B-H curve & hysteresis
Magnetic basics, virgin curve and saturation, the hysteresis loop with B_r / H_c / B_sat, soft vs hard materials, Steinmetz hysteresis-loss model P_h = K_h·f·B_max^n.
Step 1 — Magnetic basics: B = μH, μ = μ_0·μ_r
Reference notes
Magnetic materials respond nonlinearly to applied field, with both saturation and history-dependent hysteresis. Use Next → to walk through the virgin magnetization curve, the hysteresis loop, the parameters that define it (Br, Hc, Bsat), the soft-vs-hard material distinction, and the Steinmetz model for hysteresis loss in transformer and motor cores.
Magnetic basics
- B (flux density) — measured in tesla (T) or Wb/m². The "B-field" that produces force and induces voltage via Faraday's law.
- H (field strength) — measured in A/m. The "applied field" set by NI through a magnetic circuit.
- Permeability μ = μ_0 · μ_r. μ0 = 4π × 10⁻⁷ H/m (defined). μr is the dimensionless relative permeability.
- μr = 1 for vacuum / air.
- μr ≈ 1000 – 5000 for low-grade iron.
- μr ≈ 10 000 – 50 000 for high-grade transformer silicon-steel.
- μr > 100 000 for some permalloys.
Virgin magnetization curve & saturation
Starting from a fully demagnetized state, B rises with H along the virgin curve:
- At low H, B rises nearly linearly. Magnetic domains in the iron progressively align with the applied field via domain-wall motion.
- The slope dB/dH flattens as more domains align.
- Eventually almost all domains are aligned with H — B saturates at Bsat. Adding more H produces only tiny additional B.
| Material | Bsat | Use |
|---|---|---|
| Cold-rolled grain-oriented Si-steel (CRGO) | 2.0–2.2 T | Transformer cores |
| Cold-rolled non-oriented Si-steel (CRNGO) | 1.8–2.0 T | Motor laminations |
| Amorphous metal (Metglas) | 1.5–1.6 T | Premium low-loss distribution transformers |
| Cobalt-iron alloy | 2.3–2.4 T | Aerospace motor pole pieces |
| Ferrite (soft) | 0.4–0.5 T | Switch-mode supply cores |
Hysteresis loop
Reduce H back toward zero. Surprise: B does NOT trace back along the virgin curve. Domain-wall motion is partly irreversible — domains get pinned at lattice defects, impurities, grain boundaries — so the material "remembers" having been magnetized. The result is a closed loop on the B-H plane traversed each cycle of an AC excitation.
Three key loop parameters
- Bsat — peak flux density achievable.
- Br (remanence) — B retained at H = 0 after reaching saturation. Tells you how strong a residual magnetization the material holds.
- Hc (coercivity) — magnitude of negative H required to force B back to zero. Tells you how difficult the material is to demagnetize.
The area enclosed by the loop has units of joules per cubic meter — energy dissipated as heat in the iron per cycle of magnetic excitation.
Soft vs hard magnetic materials
- Soft magnetic materials — small loop, low Hc, low loss per cycle. Designed for AC use where the loop is traversed many times per second. Examples: silicon-steel (CRGO, CRNGO), permalloy, soft ferrite, amorphous metal. Used in transformers, motor stators, inductors.
- Hard magnetic materials — large loop, high Hc, high Br. Designed to RETAIN magnetization → permanent magnets. Examples: AlNiCo (high Br, modest Hc), ceramic hard ferrite (cheap, ubiquitous), samarium-cobalt (high temperature), neodymium-iron-boron (strongest, Br ≈ 1.3–1.5 T, Hc ≈ 1000 kA/m).
For a hard magnet, the specification of interest is the demagnetization curve — the second quadrant of the hysteresis loop where the magnet operates against an air-gap reluctance load.
Steinmetz hysteresis loss model
Energy lost per cycle = area enclosed by the loop, times core volume. For a sinusoidal excitation at peak Bmax and frequency f:
Three engineering implications:
- Linear in frequency. Doubling f doubles Ph at fixed Bmax. 400 Hz aerospace transformers run hot.
- Highly nonlinear in Bmax. Operating at 1.8 T instead of 1.6 T (12.5 % more flux) gives ~25 % more hysteresis loss for n ≈ 1.8. The penalty for pushing B near saturation is steep.
- Linearity allows loss separation. Total iron loss = hysteresis (∝ f) + eddy-current (∝ f²). Measuring Piron at multiple f at fixed Bmax and fitting P = a·f + b·f² separates the two components.
Eddy-current loss — the other half of core loss
Time-varying B induces voltage loops in the conductive iron, dissipating I²R as heat. Together with hysteresis, this constitutes the total core loss of a transformer or motor:
where t = lamination thickness, σ = electrical conductivity. Mitigated by (a) thin insulated laminations 0.23–0.50 mm typical, (b) high-resistivity silicon-iron alloys (3-4 % Si raises ρ by 4× over pure Fe), (c) amorphous metal cores for further improvement.
Saturation in transformer operation
Transformer designers operate Bmax below Bsat by a comfortable margin (typically 1.6–1.8 T peak vs 2.0 T saturation for CRGO). Why? Once into saturation:
- μr collapses → magnetizing current spikes 5–10× normal (inrush phenomenon).
- Current waveform becomes peaky / non-sinusoidal → harmonic injection.
- Hysteresis loss climbs steeply (Steinmetz n ≈ 1.7).
- Magnetostriction noise rises audibly.
- Local hot-spots from flux concentration.
This is why a transformer fed by a sudden 110 %-of-rated voltage event can draw enormous inrush current.
Keyboard shortcuts
- The B-H panel animates H sinusoidally; the loop area is shaded in step 6 to show the per-cycle energy loss.