Laminated Core Silicon Steel Core: CRGO Grade Selection, Stacking and Annealing
Sep 18, 2025
Leave a message
What a Laminated Silicon Steel Core Is
A laminated silicon steel core is a magnetic circuit built by stacking punched or sheared electrical steel laminations, insulated from one another, to form a three-dimensional core. The laminations are made from cold-rolled grain-oriented silicon steel with low iron loss and high permeability, normally in thicknesses of 0.23 mm, 0.27 mm and 0.30 mm. Stacking rather than casting a solid core breaks up the eddy current paths, which is the whole reason the magnetic circuit is laminated.
Material Selection for Transformer Cores
Grain-oriented strip is the standard choice because the flux path in a transformer core is fixed and aligned with the rolling direction. Typical commercial grades are named by thickness and loss, for example B23P090 for 0.23 mm high-induction strip with a maximum core loss of 0.90 W/kg at 1.7 T and 50 Hz. Grades with laser domain refinement, such as B27RK085 at 0.85 W/kg, are chosen when the no-load loss target is tight, since no-load loss runs continuously for the life of the transformer. Fully processed grain-oriented strip is ordered to ASTM A876/A876M or EN 10107.
Typical Grade Data for Core Building
| Type | Grade | Width | Max core loss at 1.7 T / 50 Hz | Lamination factor | Min induction at 800 A/m |
|---|---|---|---|---|---|
| Common type | 23Q110 | 35 - 1,200 mm | 1.10 W/kg | 97.2 % | 1.82 T |
| Common type | 23Q120 | 35 - 1,200 mm | 1.20 W/kg | 97.2 % | 1.82 T |
| Common type | 27Q120 | 35 - 1,200 mm | 1.20 W/kg | 97.5 % | 1.82 T |
| High induction | B23P085 | 35 - 1,200 mm | 0.85 W/kg | 97.2 % | 1.88 T |
| High induction | B23P100 | 35 - 1,200 mm | 1.00 W/kg | 97.2 % | 1.88 T |
| High induction | B27P100 | 35 - 1,200 mm | 1.00 W/kg | 97.5 % | 1.89 T |
| High induction | B30P120 | 35 - 1,200 mm | 1.20 W/kg | 98.0 % | 1.89 T |
| Domain refined, laser | B23R085 | 35 - 1,200 mm | 0.85 W/kg | 97.2 % | 1.88 T |
| Domain refined, laser | B27RK085 | 35 - 1,200 mm | 0.85 W/kg | 97.5 % | 1.89 T |
Core Geometry and Joint Design
Laminations are sheared to shape and stacked as a single-phase three-column core or a three-phase five-column core, depending on the transformer arrangement. The joints are stacked in the full mitre pattern, which closes the magnetic circuit at the corners so that the flux crosses the joint at 45 degrees rather than at a right angle. A right-angle joint would force the flux across the grain at the corner and raise loss and magnetising current sharply. Alternating the mitre from layer to layer, the step-lap principle, further reduces the air gap effect at the joint and is standard practice in cores built for low no-load loss.
Manufacturing Route and Dimensional Control
The route is raw material sourcing, slitting to the required width, punching or shearing to lamination shape, stacking, core forming and final testing. High-precision slitting and cross-cutting lines hold the strip width and the cut length within tight limits, because a lamination cut slightly oversize or out of square will not close the mitre properly and will leave a gap that shows up as increased no-load loss. Burr height is controlled for the same reason, since burrs create short circuits between laminations and destroy the insulation effect that the lamination depends on.
Annealing and Loss Reduction
Shearing and punching strain the material near the cut edge, and that strain raises hysteresis loss. Where the grade and the coating permit, finished cores are stress-relief annealed to remove the strain and to restore the magnetic properties of the cut edge zone. Annealing also relaxes the mechanical stress introduced by clamping. The gain is largest in small cores, where the cut perimeter is a high proportion of the total core area, and it is one reason a properly annealed core can measure significantly lower no-load loss than the same core assembled without annealing.
Where Laminated Cores Are Used
Laminated silicon steel cores are used in oil-immersed distribution transformers, resin-insulated dry-type transformers, on-load tap-changing transformers, mine and earthing transformers, and special-purpose units. Dry-type designs with cast-resin windings and oil-immersed designs of the sealed-tank type both rely on the same core construction, with the choice of grade set by the permitted no-load loss and by the noise requirement of the installation. Small cores for instrument transformers and reactors follow the same route at a reduced scale.
Frequently Asked Questions
Q: Why is the core laminated instead of solid?
A: Lamination with inter-lamination insulation interrupts eddy current paths. A solid core would circulate large eddy currents and overheat.
Q: Which steel thickness should be used for a transformer core?
A: 0.23 mm, 0.27 mm or 0.30 mm grain-oriented strip is normal. Thinner strip lowers eddy current loss, which matters for low no-load loss targets and for higher frequencies.
Q: What is a full mitre core joint?
A: The corner joint is cut at 45 degrees so the flux turns with the grain. A square corner would force the flux across the grain and sharply increase loss.
Q: Is annealing necessary after punching?
A: It depends on the loss budget and on the coating. Stress-relief annealing removes punching strain and lowers hysteresis loss, and the benefit is largest for small cores.
Q: What does a lamination factor of 97.5 % mean?
A: 97.5 % of the stack volume is steel and the remainder is coating and air. A higher factor means a smaller core can carry the same flux.
Q: How much does core loss contribute to transformer losses?
A: No-load loss, which is almost entirely core loss, runs continuously. A grade such as B27RK085 at 0.85 W/kg instead of an ordinary grade at 1.20 W/kg cuts that component by roughly 30 %.

