Two suppliers can quote the same electromagnetic design and still deliver cores that differ in no-load loss, noise and temperature rise. The difference is construction: the steel, the clamping, the cooling.
In a large hydro or turbo generator, the stator core alone can weigh over a hundred tonnes of stacked steel. Long before anyone argues about the winding, that stack has decided how much input energy becomes heat, how loudly the machine hums, and how evenly the air gap holds over decades of service.
Construction comes down to four linked decisions: lamination grade and thickness, punching and slot geometry, stacking and clamping method, and the ventilation path. Each has a cost side and a loss side, and a thinner lamination only pays off when coating quality and stacking factor are good enough to keep the slot area you paid for.
The build sequence is short, but each step constrains the next one.
Small cores merge these steps: a progressive die punches and interlocks the laminations, and the stack is welded or bonded on the outside diameter. That removes cost rather than risk, moving the risk into tooling wear and coating consistency.
Electrical steel grades are named for their loss. In a designation such as M400-50A, the 400 means a maximum specific loss of 4.00 W/kg at 1.5 T and 50 Hz, and the 50 means 0.50 mm thickness. Change the trailing number and you change the eddy current component, which grows with the square of thickness and frequency.
| Thickness | Grade examples | Where it fits | What you trade |
|---|---|---|---|
| 0.65 mm | M600-65A, M800-65A | Small, low-frequency machines and light-duty alternators | Highest eddy loss per kilogram, but the lowest material and tooling cost |
| 0.50 mm | M400-50A, M470-50A | Mainstream 50/60 Hz structures, hydro, industrial and standby generators | Balanced loss and cost; the default unless efficiency targets demand more |
| 0.35 mm | M350-35A, M300-35A | High-efficiency and higher-frequency generators, wind and traction | Lower loss, higher cost per kilogram, more strokes, less forgiving coating |
| 0.20-0.27 mm | High-frequency grades | High-speed alternators and designs where frequency dominates loss | Lowest loss at frequency, but brittle edges and premium tooling |
Stepping up one grade, from M470-50A to M400-50A, removes roughly 15 percent of that lamination's specific loss. It is easy to give that gain back through a worse stacking factor, heavier coating or poorer burr control, which is why material and process should be quoted together. The material side is covered in this guide to choosing the right electrical steel for a generator stator core.
Laminations do not hold themselves in place. In large generator cores the stack sits in dovetail slots on key bars inside the frame, compressed by pressure fingers at both ends, with through-bolts or clamping plates maintaining end pressure. Smaller cores use interlocking, welding or bonding instead.
Two numbers show whether stacking was done properly: stacking factor, normally guaranteed between 0.94 and 0.98 for coated electrical steel, and the flatness and squareness of the finished stack. Below 0.94 you are buying air where the design assumed iron.
Cores that travel by road or sea are re-checked and re-torqued after the winding is fitted, because clamping has to survive transport as well as operation.
Heat leaves a stator core by conduction along the laminations into the frame, convection into the air gap, and forced flow through the core itself. Construction decides how much of each route you get.
In large machines the stack is divided into packs by radial ventilation ducts formed from duct spacers, commonly every 40 to 60 mm of stack length, with axial channels through the yoke where the cooling circuit needs them.
The end region is the difficult part. Stray flux from the end winding enters the core ends and the clamping structure, so end packs are stepped, built from progressively shorter laminations to reduce loss density. Get that detail wrong and the hottest point in the machine sits where no thermocouple was installed.
In most generator cores, the limiting temperature is not the average temperature of the stack. It is the tooth tip beside the air gap and the end lamination behind the pressure finger.
Construction quality can be verified in the plant, before the winding is fitted. These are the checks that separate a core matching its calculation from one matching only its drawing.
| Check | Typical method | What it protects against |
|---|---|---|
| Stacking factor | Lamination count and stack height measurement | Lost slot area, higher flux density, reduced output |
| Bore diameter and roundness | Bore micrometer or laser measurement at several axial positions | Uneven air gap, unbalanced magnetic pull, vibration |
| Slot dimensions and pitch | Slot gauges and template checks across several teeth | Winding fit, thermal contact and stator stiffness |
| Interlaminar resistance | Franklin or surge test at a defined voltage | Shorts between laminations and local eddy current heating |
| Core loss | Ring flux or magnetising test on the assembled core | No-load loss higher than the design promised |
| Clamping tightness | Bolt torque check, tap test, vibration measurement | Loosening in service, fretting and progressive core damage |
Coating batch, punch tool life, stack pressure, curing cycle and ring test data together describe a process. A certificate on its own describes a document.
The same rules lead to different answers depending on the machine.
Slot geometry, duct pitch and stack pressure are cheap to adjust on a drawing and expensive to adjust in a die.
When a generator core goes out for quotation, the specifications that move the price most are rarely the ones in the drawing title block. Grade and thickness, stack length and stacking factor, clamping method, duct layout, bore and slot tolerances, and the expected duty cycle decide whether two quotes for the same core differ by twenty percent.
The practical sequence: fix the loss budget before the slot area, confirm the stacking factor you are buying, witness the core test, and keep the inspection records with the machine.
If the generator is still being designed, bring the core supplier in early. It is easier to agree a stacking factor, a duct pitch and a bore tolerance while the frame is a drawing than once the first prototype core is on the test bed. You can talk to our core engineers at any stage.