Motor Computational Tool
2D nonlinear FE solver
STATOR SLOT
ROTOR POLE
SYMMETRY SECTOR
FULL MOTOR
drag to pan · scroll to zoom · double-click to reset
Active geometry: toy_spm 🔒 idle
Newton–Raphson runtime log

          
Symmetry sector · geometry to be simulated
drag to pan · scroll to zoom · double-click to reset
Electromagnetic torque
Mechanical power
Phase currents
Phase voltages (recovered)
Pareto front · live (grey = population, coloured line = non-dominated front)
Convergence · best objective vs generation
Validation · predicted vs true (held-out test set)
Live prediction (instant — no FE solve)
TEMPERATURE ZONES (lumped — representative)
Validated thermtool reluctance network (Mellor cylindrical components) — docs/BENCHMARKS.md.
3D VIEW — TRUE KINEMATICS (drag to orbit · wheel to zoom · double-click to reset)
rate the reducer to see it in 3D
MESH IN THE CARRIER FRAME — TRUE INVOLUTES · LINE OF ACTION A–E · LOAD PER TOOTH PAIR
MEMBERS — STATIC AND LOAD-SPECTRUM SAFETY
TOOTH CONTACT ANALYSIS — ONE MESH CYCLE AND ONE TOOTH PAIR FROM A TO E
Static transmission error (µm along the line of action)
Load share of one tooth pair
Contact stress along the path
Specific sliding
Contact temperature (bulk + flash)
S-N
ISO 6336 calculation sheet
Gear geometry
ISO 6336:2019 method B (pitting part 2, bending part 3 with Y_F/Y_S from the 30° tangent of the fillet, load at the outer point of single contact; internal gear: rack substitute), K_V and K_Hβ by method C, life factors and Miner's rule per part 6, material values MQ (part 5). TCA: quasi-static load sharing with ISO single stiffness, parabolic along the path, linear tip relief; Blok flash temperature. Spur gears.
Drive train — pick a machine and a reducer, set the duty at the joint and simulate them together. The reducer is rated at the duty with its own model; the machine is solved by FE at every operating point (the current for the required torque, then a transient); the field and the gearbox animate together.
CAMPBELL DIAGRAM — ring modes (—) vs force-order excitation (rising) · crossings = resonance
STATOR RING MODES
FORCE HARMONICS
Analytic ring-mode NVH screening (2D). EM Maxwell-stress force orders vs stator natural frequencies; a 3D modal POC with eigenvalue reuse is on the roadmap (docs/ROADMAP.md Phase G).
Individual = each discipline solved standalone from the current design. Coupled = outputs feed inputs (EM losses → thermal with resistivity feedback → structural; EM air-gap harmonics → NVH). Elmer rungs run real 3D FEM when installed; otherwise the analytic/LPTN rung, flagged.
Victor Mukherjee VM

Victor Mukherjee

Creator of Helsinki — 2D finite-element solver for electric machines

This is part of my hobby project — a 2D nonlinear finite-element solver for electric machines, built for fast, accurate motor-design exploration right in the browser. It covers a range of topologies, windings, cooling layouts and analysis modes:

3machine geometriessurface PM · V-type IPM · double-V IPM
2winding schemesconcentrated · distributed
2cooling mountsyoke holes · tooth-centre notch
7calculation profilessingle-point, no-load, cogging, back-EMF, torque sweeps…

✓ Validated

Cross-checked with Ansys Maxwell 2D — within 2% on torque and operating KPIs for identical inputs.

☁ Cloud-ready

A browser-based calculation tool, fully cloud-ready and scalable on a microservice-based architecture.

⚙ Applications

Rotating actuators, electric vehicles, and other high-performance electric drives.

Advanced version (not in this production build)

A more advanced version performs higher-order statistical analysis such as Uncertainty Quantification, and builds surrogate models via order reduction or machine-learning models (XGBoost, RNN). Pre-built FMU generation is also on the roadmap.

This is a private tool I use for my own research and for educational purposes with my students. I'm glad to discuss the different technical layers behind it — feel free to reach out.

in  Connect on LinkedIn

Demo access — API key