PFMEA For Lamination Stacks: Key Controls

Key Takeaways

The highest-risk failures usually connect to burrs, insulation damage, misalignment, missing or double laminations, poor joining, and hidden interlaminar shorts.
Stack height inspection is useful, but it cannot prove electrical integrity.
Modern FMEA practice should not rely only on old-style RPN ranking. Action Priority, or AP, is now widely used in the harmonized automotive FMEA method to guide whether action is high, medium, or low priority.
A strong lamination stack PFMEA should feed directly into the control plan, inspection method, reaction plan, and process audit checklist.

What this PFMEA covers

A lamination stack is built from repeated thin metal sheets, usually electrical steel, stacked and fixed into a magnetic core. It may become part of a stator, rotor, transformer core, actuator, sensor, or other electromagnetic assembly.

This article focuses on the manufacturing process:

Incoming lamination material or blanks
Punching, cutting, or profiling
Cleaning and handling
Stacking
Compression
Joining, bonding, welding, riveting, or interlocking
Final inspection before winding, magnet insertion, shaft assembly, or shipment

That means this is mainly a Process FMEA, or PFMEA.

Some design issues are mentioned because they affect manufacturing risk. But they should not be buried inside the PFMEA without discipline. Lamination thickness, coating type, slot geometry, skew angle, joining concept, and stack factor targets usually begin as design decisions. Once frozen, the PFMEA asks a different question:

How can the process fail to make that design correctly?

Small distinction. Big audit difference.

DFMEA vs PFMEA for lamination stacks

Topic	DFMEA concern	PFMEA concern
Lamination thickness	Is the selected thickness suitable for loss, cost, strength, and manufacturability?	Is the correct thickness loaded, verified, and protected from mix-up?
Insulation coating	Does the coating meet electrical, thermal, and process requirements?	Is the coating scratched, crushed, contaminated, or damaged during production?
Slot geometry	Does the slot design support winding, flux path, noise, and fill factor?	Are slots distorted, burred, misaligned, or out of profile after cutting and stacking?
Stack height	Is the nominal stack height suitable for electromagnetic and mechanical design?	Is the actual stack height wrong because of missing sheets, double sheets, debris, burrs, or compression error?
Joining method	Is welding, bonding, interlocking, or riveting suitable for strength and magnetic performance?	Are joining parameters controlled, verified, and contained when out of window?
Skew	Does skew reduce torque ripple, noise, or cogging as intended?	Is the skew angle built correctly and repeatably?

A clean PFMEA does not pretend to redesign the motor or transformer. It controls the process that builds the stack.

Why lamination stack PFMEA needs more than dimensional checks

The hard part is that many lamination stack defects hide well.

A stack can pass height inspection and still have an electrical fault. A bore can measure correctly while several internal slots are slightly rotated. A weld can look acceptable and still leave the stack loose after thermal cycling. A burr can be tiny by eye and still cut through insulation when compressed.

Electrical steel laminations are insulated from each other to restrict eddy currents; interlaminar faults can increase core losses and cause damage in electrical machines. Burrs from punching or cutting can impair insulation between adjacent sheets and create random conductive contact during stack pressing.

That is why a useful PFMEA does not ask only, “Is the part in tolerance?”

It asks:

What did the process do to the stack that the drawing cannot easily show?

PFMEA table for lamination stack manufacturing

Use this as a working template. The AP column is not a fixed rating. It shows where action priority usually deserves special attention. Your actual AP should come from your internal S/O/D tables, customer requirements, and risk method.

Process step	Failure mode	Likely cause	Effect	Prevention control	Detection control	AP focus
Incoming material	Wrong material grade, thickness, or coating	Lot mix, labeling error, supplier escape	Loss, heat, stack height error, poor joining	Lot segregation, barcode control, approved material list	Thickness check, certificate review, coating verification	High if safety or performance-critical
Cutting / punching	Excessive burr height	Tool wear, wrong die clearance, dull punch, material variation	Interlaminar short, insulation damage, false stack height, winding damage	Tool life limits, die clearance control, sharpening schedule	Burr measurement, edge microscopy, vision inspection	High
Cutting / punching	Slot or tooth profile error	Feed error, tool damage, poor strip control	Winding interference, torque ripple, noise, reduced fill	First-off approval, SPC, tool maintenance	Optical profile check, slot gauge, CMM sampling	Medium to high
Cutting / punching	Edge stress or magnetic degradation	Aggressive cutting, heat-affected edge, poor process window	Higher core loss, local heating	Qualified cutting window, controlled tool condition	Core loss test, thermal scan, sample magnetic test	Medium to high
Cleaning / handling	Debris between laminations	Slivers, dust, oil sludge, coating flakes	Stack tilt, local short, height error, loose region	Cleaning standard, covered WIP, clean containers	Visual check, height map, teardown audit	Medium
Stacking	Missing lamination	Feeder skip, manual count error, pickup failure	Low stack height, magnetic performance shift, loose assembly	Sheet counter, feeder interlock, kitted stack quantity	Weight check, stack height under load	High
Stacking	Double lamination	Oil adhesion, poor separation, magnetic pickup, vacuum error	Excess height, compression distortion, slot mismatch	Air separation, pickup tuning, double-sheet prevention	Double-sheet sensor, weight check, force-distance curve	High
Stacking	Angular misalignment	Worn pins, loose nest, part bounce, poor datum	Slot drift, winding issue, skew error, torque ripple	Hardened datums, anti-rotation feature, nest maintenance	Vision check, angular gauge, end-face inspection	High
Stacking	Radial misalignment / concentricity error	Dirty datum, clamp imbalance, fixture wear	Air-gap variation, vibration, rotor imbalance	Datum cleaning, controlled clamping, fixture inspection	Runout check, bore-to-OD measurement	High
Compression	Over-compression	Wrong press setting, recipe error, attempt to force height	Coating damage, interlaminar short, slot distortion	Press recipe lock, mechanical stop, force limit	Force-displacement monitoring, insulation test	High
Compression	Under-compression	Low force, short dwell, fixture springback	Loose stack, height instability, poor joining	Press force control, dwell control, calibrated stops	Stack height under defined load, resonance check	Medium to high
Joining	Weak weld, bond, rivet, or interlock	Contamination, wrong energy, poor cure, worn tooling	Stack looseness, vibration, dimensional drift	Parameter window, surface cleanliness, cure control	Pull test, cross-section, visual check, process data review	High
Joining	Excess heat or local damage	Weld energy too high, poor fixture heat control	Magnetic loss, distortion, coating damage	Heat input limits, fixture cooling, parameter lock	Dimensional check, core loss check, thermal inspection	Medium to high
Final inspection	Electrical short not detected	Wrong test method, skipped test, poor sampling	Heat, loss, field failure	Mandatory test plan, reaction plan, audit of test bypass	Interlaminar resistance, core loss test, thermal scan	High
Packaging / storage	Corrosion or coating degradation	Humidity, long WIP time, poor packaging	Poor insulation, poor bond, contamination	Humidity control, FIFO, sealed packaging	Surface inspection, storage audit	Medium

Quality lab inspection setup for lamination stack

How to think about AP, not just S/O/D numbers

Many older FMEA files still multiply Severity × Occurrence × Detection into an RPN. The problem is simple: different risk combinations can create the same number, even when one is clearly more serious.

The newer AP approach is more useful because it forces a question before the math becomes decorative:

Given severity, occurrence, and detection, how urgent is action?

For lamination stacks, AP should usually rise when:

The effect involves heat, electrical loss, shorting, winding damage, imbalance, or field failure.
The cause is linked to tool wear, fixture wear, or a process that can drift silently.
Detection happens late, after joining, winding, magnet insertion, or shaft pressing.
The inspection method cannot reliably see the real defect.
The defect is intermittent.

One awkward truth: intermittent defects are often worse than constant ones. Constant failures get noticed. Intermittent burrs, skipped sheets, scratched coatings, or loose stacks can escape because the process still “mostly works.”

That is exactly the kind of risk PFMEA should catch.

Critical controls by failure family

1. Burr and edge-condition controls

Burrs are not just cosmetic. They can lift laminations, damage coating, create conductive bridges, and affect stack height.

Good controls include:

Burr height limit by feature, not only by general part condition
Burr direction requirement
Tool wear tracking by station or cavity
Sharpening interval based on measured trend, not guesswork
Edge inspection after tool maintenance
Reaction plan when burr trend approaches limit

Avoid using final stack height as the main burr control. It is too late and too indirect.

2. Insulation and interlaminar short controls

Interlaminar insulation can be damaged by burrs, scratches, debris, excessive compression, poor handling, and aggressive joining.

Useful detection methods may include:

Interlaminar resistance check
Core loss test
Low-flux excitation test
Thermal scan during energized test
Sample teardown after compression or joining
Fault isolation on suspect stacks

Not every product needs every test. But if core loss or heat is a key product risk, the PFMEA should include a functional electrical check somewhere before the stack becomes expensive to scrap.

3. Stack count and height controls

Missing and double laminations are basic errors, but they still happen.

Use layered controls:

Sheet counting
Double-sheet detection
Weight check
Stack height under defined load
Force-displacement curve during compression

Height alone can lie. A double sheet may be partly masked by compression. A missing sheet may be masked by burrs, coating buildup, or debris. Pair the measurements.

4. Alignment and runout controls

For stator and rotor stacks, small alignment errors can become air-gap variation, winding trouble, magnet pocket issues, torque ripple, noise, or imbalance.

Controls should include:

Datum cleaning
Pin and nest wear checks
Anti-rotation features
Vision inspection of slot orientation
Bore-to-OD runout measurement
Skew angle verification if skew is designed into the stack

The PFMEA should list fixture wear as a cause. Not just “operator error.” Operator error is sometimes real. Often it is just a lazy label for a weak process.

5. Joining controls

Welding, bonding, interlocking, and riveting all create different risks.

A welded stack may be strong but can suffer local heat damage or distortion. A bonded stack may be clean but depends on surface condition, cure, and adhesive control. An interlocked stack may be efficient for high-volume production but can introduce local deformation if not controlled.

The PFMEA should connect joining risks to actual controls:

Weld energy, speed, position, and penetration checks
Bond material shelf life, mix ratio, cure time, and cure temperature
Rivet or interlock force monitoring
Pull, shear, or separation test
Cross-section audit
Dimensional check after joining

A nice-looking joint is not always a good joint.

Control plan linkage

A PFMEA is not finished until it drives the control plan.

PFMEA risk	Control plan item	Reaction plan
Burr exceeds limit	Burr check at defined frequency; tool wear trend	Stop, segregate suspect parts since last good check, inspect tool
Double lamination	Double-sheet sensor and weight check	Hold stack lot, verify feeder, audit recent stacks
Angular misalignment	Vision or mechanical angular gauge	Stop stacking cell, inspect datum pins and nest
Interlaminar short	Resistance or core loss test	Contain batch, review burr/coating/compression history
Weak joining	Parameter monitoring and pull test	Quarantine joined stacks, verify equipment settings
Stack height drift	Height under defined load and press force curve	Check material thickness, debris, burrs, press stop, sheet count
Corrosion	Storage humidity and WIP age check	Sort affected WIP, review packaging and storage condition

This is where many FMEAs fail. They list risks, then the control plan lives somewhere else and says “visual inspection.” That gap is where escapes happen.

Measurement examples for lamination stack PFMEA

Do not copy these as universal limits. They are examples of measurement types, not default specifications.

Characteristic	Possible measurement method	Why it matters
Burr height	Contact profilometer, optical microscope, vision system	Helps prevent insulation damage and stacking interference
Stack height	Height gauge under defined load	Confirms built height under repeatable condition
Stack mass	Precision scale	Helps detect missing or double laminations
Alignment	Vision system, slot gauge, CMM sampling	Confirms slot, tooth, bore, and OD relationship
Runout	Dial indicator, roundness system, CMM	Controls air-gap and balance risk
Interlaminar resistance	Electrical resistance test	Detects conductive paths between laminations
Core loss	Magnetic test fixture	Checks functional magnetic loss behavior
Compression profile	Press force-displacement monitoring	Finds debris, double sheets, under-compression, over-compression
Joining strength	Pull, shear, peel, or separation test	Confirms mechanical stack integrity

Specific limits should come from design requirements, customer specifications, capability studies, material data, and validation results. Guessing limits for SEO would be bad engineering.

Exploded view of lamination stack assembly

Practical PFMEA review questions

Use these in the meeting. They work better than staring at the spreadsheet.

Can this defect be created by tool wear?
Can it be hidden by compression?
Can it pass visual inspection?
Can it become worse after joining?
Can it damage winding, magnets, shaft fit, or housing assembly?
Can it increase core loss or local heating?
Can the current detection method actually find it?
What is the first point where this defect becomes expensive?
What is the reaction plan if the control fails?
Is this a design risk pretending to be a process risk?

The last question saves time. Sometimes the process is blamed for a design that has no margin.

Common PFMEA mistakes for lamination stacks

Mistake 1: Treating stack height as proof of stack quality

Stack height matters, but it is not proof of insulation health, correct count, clean layers, good alignment, or good joining.

Mistake 2: Hiding everything under “poor stack quality”

That phrase is too wide. Split it into actual failure modes: double lamination, missing lamination, burr short, angular misalignment, low stack factor, weak bond, corrosion, debris, distorted slot.

Mistake 3: Using final product test as the main detection control

Final test is important. It is also late. If a stack defect is found after winding or assembly, the PFMEA should ask why the stack was allowed to move forward.

Mistake 4: Not linking PFMEA to maintenance

Punches wear. Dies wear. Pins wear. Sensors drift. Press stops move. Fixtures get dirty. These are not side issues. For lamination stacks, they are normal causes.

Mistake 5: Mixing DFMEA and PFMEA without saying so

Design choices create the risk environment. Process controls manage production risk. Keep both visible, but do not mix them into one vague table.

FAQ

What is the most important failure mode in lamination stacks?

There is no universal single failure mode. In many applications, the most serious risks are interlaminar shorts, excessive burrs, misalignment, missing or double laminations, poor joining, and stack looseness.

Why are burrs such a big issue?

Burrs can damage insulation between laminations, create conductive contact, affect stack height, and interfere with winding or assembly. In magnetic cores, that can increase losses and local heating.

Is visual inspection enough for lamination stacks?

Usually not. Visual inspection may catch obvious damage, missing features, rust, or severe burrs. It is weak for internal shorts, count errors masked by compression, subtle angular drift, and joining weakness.

Should RPN still be used?

Some organizations still keep RPN for legacy systems, but modern automotive-style FMEA practice gives more weight to AP-based action decisions. AP helps prevent teams from treating very different risks as equal just because they produce the same multiplication result.

What should trigger a PFMEA update?

Update the PFMEA after tool changes, material changes, coating changes, new joining parameters, new inspection methods, fixture replacement, customer complaints, field returns, repeated scrap trends, or any defect that escaped the existing controls.

What is the best early warning sign of lamination stack trouble?

Burr trend, press force-displacement drift, double-sheet sensor faults, fixture wear, abnormal stack height variation, and rising electrical test failures are strong early signals. One bad stack is a defect. A drifting trend is a process talking.

How should PFMEA connect to the control plan?

Every high-risk failure mode should have a prevention control, detection control, inspection frequency, owner, record method, and reaction plan. If the PFMEA says “interlaminar short” but the control plan only says “visual check,” the system has a hole.

Final note

A lamination stack is a repeated part, but not a simple one.

The process can make the same tiny mistake hundreds of times in one core. Burrs repeat. Scratches repeat. Alignment error repeats. Compression damage repeats. Then the finished stack behaves like the defect was designed into it.

That is the point of PFMEA here: not to fill a form, not to impress an auditor, and not to rank risks until the numbers look tidy.

It is to catch the small process failures while they are still small.