Estimating Core Loss for a Non-Oriented Electrical Steel Stack in Python (50W470 Example)

electricalsteel — Thu, 12 Nov 2026 09:18:21 GMT

A non-oriented electrical steel grade like 50W470 already contains two numbers you can use: the strip thickness and the guaranteed core loss. This post turns them into a quick estimate of stack mass, lamination count and core loss, so you can sanity-check a motor core before asking for a quote.

What the grade code tells you

In the Chinese standard (GB/T 2521), 50W470 reads like this:

50 → 0.50 mm thickness
W → non-oriented
470 → maximum core loss of 4.70 W/kg at 1.5 T and 50 Hz (P1.5/50)

The same steel appears as M470-50A in EN 10106 and 50A470 in JIS. If you need to compare codes across standards, our guide to electrical steel grades and standards lists the equivalents.

The calculation

Three steps:

Net iron volume = lamination area × stack height × stacking factor
Mass = volume × density
Core loss = mass × P1.5/50 × building factor

import re

DENSITY_G_CM3 = 7.65  # typical for non-oriented grades; check your mill data sheet

def parse_gb_grade(code: str) -> dict:
    m = re.fullmatch(r"(\d{2,3})W(\d{3,4})", code.strip().upper())
    if not m:
        raise ValueError(f"Not a GB non-oriented grade: {code}")
    return {
        "thickness_mm": int(m.group(1)) / 100,
        "p15_50_w_per_kg": int(m.group(2)) / 100,
    }

def estimate_core_loss(grade, lamination_area_cm2, stack_height_mm,
                       stacking_factor=0.96, density=DENSITY_G_CM3,
                       building_factor=1.0):
    g = parse_gb_grade(grade)
    volume_cm3 = lamination_area_cm2 * (stack_height_mm / 10) * stacking_factor
    mass_kg = volume_cm3 * density / 1000
    return {
        "iron_mass_kg": round(mass_kg, 2),
        "laminations": round(stack_height_mm * stacking_factor / g["thickness_mm"]),
        "core_loss_w": round(mass_kg * g["p15_50_w_per_kg"] * building_factor, 1),
    }

print(estimate_core_loss("50W470", lamination_area_cm2=150, stack_height_mm=100))
# {'iron_mass_kg': 11.02, 'laminations': 192, 'core_loss_w': 51.8}

print(estimate_core_loss("50W470", 150, 100, building_factor=1.3))
# {'iron_mass_kg': 11.02, 'laminations': 192, 'core_loss_w': 67.3}

For a 150 cm² lamination stacked 100 mm high, you get about 11 kg of iron, roughly 192 sheets, and about 52 W at the catalog value.

Where the estimate stops being accurate

P1.5/50 is a guaranteed maximum measured on test strips at 1.5 T and 50 Hz. Your motor may run at a different flux density and frequency, so treat this as a comparison number between grades, not a prediction.
Building factor is not 1.0. Punching, burrs, clamping stress and harmonics usually push real core loss above the catalog figure. Use a factor from your own test data.
Stacking factor depends on your process. Coating type, burr height and stacking pressure all change it. 0.96 is a placeholder, not a rule.
Density varies slightly by grade. Take the value from the mill certificate when you have it.

Using it to compare grades

Run the same stack through a few grades and look at the loss difference against the price difference per kg. A thinner, lower-loss grade does not always pay back, especially on low-duty-cycle motors. Specs for our non-oriented range are on the CRNGO steel coil page.

If you are comparing grades for a specific motor, send us the lamination drawing and the loss target and we will tell you which stock grades fit.What the grade number guarantees, and how to turn it into a first-pass loss figure for a lamination stack

How to Write a Slitting Spec for Electrical Steel (CRGO and CRNGO)

electricalsteel — Fri, 06 Nov 2026 01:56:18 GMT

A master coil of electrical steel comes off the mill 1,000 mm wide or more. A transformer core leg or a motor stator die needs something much narrower, so almost every coil of CRGO or CRNGO gets slit before it is cut, stamped or stacked.

The slitting order is usually a single line in a purchase order, something like "slit to 230 mm". That line leaves out most of what decides whether the strip will run on your line. This post covers the fields a slitting spec should include, why each one matters, and the defects you get when one of them is left blank.

First, know which steel you are slitting

Grain-oriented (CRGO) and non-oriented (CRNGO) steel are slit on the same kind of line, but they are bought for different jobs, and the job changes the spec.

CRGO goes into transformer cores. The flux has to run along the rolling direction, so strip width maps directly onto core steps. Width error shows up as gaps in the step-lap joints, and those gaps raise no-load loss.
CRNGO goes into motors and generators. The flux rotates, so direction matters less. What matters is that the strip feeds a progressive die cleanly. Camber and burr cause the trouble here.

If you are not sure which one your design calls for, this comparison of grain-oriented vs non-oriented electrical steel explains the three differences that drive the choice.

The 8 fields of a slitting spec

1. Input grade and thickness

Give the full grade designation, for example 27Q120, B23P085 or 50W470. Do not write just "CRGO 0.27". Slitting does not change the grade, but your supplier has to slit from a coil that meets it, and the coil number on each slit coil should trace back to that master coil's test data.

2. Slit width(s)

List every width you need. Transformer cores built with step-lap joints often need four to eight different widths for one core. Those widths can usually come off the same master coil in one or two passes, as long as the widths, the knife allowance and the edge trim add up to no more than the master width.

A worked example for a 4-step limb:

Step	Width	Strips per pass
1 (centre)	240 mm	1
2	215 mm	2
3	185 mm	2
4 (outer)	145 mm	4

All of them coming from one master coil keeps the grade and heat consistent across the whole core, which is worth asking for.

3. Width tolerance

Never leave this field empty. A supplier will fill it with whatever their line normally holds.

±0.3 mm is a general tolerance and fine for simple distribution cores.
±0.1 mm suits most step-lap transformer cores.
±0.05 mm is a precision line tolerance. Ask for it for step-lap cores in high-efficiency designs and for EV motor stator dies.

Tighter costs more, because the knives have to be set and checked more often. Specify what the design needs, not the tightest number you have seen.

4. Burr height

Burr is the small lip a slitting knife leaves on the cut edge. In a stacked core it can bridge the insulation between laminations and create local eddy-current paths. In a motor die it shows up as stacking-factor loss.

Put a number on it, in mm or µm. Burr depends on how sharp the knives are, so a low limit means the supplier has to regrind knives on schedule. That is exactly the discipline you want them to have.

5. Camber (edge straightness)

Camber is how far the strip bows sideways, usually measured over 2 m. A strip with too much camber walks across the feeder of a progressive die and produces off-centre laminations. For motor stamping, a tight camber limit, often 0.5 mm per 2 m or better, matters more than a tight width tolerance.

6. Coil inner diameter

Match it to your decoiler mandrel, typically 500 mm or 508 mm. Get this wrong and the coil will not go on the machine at all.

7. Maximum coil weight and outer diameter

Your decoiler, crane and floor loading set the limit, and so does how often you want to change coils. Narrow strips wound to the full master coil diameter can be heavy and unstable, so cap the weight (for example ≤ 2,000 kg) or the OD.

8. Packing

Electrical steel coating is sensitive to moisture and to strap damage. State what you want: VCI paper or film, edge protectors, eye-to-sky or eye-to-wall stacking, and whether you want plastic or steel strapping. Narrow strips also need separators between coils so they do not telescope in transit.

A one-line spec vs a usable spec

One line:

CRNGO 0.5 mm, slit 230 mm, 20 t.

Usable:

50W470, 0.50 mm. Slit width 230 mm ±0.05 mm. Burr ≤ 0.02 mm. Camber ≤ 0.5 mm / 2 m. ID 508 mm, coil weight ≤ 1,500 kg. VCI wrapped, eye-to-sky on wooden pallet, plastic strapping. 20 t total. Coil labels to show master coil number.

The second version costs a supplier five minutes to quote, and it removes most of the arguments that come up when coils arrive.

What to check when the coils arrive

Width at five or more points over the first few metres of each coil.
Labels: each coil's number should trace to the master coil's test data.
Edges: burr checked with a micrometer at a few points per edge.
Camber: unroll about 2 m on a flat floor and measure the bow against a straight edge.
Surface: look for coating scratches, knife marks and rust.

Does slitting hurt the magnetic properties?

It does, but only at the cut edge. Mechanical cutting work-hardens a narrow band along the edge, and the steel there has slightly worse magnetic properties. For wide laminations the effect is negligible. For very narrow strips or small motor teeth it becomes a larger share of the cross-section, which is one reason designers avoid extremely narrow laminations, and one reason some motor makers stress-relief anneal their stampings.

Where to go from here

If you are working out a slitting order for CRGO or CRNGO, you can find a full walkthrough of the slitting process, tolerance classes and order fields on this electrical steel slitting service page.

Written by CN Electrical Steel, an electrical steel supplier in Wuxi, China. We slit CRGO and CRNGO to width for transformer and motor core makers.

CN Electrical Steel – CRGO & CRNGO Notes