Computational validation of Differential Flux Steering
FEMM 4.2 modelling demonstrates the Magnetic Transistor proof of principle: a comparatively small, reversible steering excitation redirects a larger established magnetic flux between parallel output paths.
Computational proof of principlePublished for independent replication
Established flux redirects between the left and right output paths as steering polarity changes. Download MP4
Simulation overview
What the animation demonstrates
01
Core operating effect
The animation above demonstrates the core operating effect of the FluxWorx Magnetic Transistor: a small control excitation progressively redirects an established magnetic flux of approximately 1.2 tesla between two output paths. That is a substantial working flux density, comparable in magnitude to the field strength of a 1.5 T clinical MRI system, although here it is concentrated within the steel magnetic circuit. As the steering current moves from −20 mA through zero to +20 mA, the preferred path reverses smoothly and controllably. The FEMM modelling therefore provides strong computational validation of the Magnetic Transistor effect: a compact control input governing a much larger magnetic transfer inside the core itself.
02
Engineering pathway
That capability opens a new engineering pathway across data-centre power infrastructure, direct-DC conversion, motors, grid equipment, industrial power systems and emerging magnetic-domain computing. DFS may enable fewer conversion stages, lower heat, reduced EMI, simpler cooling and new forms of active magnetic logic and multi-state control. The technology is protected under PCT/AU2026/050419, and the results are being published to support the next decisive step: independent laboratory validation and collaborative commercial development with institutions and industry partners capable of turning a validated Australian invention into a globally significant technology.
Evidence position
This FEMM work is computational evidence, not physical certification. Dynamic operation, thermal performance, conversion efficiency and commercial-product performance remain subjects for independent laboratory validation.
Reading the animation
A smooth, reversible steering sequence
The model starts with an established magnetic flux of approximately 1.2 T. As steering current moves from −20 mA through zero to +20 mA, the preferred output path reverses smoothly and controllably.
≈ 1.2 TEstablished magnetic flux−20 to +20 mASteering-current sweepReversibleLeft/right path redirection
A carefully bounded MRI comparison
The approximately 1.2 T flux density is comparable in magnitude to a 1.5 T clinical MRI field. This is a magnitude comparison only: in the FEMM model the field remains concentrated within the steel magnetic circuit, and the systems and field environments are not equivalent.
Reference model 1.9
Model configuration
Inputs recorded in the revised validation appendix for reproducible institutional review.
Solver and geometry
Software
FEMM 4.2, DC magnetostatic
Problem
2D planar magnetostatic, Cartesian
Depth
25.4 mm
Precision
1 × 10−8
Material and mesh
Core
M-19 electrical steel
B-H data
Nonlinear 33-point built-in curve
Meshing
FEMM Smart Mesh
Boundary
A = 0 circular external air region
Excitation
Source
200 turns × 0.2 A = 40 At
Steering
500-turn windings
Test sequence
5, 10, 15 and 20 At
Polarity
Reversed to exchange the preferred path
Recorded steering results
Local flux-density readings
These values were read manually from the FEMM post-processor cursor at consistent reference locations. They are local, human-read values—not integrated path flux or laboratory measurements.
FEMM post-processor cursor readings for neutral and left/right-favouring conditions
Condition
Current (A)
MMF (At)
Centre B (T)
Left B (T)
Right B (T)
Neutral
0
0
1.275
0.639
0.633
Right-favouring
0.005
5
1.260
0.475
0.781
Left-favouring
0.005
5
1.260
0.785
0.477
Right-favouring
0.010
10
1.270
0.350
0.920
Left-favouring
0.010
10
1.270
0.920
0.350
Right-favouring
0.015
15
1.210
0.180
1.020
Left-favouring
0.015
15
1.210
1.010
0.190
Right-favouring
0.020
20
1.190
0.030
1.160
Left-favouring
0.020
20
1.190
1.160
0.030
Observed outcome: the neutral condition is closely balanced. At the largest recorded steering step, the local branch readings are approximately 1.16 T and 0.03 T, with the preferred side reversing with steering polarity.
Evidence interpretation
What the simulation supports
The sequence shows progressive differential control, closely mirrored reversal and a substantial centre-limb bias throughout the sweep. The branch separation is much larger than reasonable cursor-placement variation.
Clear differential control
The response progresses from near balance to a pronounced high-flux/low-flux state as steering MMF increases.
Directional symmetry
Equal-magnitude steering in opposite directions produces closely mirrored left/right results.
Established bias remains
The centre-limb reading stays substantial while the output limbs exchange magnetic loading, consistent with a flux-routing control element.
Limits of the present evidence
The DC magnetostatic model does not establish dynamic switching behaviour, thermal rise, control power, load transfer, conversion efficiency, manufacturability, reliability or independent physical validation. Formal mesh convergence, boundary-size sensitivity, steering-only analysis, integrated flux and tolerance studies also remain recommended.
Independent verification
Recommended laboratory programme
The next decisive step is qualified institutional replication using controlled materials, calibrated instrumentation and documented measurement protocols.
Rebuild or import the reference geometry and reproduce the neutral and reversible steering sequence.
Record the FEMM build, mesh counts and external-domain dimensions.
Calculate integrated flux across defined left and right limb sections.
Run mesh-refinement, enlarged-boundary, bias-only, steering-only and combined comparisons.
Construct an M-19 or equivalent test article and measure branch flux with calibrated Hall or flux sensors.
Progress to dynamic excitation, thermal rise, control power, load transfer and efficiency characterisation.
Revised technical appendix
Review the complete FEMM record
Download the model inputs, steering sequence, evidentiary position, validation record and recommended replication programme.
The model provides computational proof of principle that a reversible steering excitation can redistribute established magnetic flux between two output paths. It supports independent replication; it is not physical certification or validation of a finished product.
Are the published flux-density values laboratory measurements?
No. They are consistent, local human-read cursor values from the FEMM post-processor. They show the scale, direction and repeatability of the modelled response, but calibrated sensors and integrated cross-sectional measurements are required for physical verification.
What remains to be independently validated?
Independent work should reproduce the model, test mesh and boundary sensitivity, calculate integrated path flux, construct a physical test article, and characterise dynamic behaviour, thermal rise, control power, load transfer and efficiency.
How does the approximately 1.2 T field compare with MRI?
The comparison is one of magnitude only: approximately 1.2 T is in the range of a 1.5 T clinical MRI field. In this model the field remains concentrated within the steel magnetic circuit, so the systems and field environments are not equivalent.
Protected under PCT/AU2026/050419
Partner in independent replication
FluxWorx is seeking qualified institutions and industry partners to reproduce, measure and extend this computational proof of principle through physical laboratory evaluation and collaborative engineering.