck Br/BHmax impact on torque density? | Higher-temperature grades can carry lower Br/BHmax; thermal gain can come with magnetic loading loss. | Mark output as provisional and run geometry-current re-optimization before RFQ lock. |
| Can your sourcing plan survive a 2025-style export-control shock? | Supply concentration and controls can create large regional price and lead-time spread in months, not years. | Create dual-source and alternate-grade path before committing launch timing. |
| If this is a five-phase design, is five-phase inverter/control support locked? | Five-phase topology needs matching modulation, fault handling, diagnostics, and service tooling. | Treat the result as boundary state and validate control-chain readiness before RFQ freeze. |
| Are you pairing the 10.5 kW / 45 Nm motor with a 48V battery system? | At nominal 48 V, 10.5 kW corresponds to 218.75 A ideal DC input current before losses and bus-voltage sag. This is not phase current and does not define a universal safe current limit. | Request the battery voltage window, efficiency, duty cycle, and continuous/peak ratings for the battery, inverter, and wiring before choosing a bus voltage. |
| ID | Source | Key data | Use in this page | Date |
|---|---|---|---|---|
| S1 | USGS Mineral Commodity Summaries 2026 (Rare Earths chapter) | 2025 U.S. apparent consumption of rare-earth compounds/metals: 27,000 t REO (vs 9,010 in 2024); net import reliance: 67%; China share of 2021-24 U.S. imports: 71%; NdPr oxide average: $69/kg in 2025 (from $55/kg in 2024). | Supply chain and cost-risk context for NdFeB dependent motor designs. | February 2026 |
| S2 | UN/ECE Regulation No. 85 (measurement of net and maximum 30-minute power) | "Maximum 30 minutes power" is defined as 30-minute average net power. Test speed must be where net power is above 90% of maximum, with 25 +/- 5 C conditioning requirements. | Boundary for interpreting whether 10.5 kW should be treated as continuous/rated power or as short-duration peak. | Accessed October 3, 2026 |
| S3 | EMRAX 228 datasheet v1.6 | Headline values: 124 kW peak / 75 kW continuous, 220 Nm peak / 130 Nm continuous. Table explicitly labels peak as S2 (2 min) and continuous as S1. | Concrete reminder that peak and continuous ratings are not interchangeable inputs for this checker. | Version 1.6 (March 2025), accessed October 3, 2026 |
| S4 | Regulation (EU) 2024/1252 (Critical Raw Materials Act) | By 2030 benchmarks for strategic raw materials: 10% extraction, 40% processing, 25% recycling, and no more than 65% dependence on one third country at relevant processing stage. | Procurement strategy boundary for region-sensitive supply planning and single-country concentration risk. | April 11, 2024 (in force May 23, 2024) |
| S5 | EU 2024/1252 Articles 28-29 (permanent magnet obligations) | Article 28 applies to MRI devices, motor vehicles, and type-approved category L vehicles from May 24, 2029; its data-carrier requirements are staged from the relevant implementing act. Article 29 recycled-content disclosure applies to covered products with >0.2 kg eligible magnets by May 24, 2027 or two years after the relevant delegated act takes effect, whichever is later. Article 29(3) requires the Commission to adopt delegated acts setting future minimum shares by December 31, 2031; that date is not a manufacturer compliance deadline for those shares. | Compliance timeline and redesign risk for EU-bound products that contain NdFeB/SmCo/AlNiCo magnets. | April 11, 2024 |
| S6 | MDPI Energies 2024 topology comparison (YASA, AFIR, offset AFIR) | Under shared analytical + 3D FEA comparison conditions: YASA minimized core losses, AFIR achieved highest torque density, and offset AFIR showed highest efficiency and higher power factor. | Independent evidence that topology choice depends on objective function; there is no universal best axial-flux layout. | January 13, 2024 |
| S7 | OSTI entry for IEEE TIA review of AFPM machines | Review note: AFPM and radial PM machines share principles, but AFPM geometry introduces additional analysis complexity across design and optimization. | Supports why this page is a screening tool and not a substitute for full electromagnetic + thermal + mechanical model closure. | March 2023 |
| S8 | YASA 2025 prototype release (vendor claim) | Vendor-reported short-term peak benchmark: 750 kW from 12.7 kg (59 kW/kg), with estimated continuous power 350-400 kW. | Kept as clearly labeled vendor benchmark only; not treated as normalized cross-vendor lab baseline. | October 22, 2025 |
| S9 | IEA Renewables 2025 executive summary | In 2024, China controlled about 60% of rare-earth mining, nearly 90% of refining, and around 90% of rare-earth magnet production. | Baseline concentration risk used to justify dual-source procurement logic. | October 2025 |
| S10 | IEA commentary on 2025 export controls | China exported about 58,000 tonnes of rare-earth magnets in 2024. Following April 2025 controls, prices in Europe rose to up to six times China levels. | Provides dated shock scenario for lead-time and pricing stress tests. | December 11, 2025 |
| S11 | Arnold Magnetic Technologies NdFeB grade table | Example catalog values: N48SH about 1.39 T Br / 47 MGOe and 150 C max operating temperature, while N30UH is about 1.125 T Br / 31 MGOe and 180 C max. | Quantifies thermal-grade upgrades versus magnetic-loading tradeoff. | Accessed October 3, 2026 |
| S12 | Arnold Magnetic Technologies RECOMA SmCo data table | Example Sm2Co17 35E catalog line: about 1.19 T Br, 33.3 MGOe, and up to 300 C max operating temperature. | Supports high-temperature fallback positioning and related tradeoff comments. | Accessed October 3, 2026 |
| S15 | BIZ Karts EcoVolt GT manufacturer specification | The EcoVolt GT product page lists a 10.5 kW / 45 Nm permanent-magnet brushless motor; it does not identify the motor as axial flux. | This is one published product example of the requested power/torque point, not evidence that it is an industry standard or an axial-flux design. | Accessed October 3, 2026 |
| S16 | BIZ Karts EcoVolt GT ECU parts listing | A separate official parts listing names a Curtis ECU as 48 V / 350 A; this does not establish motor phase current or whether that ECU rating is continuous or peak. | Keeps the controller part description separate from the motor rating; do not use it as a universal current-sizing value. | Accessed October 3, 2026 |
| S18 | Development of Mathematical Models in Explicit Form for Design and Analysis of Axial Flux PMSMs | The paper describes the absence of cogging torque as an advantage of coreless axial-flux topologies; it does not establish a universal 100 rpm generator design rule. | Supports the coreless/cogging trade-off only. A 100 rpm cut-in target still requires application-specific torque and electrical analysis. | Published 2020; accessed October 3, 2026 |
Universal public threshold for shear stress (kPa) that guarantees AFPM durability
No reliable universal threshold is applied. The checker reports an estimate for context, without a pass/fail band.
Validate against project-specific FEA + thermal tests before release decisions.
Open lot-level demagnetization curves for commercial NdFeB/SmCo suppliers
Public data are typically catalog-level; lot-level high-temperature curves are usually supplier-confidential.
Require lot-level B-H curve package in RFQ/PPAP gate.
Uniform public benchmark set for axial vs radial power density under same duty cycle
No reliable openly maintained dataset found that normalizes cooling, voltage window, and duty cycle across vendors.
Use controlled internal benchmarking and document test protocol.
Open dataset linking NdFeB lot chemistry directly to irreversible-demag onset for AFPM duty profiles
No reliable public cross-supplier lot-level dataset found; available data are mostly catalog grade-level.
Treat demagnetization threshold as pending confirmation until supplier lot data + own thermal test are both complete.
Evaluated grade: N48SH
Profile reference temperature: 150 C
Evaluated hotspot: 120 C
Difference to profile reference: 30 C
A positive difference does not prove demagnetization safety; check the supplier’s exact grade and operating conditions.
Comparison Layer: materials, options, and tradeoffs
Comparison rows are decision-oriented. If a metric is not normalized across vendors, it is marked explicitly.
| Option | Typical Br | Typical BHmax | Example catalog temperature reference | Tradeoff and decision implication | Source |
|---|---|---|---|---|---|
| N48SH NdFeB | 1.39 T | 47 MGOe | 150 C | Good magnetic loading for compact torque targets, but lower thermal headroom than UH/SmCo. | S11 |
| N30UH NdFeB | 1.125 T | 31 MGOe | 180 C | Higher thermal class can reduce magnetic loading; often needs geometry/current compensation. | S11 |
| Sm2Co17 (35E example) | 1.19 T | 33.3 MGOe | 300 C | Strong thermal stability for hot duty cycles, typically with cost and sourcing penalties. | S12 |
| Ferrite | Public value varies by grade | No single normalized value used here | Usually thermally robust | Lower rare-earth dependency but often larger magnetic volume is required for same torque. | Pending confirmation |
| Reference | Quoted figure | Normalization status |
|---|---|---|
| YASA 2025 prototype release | 750 kW short-term peak from 12.7 kg (59 kW/kg), with estimated 350-400 kW continuous | Explicit vendor claim; not normalized by independent duty-cycle parity |
| EMRAX 228 datasheet v1.6 | 124 kW peak (S2 2 min) / 75 kW continuous (S1); 220 Nm peak / 130 Nm continuous | Published peak and continuous are separated; basis must match your design question |
| Generic radial-flux competitor set | N/A | Public comparable dataset not uniform across duty cycles |
Risk Layer: misuse, cost, and scenario mismatch
Risks are mapped with mitigation actions so the page remains operational, not just descriptive.
Misuse risk: over-trusting a fast checker
Mitigation: require thermal-coupled FEA sign-off and lot-level demag data before release.
Cost risk: rare-earth price volatility
Mitigation: qualify an alternate grade and define procurement trigger points by price band. USGS 2026 shows NdPr oxide averaging $69/kg in 2025 vs $55/kg in 2024, and IEA reported regional price spikes after 2025 controls.
Concentration shock risk: export controls and lead-time jumps
Mitigation: do not rely on one-country magnet path only. Build dual-source + fallback grade into RFQ and controller derating plans before SOP commitment.
Scenario mismatch: duty cycle drift after prototype
Mitigation: re-run checker + thermal model for each duty profile revision, not only nominal point.
Compliance risk: EU permanent-magnet obligations
Mitigation: map Article 28/29 timelines into sourcing gates for covered products, especially if your launch window is near 2029-2031.
| Trigger | Why it matters | Minimum action | Source basis |
|---|---|---|---|
| Program input uses peak/S2 power as baseline | Peak and continuous ratings can differ materially (example: EMRAX 228 publishes both S2 and S1 values). | Re-run with continuous/rated basis before architecture decisions. | UN/ECE R85 + EMRAX v1.6 |
| EU-bound product roadmap crosses 2029 | Article 28 applies to listed vehicle/product groups from May 24, 2029. | Add magnet labeling and data-carrier fields to BOM and digital traceability. | EU 2024/1252 Art. 28 |
| Product uses >0.2 kg eligible permanent magnets | Recycled-content disclosure requirements are staged before minimum-share mandates. | Prepare disclosure method now; track minimum-share updates toward 2031. | EU 2024/1252 Art. 29 |
| NdPr price and import dependency shock | U.S. import reliance and price changed sharply between 2024 and 2025. | Use dual-grade procurement and scenario-based cost reserves before volume lock. | USGS MCS 2026 |
| Single-source magnet plan assumes stable export rules | IEA reports around 58,000 t of Chinese rare-earth magnet exports in 2024, plus 2025 controls that drove sharp regional price divergence. | Add dual-source allocation, alternate-grade qualification, and launch-buffer lead-time scenario before freeze. | IEA 2025 commentary + USGS 2026 |
5-phase dual-rotor 100 rpm direct-drive screening
Assumptions: 10 kW target at 100 rpm implies about 955 Nm torque before drivetrain losses, with dual-surface rotor, 5 phases, and 30 poles.
Outcome: The speed/torque relation exposes a high-torque design point. This tool reports calculations only; use application-specific electromagnetic, structural, and thermal analysis to assess feasibility.
Baseline 10.5 kW / 45 Nm commuter e-drive
Assumptions: Operating at ~2,228 rpm. OD 220 mm, dual-surface rotor, and a 120 C hotspot target are illustrative assumptions. At nominal 72 V, 10.5 kW is 146 A ideal DC input before losses; this is not phase current.
Outcome: The checker does not validate a fit or current limit. Use actual bus voltage under load, conversion efficiency, duty cycle, and supplier thermal/demagnetization data for the design review.
Higher ambient + aggressive current derating
Assumptions: Same torque target, but ambient 55C and internal hotspot forecast near 145C.
Outcome: Compare the supplier-specific temperature-dependent B-H curves and re-run the thermal/electromagnetic model; the grade label alone cannot select a safe material.
Packaging squeeze (OD under 170 mm)
Assumptions: Torque unchanged, but rotor diameter reduced to fit a compact gearbox envelope.
Outcome: The calculated average shear estimate increases as the active annulus shrinks. No generic pass band is applied; check the detailed electromagnetic and thermal design.
High-pole control at same mechanical speed
Assumptions: Pole count increased to improve torque ripple behavior while maintaining 10.5 kW / 45 Nm target.
Outcome: Electrical frequency climbs, elevating AC losses and inverter stress; evaluate copper/steel loss budget early.
Commercial 48V electric karting (e.g. EcoVolt GT class)
Assumptions: BIZ Karts lists a 10.5 kW / 45 Nm PM brushless product and separately lists a 48 V / 350 A Curtis ECU. Neither source identifies axial-flux topology or explains the controller rating basis.
Outcome: At nominal 48 V, 10.5 kW is 218.75 A ideal DC input before losses. Treat the product as one example only and verify battery, inverter, and phase-current ratings for the actual duty.
FAQ and conversion path
Questions are grouped by decision intent. This avoids glossary-style filler and keeps actionability high.