Magnetic-Circuit Design and Electromagnetic Loading

Introduction

Electrical-machine design connects electromagnetic performance with geometry, materials, cooling, manufacturability, cost, and reliability. The magnetic circuit establishes the flux required for energy conversion, while the electric loading establishes the current-producing capability distributed around the air gap. Their interaction determines torque, voltage, losses, temperature rise, power factor, and principal dimensions.

A useful preliminary design must therefore answer three linked questions:

1. What magnetomotive force (mmf) is required to drive the specified flux through iron, air gaps, joints, and leakage paths?
2. What magnetic and electric loadings can be used without unacceptable saturation, loss, heating, demagnetization, or mechanical stress?
3. What active diameter and core length are required for the rated output?

This chapter develops magnetic-circuit models, treats saturation, leakage and fringing, derives the three-phase output equation, and applies the resulting principles to machine sizing.

Learning Objectives

After studying this chapter, a student should be able to:

• model series and parallel magnetic paths using flux, reluctance, and mmf;
• interpret nonlinear B–H data and identify saturation;
• estimate air-gap, leakage, and fringing effects;
• distinguish a section mmf drop from the total winding mmf;
• select preliminary values of average air-gap flux density, specific electric loading, and conductor current density;
• derive the output equation of a sinusoidal three-phase machine;
• determine principal dimensions and assess the resulting design trade-offs.

1. Magnetic-Circuit Foundations

1.1 Flux, flux density, and field strength

Magnetic flux Φ is measured in webers (Wb). If flux is approximately uniform and normal to a cross-sectional area A, the flux density is

where B is in teslas (T), Φ is in Wb, and A is in m². This expression is a section-average approximation; local tooth tips, corners, joints, and ventilation ducts can have substantially different flux densities.

Magnetic field strength H, measured in A/m, is related to flux density by

for a linear, isotropic material. Here μ₀ = 4π × 10^-7 H/m is the permeability of free space and μ_r is relative permeability. Ferromagnetic electrical steels are nonlinear, hysteretic materials, so their operating H must normally be obtained from the relevant B–H curve rather than from a constant value of μ_r.

1.2 Ampère’s law and mmf drops

For a closed contour enclosing a winding of N turns carrying current I, Ampère’s circuital law is

The product NI is the total applied winding mmf, measured in ampere-turns (At). Under a lumped magnetic-circuit approximation,

where H_j l_j is the mmf drop of section j. Thus, H_j l_j is not generally equal to NI for an individual section. It equals the total winding mmf only if that section constitutes the complete modeled magnetic path or all other drops are negligible.

Figure: A winding supplies the total mmf required by all sections of a series magnetic path.

1.3 Reluctance and permeance

For a uniform linear section of length l, area A, and permeability μ, its reluctance is

Reluctance ℜ has units At/Wb. The section mmf drop is

This relation resembles Ohm’s law, but the analogy has limits. Flux is not a material flow, magnetic permeability may be nonlinear, and leakage paths are distributed rather than perfectly confined.

For series sections carrying the same flux,

For ideal parallel branches subjected to the same mmf difference,

At a junction, flux continuity gives Φ_in = ∑Φ_out. Permeance is Λ = 1/ℜ, measured in Wb/At.

2. Nonlinearity, Saturation, and Core Construction

2.1 The B–H characteristic

At low and moderate flux densities, a small increase in H produces a relatively large increase in B. Near the knee of the B–H curve, incremental permeability declines. Beyond the knee, much larger mmf is required for a small increase in flux density. This is magnetic saturation.

Saturation is commonly most severe in stator teeth, rotor teeth, pole shoes, and narrow yoke sections. Its consequences include:

• increased magnetizing current and reduced power factor;
• waveform distortion and harmonic loss;
• reduced incremental inductance;
• local heating and acoustic noise;
• limited overload capability.

The correct design procedure divides the path into sections, calculates each section’s flux density, reads H from material data at that flux density, and sums the products H_jl_j. A single relative permeability should not be used over a path containing strongly saturated steel.

2.2 Net iron area and stacking factor

A laminated core does not contain solid iron over its gross cross-section because of insulation coatings, burrs, and manufacturing gaps. The net iron area is

where k_s is the stacking factor, A_g is gross laminated area, and A_i is net iron area. Typical stacking factors are approximately 0.90–0.97, depending on lamination thickness and manufacturing quality. Flux density in the steel must be calculated using net area.

2.3 Core losses

Alternating flux produces hysteresis and eddy-current losses. Manufacturers normally provide specific loss data in W/kg at specified frequency and peak flux density. Preliminary empirical representations may separate the loss as

where p_c is specific core loss, f is frequency, and the coefficients are fitted to a particular material and lamination thickness. This equation is not universal and should not replace supplier data, especially under rotational flux or harmonic excitation.

3. Air Gaps, Fringing, Slotting, and Leakage

3.1 Air-gap mmf

For a uniform air gap with negligible fringing,

where g is physical gap length. Because air has much lower permeability than iron, the air-gap mmf often dominates the total excitation.

Air-gap length involves a major compromise. A larger gap improves mechanical clearance, reduces sensitivity to eccentricity, and can reduce some saturation effects, but it requires greater magnetizing mmf. Induction motors also suffer reduced magnetizing reactance and power factor as the gap increases. Permanent-magnet machines may use an effective magnetic gap that includes physical clearance, magnet recoil permeability, sleeves, and slotting effects.

3.2 Fringing

At an air-gap boundary, flux spreads beyond the nominal pole-face area. This fringing increases effective area and therefore reduces average gap flux density for a given flux. For a simple isolated rectangular pole face of dimensions w × l, a rough estimate is

This approximation is only suitable for preliminary estimates when the gap is small relative to pole dimensions and neighboring poles do not strongly interact. Finite-element analysis or calibrated permeance methods are preferable for complex geometries.

3.3 Slotting and Carter’s coefficient

Slot openings make a slotted air gap magnetically less effective than a smooth gap. Their effect is often represented by an effective gap

where Carter’s coefficient k_c ≥ 1. If both surfaces are slotted, separate stator and rotor corrections may be combined using a documented Carter-factor convention. The corresponding average gap mmf is approximately

Carter-factor formulas depend on slot geometry and assumptions; values should not be transferred indiscriminately between open, semi-closed, and closed slots.

3.4 Leakage flux

Not all produced flux links the intended windings or crosses the useful air-gap region. Leakage includes slot leakage, tooth-tip leakage, end-winding leakage, zigzag leakage, pole leakage, and magnet end leakage.

A leakage coefficient may be defined as

so that λ ≥ 1. This convention must be stated because some references define leakage factors differently. Leakage flux affects reactance and mmf but does not contribute fully to useful induced emf or torque.

4. Worked Magnetic-Circuit Example

A laminated core has two series iron sections. Each section has net area 0.010 m² and carries flux 0.0132 Wb. Their mean lengths are 0.45 m and 0.30 m. The uniform air gap is 1.20 mm long and has the same effective area; neglect leakage, fringing, and slotting. At 1.32 T, the specified steel’s measured B–H data give H = 420 A/m. A 500-turn coil produces the excitation. Determine the required current.

Because both iron sections carry the same flux and have the same area,

The total iron mmf drop is

The air-gap mmf is

Hence the winding must provide

and

The air gap accounts for about 80% of the total mmf. The example also illustrates that the two iron flux densities are both 1.32 T; assigning different values would violate B = Φ/A under the stated geometry.

5. Electromagnetic Loading

5.1 Specific magnetic loading

The specific magnetic loading, or average air-gap flux density, is

where P is the number of poles, Φ is useful flux per pole, D is air-gap diameter, and L is active core length. Equivalently,

The quantity πD/P is pole pitch. Increasing B_av reduces required active volume but raises tooth and yoke flux densities, core loss, magnetizing current, saturation, and potentially noise.

5.2 Specific electric loading

For an m-phase armature, define specific electric loading as total RMS ampere-conductors per metre of armature periphery:

where I_ph is RMS phase current and Z_ph is the number of series conductors per phase. Since Z_ph = 2T_ph,

The unit of q is A/m. A higher electric loading improves output per unit volume but increases copper loss, temperature rise, leakage reactance, electromagnetic force, and cooling demand.

5.3 Conductor current density

Conductor current density is

where I_c is conductor current and A_c is conductor copper area. Current density governs local copper loss density, whereas specific electric loading measures circumferential current loading. They are related through slot geometry, number of conductors, winding arrangement, and copper fill factor; neither can replace the other.

5.4 Representative preliminary ranges

The following ranges are illustrative starting points, not guaranteed design limits. Final values require thermal, electromagnetic, mechanical, insulation, duty-cycle, and standard-compliance checks.

Peak tooth flux density commonly reaches roughly 1.5–2.0 T and yoke density about 1.2–1.7 T in conventional electrical steels, but allowable values depend strongly on material grade, frequency, loss target, and waveform. High-frequency machines generally require lower flux density. High-speed machines may be constrained more by rotor stress, windage, and loss than by the basic output equation.

6. Derivation of the Three-Phase Output Equation

Consider a balanced sinusoidal m-phase AC machine. Let winding factor be k_w, phase turns T_ph, electrical frequency f, and useful fundamental flux per pole Φ. The RMS induced phase emf is

This expression assumes approximately sinusoidal time variation of flux linkage; non-sinusoidal waveforms require harmonic treatment.

The electromagnetic apparent power associated with the internal phase emf is

Using the specified definition of electric loading,

For a P-pole machine rotating at synchronous mechanical speed n revolutions per second,

and from average magnetic loading,

Substitution gives

Therefore,

or numerically,

where

For this convention, S is in VA when B_av is in T, q in A/m, D and L in m, and n in r/s. Thus C₀ has units VA·s/m³.

For a three-phase machine, m = 3 has already been absorbed through the definition of q. The result is valid for either star or delta connection when consistent phase values are used. It is not a universal mechanical-output equation. A preliminary rated shaft-output relation is

for a motor, where η is efficiency and cosφ is power factor. This conversion is approximate because the derivation uses internal induced emf rather than terminal voltage and neglects detailed voltage drops.

The product D²L is the output volume parameter. The output equation fixes that product but not D and L separately. Their ratio is selected using pole pitch, ventilation, end-winding length, rotor dynamics, leakage reactance, transport limits, and manufacturing constraints.

7. Principal Dimensions and Design Iteration

A preliminary sequence is:

1. Convert rated shaft output to the relevant electromagnetic apparent power.
2. Select defensible initial values of B_av, q, and k_w.
3. Calculate D²L = S/(C₀n).
4. Select an aspect ratio or relation between core length and pole pitch.
5. determine slot, tooth, yoke, and conductor dimensions;
6. recalculate saturation, losses, reactance, temperature rise, efficiency, power factor, noise, and mechanical stress;
7. iterate until all constraints are satisfied.

The sizing coefficient is not an independent material constant: it packages design choices. Increasing it produces a smaller electromagnetic core but usually increases loss density and cooling difficulty.

8. Illustrative Design Case: Preliminary Motor Sizing

Background and problem

An illustrative design team is developing a 15 kW, 400 V, three-phase, four-pole, 50 Hz induction motor. The target efficiency is 0.90 and power factor is 0.84. Preliminary choices are B_av = 0.50 T, q = 30,000 A/m, and k_w = 0.93. Determine a first estimate of bore diameter and core length if active core length is chosen equal to pole pitch.

Analysis

The synchronous speed is

The required apparent input associated with rated shaft output is approximately

The output coefficient is

Therefore,

For four poles, pole pitch is τ_p = πD/4. With L = τ_p,

Thus,

Approach and outcome

The preliminary active dimensions are therefore a bore diameter of about 188 mm and an active core length of 148 mm. These are not final manufacturing dimensions. The team must next choose slots and turns, check tooth and yoke flux densities, calculate the Carter-corrected magnetizing current, determine rotor geometry, and perform thermal and mechanical verification.

Learning points

• The output equation determines an active-volume parameter, not a finished machine.
• Efficiency and power factor materially affect the apparent-power basis used for sizing.
• Moderate loadings provide a plausible starting point, but detailed checks may force larger dimensions.
• Increasing q alone could reduce size while worsening copper loss and temperature rise.

9. Computational Tools and Contemporary Practice

Analytical magnetic circuits remain valuable for understanding, initial sizing, sensitivity analysis, and checking simulations. Two-dimensional finite-element analysis is widely used to calculate local saturation, slotting harmonics, leakage, cogging torque, forces, and magnet operating points. Three-dimensional analysis is required when end leakage, skew, axial-flux paths, or end-region fields are important.

Optimization software can vary dimensions, materials, and loadings against efficiency, mass, cost, and thermal limits. Surrogate and machine-learning models can accelerate repeated evaluations, but they inherit the validity limits and biases of their training data. Best practice is to verify computational results against conservation laws, mesh-convergence studies, material data, analytical estimates, and test measurements.

10. Engineering, Safety, and Sustainability Considerations

Designers must consider applicable standards for efficiency classification, temperature rise, insulation systems, dielectric strength, vibration, balancing, enclosure protection, and hazardous environments. Examples include relevant IEC 60034 or NEMA MG 1 requirements, subject to jurisdiction and product category.

Aggressive loading can reduce material use but may lower efficiency or shorten insulation life. Sustainable design therefore requires lifecycle evaluation rather than minimizing initial mass alone. Responsible material selection also considers copper and electrical-steel production, permanent-magnet supply chains, recyclability, repairability, and accurate declaration of efficiency. Deliberately overstating efficiency or using unverified AI-generated designs presents professional and public-safety risks.

11. Design Limitations and Best Practices

Lumped magnetic circuits neglect distributed fields, local saturation, harmonics, anisotropy, and three-dimensional effects. Output equations neglect detailed voltage drops, end windings, cooling passages, structural material, and inactive volume. Representative loading ranges cannot substitute for validated material and thermal data.

Sound practice includes:

• using net iron area and the actual material B–H curve;
• stating leakage-factor, Carter-factor, and electric-loading conventions;
• separating useful flux from total flux;
• checking both average and local peak flux densities;
• coordinating B_av, q, and J rather than maximizing each independently;
• validating analytical, numerical, and experimental results against one another;
• retaining design margins for supply variation, overload, manufacturing tolerances, and cooling degradation.

Chapter Summary

Magnetic-circuit design applies Ampère’s law to the complete flux path. Each section contributes an mmf drop H_jl_j, while the winding supplies their total. Reluctance models are useful for linear or locally linear sections, but ferromagnetic saturation requires measured B–H data. Air gaps frequently dominate mmf, while fringing, slotting, leakage, net iron area, and core loss modify ideal predictions.

Specific magnetic loading describes useful flux per air-gap surface area, specific electric loading describes RMS ampere-conductors per metre of periphery, and conductor current density describes current per copper area. Their selection controls machine size, loss, heating, power factor, and reliability.

For the stated sinusoidal multiphase convention, combining the phase-emf equation with the electric- and magnetic-loading definitions gives S = 10.95k_wB_avqD²Ln. This equation provides a first estimate of active dimensions, not a completed design. Final dimensions require iterative electromagnetic, thermal, mechanical, acoustic, economic, and standards-based verification.

Key Terms

Air-gap mmf: Magnetomotive-force drop required to establish flux across an air gap.

Ampère’s circuital law: Relation equating the closed-path integral of magnetic field strength to enclosed current-turns.

Average air-gap flux density: Useful pole flux divided by total active air-gap area.

B–H curve: Material characteristic relating flux density to magnetic field strength.

Carter’s coefficient: Factor representing the increased effective air gap caused by slot openings.

Core loss: Hysteresis and eddy-current loss produced by time-varying magnetic flux.

Current density: Conductor current divided by conductor cross-sectional area.

Electric loading: Total RMS ampere-conductors per metre of armature periphery.

Flux density: Magnetic flux per unit area normal to the flux.

Fringing: Spreading of magnetic flux at the edges of an air gap.

Leakage flux: Flux that does not follow the intended useful path or link the intended windings fully.

Magnetic loading: Average useful air-gap flux density used in machine sizing.

Magnetomotive force: Magnetic excitation measured in ampere-turns.

Output coefficient: Coefficient combining winding factor, magnetic loading, and electric loading in the output equation.

Permeance: Reciprocal of magnetic reluctance.

Principal dimensions: Air-gap diameter and active core length defining the main electromagnetic volume.

Reluctance: Opposition of a magnetic path to flux under a linear magnetic-circuit model.

Saturation: Operating condition in which increasing field strength produces a relatively small increase in flux density.

Stacking factor: Ratio of net iron area to gross laminated-core area.

Winding factor: Product accounting for coil distribution and coil pitch in induced emf.

Review Questions

Short Answer Questions

1. Distinguish between a section mmf drop and the total winding mmf.
2. Why must net iron area be used when calculating lamination flux density?
3. What physical effect is represented by Carter’s coefficient?
4. State two adverse effects of excessive specific electric loading.
5. Why does an air gap often dominate the mmf requirement?

Long Answer Questions

1. Develop a magnetic-circuit model for a path containing two iron sections and one air gap, including nonlinear material data.
2. Explain saturation and its influence on magnetizing current, loss, power factor, and waveform quality.
3. Compare specific magnetic loading, specific electric loading, and current density as design variables.
4. Derive the three-phase output equation from the RMS phase-emf equation and the stated definition of electric loading.
5. Discuss how leakage, fringing, slotting, and stacking factor alter an ideal magnetic-circuit calculation.

Analytical / Application Questions

1. A core carries 0.010 Wb through a net area of 0.008 m². Its iron path is 0.60 m long, with H = 300 A/m at the resulting flux density. It also contains a 1.0 mm uniform air gap of equal effective area. Calculate total mmf.
2. A six-pole, 50 Hz machine has k_w = 0.94, B_av = 0.55 T, q = 40 kA/m, and D²L = 0.012 m³. Estimate its electromagnetic apparent power.
3. Evaluate the likely electromagnetic and thermal consequences of increasing q by 25% while keeping machine dimensions, cooling system, and flux density unchanged.

Multiple Choice Questions with Answers

Question 1

A uniform magnetic section carries 0.012 Wb through an area of 0.008 m². What is its average flux density?

A. 0.667 T

B. 1.50 T

C. 8.00 T

D. 96.0 T

Correct Answer: B. 1.50 T

Question 2

Which statement correctly distinguishes section mmf from winding mmf?

A. Every section has an mmf drop equal to NI.

B. NI is the total mmf around the modeled closed path, while Hl is one section's mmf drop.

C. NI applies only to air gaps, while Hl applies only to iron.

D. A section mmf is measured in webers, while NI is measured in teslas.

Correct Answer: B. NI is the total mmf around the modeled closed path, while Hl is one section's mmf drop.

Question 3

A 1.0 mm air gap carries a uniform flux density of 1.0 T. Neglecting fringing and slotting, its mmf drop is approximately F_g = Bg/μ₀.

A. 79.6 At

B. 398 At

C. 796 At

D. 1257 At

Correct Answer: C. 796 At

Question 4

What is the primary purpose of Carter's coefficient in preliminary machine design?

A. To correct conductor resistance for temperature

B. To represent the increased effective air gap caused by slot openings

C. To convert peak current into RMS current

D. To calculate lamination stacking factor

Correct Answer: B. To represent the increased effective air gap caused by slot openings

Question 5

For the electric-loading convention used in the chapter, which expression is correct?

A. q = mI_phZ_ph/(πD)

B. q = πD/(mI_phZ_ph)

C. q = B_avπDL/P

D. q = I_c/A_c

Correct Answer: A. q = mI_phZ_ph/(πD)

Question 6

Under the stated sinusoidal multiphase assumptions, the output coefficient is

A. C₀ = 4.44k_wB_avq

B. C₀ = 10.95k_wB_avq

C. C₀ = πk_wB_av/q

D. C₀ = 2fB_avq

Correct Answer: B. C₀ = 10.95k_wB_avq

Question 7

A machine has k_w = 0.90, B_av = 0.50 T, and q = 30,000 A/m. What is its output coefficient to three significant figures?

A. 14,800 VA·s/m³

B. 73,900 VA·s/m³

C. 148,000 VA·s/m³

D. 329,000 VA·s/m³

Correct Answer: C. 148,000 VA·s/m³

Question 8

Which change most directly increases copper-loss density if conductor material and temperature are initially unchanged?

A. Increasing conductor current density

B. Increasing stacking factor

C. Reducing Carter's coefficient

D. Reducing core-loss frequency

Correct Answer: A. Increasing conductor current density

Question 9

Two series iron sections each carry 0.0132 Wb and each has an area of 0.010 m². Their flux densities are

A. 1.32 T and 1.32 T

B. 1.20 T and 1.30 T

C. 0.758 T and 0.758 T

D. 13.2 T and 13.2 T

Correct Answer: A. 1.32 T and 1.32 T

Question 10

Why can the output equation not by itself determine both bore diameter and core length?

A. It contains no rotational speed term.

B. It determines only the product D²L, so an additional geometric criterion is required.

C. It applies only after diameter and length are known exactly.

D. It omits specific electric loading.

Correct Answer: B. It determines only the product D²L, so an additional geometric criterion is required.