Restricted Access

This vault contains proprietary research, mathematical derivations, and design data.
Access key required.

Access denied — incorrect key or missing acknowledgment.

The Vault

Mathematical foundations, research archive, and parametric design simulation for MagFlux electromagnetic systems.

Magnetic Flux Fundamentals

The theoretical backbone of MagFlux designs — from Maxwell's field equations through to applied torque and power density derivations for axial-flux and Halbach-array configurations.

Maxwell's Flux Law (Gauss)

∇ · B = 0

The divergence of the magnetic flux density B is always zero — magnetic monopoles do not exist. All field lines are closed loops. This is the governing constraint for every core geometry we design.

Faraday's Law of Induction

ε = -N · dΦ/dt = -N · d(B·A·cos θ)/dt

The induced EMF (ε) in a winding of N turns is proportional to the rate of change of flux linkage. For axial-flux motors, maximizing B and the effective pole-face area A directly drives voltage output per revolution.

ε
Induced electromotive force (V)
N
Number of winding turns
Φ = B·A·cos θ
Magnetic flux linkage (Wb)
B
Flux density (Tesla)
A
Effective pole area (m²)
θ
Angle between B and surface normal

Ampere's Law (Magnetomotive Force)

MMF = N·I = H·l = Φ·R_m

Magnetomotive force drives flux through a magnetic circuit. The reluctance R_m of the gap geometry determines how much flux a given MMF produces — minimizing air-gap reluctance is the central optimisation problem in our actuator designs.

MMF
Magnetomotive force (Ampere-turns)
H
Magnetic field intensity (A/m)
l
Magnetic path length (m)
R_m = l/(μ·A)
Magnetic reluctance (A-t/Wb)
μ = μ₀·μᵣ
Permeability of core material

Axial-Flux Motor Torque

T = (π/4) · B_avg · K_s · (D_o³ - D_i³) · p

The electromagnetic torque of an axial-flux machine scales with the cube of the outer-to-inner diameter ratio. This favours pancake geometries with large diameter and short axial length — exactly the topology our designs target.

T
Electromagnetic torque (N·m)
B_avg
Average air-gap flux density (T)
K_s
Linear current density (A/m)
D_o, D_i
Outer & inner stator diameters (m)
p
Pole pairs

Halbach Array Flux Density

B_peak = B_r · (1 - e^(-2πh/λ)) · sin(π/n) / (π/n)

A Halbach array concentrates flux on one side while cancelling it on the other. B_peak is maximised by increasing magnet thickness (h) relative to pole pitch (λ), and by using high-remanence magnet grades. N≥4 segments per pole approximates the ideal sinusoidal case.

B_r
Remnant flux density of magnet (T)
h
Magnet array thickness (m)
λ
Pole pitch (m)
n
Segments per pole (≥4)
B_peak
Peak air-gap flux (T)

Power Density & Efficiency

P = T · ω = T · (2π · n/60)
η = P_out / P_in = P_out / (P_out + P_cu + P_fe + P_mech)

Power density (W/kg) is the primary metric for our actuator designs. Copper losses (P_cu = I²R) and iron losses (P_fe = P_hysteresis + P_eddy) must be minimised. Axial-flux topologies excel here by using thin laminations or SMC cores to suppress eddy currents at high frequencies.

ω
Angular velocity (rad/s)
P_cu
Copper (winding) losses (W)
P_fe
Iron core losses (W)
η
Overall efficiency (%)

Research Archive

Papers, simulation reports, and technical notes produced by Amon Hen research agents and reviewed by the principal investigator.

📄
2025-Q4 · Technical Paper · Amon Hen Research
Axial-Flux Motor Topology Survey: Performance Benchmarking Across YASA, TORUS, and Coreless Configurations
Comprehensive comparative analysis of axial-flux motor architectures. Evaluates torque ripple, power density (kW/kg), and thermal performance across three leading topologies. Simulation data generated via FEA agent; conclusions validated against published literature.
Axial-Flux FEA Benchmarking
📄
2025-Q4 · Technical Paper · Amon Hen Research
Halbach Array Optimization for High-Flux Linear Actuators: Segmentation, Orientation, and Material Trade-offs
Parametric study of Halbach configurations across neodymium (N52, N48H) and samarium cobalt grades. Air-gap flux density improvements of 38–62% over conventional arrays. Key finding: diminishing returns beyond n=8 segments per pole at λ < 20 mm.
Halbach Linear Actuator NdFeB
🔬
2025-Q3 · Simulation Report · FEA Agent
Air-Gap Sensitivity Analysis: Effect of Tolerance Stack-Up on Flux Linkage in Pancake Motor Designs
Monte Carlo simulation across ±0.05–0.5 mm air-gap variation in a 12-pole, 180mm OD axial-flux machine. Results show 2.3% torque degradation per 0.1mm gap increase. Mechanical tolerance requirements derived for prototype manufacturing specs.
Simulation Tolerance Analysis Manufacturing
📝
2025-Q3 · Design Note · Amon Hen Orchestrator
Soft Magnetic Composite (SMC) vs. Laminated Steel: Core Loss Comparison at 400–2000 Hz Operating Frequencies
Design decision document comparing SMC powder cores against 0.2mm silicon steel for high-frequency axial-flux stators. SMC eliminates the lamination challenge for 3D flux paths and reduces eddy losses by 70–85% above 800 Hz, at the cost of 15–20% lower saturation flux density.
Materials SMC Core Loss
📄
2025-Q2 · Technical Paper · Amon Hen Research
Winding Configuration Optimization for Fractional-Slot Concentrated Windings in Axial-Flux Machines
Analysis of slot/pole combinations (12s/10p, 12s/14p, 9s/8p) for concentrated winding axial-flux motors. Winding factor calculations, cogging torque comparison, and short-circuit fault tolerance assessment. 12s/10p selected as primary configuration for Phase 1 prototype.
Windings Slot-Pole Cogging
🔬
2025-Q2 · Simulation Report · FEA Agent
Thermal Modeling of Coreless Axial-Flux Stator Under Continuous 80% Load: Hotspot Identification and Cooling Pathway Design
Computational thermal analysis of a coreless PCB stator at rated current density. Identifies winding hotspots, derives steady-state temperature distribution, and evaluates aluminium heat-spreader geometry for passive cooling to achieve T_max < 130°C at ambient 40°C.
Thermal Coreless PCB Stator
📝
2025-Q1 · Design Note · Amon Hen Orchestrator
Agent-Driven Design Space Exploration: Automating FEA Parametric Sweeps with the Aviary AI Fleet
Internal methodology paper describing how the Aviary cluster's 12-agent AI fleet performs automated parametric sweeps over motor design variables. Covers agent specialization, tool-calling pipelines, result aggregation, and human-in-the-loop review gates for design sign-off.
AI Agents Automation Methodology

Parametric Design Simulator

Adjust each variable to explore the electromagnetic design space. Results are computed from the core analytical models — use as a first-pass tool before agent-driven FEA sweeps.

⚙️ Design Variables

Geometry
Stator outer diameter. Larger D → higher torque (cubic scaling).
Stator inner diameter. Ratio D_i/D_o ≈ 0.6 is typical optimal.
Minimise for max flux linkage; limited by mechanical tolerance.
Thicker magnets → higher flux density at gap. Diminishing returns.
Electromagnetics
More poles → lower speed, higher torque at same power.
N52=1.45T, N48H=1.38T, SmCo28=1.05T typical values.
Segments per pole. n=1 = conventional array. n≥4 recommended.
Linear current density. Higher → more torque but more heat.
Operating Point
Estimated overall efficiency (copper+iron+mech losses combined).
Peak Air-Gap Flux
Tesla
EM Torque
N·m
Shaft Power
kW
Power Density
kW/kg (est.)

Torque vs. Outer Diameter (at current settings)

Air-Gap Flux Density vs. Magnet Thickness

Power vs. Speed Curve

Configuration Readout

Press "Compute Results" to generate output...