Transformer Silicon Steel Magnetic Property Testing: Standards, Methods, and Key Parameters

Transformer silicon steel magnetic property testing measures the core loss, magnetic polarization, permeability, and apparent power of electrical steel used in transformer cores, verifying conformity with international standards such as IEC 60404, ASTM A343, EN 10107, and GB/T 2521. Accurate magnetic characterization enables transformer manufacturers to control no-load losses, confirm material grades, and comply with increasingly strict energy-efficiency regulations. This article explains the tested sample types, key parameters, measurement methods, and typical applications of magnetic property testing for transformer silicon steel.

Technical Overview

Transformer silicon steel, widely known as electrical steel, is a soft magnetic iron-silicon alloy containing approximately 1% to 4.5% silicon by weight, produced as thin laminations to suppress eddy-current losses. It is supplied in two principal forms: grain-oriented (GO) electrical steel, in which a controlled Goss crystallographic texture aligns the easy axes of magnetization with the rolling direction to achieve very low core loss and high permeability, and non-oriented (NO) electrical steel, which exhibits nearly isotropic magnetic behavior and is used mainly in rotating machines and small transformers.

Because the magnetic performance of the steel directly determines the no-load loss, excitation power, and acoustic noise of a transformer, magnetic property testing is essential at every stage of the material and product life cycle. Testing is typically performed by accredited third-party laboratories operating under ISO/IEC 17025 quality management systems, using standardized methods defined by IEC, ASTM, EN, GB, and JIS standards organizations. The principal purposes of testing include:

• Verification of delivery guarantees and grade conformity for grain-oriented and non-oriented electrical steel
• Provision of reliable input data for transformer core design, no-load loss calculation, and noise prediction
• Compliance demonstration against energy-efficiency regulations for transformers
• Support for research and development of thinner gauges, high-permeability grades, and improved insulation coatings

Testing Samples

Grain-Oriented Electrical Steel Strip

Grain-oriented silicon steel is supplied in thicknesses of typically 0.18 mm to 0.35 mm, with the best magnetic properties along the rolling direction. Epstein frame specimens are cut parallel to the rolling direction as strips of 305 mm length and 30 mm width, while Single Sheet Tester specimens are square sheets of 500 mm by 500 mm. Conventional grades achieve magnetic polarization values of about 1.88 T to 1.92 T at 800 A/m, whereas high-permeability (HiB) grades reach approximately 1.92 T to 1.97 T at the same field strength.

Non-Oriented Electrical Steel Sheet and Strip

Non-oriented silicon steel is commonly supplied in thicknesses of 0.35 mm to 0.65 mm. Because its properties differ slightly between the rolling and transverse directions, Epstein specimens are prepared with half of the strips cut parallel and half perpendicular to the rolling direction, so that the measured result represents the average in-plane magnetic behavior.

Ring (Toroidal) Specimens

Ring specimens form a fully closed magnetic circuit without air gaps and are used for DC hysteresis characterization and for AC measurements at elevated frequencies. The rings are wound directly with primary magnetizing and secondary measuring windings, and stacked rings may be used to increase the effective cross-section.

Specimen Preparation and Conditioning

To obtain reference-quality results, specimens are usually stress-relief annealed in a protective atmosphere to remove the shearing and punching stresses introduced during cutting, since mechanical stress measurably degrades permeability and increases core loss. Before each measurement the specimen must be demagnetized by a gradually decreasing alternating field, and the specimen mass, dimensions, and density (determined in accordance with IEC 60404-13) must be recorded accurately for the loss calculations.

Testing Items

A comprehensive magnetic property testing program for transformer silicon steel covers the following parameters:

• Specific total loss (core loss), expressed in W/kg at defined induction and frequency, for example P1.5/50 for non-oriented grades and P1.7/50 for grain-oriented grades, together with their 60 Hz equivalents
• Magnetic polarization at specified magnetic field strengths, such as J800 (commonly written B8) for grain-oriented steel and J5000 (B50) or J10000 (B100) for non-oriented steel
• Specific apparent power S in VA/kg at defined induction levels, which determines the magnetizing current requirement of the core
• Amplitude permeability under alternating magnetization and relative permeability under DC conditions
• DC hysteresis parameters, including coercivity Hc, remanence Br, and maximum permeability
• Magnetostriction coefficient for noise-critical transformer applications, measured in accordance with IEC 60404-17
• Supporting material properties: lamination (stacking) factor per IEC 60404-13 or ASTM A719, and surface insulation resistance per ASTM A717

Testing Methods

Epstein Frame Method

The Epstein frame method specified in IEC 60404-2, ASTM A343/A343M, ASTM A804/A804M, GB/T 3655, and JIS C 2550 is the internationally recognized reference method. Sixteen strips of 30 mm width, or a higher multiple of four, are assembled into a 25 cm square frame with double-lapped corner joints. The frame carries primary magnetizing and secondary measuring windings, and an air-flux compensation mutual inductor corrects for air flux. A feedback-controlled amplifier maintains a sinusoidal secondary voltage with a form factor of 1.111 within tight tolerance, and the specific total loss is calculated using the standard effective magnetic path length of 0.94 m, which corrects for the non-uniform magnetization at the lapped corners.

Single Sheet Tester Method

The Single Sheet Tester (SST) method described in IEC 60404-3, ASTM A912, GB/T 13789, and JIS C 2550-2 measures a single square sheet of 500 mm by 500 mm placed between two laminated yokes. Preparation and testing are considerably faster than the Epstein method, which makes the SST the preferred tool for routine industrial acceptance testing. Because the magnetic path contains yoke sections outside the specimen, the instrument must be calibrated against Epstein reference measurements, and calibration coefficients are applied to achieve agreement with the reference method.

Ring Specimen Method

Ring specimen testing is performed in accordance with IEC 60404-4 for DC magnetic properties and IEC 60404-6 for AC properties at frequencies up to 10 kHz. The closed toroidal geometry eliminates corner effects and air gaps, giving highly accurate hysteresis loops from which coercivity, remanence, and permeability are derived by fluxmetric integration.

Measurement Circuit and Waveform Control

All methods are based on the wattmeter-ammeter-voltmeter principle, today implemented with digital sampling power analyzers. The peak magnetic polarization is derived from the mean rectified secondary voltage, while the magnetic field strength is calculated from the magnetizing current. Before measurement, the specimen is demagnetized, and measurements proceed sequentially from the highest to the lowest induction level. Corrections are applied for instrument and winding losses to ensure traceable results with typical reproducibility of approximately 1% to 2% for core loss at reference conditions.

Applications

Magnetic property testing of transformer silicon steel supports the electrical industry in the following areas:

• Incoming material inspection and supplier qualification for transformer and core manufacturers, ensuring that delivered coils meet the guaranteed loss and polarization values
• Grade conformity verification for international trade, customs clearance, and resolution of quality disputes between buyers and suppliers
• Transformer design input, including no-load loss guarantees, excitation power sizing, and acoustic noise estimation, guiding the selection of HiB, domain-refined, or thinner-gauge grades
• Compliance with transformer energy-efficiency regulations, such as the EU Ecodesign Regulation (EU) No 548/2014, the US DOE requirements in 10 CFR Part 431, and GB 20052 in China
• Research and development of new steel grades, insulation coatings, laser domain-refinement processes, and stress-relief annealing treatments
• Failure analysis of degraded, overheated, or mechanically stressed cores in service

Accredited third-party laboratories provide test reports with traceable calibrations, which are widely accepted by utilities, manufacturers, and regulatory bodies worldwide.

FAQ

What does a designation such as P1.7/50 mean?

It denotes the specific total loss, in watts per kilogram, of the material measured at a peak magnetic polarization of 1.7 T and a frequency of 50 Hz. It is the principal grading parameter for grain-oriented silicon steel, with modern grades ranging from roughly 0.85 W/kg for thin high-permeability material to about 1.35 W/kg for conventional 0.35 mm strip.

Which method is more authoritative, the Epstein frame or the Single Sheet Tester?

The Epstein frame method is the international reference method against which all other methods are calibrated. The Single Sheet Tester is faster and better suited to routine production control, but its accuracy depends on calibration against Epstein reference specimens.

Why must the magnetic flux waveform be sinusoidal during testing?

Core loss depends strongly on the waveform shape. Harmonic distortion of the flux waveform changes the measured loss significantly, so standards require the secondary voltage form factor to be maintained at 1.111 within a tolerance of about 1% using waveform feedback control.

Do specimens need stress-relief annealing before testing?

For reference-grade measurements, yes. Cutting and punching introduce internal stresses that degrade magnetic performance, so Epstein specimens are normally stress-relief annealed in a protective atmosphere to restore the intrinsic properties of the material.

Can testing be performed at 60 Hz or at higher frequencies?

Yes. The standard methods cover 50 Hz and 60 Hz power frequencies, and IEC 60404-6 extends ring and single-sheet techniques to frequencies up to 10 kHz for applications such as high-speed rotating machines and special transformers.

How are the results used in transformer engineering?

Measured values of specific total loss, polarization, and apparent power are substituted directly into core design calculations to predict no-load loss, magnetizing current, and noise, allowing manufacturers to guarantee performance and meet minimum energy-efficiency requirements.