3D-Printed Copper Motor Windings: Slot Fill, AC Loss, Insulation, and Cooling

Decision first: a 3D printed copper motor winding is not automatically a lower-loss winding. Additive manufacturing can raise geometric slot utilization, shorten end turns, vary conductor section, transpose or segment current paths, and place cooling close to the copper. Those gains matter only after the finished conductor's resistivity, AC loss, insulation, joints, temperature rise, mechanical support, and production repeatability have been demonstrated over the motor's real duty map.

The central mistake is to treat high slot fill as proof of high motor efficiency. More copper can reduce DC resistance, but a larger solid conductor can suffer greater skin, proximity, slot-leakage, and circulating-current losses as frequency rises. Cooling voids and insulation consume slot area, rough AM surfaces complicate coating, and a short end winding may create a difficult interface. Compare total loss and temperature at defined operating points, not one geometric percentage.

Published by COPPER 3DP / Suzhou Como. This article provides general engineering decision guidance. Motor architecture, manufacturability, insulation, qualification, and acceptance criteria require project-specific confirmation.

Define a Motor Winding Before Comparing Copper Coils

Here, a motor winding means the current-carrying stator conductors, end turns, phase connections, terminals, insulation, impregnation, supports, and any integrated thermal features that operate inside a rotating electrical machine. It is not a plastic hobby motor whose housing happens to be printed. It is also not an induction-heating coil or a busbar.

An industrial induction coil creates a field to heat an external workpiece; workpiece coupling, heating frequency, active-face geometry, coolant, and cycle repeatability dominate that system. A copper busbar distributes current between defined terminals; voltage drop, contact resistance, enclosure, and assembly temperature dominate. A motor winding must additionally work in the stator-slot magnetic environment, establish the required turns and phase connections, survive converter impulses and vibration, and meet a torque-speed thermal envelope.

Freeze the topology, pole-slot combination, stack and slot geometry, turns, parallel paths, voltage, current waveform, speed, switching, cooling, ambient range, and required life before optimizing a printable coil. Without them, "high fill," "low loss," and "high power density" are untestable.

Choose the Winding Architecture Before the AM Process

Round wire, conventional hairpin, a hybrid printed end winding, a full-AM shaped conductor, and an internally cooled AM winding solve different production problems. The best baseline is the best credible conventional design for the same motor and duty map, not an intentionally weak reference. A 2020 University of Bristol study experimentally demonstrated shaped-profile AM windings for a concentrated-wound open-slot permanent-magnet machine. Its result supports geometry tailored to the local magnetic field, but it also shows that loss behavior is operating-mode dependent.

Architecture	Where it can create value	Dominant penalty	Evidence required before selection
Round-wire baseline	Mature insulation, winding, repair, and volume production	Conductor packing and end-turn geometry are constrained by winding practice	Measured resistance, loss, temperature, manufacturability, and cost baseline
Conventional hairpin	Rectangular packing and repeatable automated assembly	High-frequency loss, many joints, and variant-specific tooling	Parallel-path balance, AC-loss map, weld quality, and drive-cycle result
Wrought slot conductors plus AM end winding	Shorter end region and tool-less variant flexibility	Alignment, printed-to-wrought interface, coating, and low productivity	Interface resistance, geometry, insulation, thermal cycling, and fatigue
Full-AM shaped or transposed winding	Variable section, three-dimensional routing, segmentation, and transposition	Process-dependent conductivity, roughness, supports, and insulation access	Four-wire DC resistance, AC-loss map, coating qualification, CT, and machine test
AM winding with internal cooling	Reduced thermal path and direct removal of conductor heat	Trapped powder, pressure drop, leakage, corrosion, and dielectric consequences	Channel CT, cleanliness, flow, proof/leak, thermal, and durability tests

A 2024 Strathclyde study used algorithmic optimization and 2D/3D models to seek minimal total machine loss with AM winding geometry. Its reported 97% efficiency belongs to that modeled machine and search space; it is not a transferable AM-winding efficiency.

Define Slot Fill Factor Before Quoting a Percentage

Slot fill factor may mean bare copper area divided by gross slot area, insulated conductor area divided by usable slot area, or another project convention. The number changes when slot liner, turn coating, phase insulation, impregnation resin, assembly clearance, cooling voids, and tolerances are included. State the numerator, denominator, section plane, and measurement method before comparing designs.

One peer-reviewed LPBF study reported approximately 80 µm minimum feature spacing, 79% slot fill for a square-conductor coil, 63% for a round-conductor coil, and 87% IACS after its heat treatment. Those values came from the paper's own machine, powder, scan strategy, coil geometry, and definition in On the Fabrication of High-Performance Additively Manufactured Copper Winding. They are not general copper-AM design limits.

More copper can lower resistance when material state, temperature, and current distribution remain favorable. It can also reduce room for insulation or cooling and enlarge the conductor exposed to slot leakage flux. Review the slot as an electromagnetic, dielectric, thermal, and assembly volume. The copper AM design-rules guide explains why walls, gaps, channels, and tolerances remain process-specific.

Shorten End Windings Without Moving Risk Into the Interface

End turns add copper length and DC resistance without producing useful air-gap force over the active stack. AM can sculpt the end region, avoid some bending and twisting constraints, and create a more direct connection. That makes a hybrid architecture attractive when the slot conductors are still better produced as conventional wire or bar.

An RWTH Aachen study printed pure-copper hairpin winding heads directly onto conventional conductors. In its stator, the printed twist-side end winding saved 12.5 mm, about one third of the conventional 37 mm winding-head height, and approximately 0.564 kg of copper. The same paper reported lower process productivity, demanding alignment, a downskin surface roughness around 32 µm Sa, and the need for further insulation breakdown testing. Read the hybrid PBF-LB/M winding-head study as one architecture demonstration, not as a universal percentage reduction.

A shorter end region is useful only if the transition avoids current crowding, heat-damaged insulation, coating gaps, excessive residual stress, and weak support. Specify its CAD, locating method, surface preparation, resistance, insulation process, and fatigue load case.

Separate IACS, DC Resistance, and Complete-Motor Efficiency

Material conductivity, coil resistance, copper loss, and motor efficiency answer different questions. IACS describes a controlled specimen and state. DC resistance adds path length, effective section, defects, temperature, and joints. Copper loss adds nonuniform AC current. Motor efficiency also includes core, magnet, mechanical, aerodynamic, stray, inverter-related, and thermal effects.

The literature itself shows why a single AM-copper conductivity assumption is unsafe. A 2022 400 W L-PBF study of Cu, Ag, and Cu-Ag variants reported a highest density of 88% and a highest conductivity of 73% IACS, while also investigating electropolishing for later insulation. A separate 2025 electron-beam study reported more than 99% bulk density and conductivity equal to 91.7 ± 1.8% of standardized pure copper. Compare the actual scopes in Robinson et al. and the EBM copper-winding feasibility study; neither result defines all AM copper.

Specify resistivity or conductivity by material, orientation, final heat treatment, specimen geometry, surface condition, test method, reference temperature, sampling location, and lot. Then measure the finished phase resistance by a suitable four-terminal method and record conductor temperature. The IACS specification guide gives the required property boundaries. Never convert a coupon IACS result directly into an efficiency claim.

Map Skin, Proximity, Slot-Leakage, and Circulating-Current Losses

At DC, increasing conductor area generally reduces resistance for the same length and material state. At AC, current distribution depends on frequency, conductor dimensions, neighboring conductors, local leakage field, slot opening, lamination geometry, parallel connections, temperature, and PWM harmonics. Conductors near the slot opening may see a different field from conductors deeper in the slot. A design optimized for peak torque at low speed may be poor in high-speed field weakening.

A peer-reviewed study of hairpin machines separates four mechanisms: skin effect from the conductor's own changing field, proximity effect from nearby conductors, eddy current caused by external field, and circulating current in connected strands or parallel paths. Its high-frequency loss mitigation analysis supports strand-level modeling and measurement; it does not supply a universal frequency or conductor-thickness limit.

AM creates more options than merely making a solid rectangle larger. The profile can vary with slot position, strands can be segmented, and transposition can be moved into the end region. A 2024 IEEE study manufactured two- and three-layer additively manufactured end-transposed rectangular windings and experimentally verified AC-loss suppression in its doubly salient brushless DC generator. That result remains generator-, geometry-, connection-, and excitation-specific.

Request DC resistance at defined temperature and an AC-loss or AC/DC-resistance map over intended current, frequency, switching, and temperature. One impedance point, unloaded spin, or uncorrelated FEA contour is insufficient.

Design Insulation and PDIV Before Freezing the Copper

AM texture, edges, downskin, pores, and oxide influence coating continuity and local electric field. A 2024 Bristol study compared powder, spray, and dip coatings on AM CuCrZr and a commercial conductor. Surface roughness and coating route mattered; basic smoothing was necessary, and spray gave the most conformal coverage in that comparison. The insulation-coating study measured breakdown behavior, not complete motor-insulation qualification.

Lock the surface state, edge radii, cleaning, coating chemistry, thickness distribution, cure, impregnation compatibility, repair method, and inspection before reducing conductor gaps. The surface-roughness and post-processing guide explains why one Ra or Sa number cannot represent every winding face.

For voltage-source-converter-fed AC machines within the relevant IEC scope, identify whether the winding insulation is Type I, intended to remain partial-discharge-free, or Type II, intended to resist partial discharge. IEC 60034-18-41 addresses Type I systems, while IEC 60034-18-42 covers Type II systems expected to withstand PD activity. IEC TS 60034-27-5 provides an offline method for PD inception and extinction voltage under repetitive impulses. None supplies one universal PDIV target for every voltage, rise time, cable, altitude, environment, winding geometry, or application.

Qualify Connections and Mechanical Support as Part of the Winding

A printed coil still needs phase connections, terminals, assembly references, and restraint. A monolithic CAD model does not eliminate every joint; full-AM windings may still need terminals, sensors, fluid couplings, or service connections.

For conventional hairpins, laser welds are numerous and function-critical. A 2025 open-access fatigue study notes that loss of electrical contact at one hairpin weld can make the machine inoperable and that established fatigue guidance for pure-copper weld seams is limited. Its notch-stress assessment of Cu-ETP hairpin welds is valuable evidence for treating joint geometry and cyclic load seriously, not a fatigue allowable for a printed joint.

Define electromagnetic force, vibration, handling, coolant reaction, thermal expansion mismatch, acceleration, and thermal-cycle loads. Inspect joint geometry and pores, measure contact resistance, and test the joined assembly mechanically and electrically before potting hides the interface. Where form-wound insulation is relevant, IEC 60034-18-34 provides comparative thermomechanical-endurance procedures for insulation systems; it does not replace application vibration or joint fatigue tests.

Integrate Cooling Only When the Channel Can Be Cleaned and Proved

Direct conductor cooling can reduce the thermal path from copper to coolant and may permit higher continuous loading. It also removes conductive section, changes AC current paths, adds fluid pressure, creates powder-removal obligations, and introduces a leak beside electrical insulation. Coolant chemistry, corrosion, electrical conductivity, gas management, pressure drop, pumping power, freeze behavior, and service cleaning belong in the design.

A 2018 non-AM study developed and experimentally assessed directly cooled tooth-coil concepts based on cast coils while accounting for current displacement. Its reported high current density and additional-loss reduction belong to those coil designs and cooling conditions; it does not validate AM manufacture or channel cleanability. Use the direct liquid-cooling paper to justify coupled electromagnetic and thermal work, not to quote a transferable current-density limit.

The strongest caution comes from Oel et al.'s 2023 multi-material PBF-LB stator-tooth study. The CuCr1Zr/316L component distorted thermally; electrical testing found short-circuit currents involving the coils and teeth, and intended cooling channels could not be completely freed of powder. The authors documented the result in Additive Manufacturing Letters. A channel visible in CAD, or even visible at an opening, is not yet a clean, electrically safe, hydraulically repeatable cooling circuit.

Require access and orientation review, a depowdering procedure, retained-powder acceptance, justified CT, cleanliness, pressure-drop, proof/leak, coolant-compatibility, and thermal tests. Retest after any process that can block, crack, distort, or contaminate the circuit.

Use a Claim-to-Evidence Qualification Ladder

Qualification should move from material to conductor, coil, motorette, stator, and machine. A 2026 Warwick study experimentally compared an LPBF edgewise coil mounted on a stator with a commercially manufactured edgewise coil and found comparable loss behavior for that geometry. The authors still required insulation and thermal/dielectric qualification and used Kapton interlayer tape as a laboratory solution. Keep that boundary attached to the loss-characterization result.

Claim	Minimum useful evidence	Conditions to record	Invalid substitute
Conductive material	Calibrated resistivity or conductivity on controlled specimens	Chemistry, orientation, heat treatment, surface, density, temperature, and method	Powder purity or a generic material-page value
Low winding resistance	Four-terminal phase resistance and balance on the joined winding	Conductor temperature, terminal definition, current, stabilization, and uncertainty	Coupon IACS alone
Low AC copper loss	Measured and correlated loss map on coil, motorette, stator, or machine	Current, fundamental and PWM spectrum, slot field, speed, temperature, and connection	One DC point or uncorrelated FEA image
Qualified insulation	Turn, phase, and ground-wall tests plus applicable PD and endurance work	Waveform, rise time, repetition, altitude, temperature, humidity, and thermal cycles	Coating thickness or coupon breakdown voltage
Controlled temperature	Steady and transient hotspot tests across required operating points	Coolant state, ambient, flow, pressure, inlet temperature, duty, and sensor locations	Room-temperature flow test
Reliable cooling channel	CT/cleanliness, pressure drop, proof, leak, corrosion, and cycling evidence	Fluid, contamination, temperature, pressure cycles, and post-process state	A successfully printed opening
Higher motor efficiency	Comparable complete-machine efficiency and temperature maps	Control, inverter, cooling, torque, speed, temperature, and loss segregation	High slot fill, short end turns, or conductor-only simulation

IEC 60034-2-3:2024 specifies test methods and interpolation for losses and efficiency of converter-fed AC motors, including defined regions of operation. IEC 60034-1:2026 supplies the general rating-and-performance framework but excludes machines covered by other application standards, including road and rail vehicle machines. Naming either document does not establish automotive, aerospace, or another sector-specific qualification.

Put These 14 Items in the Motor-Winding RFQ

1. Machine and program definition: motor topology, application, pole-slot combination, active length, prototype and production quantities, and target production gate.
2. Controlled geometry: stator-stack and slot CAD, slot opening, end-turn envelope, assembly datums, terminal locations, keep-outs, tolerances, and machining allowances.
3. Electrical duty: phases, turns, connection, parallel paths, DC-link and phase voltage, RMS/peak current, fundamental frequency, PWM strategy, switching frequency, rise time, maximum speed, and duty map.
4. Slot-fill definition: target plus the exact numerator and denominator, including treatment of liner, conductor coating, impregnation, gaps, cooling voids, and usable slot area.
5. End-region requirements: conductor path, maximum axial/radial envelope, minimum spacing, cross-section transitions, support, access, and printed-to-wrought alignment.
6. Material and final state: pure copper or alloy chemistry, powder limits and reuse, density, build orientation, heat treatment/HIP, surface state, and traceability.
7. IACS or resistivity acceptance: minimum conductivity or maximum resistivity, test temperature, direction, location, specimen geometry, method, correction, sampling, and uncertainty.
8. DC and AC loss acceptance: phase resistance and balance at stated temperature plus the required AC-loss or AC/DC-resistance map over current, frequency, speed, harmonic, and thermal conditions.
9. Insulation system: Type I or Type II strategy where applicable, turn/phase/ground insulation, edge preparation, coating process and thickness, cure, impregnation compatibility, thermal class, and repair rule.
10. Dielectric and PD evidence: insulation resistance, withstand, PDIV/PDEV, repetitive-impulse waveform, rise time, repetition rate, altitude, environment, thermal-cycle conditioning, standards, and pass limits.
11. Thermal acceptance: continuous and transient loss points, allowable conductor and insulation hotspots, cooling boundary conditions, sensor locations, temperature-rise method, and model-to-test correlation.
12. Cooling-circuit evidence: coolant and chemistry, inlet range, flow, pressure, allowed pressure drop, channel-access plan, powder-removal and cleanliness limits, proof/leak criteria, corrosion, and service flushing.
13. Connections and mechanics: terminals, weld/braze/printed interface, contact resistance, joint inspection, electromagnetic force, vibration, handling, thermal expansion, fatigue, and restraint requirements.
14. Release package: first-article dimensional/XCT and minimum-area evidence, process records, coupons, material results, phase resistance/inductance, AC loss, insulation/PD, flow/leak, thermal and complete efficiency maps, raw data, sampling plan, and production-capability thresholds.

Make the Decision on Total System Evidence

A defensible program answers four questions: does three-dimensional geometry beat the best conventional baseline; does the final material state meet electrical and mechanical needs; can the real surface be insulated, connected, cooled, inspected, and assembled; and does the stator or motor improve the specified duty map with acceptable production risk?

Reject five shortcuts: do not transfer one paper's slot fill or end-turn reduction to another motor; do not assign one IACS result to all printed copper; do not convert comparable coil loss or simulated efficiency into complete-machine performance; do not treat a printable cooling channel as a qualified circuit; and do not treat coating thickness, coupon breakdown voltage, or laboratory tape as a released insulation system.

If a motor winding is at concept, redesign, or sourcing stage, send the controlled stator/slot CAD, electrical duty map, end-region envelope, material preference, insulation and converter stress, cooling conditions, quantity, and acceptance plan through the COPPER 3DP engineering RFQ page. The first review should decide whether full AM, an AM end-winding hybrid, or a conventional winding is the lowest-risk route to the required measured performance.

Disclosure: This article was prepared with AI-assisted research and editorial review.