Motor Lamination Stamping: Stator & Rotor Cores That Hold…

Every fractional-horsepower motor, EV traction machine, and servo actuator starts with the same precision part: laminations. A stator or rotor core is not one solid block of steel - it is dozens to thousands of thin stamped silicon-steel sheets stacked into a tight, aligned package. Stamp those sheets a micron off and the motor runs hot, noisy, and below its rated efficiency. This guide covers how we stamp motor laminations that actually hold tolerance - grades, stack build, burr control, and the annealing step that most shops skip - and it gives buyers the selection logic, tolerance tables, and supplier audit questions that turn a lamination RFQ into a motor that performs.

Lamination stamping sits at the intersection of three disciplines that rarely live in the same shop. The die designer must control strip geometry, the press operator must manage burr and stack flatness at speed, and the metallurgist must preserve the magnetic properties that the steel mill built in. A shop that treats laminations like ordinary brackets will ship stacks that meet the dimensional print but fail the efficiency target. The sections below walk through each control point, with the numbers to put in your RFQ and the checkpoints to enforce at first article.

Why Motors Use Laminated Cores

Eddy currents form inside any solid conductive rotor or stator exposed to a changing magnetic field. Those currents waste energy as heat and drag efficiency down. Laminating - stacking insulated thin sheets - interrupts the eddy-current path perpendicular to the flux. Typical lamination thickness is 0.10 mm to 0.65 mm, with 0.20 mm and 0.35 mm the most common for fractional-horsepower and EV traction motors. Thinner steel = lower eddy loss but more pieces per stack and tighter burr control.

The physics explains why the trend in motor design is toward thinner steel. Eddy-current loss scales with the square of lamination thickness, so halving the sheet thickness cuts the eddy component of core loss by roughly a factor of four. That is why a 0.27 mm grade can carry an efficiency class that a 0.50 mm grade cannot reach, and why premium EV traction motors specify 0.20 mm or thinner. The trade-off is economic: a thinner stack needs more sheets, more press strokes, and tighter burr limits to reach the same stack height, so the material and processing cost per core rises.

The second loss component, hysteresis, depends on the steel chemistry and grain structure rather than the sheet thickness. Silicon content of 2-3.5% raises resistivity and lowers both loss components, which is why motor laminations are almost always silicon steel rather than plain carbon steel. The combination of silicon content, thickness, and processing route (fully processed versus semi-processed) is what the grade designation encodes.

Silicon Steel Grades and Thickness Selection

Motor laminations are almost always non-oriented silicon steel (NO steel), chosen for low core loss rather than grain direction:

50W470 / 50W600 - 0.50 mm, general-purpose fractional-HP and appliance motors (IE2-IE3).
35W300 / 35W440 - 0.35 mm, higher-efficiency drives, servo and pump motors (IE4).
27ZW1400 / 27ZW1500 - 0.27 mm, laser-scribed, premium EV traction and aerospace (IE5 / ultra-premium).
M19 / M36 (ASTM) - legacy fully-processed grades still specified on legacy draws.

Core loss is quoted at 1.5 T / 50 Hz - e.g. 35W300 ≈ 3.0 W/kg, 27ZW1400 ≈ 1.4 W/kg. We stock the common grades and can laminate to a customer-specified melt.

The grade choice is a cost-performance decision, not a materials preference. The table below shows how the selection logic works across the motor classes we stamp:

Motor class	Typical grade	Thickness	What drives the choice
Appliance / fan / pump (IE2-IE3)	50W470, 50W600	0.50 mm	Lowest material cost, adequate loss for fixed-speed duty
Industrial drives / servo (IE4)	35W300, 35W440	0.35 mm	Lower eddy loss at inverter switching frequencies
EV traction / premium (IE5)	27ZW1400, 27ZW1500	0.27 mm or 0.20 mm	Minimum core loss, laser-scribed domains, thin stack
Legacy / replacement programs	M19, M36 (ASTM)	0.35-0.64 mm	Drawing lock-in, replacement of existing core design

Buyers commonly over-specify the grade. If the motor runs at fixed line frequency and the efficiency class is IE2, a 35W300 grade adds material cost without measurable benefit over 50W470. If the motor is inverter-fed, the switching frequency generates additional high-frequency loss, and the thinner grade earns its premium. Send the operating point - frequency, duty cycle, efficiency target - with the RFQ, not just the drawing, so the grade recommendation is grounded in the actual loss profile.

Semi-processed versus fully-processed steel is the second axis. Fully-processed grades arrive with the final magnetic properties and need only stress-relief annealing after stamping to restore what blanking degraded. Semi-processed grades are stamped, annealed, and then the finished core is heat-treated by the motor builder to develop the final permeability. If the drawing does not state which route is intended, ask - the annealing schedule and the cost structure are completely different.

Stack Tolerances That Matter

A motor core is judged by how the whole stack behaves, not one blank:

Parameter	Typical capability	Why it matters
Blank thickness tolerance	±0.005 mm	Holds flux density per sheet
Blanking burr (cut edge)	≤ 0.03 mm (≤ 5% t)	Burr shorts inter-laminar insulation → eddy loss
Slot / tooth pitch	±0.01 mm	Air-gap uniformity, cogging torque
Stack height (via weld / rivet)	±0.05 mm per 25 mm	Active steel length = torque
Flatness of finished stack	≤ 0.10 mm over 100 mm	Rotors must seat true in the bore
Inner-diameter roundness (rotor)	≤ 0.02 mm	Bearing fit, runout

The stack height tolerance deserves special attention because it compounds. Each lamination carries its own thickness variation from the coil; if the strip thickness drifts by a few microns across the coil, and the stack count is fixed, the finished height drifts with it. Two controls fix this. The first is a coil thickness specification with the tolerance band communicated to the stamper, not just the nominal value. The second is stack-height compensation: the press counts sheets, measures the growing stack, and adjusts the sheet count or the rivet/weld program so the final height lands inside the window even when the coil drifts.

Slot and tooth pitch is the feature that determines cogging torque and air-gap uniformity. A pitch error of a few hundredths of a millimetre accumulates across the circumference and shows up as torque ripple and audible noise. Because pitch is set by the die and the piloting system, it is stable once the tool is proven - which is why the first-article measurement of slot pitch is a stronger quality signal than any downstream inspection.

Roundness and flatness failures come from the stacking method, not the blanking. Welding shrinks material locally and can pull the stack oval; riveting displaces material and can bow the stack if the tabs are placed asymmetrically. The stack method has to be matched to the geometry - a thin-walled stator with tight roundness cannot take the same weld pattern as a chunky rotor core. See the stacking section below for the trade-offs.

Burr Control Is the Whole Game

A burr on a lamination edge is the fastest way to wreck a motor. Burrs pierce the inter-laminar oxide or coating and create a short circuit between sheets - exactly the eddy path lamination is meant to break. We hold burr to ≤ 5% of thickness by:

Progressive dies with shaved / fine-blanked edges on critical cuts.
Monitoring punch-die clearance at 5-8% of thickness and replacing steel before it mushrooms.
De-burring or tumbling only where the design allows (rotor bores usually cannot be tumbled).

Burr grows with stroke count. A fresh die on 0.35 mm silicon steel may hold burr at 0.01 mm for the first hundred thousand strokes; as the cutting edges wear, burr climbs toward the 0.03 mm limit, and past that the insulation is compromised. The practical control is scheduled die maintenance: a maintenance interval defined in strokes, not weeks, with a burr-height check at each service. A stamper that cannot tell you the planned sharpening interval is running burr on hope.

Burr direction matters as much as burr height. In a stack, all laminations should present the burr on the same side relative to the stack axis, so the burr of one sheet does not bridge to the adjacent sheet. The die is designed so the burr lands on the designated side; flipping laminations during stacking defeats this and creates local shorts. Specify the burr side on the drawing, and verify at first article that the stack orientation matches.

Fine blanking and shaving add cost per stroke but buy a clean edge with a much smaller shear zone. For rotor bores that will carry a bearing, and for slot edges that must not short, the extra station is usually justified. For the outer contour of a stator where the edge is not electrically critical, a conventional blank with controlled burr is enough. The decision belongs in the die design review, not discovered at PPAP.

How the Stack Is Held Together

Individual laminations must become one rigid core:

Self-riveting (notching + protrusion) - cold-formed tabs lock sheets with no extra parts; best for high-volume stator and rotor stacks.
Laser / spot welding - along the yoke or at weld tabs; common for rotors and skewed stacks.
Adhesive bonding (epoxy / phenolic) - full-surface bond for lowest acoustic noise in premium EV and servo motors.
Caulking / interlocking - simple mechanical interlock for low-stress applications.

Stack method	Relative cost	Acoustic / loss impact	Best fit
Self-riveting (tabs)	Lowest	Local stress, minor loss rise at tab	High-volume appliance and industrial motors
Laser / spot weld	Low-moderate	Heat-affected zone can raise loss near weld	Rotors, skewed stacks, larger OD cores
Adhesive bonding	Highest	Lowest noise and vibration	Premium EV, servo, and acoustic-sensitive motors
Caulking / interlock	Lowest	Weakest mechanical joint	Low-stress, low-cost applications

Riveting is the default for a reason: it is fast, adds no consumables, and the tab geometry is stamped into the blank in the same die. The downside is that each rivet is a local cold-worked zone where permeability is degraded and where the insulation coating is broken by the deformation. Designers limit rivet count and place tabs in the yoke where the flux density is lowest, so the loss penalty stays small.

Welding trades a different set of risks. A weld seam along the yoke is fast and strong, but the heat-affected zone anneals the steel locally and can raise core loss at the weld line; a poor weld can also distort the stack into an oval. Laser welding with low heat input and a controlled pattern minimizes both effects. For skewed rotors, welding is often the only practical way to hold the twist, which is why it dominates that niche.

Bonded stacks are the premium answer for noise-sensitive applications. A full-surface epoxy bond damps vibration better than any mechanical joint and eliminates the local stress concentrations of rivets and welds. The cost is a curing step and a coating system that must be compatible with the adhesive. For EV traction motors where NVH is a customer-facing metric, bonded stacks are increasingly the specification of record.

The Annealing Step Most Shops Skip

Blanking and progressive forming work-harden silicon steel and degrade its magnetic properties. Stress-relief annealing (typically 750-850 °C in a decarburizing atmosphere) restores permeability and cuts core loss - often by 10-20%. We anneal motor laminations after stamping when the draw calls for IE4/IE5 efficiency or thin (≤0.27 mm) grades, then apply the insulating coating (C3/C5 phosphate or epoxy) after anneal.

Blanking is a cutting operation, but the material around the cut edge is plastically deformed, and that cold work raises the hysteresis loss. On thin, high-grade steel the effect is proportionally larger because the damaged zone is a bigger share of the sheet. Annealing recrystallizes the damaged structure and restores the permeability the mill built in. The result is a measurable core-loss reduction on the finished stack - the reason premium motor programs specify anneal as a mandatory step rather than an option.

The annealing atmosphere matters as much as the temperature. A decarburizing atmosphere removes carbon from the surface and grows a thin oxide layer that improves the inter-lamellar insulation; a neutral atmosphere leaves the surface clean but does not add the insulating benefit. The coating is applied after annealing because a coating applied before the furnace cycle would not survive it. The process order - stamp, anneal, coat, stack - is fixed by the metallurgy, and a supplier that skips anneal to save a day should not be quoting IE4+ programs.

Not every program needs annealing. A 0.50 mm appliance stator at IE2 does not justify the furnace cost, and the loss penalty from blanking is small relative to the total loss budget. The decision rule we use: anneal when the grade is 35W300 or thinner, when the efficiency class is IE4 or higher, or when the customer's loss measurement at first article shows the stamped stack missing the target. Buyers should state the loss target in the RFQ, not just the grade, so the anneal decision is made on data.

Inter-Lamellar Insulation

Each sheet needs a thin insulating layer so eddy currents cannot jump between laminations. Standard finishes:

C3 (organic) - 0.5-2 µm film, standard rotors and stators.
C5 / C6 (inorganic-phosphate) - higher temperature rating, EV and traction duty.
Epoxy / F-class - bonded stacks and high-thermal-grade motors.

The insulation class is specified against the motor's thermal class. C3 organic coatings serve most industrial motors and are the cheapest; C5 inorganic coatings survive higher operating temperatures and the aggressive environments of EV traction; C6 adds a resin component for improved punchability and stacking. The coating must survive both the stamping (the punch sliding against the coated surface) and the stack bonding (if adhesive is used), so the coating and stack methods are chosen together, not independently.

Coating integrity is the quiet failure mode. A coating that looks fine under a cursory inspection but is cracked at the cut edge lets the burr or the exposed steel short adjacent sheets. The inter-lamellar resistance test - applying a voltage across the stack and measuring leakage - is the standard verification, and it should be part of the first-article report for any program where loss matters. Ask your supplier for the test result at PPAP, not a coating certificate alone.

Geometry and Die Design We Can Hold

Outer diameter: up to 400 mm on our press range.
Slot shapes: round, wedge, and custom tooth geometries on a progressive die.
Skew: stepped or twisted rotors for low cogging (built into the stack tooling).
Shaft holes, keyways, and locating notches as part of the blank.

Lamination dies are progressive dies with a specific rhythm. The strip enters, a pilot hole is pierced, the slot and tooth geometry is cut progressively across stations (because a full stator blank with a dozen slots cannot be pierced in one hit without weakening the strip), the outer contour is blanked, and the stacking feature - rivet tab, weld tab, or bonding surface - is added in the final stations. Station count scales with slot count and geometry complexity; a typical stator die runs more stations than a connector die of the same physical size because every slot group needs its own cut.

The strip layout is where utilization is won or lost. Multi-up layouts, nesting two or more blanks across the strip, and using the scrap skeleton for smaller parts are the levers. On silicon steel, which is bought by the kilogram and priced on grade, utilization is a direct cost line: a five-point utilization gain on a high-volume stack program is worth real money over the life of the die. Ask for the strip layout and utilization figure with the quote, and treat a supplier that hides the layout as a red flag.

Skew is a geometry that most stamping shops cannot hold. A skewed rotor has its laminations rotated progressively around the axis, which cancels cogging torque and reduces noise. The skew is built either as stepped laminations (a small rotation between groups, produced by indexing the die) or as a continuous twist (produced by rotating the stack during riveting or welding). Both add die or stacking complexity, and both are capabilities worth confirming before awarding a skewed-rotor program.

The die steel and surface treatment matter on thin silicon steel. Silicon steel is abrasive - the silicon content that helps the magnetics wears cutting edges faster than mild steel. Tool steels with high wear resistance, carbide inserts on high-volume stations, and coatings that reduce galling extend the interval between sharpening and keep burr inside the limit across the whole run. Ask how the die is built for wear, because that answer predicts both burr stability and maintenance cost.

Failure Modes and Root Causes

Lamination programs fail in a small set of predictable ways. Knowing the root causes turns a mysterious efficiency miss into a fixable process change:

Symptom	Most likely root cause	Where it shows up
Core loss above target	Burr shorting sheets, missing anneal, wrong grade	Efficiency test, inter-lamellar resistance test
Stack height out of window	Coil thickness drift, no stack-height compensation	Height gauge at assembly
Torque ripple / noise	Slot pitch error, skew lost in stacking	NVH test, cogging measurement
Oval rotor bore	Weld distortion, rivet tab placement	Roundness measurement, bearing fit
Inter-lamellar shorts	Burr direction flipped, coating cracked, weld spatter	Resistance test, running temperature

The most common failure is the cheapest to fix: burr. Burr grows gradually with die wear, so the fix is a sharpening schedule tied to stroke count, with a burr-height gauge at the press. The most expensive failure is the one discovered at the motor builder's efficiency test - at that point the entire stack lot is suspect, and the rework decision (reject, strip, or accept with derating) is made under schedule pressure. Every control in this article exists to move that discovery earlier, to first article, where it costs a die tweak instead of a lot rejection.

Magnetic property drift between coils is a supplier-quality issue that shows up only at the efficiency test. Steel mills run to a loss band, not a point, and two coils of the same grade can sit at opposite ends of the band. The controls are a material certificate per coil, a loss measurement on the first stack of each coil, and a process that flags a coil whose loss sits at the high end of the band before it becomes fifty thousand finished stacks. Ask your supplier how they manage coil-to-coil variation; the answer separates a lamination specialist from a general stamper.

Supplier Evaluation Checklist

Use this checklist when auditing a lamination stamper, and put the answers in the RFQ response:

Burr control plan: What is the target burr, the measurement method, and the sharpening interval in strokes?
Annealing capability: In-house furnace or outsourced? Atmosphere control? Where is the coating applied in the sequence?
Stack-height control: How is coil thickness drift managed? Is there in-process stack-height measurement?
Skew capability: Stepped or continuous? Proven on similar OD and stack length?
Inter-lamellar test: Is the resistance test in the control plan? At what frequency?
Coil traceability: Per-coil material certificates, heat numbers, and loss data available?
Strip layout visibility: Will the supplier share the layout and utilization with the quote?
Quality system: IATF 16949:2016 certification current? PPAP documentation standard for automotive and EV programs?

The facility operates under IATF 16949:2016 and ISO 14001:2015, with first articles measured on CMM and optical systems, in-line vision on the high-speed lines, and per-lot records that trace every coil through stamping, annealing, coating, and stacking. Those systems are not decoration; they are the difference between a stack that meets the efficiency target on the first build and a program that spends its first three months chasing a loss number.

What We Need to Quote a Lamination Run

Send the drawing or sample and we return a tooling + piece price fast:

Steel grade and thickness (or we recommend from efficiency target).
OD / ID, stack height, slot count and geometry.
Burr and flatness limits (default: ≤5% t burr, ≤0.10 mm flatness).
Volume and annual forecast (drives die style and anneal decision).
Coating / insulation class.

The single most useful extra item is the motor's efficiency class or loss target. With that number, the grade, thickness, anneal decision, and stack method are all determined by engineering logic instead of guesswork. Without it, we quote to the drawing and may under- or over-specify the metallurgy.

Frequently Asked Questions

Can laminations be stamped from any silicon steel grade? Yes, within the strip envelope. The fleet stamps material from 0.05 mm to 3.0 mm thick on strip up to 650 mm wide, which covers the 0.20-0.65 mm lamination range comfortably. The grade choice is driven by the loss target and efficiency class, not by press capability.

Is annealing always required? No. It is required when the efficiency class is IE4 or higher, when the grade is 35W300 or thinner, or when the loss measurement misses target. For IE2-IE3 appliance duty on 0.50 mm grades, the furnace cost is usually not justified.

What is the difference between riveted and welded stacks? Riveting is faster and cheaper, adds no consumables, but creates local cold-worked zones; welding is stronger for skewed and larger cores but risks heat-affected zone loss and distortion. Bonded stacks cost most but deliver the lowest noise.

How is burr controlled at high volume? By die design (clearance 5-8% of thickness), scheduled sharpening tied to stroke count, and burr-height gauging at the press. Burr is measured at first article and monitored through the run, not discovered at the customer.

Can you hold a skewed rotor geometry? Yes - both stepped and continuous skew are built into the stacking tooling. Confirm the skew type and stack length with the RFQ so the die and stacking method are matched to the geometry.

Ready to Spec Your Core?

Whether it is a 20 mm servo rotor or a 300 mm EV traction stator, we stamp motor laminations that hold ±0.01 mm slot pitch and ≤0.03 mm burr - then prove it with a stack build report that includes the inter-lamellar resistance test and, where required, the anneal certificate. Send your lamination drawing for a quote and we will flag the grade, stack method, and anneal call before any tooling is cut.

Motor Lamination Stamping: Stator & Rotor Cores That Hold Tolerances