What Silicon Steel Is
Silicon steel, also called electrical steel, is a low-carbon iron-silicon soft magnetic alloy and is the highest-volume functional metal material in production, accounting for roughly 1% of world steel output. It is produced with a silicon content of 0.8-4.8% and rolled hot and cold to sheet and strip thinner than 1 mm. Alloying with silicon raises the resistivity and the maximum permeability of iron and reduces coercivity, core loss and magnetic ageing, which is why the material is used for the cores of motors, generators and transformers.
Silicon steel is normally classified in two ways: by silicon content, into low-silicon grades below 2.8% and high-silicon grades from 2.8% to 4.8%, and by production route, into hot-rolled and cold-rolled product. Cold-rolled strip has uniform thickness, better surface quality and higher magnetic performance, and has largely displaced hot-rolled product. Cold-rolled material is itself divided into two families, and that division is the subject of this article: non-oriented and grain-oriented.
The Structural Difference: Random Grains Versus a Controlled Texture
Non-oriented silicon steel is produced so that the crystal orientation of the grains is essentially random through the thickness of the sheet. The magnetic properties are therefore similar in all directions in the plane of the strip, which suits rotating machines where the flux path continually changes direction.
Grain-oriented silicon steel is produced with a strong recrystallisation texture in which the easy direction of magnetisation of the grains lies along the rolling direction. Magnetisation along the rolling direction therefore requires far less field than along any other direction; the permeability perpendicular to the rolling direction is much lower, and the material is anisotropic by design rather than isotropic. That is why grain-oriented strip must be sheared and stacked so that the flux path follows the rolling direction, and why an incorrectly laid-out lamination can lose most of the advantage the grade was bought for.
Grade Designations and Thickness Range
Grade designations encode thickness, product type and guaranteed maximum core loss, which makes them a fast way to compare catalogue entries.
| Family | Typical designation | Meaning | Typical thickness |
| Non-oriented | 50W470 | 0.50 mm thick, non-oriented, maximum core loss 4.70 W/kg at 1.5 T and 50 Hz | 0.35, 0.50, 0.65 mm |
| Non-oriented, high grade | 35W270 | 0.35 mm thick, non-oriented, maximum core loss 2.70 W/kg at 1.5 T and 50 Hz | 0.35 mm |
| Grain-oriented, conventional | 30Q120 | 0.30 mm thick, grain-oriented, maximum core loss 1.20 W/kg at 1.7 T and 50 Hz | 0.27, 0.30, 0.35 mm |
| Grain-oriented, high permeability | 27QG100 | 0.27 mm thick, high-permeability grain-oriented, core loss index 1.00 | 0.20, 0.23, 0.27 mm |
The leading number is the nominal thickness in millimetres multiplied by 100, the letter identifies the product family, and the trailing number is the guaranteed maximum core loss in watts per kilogram multiplied by 100. The reference induction is 1.5 T for non-oriented grades and 1.7 T for grain-oriented grades, which reflects the higher flux density at which transformer cores normally operate. Products are supplied in the fully processed condition and are usually specified to an insulating coating type, since the coating provides the interlaminar resistance of the finished stack.
Comparison of the Two Families
| Feature | Non-oriented | Grain-oriented |
| Crystal texture | Random orientations | Strong rolling-direction texture |
| Anisotropy | Nearly isotropic in the plane | Strongly anisotropic |
| Silicon content | About 0.8-3.2% | Typically about 2.9-3.5% |
| Coating requirement | Insulating coating, frequently semi-processed grades | Insulating coating applied over a forsterite base layer |
| Typical core loss at 0.50 mm | 3.0-6.0 W/kg at 1.5 T, 50 Hz | Not normally produced at this thickness |
| Main applications | Motors, generators, welding transformers, small power transformers | Power transformers, distribution transformers, pulse transformers, magnetic amplifiers |
The important commercial consequence is that grain-oriented grades reach core losses roughly one third of those of comparably thick non-oriented grades when the flux runs along the rolling direction, but they are more costly and their advantage disappears in any application where the flux path is not aligned.
Manufacturing Route
Both families start from a hot-rolled or continuously cast feedstock of about 2.0-2.3 mm, which is pickled and then cold rolled. Non-oriented grades are cold rolled to final gauge in one or two stages and are then annealed, either fully processed at the works or supplied semi-processed so that the customer performs the final decarburising anneal after punching. Grain-oriented grades follow a longer route: two-stage cold rolling with an intermediate anneal, a decarburising anneal, application of a magnesium oxide coating, a high-temperature secondary recrystallisation anneal that develops the texture, and finally the application of an insulating coating. The additional rolling, annealing and coating steps explain most of the price difference between the two families.
Core Loss and Its Effect on Efficiency
Core loss has two components: hysteresis loss, related to the energy needed to move domain walls during each magnetisation cycle, and eddy-current loss, related to circulating currents induced in the laminations. Silicon additions reduce both by raising resistivity and lowering coercivity, and thinner strip reduces eddy-current loss further, which is why high-frequency applications move towards thinner material. Domain-refined high-permeability grades, produced by laser scribing or by mechanical micro-stress treatment, reduce hysteresis loss further by subdividing the magnetic domains. The cumulative effect of texture control, higher silicon content, thinner gauge and domain refinement is a transformer core that loses a fraction of the energy of an early hot-rolled core at the same flux density, which is the reason grain-oriented steel became the standard material for transformer cores.
Development of Grain-Oriented Grades
The physical basis of grain-oriented steel was established in the laboratory work of the 1920s, when measurements on single crystals of iron-silicon showed that magnetisation is easiest along the cube-edge direction of the crystal lattice. In 1934 a laboratory process combining cold rolling with high-temperature heat treatment produced the first sheet in which the grains were arranged in an orderly manner along the rolling direction, and a patent covering the process was filed in 1935. Industrial production of cold-rolled grain-oriented sheet followed in the mid-1930s and expanded through the 1940s. The technology spread to other producing countries in the 1950s, and continuous improvement after that period raised the magnetic performance of cold-rolled grain-oriented sheet to its present level. In the 1960s the first high-permeability grades appeared, in the 1970s and 1980s grain size control was refined, and surface laser irradiation and mechanical micro-stress techniques were introduced in the 1980s to reduce iron loss further by domain refinement.
Frequently Asked Questions
Q: What is the main difference between oriented and non-oriented silicon steel?
Grain-oriented steel has a controlled rolling-direction texture so that magnetisation is easy along the strip direction, while non-oriented steel has a random grain orientation and therefore nearly uniform magnetic properties in all directions.
Q: Which one is used in motors and which in transformers?
Non-oriented grades are used for the cores of motors and generators and for welding transformers; grain-oriented grades are used for power and distribution transformer cores, pulse transformers and magnetic amplifiers.
Q: How are silicon steel grades designated?
The designation carries thickness in millimetres multiplied by 100, a family letter, and the guaranteed maximum core loss in watts per kilogram multiplied by 100; for example 50W470 is a 0.50 mm non-oriented grade with a core loss of 4.70 W/kg.
Q: What silicon content is used?
Silicon steel contains 0.8-4.8% silicon. Low-silicon grades below 2.8% are used mainly for motors and high-silicon grades from 2.8% to 4.8% are used for transformer cores; grain-oriented strip is typically about 2.9-3.5% silicon.
Q: Why is grain-oriented steel more expensive?
Because it requires two-stage cold rolling, a decarburising anneal, a magnesium oxide coating and a high-temperature secondary recrystallisation anneal to develop the texture, none of which are needed for a non-oriented grade.
Q: What thicknesses are available?
Non-oriented strip is commonly supplied at 0.35, 0.50 and 0.65 mm, and grain-oriented strip at 0.20, 0.23, 0.27, 0.30 and 0.35 mm, with thinner gauges used where higher operating frequency makes eddy-current loss significant.





