For rotating-machine cores, the choice between 0.27 mm and 0.35 mm non-grain-oriented (NGO) silicon steel is fundamentally a trade between eddy-current loss and manufacturing cost, and the breakeven sits near 400 Hz fundamental frequency per recent review work [S1].
0.35 mm NGO (e.g. grades M270-35A, M300-35A) remains the default thickness for industrial AC motors and generators, with AISI silicon stock widely supplied in the 0.35-0.5 mm range [S2]. 0.27 mm NGO enters the picture for traction, aerospace, and high-speed PMa-SynRel machines where the extra 23% reduction in sheet thickness yields a measurable drop in iron loss.
Thickness, Grade Coding, and Frequency Map
NGO steel is offered in a wider thickness range than grain-oriented (GO) because motor designs span very different frequencies and efficiency targets [S3]. Common NGO thicknesses are 0.35, 0.50, and 0.65 mm, while thin-gauge NGO at 0.20 mm and below is reserved for high-speed motors, EV traction, and high-frequency inductors [S3].
The numeric suffix in a European designation such as M270-35A carries two pieces of information: 2.70 W/kg max core loss at the standard test condition, and 0.35 mm nominal thickness [S3]. For 0.27 mm NGO the analogous grades sit in the M250-35A and M235-35A family; AISI equivalents in the M-15 to M-19 range correspond to the thinnest 0.35 mm-class NGOES used in rotating machines [S2]. Operating frequency is the dominant driver: as electrical frequency rises, eddy-current loss scales with the square of thickness, and the thinner lamination starts to pay back its processing penalty above roughly 400 Hz fundamental, or above 800 Hz electrical in a high-pole-count machine [S1].
Core Loss and Magnetic Performance Numbers
Cutting-process data on M270-35A at 0.35 mm show that laser cutting on thin or narrow magnetic regions can nearly double iron loss and reduce developed torque by more than 6% compared with the parent strip, because heat-affected zones in laser-cut 0.27 mm NGO expand the damaged fraction of the lamination cross-section [S1].
WEDM (wire electrical discharge machining) preserves magnetic properties best across both thicknesses, with minimal torque and efficiency penalties in the same PMa-SynRel test campaign, while punching lands in the middle, with degradation scaling with burr height, tool wear, and plastic strain [S1]. Beyond loss, the headline market-level figures: GO delivers magnetic permeability above 1.90 Tesla, and NGOES supports motor efficiencies exceeding IE4/IE5 classes, which is precisely why high-efficiency rotating machines continue to standardise on NGO despite the lower peak B of GO [S6].
Manufacturing and Cost Trade-Offs at 0.27 mm vs 0.35 mm
The 0.27 mm option buys lower eddy-current loss but raises three manufacturing penalties. First, tooling wear: thinner stock demands tighter clearance control and shortens punch life. Second, stacking factor: with the same interlaminar insulation build-up, a 0.27 mm stack needs roughly 30% more sheets than a 0.35 mm stack for the same core height, increasing handling, stacking, and clamping labour. Third, scrap and burr sensitivity rise because the cut-edge damaged zone is a larger fraction of the magnetic cross-section in thin stock [S1][S3].
For a 3.75 kW PMa-SynRel demonstrator, geometry-dependent degradation was significant enough that the published recommendation is to include cutting-process effects in electromagnetic models rather than assume virgin-strip B-H curves [S1]. For a 225 kW PMa-SynRel, the same study confirms that the lamination choice and the cutting method must be co-selected, since the iron-loss gap between laser-cut and WEDM-cut 0.27 mm parts exceeds the loss gap between 0.27 mm and 0.35 mm cut by the same process [S1]. In a high-volume motor grader drivetrain or industrial pump build, the unit economics still favour 0.35 mm NGO unless the electrical frequency forces the issue.
Decision Matrix: 0.27 mm vs 0.35 mm NGO by Use Case
Cost (per kg, baseline): 0.35 mm NGO is the cheaper stock; 0.27 mm NGO commands a 10-25% premium driven by rolling and annealing yield [S2][S3]. Tooling wear: 0.27 mm shortens punch life and requires finer clearance; 0.35 mm tolerates worn tooling with less magnetic penalty [S1][S3]. Frequency fit: 0.27 mm wins above ~400 Hz fundamental; 0.35 mm is optimal up to ~200 Hz, with the crossover band from 200-400 Hz driven by duty cycle and efficiency class [S1]. Cut-edge sensitivity: 0.27 mm is roughly 1.3x more sensitive to the same laser HAZ fraction because the damaged depth is a larger share of thickness [S1].
Use 0.27 mm NGO for EV traction motors, aerospace actuators, high-speed PMa-SynRel machines above 10 krpm, and any design chasing IE5 efficiency at elevated fundamental frequency [S1][S3]. Use 0.35 mm NGO for general-purpose industrial AC motors, standard drive motors below 10 kW, induction generators, and servo motors where unit cost, supplier availability, and predictable stamping behaviour dominate [S2][S3].
Standards, Sourcing, and Specification Discipline
Specification discipline matters more than thickness alone. A 0.27 mm M250-35A sheet purchased as a 0.35 mm equivalent will pass dimensions but fail the loss class because the core-loss number in the designation is thickness-dependent [S3]. Mill test certificates should be checked for both the W/kg figure at 1.5 T/50 Hz and the actual measured thickness, since some stock is supplied at 0.30 mm under a 0.27 mm label.
Selection should also account for the GO versus NGO crossover: GO thicknesses of 0.18, 0.23, 0.27, and 0.30 mm are produced, and the GO market segments by thickness into 0.23, 0.27, 0.30, 0.35 and above, and 0.20 and below buckets, but these are transformer-oriented grades and will underperform a rotating machine unless the flux path is heavily directional [S3][S4]. For the same reason, GO permeability above 1.90 Tesla is not a reason to specify GO in a random-flux motor; the isotropic 1.50-1.70 T range of NGOES is what keeps torque ripple and radial forces predictable in a squirrel-cage AC motor [S6]. Designers chasing weight reduction in linear motors for aerospace or semiconductor stages should pair 0.27 mm NGO with WEDM or fine blanking rather than laser to hold the loss benefit, and budget for the stacking-factor hit up front in the FEA model.
Limitations and Failure Modes to Watch
Three failure modes recur in 0.27 mm NGO builds. First, interlaminar short: thinner sheets amplify the effect of any burr that bridges the insulation coating, so stacking-factor targets must be set 1-2% below 0.35 mm designs. Second, welding and joining: the smaller thermal mass makes laser welding more likely to cause local magnetic annealing; WEDM is safer but slower. Third, supply chain risk: 0.27 mm NGO is a thinner commercial market than 0.35 mm, with fewer qualified mills and longer lead times, so a dual-source qualification on 0.35 mm is a common hedge for volume programs [S1][S3].
For plants standardising on motor accessories such as laminating fixtures, stacking jigs, and burst-test rigs, the geometry rarely changes between 0.27 mm and 0.35 mm, but the press tonnage, stack height, and clamping pressure do, and these should be re-validated on the new thickness before production release.
Trackable signals for the next design cycle: watch for more published data on 0.20 mm and below NGO grades moving from aerospace into high-volume EV programs, and for press-shop disclosures on punch life deltas between the two thicknesses.
This topic is covered further in Hooked-End vs Straight Steel Fiber: Pull-Out Mechanics and Spec Tradeoffs.