Eddy Current Loss Interactive Calculator

Classical Eddy Current Loss Formula

P_e = k_e · B_max² · f² · t² · m

Where:
P_e = Eddy current power loss (W)
k_e = Eddy current coefficient (material-dependent, dimensionless or with appropriate units)
B_max = Peak magnetic flux density (T - Tesla)
f = Frequency of magnetic field variation (Hz)
t = Lamination thickness (m)
m = Mass of core material (kg)

Steinmetz Total Core Loss Equation

P_total = k_h · f · B_max^α · m + k_ec · f² · B_max^β · m

Where:
P_total = Total core loss (W)
k_h = Hysteresis loss coefficient (material-specific)
k_ec = Eddy current loss coefficient (material-specific)
α = Steinmetz exponent for hysteresis (typically 1.6-2.2)
β = Steinmetz exponent for eddy currents (typically 2.0)
f = Frequency (Hz)
B_max = Peak flux density (T)
m = Core mass (kg)

Volumetric Loss Density

p_v = P_e / V

Where:
p_v = Volumetric loss density (W/m³)
P_e = Total eddy current loss (W)
V = Volume of magnetic core (m³)

Specific Loss (Per Unit Mass)

p_m = P_e / m

Where:
p_m = Specific power loss (W/kg)
P_e = Eddy current power loss (W)
m = Core mass (kg)

Scenario: Electric Vehicle Motor Designer

Akiko, a senior engineer at an electric vehicle manufacturer, is optimizing a 150 kW permanent magnet synchronous motor operating at variable frequencies from 100-600 Hz. Using the eddy current loss calculator with measured flux densities of 1.62 T and 0.27 mm silicon steel laminations, she determines losses range from 284 W at low speed to 10.2 kW at maximum RPM. This analysis reveals that eddy currents consume 6.8% of total power at highway speeds, prompting her to specify 0.20 mm laminations and adjust the cooling system to handle 14.3 kW total core losses. The thermal modeling enabled by accurate loss calculations ensures the motor stays below 155°C even during sustained high-speed operation on mountain grades.

Scenario: Utility Transformer Procurement Specialist

James works for a municipal utility evaluating bids for 47 new 2.5 MVA distribution transformers. Manufacturers provide conflicting loss data, so he uses the Steinmetz calculation mode with vendor-supplied material coefficients to independently verify no-load losses. For one bid claiming 2.8 kW core loss, his calculation with stated M-3 steel properties at 1.53 T and 60 Hz yields 3.47 kW—a 24% discrepancy indicating either measurement error or optimistic assumptions. He requests third-party tested loss data, ultimately saving the utility an estimated $187,000 in excess energy costs over the transformers' 30-year service life. This due diligence, enabled by understanding the fundamental physics of eddy current losses, justified his analysis to the procurement committee.

Scenario: Industrial Induction Heating System Designer

Maria designs induction heating coils for a steel mill's billet heating furnace operating at 3.2 kHz. She uses the volumetric loss density calculator to determine that with the 47 kg ferrite core and measured flux density of 0.38 T, eddy current losses reach 1,750 W in the coil assembly itself—separate from intentional heating in the steel workpiece. This parasitic loss requires dedicated cooling, so she specifies a water-cooled aluminum housing with 0.95 L/min flow to maintain core temperature below 120°C. By calculating the 37,234 W/m³ volumetric loss density, she properly sizes the cooling channels and predicts the 6.2°C temperature rise across the heat exchanger, ensuring reliable continuous operation in the demanding foundry environment where equipment failures cost $18,000 per hour in lost production.

▼ Why do eddy current losses increase with the square of frequency?

The f² relationship arises from two separate frequency dependencies. First, Faraday's law dictates that induced voltage (and thus current) is proportional to the rate of flux change dB/dt, which scales linearly with frequency for sinusoidal excitation. Second, the power dissipated follows P = I²R, introducing another factor of frequency from the current term. Mathematically, if B = B_maxsin(ωt), then dB/dt ∝ ω = 2πf, the induced current I ∝ f, and power P ∝ I² ∝ f². This quadratic scaling makes eddy currents the dominant loss mechanism in high-frequency applications—doubling frequency from 60 Hz to 120 Hz quadruples eddy losses while hysteresis losses only double. For variable-frequency drives operating motors from 0-400 Hz, eddy current losses at maximum speed can exceed low-speed values by factors exceeding 40, fundamentally altering thermal design requirements and necessitating active cooling systems that would be unnecessary at fixed 60 Hz operation.

▼ What determines the eddy current coefficient ke for a specific material?

The eddy current coefficient encapsulates electrical resistivity (ρ), material density, and geometric/dimensional constants that convert the fundamental physics into a practical engineering parameter. For laminated cores, k_e is typically determined empirically by manufacturers through standardized testing per ASTM A677 or IEC 60404 standards, measuring total losses at specified frequencies and flux densities, then curve-fitting to separate hysteresis and eddy components. The coefficient inversely relates to resistivity—materials with higher resistivity exhibit lower k_e values and reduced losses. Silicon content in electrical steel directly increases resistivity: pure iron has ρ ≈ 10 μΩ·cm, while 3% silicon steel achieves ρ ≈ 45 μΩ·cm, reducing k_e by approximately 78%. Temperature significantly affects k_e through resistivity changes; most electrical steels exhibit 0.3-0.5% resistivity increase per °C, meaning a core operating at 80°C versus 20°C experiences roughly 8-12% lower eddy losses. Manufacturers typically specify coefficients at 20°C or 25°C, requiring temperature correction for accurate thermal modeling.

▼ How does lamination thickness affect mechanical and magnetic properties beyond loss reduction?

Thinner laminations introduce several tradeoffs beyond the obvious eddy current reduction. The stacking factor—ratio of active steel to total core volume—decreases as insulation coating represents a larger percentage of total thickness; 0.65 mm laminations achieve ~0.97 stacking factor while 0.18 mm laminations typically reach only ~0.92, requiring 5.4% more core volume for equivalent flux capacity. Mechanical rigidity decreases with thinner laminations, necessitating more substantial clamping structures and potentially increasing core assembly costs by 15-30%. Manufacturing constraints limit practical minimum thickness; cold-rolled silicon steel below 0.15 mm becomes difficult to handle without damage, and punching operations cause proportionally more edge burrs and increased interlaminar shorts. The insulation coating itself (typically C-5 phosphate or organic varnish) must remain defect-free; a single short between laminations creates a low-resistance path allowing eddy currents across multiple sheets, potentially increasing local losses by 300-800%. Quality control becomes progressively more critical as lamination count increases—a 100 kg core uses 476 laminations of 0.65 mm steel but 1,587 laminations of 0.18 mm steel, multiplying opportunities for handling damage or coating defects.

▼ When should I use the Steinmetz equation instead of the classical eddy current formula?

The classical formula isolates only macroscopic eddy current losses, assuming uniform flux penetration and ignoring hysteresis and anomalous loss mechanisms. Use it for theoretical analysis, understanding fundamental scaling relationships, or comparing relative effects of design changes (thickness, frequency) where absolute accuracy is secondary. The Steinmetz equation provides total core losses including hysteresis, classical eddy currents, and anomalous excess losses arising from domain wall dynamics and microstructural discontinuities. Use Steinmetz when you need actual power dissipation for thermal design, efficiency calculations, or performance specifications. The equation requires empirically-determined coefficients from manufacturers (often tabulated at standard test conditions like 1.5 T and 60 Hz), and accuracy depends on operating near these characterization points. For non-sinusoidal waveforms (PWM, pulsed DC), neither approach is fully accurate without harmonic decomposition or modified coefficients; finite element analysis with complex permeability models becomes necessary for precision work. As a practical guideline, classical formula errors typically range 30-60% low for total loss prediction but accurately capture thickness/frequency scaling trends, while Steinmetz achieves ±10% accuracy when using manufacturer coefficients at similar operating conditions.

▼ What cooling methods are most effective for high eddy current loss scenarios?

Cooling strategy depends critically on volumetric loss density (W/m³) and permissible temperature rise. Natural convection suffices for p_v below 15,000 W/m³ with adequate surface area (minimum 0.15 m²/kW dissipated); this covers most 50/60 Hz power transformers using ONAN (oil natural, air natural) cooling where oil circulation by thermal convection transfers heat to radiators. Forced air cooling becomes necessary from 15,000-80,000 W/m³; high-efficiency motors and small transformers use ONAF (oil natural, air forced) or direct fan cooling achieving 8-15°C temperature rise per kW with 200-400 CFM airflow per kilowatt of core loss. Liquid cooling is essential above 80,000 W/m³—typical in high-frequency inductors, induction heating equipment, and electric vehicle motor cores where concentrated losses demand water-cooled housings or direct oil spray cooling (ODAF - oil directed, air forced). The thermal resistance from core center to coolant typically ranges 0.08-0.15 °C·m²/W for laminated structures, meaning a 50 kW/m³ volumetric loss produces 4-7.5°C temperature differential through the core itself before considering interface and coolant resistances. Effective cooling requires addressing all thermal path components: core-to-housing interface (thermal grease or potting compound eliminates air gaps), housing-to-ambient (fins increase surface area 3-8×), and coolant-side heat transfer (turbulent flow maintains low film coefficients). Engineers frequently discover that inadequate core clamping pressure creates air gaps with 50-100× higher thermal resistance than solid contact, causing hotspots despite apparently adequate cooling system capacity.

▼ How do I account for eddy current losses in non-sinusoidal or PWM applications?

Non-sinusoidal waveforms require harmonic decomposition using Fourier analysis to represent the actual flux waveform as a sum of sinusoidal components at fundamental frequency and harmonics. Apply the eddy current loss formula to each harmonic individually (using its specific frequency and amplitude), then sum the results—losses are additive since power dissipation is an average quantity. For a typical PWM motor drive with 10 kHz switching frequency, significant harmonic content exists at odd multiples of the fundamental up to and beyond the switching frequency. A 180 Hz fundamental (3600 RPM motor) might have 3rd harmonic (540 Hz) at 15% amplitude, 5th (900 Hz) at 8%, and switching frequency harmonics at 2-5%. Since loss scales with f² and B², even small high-frequency components contribute disproportionately: the 10 kHz harmonic at just 3% flux amplitude contributes loss equal to the fundamental at 100% amplitude because (10,000/180)² × (0.03)² ≈ 92.6 versus 180² × 1² = 32,400, normalized gives approximately equivalent contribution. This explains why motors on PWM drives typically exhibit 40-85% higher core losses than sinusoidal operation at the same RMS voltage. More sophisticated approaches use the Bertotti three-component model separating hysteresis, classical eddy, and excess losses with frequency-dependent coefficients, or employ time-stepping finite element analysis with rate-dependent B-H curves. For practical engineering, manufacturers increasingly provide "inverter-duty" loss coefficients characterized under representative PWM waveforms, offering acceptable accuracy without extensive analysis.

Free Engineering Calculators

Explore our complete library of free engineering and physics calculators.

🔗 Explore More Free Engineering Calculators

• Magnetic Torque Current Loop Interactive Calculator
• Phase Angle AC Interactive Calculator
• Electromagnetic Induction General Interactive Calculator

Browse All Engineering Calculators →

About the Author

Robbie Dickson — Chief Engineer & Founder, FIRGELLI Automations

Robbie Dickson brings over two decades of engineering expertise to FIRGELLI Automations. With a distinguished career at Rolls-Royce, BMW, and Ford, he has deep expertise in mechanical systems, actuator technology, and precision engineering.

Wikipedia · Full Bio