Recorded · Reviewed

AFPM Core Manufacturing Process

Compare three core routes and export a preliminary RFQ brief.

Manufacturing route evaluator

Instant results; illustrative defaults.

1. Configure your stator (all fields required)

60–500 mm

30–400 mm

8–150 mm

6–96 slots

50–2000 Hz

0.02–0.2 mm; full-width budget

2. Review the preliminary result

Preparing interactive controls…

Leading process-fit candidate

SMC die compaction

SMC die compaction leads this process-fit shortlist for yokeless / segmented teeth and series: 1,001–50,000 stators/year. Three-dimensional tooth shaping with insulated powder. Validate actual losses, tooling and assembly variation before selection.

Points are a screening heuristic. No measured efficiency, price or production capability is predicted. View assumptions and scoring.

Review this configuration
SMC die compaction95/100 points · Candidate

Tool investment, ejection damage, density gradients and material-specific losses.

Next check: Request a fill/ejection study and heat-treated magnetic and strength coupons.

Segmented lamination stacking85/100 points · Candidate

Cut-edge degradation, flux orientation, stack bonding and carrier alignment.

Next check: Specify lamination planes, rolling direction, coating and bonded-stack qualification.

Tape-wound slotted core0/100 points · Topology excluded

A continuous-yoke ring contradicts this yokeless topology.

Next check: Compare pre-slotted strip with post-winding slotting and test insulation and core loss.

Illustrative variation budget

Review budget: worst-case sum 0.047 mm versus 0.05 mm budget.

Example full-width contributions: tooth 0.02, bond 0.015, carrier 0.012 mm. These are assumptions, not process capabilities. RSS 0.028 mm requires independent, consistently defined spreads and is not measured TIR.

Replace the example contributions with measured production data and verify assembled runout against the drawing datum axis.

Geometry & force assumptions
Mean diameter
160 mm
Radial span
40 mm
Assumed tooth width
11.519 mm
Height / width
3.038:1
Projected face area
460.767 mm²
Tooth-only volume
387.044 cm³
Ideal force at 700 MPa
322.537 kN
With 30% sizing allowance
43 metric tf

55% tooth-to-pitch fraction; axial pressing, single cavity. This force estimate excludes ejection, friction, tooling loads and press suitability. No lamination count is inferred from axial tooth length.

Limits and missing evidence
  • Process-fit points are editorial heuristics, not measured efficiency, success probability or a supplier quote. Frequency alone cannot select a material.
  • Tooth width assumes 55% of mean slot pitch. Tooth-only volume excludes yokes, shoes, slots and allowances; actual CAD and pressing direction can change the result.
  • NRE, scrap, cycle time, lead time, steel gauge, lamination count and thermal performance require supplier and drawing data and are not calculated.

Read the shortlist and its assumptions

Process-fit comparisonEditorial points for SMC, segmented laminations and wound strip. A yokeless topology excludes a continuous-yoke ring.SMC compaction95CandidateSegmented laminations85CandidateWound ring0Topology excluded
Process fit, not efficiency or probability. Reproduce the scoring.
Projected area to ideal forceAxial single-cavity assumption at 700 MPa. The allowance is illustrative; toolmaker review is required.1. Estimate one tooth face460.767 mm²; 55% pitchfraction2. Ideal force = area ×pressure322.537 kN at 700 MPa3. Add a 30% sizing allowance43 metric tf; not pressapproval
Projected area to ideal force. Axial single-cavity assumption at 700 MPa. The allowance is illustrative; toolmaker review is required.
Illustrative variation stackFull-width contributions are examples. RSS needs independent comparable spreads; neither number is measured assembled TIR.1. Tooth, bond and carrier0.02 + 0.015 + 0.012 mm2. Worst-case sum0.047 mm vs 0.05 mm budget3. RSS for comparison only0.028 mm; verify actual data
Illustrative variation stack. Full-width contributions are examples. RSS needs independent comparable spreads; neither number is measured assembled TIR.
Grade-specific thermal routeSource: Höganäs product guidance and 3P material data. Confirm the exact powder grade and schedule; maxima are not recipes.1. Somaloy 700 3P data example530°C steam; 800 MPa testcondition2. Somaloy 5P product guidanceNitrogen; maximum 650°C3. Steel and bondingtreatmentsUse the steel/coatingsupplier’s window
Grade-specific thermal route. Source: Höganäs product guidance and 3P material data. Confirm the exact powder grade and schedule; maxima are not recipes.

Thermal evidence: Somaloy product guidance and 3P material data, printed page 7.

Suggested process and inspection sequence
  1. Confirm powder and die. Agree the insulated powder grade, lubricant, pressing direction and ejection geometry.

    Gate: Supplier material certificate and toolmaker fill/ejection review.

  2. Compact trial parts. Determine pressure and die conditions from the selected powder guidelines.

    Gate: Density distribution, cracks, ejection damage and dimensional repeatability.

  3. Grade-specific heat treatment. Follow the exact supplier schedule. 3P and 5P require different heat-treatment atmospheres.

    Gate: Record the thermal cycle and test strength, resistivity and magnetic loss after treatment.

  4. Insulate and wind. Use the approved bobbin/coating and winding process for the geometry.

    Gate: Agree winding resistance, dielectric tests and test voltage with the winding specification.

  5. Locate and assemble. Fixture the active faces relative to the carrier datum; qualify the chosen adhesive or impregnation resin.

    Gate: Check bond integrity, thermal cycling and dimensions after cure.

  6. Verify and release. Measure the assembled stator, retain lot traceability and validate magnetic performance.

    Gate: Use drawing-specific runout, loss and retention acceptance criteria; this tool does not release parts.

Four decisions before committing to tooling

Method, assumptions and limits

This evaluator uses transparent editorial process-fit rules, reviewed 2026-10-25. It performs synchronous local calculations; it has no quote, factory capability or motor simulation data. All inputs are required. The screening ranges are OD 60–500 mm, ID 30–400 mm, axial length 8–150 mm, 6–96 integer slots, 50–2,000 Hz and a 0.02–0.20 mm full-width variation budget. Radial span (OD − ID) / 2 must be at least 10 mm. These are model bounds, not physical manufacturing limits.

Geometry assumes tooth width = 0.55 × π × mean diameter / slot count, face area = width × radial span, and tooth-only volume = face area × axial length × tooth count. Shoes, yokes, packing and the second stator in a double-stator motor are excluded. SMC force = area × 700 / 1,000 kN, assuming axial pressing and one cavity; the displayed metric-ton-force allowance is force × 1.3 / 9.80665, rounded up. No press capacity is approved by this estimate.

Reproducible process-fit weights (SMC / segmented / wound)
ConditionPoints added
Start50 / 50 / 50
Yokeless topology+25 / +20 / excluded (0 final points)
Double rotor, continuous yoke+15 / +10 / +25
Other back-iron topology+15 / +15 / +10
Prototype / low volume−15,+20,+5 / +5,+15,+5
Series / mass volume+20,+15,−5 / +25,+15,−10
Tooling commitment priority−20 / +15 / +10
Unit-cost / automation priority+15 SMC for series/mass; otherwise +10 segmented
Turnaround priorityPrototype only: −20 SMC / +20 segmented
Efficiency priority and frequencyNo efficiency points: matched loss validation required
Height / assumed width >3.8−15 SMC; editorial toolmaker-review trigger
FinishClamp 0–100. Ties: segmented, SMC, wound. Candidate ≥60; otherwise needs development.

Variation examples appear in Table 2. The model compares the unrounded worst-case sum with the full-width budget: above budget → over; above 80% → review; otherwise → within example budget. RSS = √(tooth² + bond² + carrier²) is shown for independent, consistently defined spreads only. The fixed examples are not ± capability specifications, measured runout or statistical process evidence. Temperature, flux density, harmonics and assembly deflection are missing from the model.

AFPM core manufacturing process evaluator showing the default SMC process-fit candidate
Actual evaluator run captured 2026-10-25: yokeless stator, OD 200 mm, ID 120 mm, axial length 35 mm, 24 slots, 400 Hz, 0.05 mm variation budget, series volume and efficiency priority. Reset defaults to reproduce the SMC shortlist (95 points; segmented laminations 85; wound ring excluded). The worst-case example stack is 0.047 mm. This crop shows only the result card; site navigation is outside the capture. It records software execution, not a physical core test or customer outcome.

Next step: export the assumptions, add actual CAD and production data, and request an engineering review. For dimensions outside the model, use the same drawing-review path.

Compare routes, evidence and risk

1. Manufacturing route and topology trade-offs

Qualitative engineering screening. No general cost, scrap or loss advantage is assumed.

1. Manufacturing route and topology trade-offs
DecisionSMC compactionSegmented laminationsTape-wound slotted ring
Core formInsulated powder pressed into a toothGO/NGO sheets formed into a tooth stackStrip wound into a continuous-yoke ring
Yokeless statorCandidate: qualify retention and densityCandidate: specify planes and rolling directionExcluded as a continuous-yoke substitute
Prototype commitmentCheck existing tooling or a dedicated dieCutting plus stacking fixture can avoid a progressive dieMandrel, winding fixture and slotting trial
Series decisionQuote die life, cavities and ejection yieldQuote nesting, die life and bonding cycleQuote indexing, slotting and insulation recovery
Selection evidencePost-treatment loss and strength couponsCut and bonded-stack magnetic/insulation dataFinished slotted-ring loss and inter-turn integrity

Evidence / scope: Höganäs — Somaloy product and processing guidance · voestalpine — Backlack lamination bonding · IEC 60404-8-7:2020 — Grain-oriented electrical steel

2. Example variation budget used by the tool

Editorial full-width examples in mm, not ± limits or ISO tolerance classes. Worst-case sum decides the budget flag. RSS requires independent spreads with a consistent statistical meaning.

2. Example variation budget used by the tool
Leading routeToothBond: yokeless / otherCarrierWorst case: yokeless / other
SMC0.0200.015 / 0.0100.0120.047 / 0.042 mm
Segmented laminations0.0350.015 / 0.0100.0120.062 / 0.057 mm
Tape-wound ring0.045N/A / 0.0100.012N/A / 0.067 mm
Not includedTooth tilt and formCure shrinkage and hot driftBearing / rotor stackRunout about the datum axis and deformation

Evidence / scope: ISO 1101:2017 — Geometrical tolerancing · ISO 8015:2011 — GPS fundamentals

3. Thermal-processing evidence and boundaries

Published examples and supplier-dependent unknowns. A maximum temperature does not define ramp, soak, atmosphere quality or release criteria.

3. Thermal-processing evidence and boundaries
Material / operationKnown referenceStill required before release
Somaloy 700 3P530°C steam in the 800 MPa data-sheet exampleExact grade, lubricant, two-stage schedule and component tests
Somaloy 5P familyProduct guidance: nitrogen, max 650°CExact grade sheet, hold profile and mechanical / magnetic targets
Electrical steel stress reliefN/A: grade/coating not selectedSteel supplier approval; domain-refinement and coating compatibility
Backlack / potting resinCure window depends on system and stackSupplier window, actual part temperature, fixture pressure and cure validation

Evidence / scope: Höganäs — Somaloy product and processing guidance · Höganäs — Somaloy 3P Material Data, 2273HOG-3 · voestalpine — Backlack lamination bonding

4. Traceable material-loss reference: Somaloy 700 3P

Höganäs 2025 data, printed pp. 7–8. Values at 1.0 T on OD55 / ID45 / H5 mm ring specimens, measured under the stated IEC 60404-6:2003 method. Die temperature 80°C; steam treatment 530°C. Typical reference data, not finished-core guarantees; no interpolation is used by the tool.

4. Traceable material-loss reference: Somaloy 700 3P
Material / pressing conditionB at H=10,000 A/m100 Hz loss400 Hz loss1,000 Hz loss
700 3P, 800 MPa1.61 T10 W/kg45 W/kg132 W/kg
700 3P, 600 MPa1.55 T11 W/kg47 W/kg143 W/kg
Electrical steel comparisonN/A: grade not selectedN/AN/AN/A

Evidence / scope: Höganäs — Somaloy 3P Material Data, 2273HOG-3 · IEC 60404-6:2018 — Ring-specimen magnetic measurements

5. Cut-edge risks and a measurable recovery plan

Risk mechanisms are engineering considerations. No universal HAZ depth, burr height or loss-recovery percentage is asserted.

5. Cut-edge risks and a measurable recovery plan
OperationWhat can go wrongTrial / inspectionAcceptance evidence
StampingPlastic strain, burrs and coating damageInspect edge and track tool wearLoss and insulation on representative processed samples
Laser cuttingRecast, thermal stress and coating damageOptimize settings for selected gaugeCompare raw strip and cut-tooth magnetic coupons
EDM / grindingRecast or smear can bridge turnsInspect slot edges; trial compatible deburringInter-turn integrity and finished-core loss
Chemical treatmentCoating attack, size change or retained residueCompatibility trial, rinse and dimensional checksDemonstrated recovery without geometry or insulation failure

Evidence / scope: IEC 60404-6:2018 — Ring-specimen magnetic measurements · IEC 60404-2:1996 — Epstein-frame measurements

6. Bonding and impregnation selection checklist

N/A means the resin or process has not been selected. It must not be filled with a generic product-family value.

6. Bonding and impregnation selection checklist
Joint / treatmentDecision neededCurrent numerical valueSupplier evidence to request
Self-bonding lamination coatingHeating method and stack sizeCure / strength: N/ACoating grade, process window and bond-test method
Tooth-to-carrier adhesiveBond gap, CTE mismatch and shear dutyHot strength / Tg: N/AActual substrates, aged hot shear and thermal-cycle results
Winding impregnationPenetration, viscosity and dielectric dutyVacuum / pressure / cure: N/AResin instructions and inspection of representative assemblies
Thermal pathContact resistance and hot spotsAssembly conductivity: N/ASystem-level thermal testing with the intended cooling arrangement

Evidence / scope: voestalpine — Backlack lamination bonding

7. Manufacturing risk register and next actions

Qualitative review prompts, not a completed PFMEA. Severity, occurrence, detection and action priority need the OEM’s duty data and quality process.

7. Manufacturing risk register and next actions
Risk categoryFailure / consequenceMitigation and alternativeRelease evidence
MisuseA route score is mistaken for efficiency approvalCompare matched loss tests before material nominationProcessed-part magnetic data and motor validation
CostTooling investment is not amortized at real yieldQuote laser prototypes, existing dies and series tooling separatelyNRE, good-part yield, cavity count and volume-sensitive quote
Scene mismatchContinuous yoke is selected for yokeless teethUse modular SMC or segmented stacks; recheck full flux pathTopology, CAD and retention review
Process / qualityEjection cracks, burr shorts or weak bondsTrial coupons; revise tool, cut or bonding routeDensity, insulation, magnetic and mechanical inspection
Assembly / thermalRunout or hot distortion consumes clearanceRevise datum, fixture and joint; inspect after cure and cyclingMeasured assembly geometry and worst-duty clearance analysis

Evidence / scope: ISO 1101:2017 — Geometrical tolerancing · voestalpine — Backlack lamination bonding · IEC 60404-6:2018 — Ring-specimen magnetic measurements

Manufacturing process guide: forming to release

The following engineering recommendations explain what to qualify. Source scopes are linked beside factual claims; recommendations do not add requirements to the cited standards.

1. Orient the material to the axial flux path

In an AFPM machine the working air-gap flux is axial, but return paths and local tooth-shoe fields depend on topology. Specify lamination planes explicitly: stacking flat rings axially may force useful flux across interlaminar coatings. GO steel adds a preferred rolling direction; NGO steel still needs the correct lamination-plane orientation. SMC allows 3D shapes, with grade-specific magnetic and mechanical trade-offs.

Action: Include the flux-path sketch and lamination plane in the RFQ; do not infer them from OD, ID and axial length.

Evidence / scope: Höganäs — Somaloy product and processing guidance · IEC 60404-8-7:2020 — Grain-oriented electrical steel

2. Define the tape-wound ring correctly

For a cylindrical strip wound into an annular core, strip width sets the axial envelope; accumulated strip thickness builds the radial span. The screening geometry cannot account for slots, coatings, shoes or a real winding packing factor. A wound ring with continuous back iron also represents a different architecture from independent yokeless teeth.

Action: Specify strip width, radial build, yoke depth, slot geometry and insulation. Confirm how winding stress and inter-turn stability will be controlled.

Evidence / scope: IEC 60404-8-7:2020 — Grain-oriented electrical steel

3. Compare pre-slotted strip with post-winding slots

Pre-slotted strip requires pitch and angular indexing to track increasing circumference during winding. Cutting slots after winding trades indexing complexity for cut-edge and mechanical support risks. EDM, laser and grinding affect the edge differently; no single method can be declared the best without geometry and throughput data.

Action: Ask for a slot-forming trial with a dimension report and a magnetic comparison of the finished ring.

Evidence / scope: IEC 60404-2:1996 — Epstein-frame measurements

4. Qualify stress relief before scheduling a furnace cycle

Further heat treatment can change residual stress, coating integrity and the behavior of some domain-refined steels. A generic 800°C cycle is not suitable evidence for all strip grades. The order of annealing, bonding and machining also matters: later cutting can reintroduce damage, while an earlier organic adhesive may not survive a later high-temperature process.

Action: Obtain written grade/coating compatibility and compare loss before and after the proposed processing sequence.

Evidence / scope: IEC 60404-8-7:2020 — Grain-oriented electrical steel · IEC 60404-2:1996 — Epstein-frame measurements

5. Specify segmented GO or NGO tooth stacks

A segmented tooth can use steel sheets arranged to keep the principal flux in-plane. If GO steel is proposed, its rolling direction must follow the intended working flux, with cross-grain and tooth-tip fields reviewed separately. Bonding cannot change a steel’s intrinsic saturation characteristics or prove a fixed increase in motor air-gap flux.

Action: Nominate grade, gauge, rolling direction, coating and stacking direction. Request bonded-stack dimensions and loss data.

Evidence / scope: IEC 60404-8-7:2020 — Grain-oriented electrical steel · voestalpine — Backlack lamination bonding

Bonded-stack qualification. Use the coating supplier’s process window. Stack temperature, pressure and hold time depend on geometry.
Bonded-stack qualification. Use the coating supplier’s process window. Stack temperature, pressure and hold time depend on geometry.

6. Select SMC by properties, not by a frequency slogan

Insulated powder interrupts inter-particle current paths and permits three-dimensional core features. Total loss still includes hysteresis and eddy-current contributions, and varies with grade, compaction and component dimensions. A larger part need not match a small material ring even when nominal frequency and flux density agree.

Action: Compare supplier curves and representative geometry at the actual waveform and temperature; retain steel as a candidate until system performance is checked.

Evidence / scope: Höganäs — Somaloy product and processing guidance · Höganäs — Somaloy 3P Material Data, 2273HOG-3 · IEC 60404-6:2018 — Ring-specimen magnetic measurements

7. Check powder fill, pressing direction and ejection

Ideal force is projected area times pressure, but tool loads and powder behavior depend on pressing direction, number of cavities and ejection. The tool assumes axial pressing and 700 MPa solely to show force scale. Its 3.8:1 aspect-ratio flag is an editorial review trigger rather than a physical tooling boundary.

Action: Give the toolmaker the complete 3D part. Ask for density distribution, fill/ejection review and measured trial-part repeatability.

Evidence / scope: Höganäs — Somaloy product and processing guidance

SMC tooling validation. Qualify fill, compaction and ejection together; the schematic sets no universal tooling limit.
SMC tooling validation. Qualify fill, compaction and ejection together; the schematic sets no universal tooling limit.

8. Keep SMC thermal routes grade-specific

Somaloy families use different treatments. The 700 3P reference here uses steam at 530°C, while the 5P product guidance identifies a nitrogen route up to 650°C. This is not evidence that all powders share a resin cross-linking mechanism or that one temperature guarantees strength and resistivity.

Action: Confirm the exact powder, additive and supplier process sheet. Retest strength, resistivity and loss after the actual component cycle.

Evidence / scope: Höganäs — Somaloy product and processing guidance · Höganäs — Somaloy 3P Material Data, 2273HOG-3

9. Treat burr recovery as a qualification task

Metal bridges at cut edges can undermine insulation between sheets or wound turns. Deburring, cleaning, insulation repair and chemical treatment can each alter dimensions or coatings. Chemical etching is an option that needs compatibility evidence, not a mandatory universal acid bath. A loss test standard does not prescribe an etching formulation.

Action: Agree edge inspection, residue control and post-treatment magnetic / insulation tests before selecting the recovery route.

Evidence / scope: IEC 60404-6:2018 — Ring-specimen magnetic measurements · IEC 60404-2:1996 — Epstein-frame measurements

Cut-edge recovery decision. Deburring or etching needs compatibility trials; there is no universal acid recipe or recovery guarantee.
Cut-edge recovery decision. Deburring or etching needs compatibility trials; there is no universal acid recipe or recovery guarantee.

10. Separate variation budgets, flatness and runout

Flatness describes surface form; runout is evaluated relative to an axis under the stated measurement setup. A tooth-height spread does not by itself define either result. The tool adds illustrative full-width tooth, bond and seat contributions. RSS is shown only as a conditional statistical comparison; the worst-case sum drives the budget flag.

Action: State the datum axis and tolerance definition on the drawing. Include rotor, bearings, fixtures, thermal growth and deformation in the clearance analysis.

Evidence / scope: ISO 1101:2017 — Geometrical tolerancing · ISO 8015:2011 — GPS fundamentals

11. Qualify carrier joints and impregnation together

A carrier joint must retain the tooth through torque, magnetic attraction, vibration and thermal cycling. Resin can improve contact and reduce voids, but impregnation alone proves neither zero void content nor a fixed increase in thermal conductivity. Resin cure, substrate preparation and differential expansion need assembly-specific testing.

Action: Request hot joint strength, thermal-cycle evidence, void inspection and winding dielectric validation on representative assemblies.

Evidence / scope: voestalpine — Backlack lamination bonding

Impregnation qualification. A resin-specific process and inspection plan is needed. Impregnation alone does not prove zero voids or certification.
Impregnation qualification. A resin-specific process and inspection plan is needed. Impregnation alone does not prove zero voids or certification.

12. Compare economics per accepted stator

Tool price or press cycle alone misses cavities, tooth count, yield, maintenance, handling, bonding and inspection. A fast tooth press does not imply the same time per stator. For a quote with fixed tooling NRE and accepted quantity Q, tooling amortization is NRE / Q; use good-part yield when comparing variable cost.

Action: Request itemized NRE, repeat orders, raw material utilization, tested yield and capacity per finished stator. Obtain supplier prices and dates rather than relying on a generic crossover volume.

13. Build a staged inspection plan

Inspect incoming material, forming/cutting, thermal processing, bonding and the completed assembly. Ring methods and Epstein tests serve different material scopes; neither alone demonstrates the installed motor’s performance. Electrical test voltages, acceptance limits and sampling must follow the actual design and quality agreement.

Action: Define specimen, B/f/waveform/temperature, calibration, lot traceability, acceptance limits and handling of nonconforming parts before production.

Evidence / scope: IEC 60404-6:2018 — Ring-specimen magnetic measurements · IEC 60404-2:1996 — Epstein-frame measurements

14. Prepare a drawing-specific DFM package

The minimum useful package combines revision-controlled geometry, topology, flux and loss targets, pressing or lamination direction, datum scheme, insulation interfaces and quantity. Small tooth shoes, undercuts or inaccessible winding windows can change route feasibility even if the model’s simple dimensional limits pass.

Action: Send CAD and a controlled drawing plus prototype/annual quantities. Flag uncertain material and tolerance assumptions for joint review.

Evidence / scope: ISO 1101:2017 — Geometrical tolerancing

15. Validate cutting damage on the intended tooth

Laser heat, stamping strain and grinding smear depend on the material, edge geometry and machine settings. A universal damage depth or loss penalty is not defensible for every AFPM tooth. Narrow sections deserve representative testing because the processed-edge fraction can be large. Stamping also introduces mechanical damage even without a thermal HAZ.

Action: Compare raw-strip and processed samples under matched conditions, then test the bonded or slotted core before freezing the cutting method.

Evidence / scope: IEC 60404-2:1996 — Epstein-frame measurements · IEC 60404-6:2018 — Ring-specimen magnetic measurements

16. Check inverter duty and the complete insulation system

Electrical fundamental frequency does not describe PWM harmonics or winding voltage stress. Interlaminar insulation, inter-particle insulation and winding-to-core insulation have different functions. One cannot infer partial-discharge performance, thermal class or aviation qualification from SMC resistivity or a resin product name.

Action: Provide the inverter waveform, voltage rise time, temperature and environment. Agree a system-level dielectric and thermal validation plan with the motor team.

Evidence / scope: IEC 60404-6:2018 — Ring-specimen magnetic measurements · voestalpine — Backlack lamination bonding

Related: Core DFM guide · Magnetic circuit design · Assembly process

Discuss a production route: AFPM SMC stators · Lamination and winding services · Custom compaction tooling

Six hypothetical sourcing scenarios

Worked decision examples, not customer results or certification evidence. The first four map to evaluator presets. Values described as example stacks are model assumptions.

EV segmented-stator example

Premise: 240 / 140 mm OD / ID, 30 teeth, 500 Hz, series production, 0.04 mm variation budget.

Process: Load the EV preset: the geometric / volume heuristic leads to SMC, with segmented steel retained for a loss comparison.

Decision / next step: The assumed 0.047 mm stack exceeds the 0.04 mm budget. Next: obtain actual component variation and a revised datum/assembly plan; motor efficiency remains unknown.

Compact high-frequency drive example

Premise: 130 / 80 mm OD / ID, 18 teeth, 1,000 Hz, series automation priority.

Process: Use the compact preset to screen forming options, then compare processed SMC and thin steel at matched B and inverter waveform.

Decision / next step: The tool supplies a shortlist, not a loss winner. Next: request representative magnetic and thermal tests before material nomination.

Continuous-yoke industrial fan example

Premise: 280 / 180 mm OD / ID, 36 slots, 150 Hz, mass volume and low tooling commitment.

Process: The fan preset permits a continuous wound ring. Compare pre-slotted strip against post-winding slotting.

Decision / next step: Segmented stacks lead the current heuristic with a 0.057 mm example variation against a 0.08 mm budget; the wound-ring alternative remains eligible. Next: compare finished-ring edge integrity, assembled runout and supplier quotes before choosing.

Small servo prototype example

Premise: 90 / 50 mm OD / ID, 12 teeth, 300 Hz, 25 assumed prototype stators and 0.03 mm budget.

Process: The prototype preset favors cut segmented laminations to explore geometry with limited tooling commitment.

Decision / next step: The example 0.062 mm variation exceeds budget. Next: tighten/sort the actual contributors and evaluate finishing for insulation damage. No zero-cogging claim follows.

Aerospace duty outside model range

Premise: An engineering team requires 2,500 Hz and operation at reduced ambient pressure.

Process: The evaluator rejects frequency above its 2,000 Hz screening range. Keep the drawing and request a dedicated magnetic / dielectric review.

Decision / next step: No route or performance output is exported from the invalid configuration. Next: a qualified test plan is required; this tool provides no airworthiness or certification evidence.

Volume change before tooling commitment

Premise: A yokeless design moves from a small pilot to more than 50,000 stators/year.

Process: Compare prototype and mass-volume settings while retaining the same geometry. The continuous-yoke route remains excluded in both cases.

Decision / next step: Automation weighting can change the shortlist. Next: compare itemized tooling and accepted-stator costs with measured yield; no universal percentage saving is assumed.

Traceable sources and their limits

By AxialFluxCore. Public publisher pages and manufacturer data checked 2026-10-25. Standards catalogue scopes are used here; full procedures may require the licensed standard. They do not certify the estimates or qualify a supplier process.

  1. Höganäs — Somaloy product and processing guidance

    Live product page; accessed 2026-10-25.

    Powder families and grade-dependent compaction / heat-treatment guidance. The overview and detailed grade sections differ on the 3P maximum; use the exact grade sheet and obtain the supplier’s current schedule.

  2. Höganäs — Somaloy 3P Material Data, 2273HOG-3

    2025 edition; printed pages 7–8 (PDF pages 7–8).

    Somaloy 700 3P test conditions, magnetic curves and loss data. Ring results are material references, not finished-tooth or motor acceptance limits.

  3. voestalpine — Backlack lamination bonding

    Live manufacturer guidance; accessed 2026-10-25.

    Bonding depends on component size, heating method and fixture; achievable strength also depends on coating and test method. It does not establish one universal cure cycle.

  4. IEC 60404-6:2018 — Ring-specimen magnetic measurements

    2018 base publication; catalogue also lists Amendment 1:2021.

    AC magnetic measurements of soft metallic / powder materials, excluding electrical steels and soft ferrites, from 20 Hz to 100 kHz. It is not a tooling, acid-etch or insulation-resistance process specification.

  5. IEC 60404-2:1996 — Epstein-frame measurements

    1996 base publication; catalogue lists Amendment 1:2008.

    Electrical steel sheet and strip: sinusoidal AC measurements up to 400 Hz, and DC measurements. Higher-frequency testing needs an appropriate agreed method.

  6. IEC 60404-8-7:2020 — Grain-oriented electrical steel

    2020 edition; catalogue checked 2026-10-25.

    Fully processed GO steel at nominal 0.20, 0.23, 0.27, 0.30 and 0.35 mm gauges. This scope does not qualify a 0.10 mm grade or an AFPM manufacturing recipe.

  7. ISO 1101:2017 — Geometrical tolerancing

    2017 edition, confirmed 2022 per ISO catalogue.

    Geometrical specification symbols and interpretation. Flatness and runout require their own drawing definitions; this standard does not supply generic AFPM tolerance capability values.

  8. ISO 8015:2011 — GPS fundamentals

    2011 edition, confirmed 2021 per ISO catalogue.

    Fundamental rules for interpreting geometrical product specifications. A dimensional budget alone is not proof of assembled clearance.

AFPM core manufacturing FAQ

Route selection

What are the main AFPM core manufacturing process routes?

This guide compares SMC die-compacted teeth, segmented electrical-steel stacks and tape-wound slotted rings. Topology, flux orientation, material behavior, assembly and production volume determine which deserves a trial.

Evidence / scope: Höganäs — Somaloy product and processing guidance · IEC 60404-8-7:2020 — Grain-oriented electrical steel

Does frequency above 400 Hz automatically favor SMC?

No. Compare total loss at matched flux density, waveform, temperature and specimen geometry. Frequency does not determine hysteresis loss, process damage or the thermal limit. The tool does not predict an efficiency winner.

Evidence / scope: Höganäs — Somaloy 3P Material Data, 2273HOG-3 · IEC 60404-6:2018 — Ring-specimen magnetic measurements

Can a continuous tape-wound ring replace yokeless teeth?

A continuous stator yoke changes the magnetic architecture. This tool excludes that route for its yokeless setting. Cutting a ring into independent segments would be a different manufacturing route needing a separate review.

When should grain-oriented steel be considered?

Consider it where the intended principal flux can follow the rolling direction within each lamination plane. Review cross-grain fields and tooth shoes. No fixed motor flux or torque increase is implied by the material designation.

Evidence / scope: IEC 60404-8-7:2020 — Grain-oriented electrical steel

Tooling and economics

How does the tool estimate SMC pressing force?

It assumes a tooth width equal to 55% of mean slot pitch, axial pressing and a single cavity. Ideal force in kN is area in mm² × 700 MPa / 1,000. A separate 30% allowance shows force scale, not press approval.

Is 3.8:1 a universal SMC aspect-ratio limit?

No. It is an editorial review trigger in this model. Actual limits depend on pressing direction, powder, fill behavior, punch actions, ejection and density requirements. Ask the toolmaker to evaluate the complete geometry.

What are typical NRE, scrap and lead-time values?

They are not established for your part by the public sources used here. Request supplier quotes with cavity count, die life, nesting, accepted-part yield, inspection, volume and quote date. The tool leaves these quantities uncalculated.

Can axial tooth length determine lamination count?

Only if it is the actual stacking dimension, which cannot be assumed for an axial-flux tooth. Specify the lamination planes, stacking direction, compressed gauge and coating before calculating sheet count.

Heat and edges

Do all SMC grades use the same thermal cycle?

No. The 700 3P sheet gives a 530°C steam example; 5P product guidance identifies nitrogen up to 650°C. Obtain the exact powder and lubricant processing schedule. Maxima alone do not specify a complete cycle.

Evidence / scope: Höganäs — Somaloy product and processing guidance · Höganäs — Somaloy 3P Material Data, 2273HOG-3

Is SMC heat treatment equivalent to ordinary iron sintering?

Do not substitute a structural iron sintering cycle for the selected SMC process. Inter-particle insulation is essential to its electrical behavior and must survive the qualified treatment.

Evidence / scope: Höganäs — Somaloy product and processing guidance

Is 800°C annealing required for every wound ring?

No universal cycle is established here. Confirm the steel grade, coating, domain refinement and processing sequence with the supplier, then test the actual ring. Organic bonding must also be compatible with the chosen sequence.

Evidence / scope: IEC 60404-8-7:2020 — Grain-oriented electrical steel

What Backlack temperature and pressure should be specified?

Use the selected coating supplier’s window and validate it on the intended stack and fixture. Part size and heating method change the process. This guide does not prescribe a universal temperature, pressure or holding time.

Evidence / scope: voestalpine — Backlack lamination bonding

Is acid micro-etching always required after slotting?

No. Inspect for bridges and coating damage, then qualify compatible deburring, cleaning or insulation recovery. Any chemical treatment must demonstrate dimensional and coating compatibility as well as loss recovery.

Does laser cutting have a fixed magnetic damage depth?

No universal depth is supplied here. Material, gauge, geometry and cutting conditions matter. Compare representative processed coupons and finished stacks; a thermal HAZ is only one possible source of degradation.

Measurement and assembly

Does the variation result prove air-gap runout compliance?

No. Fixed full-width examples are used for screening. They omit surface form, tilt, rotor/bearing effects, thermal deformation and actual process distributions. Measure assembled runout relative to the drawing datum axis.

Evidence / scope: ISO 1101:2017 — Geometrical tolerancing · ISO 8015:2011 — GPS fundamentals

When can RSS be used for a tolerance stack?

Use RSS only when contributors are independent and their spreads have a consistent statistical interpretation, supported by process data. A hard dimensional bound is not automatically a standard deviation. The tool flags the worst-case example sum instead of claiming RSS guarantees clearance.

Are the listed loss values motor efficiency measurements?

No. The two numerical rows are published ring-specimen material data for Somaloy 700 3P under different pressing conditions. Actual tooth geometry, processing, inverter waveform and other motor losses require separate evaluation.

Evidence / scope: Höganäs — Somaloy 3P Material Data, 2273HOG-3

Which magnetic measurement standard applies?

IEC 60404-6 covers ring measurements for specified soft metallic and powder materials, excluding electrical steel. IEC 60404-2 covers electrical-steel Epstein measurements up to 400 Hz for sinusoidal AC conditions. Agree an appropriate method for higher frequencies or finished components.

Evidence / scope: IEC 60404-6:2018 — Ring-specimen magnetic measurements · IEC 60404-2:1996 — Epstein-frame measurements

Does impregnation guarantee zero voids or flight certification?

No. Resin selection and processing need inspection and system-level dielectric, thermal and mechanical validation. Neither this report nor its hypothetical scenarios supplies certification evidence.

Using the evaluator

What should accompany the exported RFQ brief?

Include revision-controlled CAD, topology, material candidates, pressing/stacking direction, flux density, frequency, waveform, temperature, datum/runout definition, measured variation, quantities and delivery needs. Agree NDA and file handoff before sharing sensitive drawings.

Ready to Validate Your AFPM Core Manufacturing Process?

Bring the exported brief, drawing revision, topology, material candidates, magnetic duty, datum / runout requirements and prototype / annual quantity. Agree NDA and file handoff before sharing sensitive CAD. The engineering review can identify the tooling trials, tests and quote details needed for a sourcing decision.