Post-processing

Reading the field, forces, losses and demagnetization risk once a solve finishes.

Field plots

The Postprocessing tab offers two independent choices: a colour surface field (a filled contour of one quantity across the whole model) and an overlay contour or vector type drawn on top of it — you can, for example, show a |B| colour surface with flux lines drawn over it, or a vector-field overlay instead of lines. Available quantities for a planar magnetic model: vector potential A, field strength H (magnitude, Hx, Hy), flux density B (magnitude, Bx, By), relative permeability, current density, and demagnetization risk (see Demag risk below). An axisymmetric model relabels these to the flux function r·A and Br/Bz. A thermal solve substitutes its own set: temperature, its gradient and the heat flux.

Overlay types add flux lines and B/H vector fields to the same list of quantities. Vector Field Density and Vector Field Scaling control how many arrows are drawn and how long they are. Turn on Show Mesh to see the triangulation itself under the field — useful when a result looks locally wrong and you want to check whether the mesh is unusually coarse or distorted there.

The colour surface's range is not purely automatic: two Field min / Field max fields let you type a fixed range instead of letting it auto-scale to each step's own min/max — useful for comparing two steps, or two models, on the same colour scale rather than a range that shifts every time you page through a transient result. A separate Field Colormap combo offers nine palettes (Jet, Blues, Viridis, Inferno, Plasma, Magma, Turbo, Greys, Rainbow) if the default does not suit the report or presentation you are building.

Stepping through a transient or time-harmonic result

Once a result has more than one step (any transient run, or a time-harmonic run's reconstructed Snapshots Per Period, see Solving — time integration), a step combo box shows "step: N | time: X" next to Left/Right arrow buttons that move one step at a time, and a Play button that animates through every step at a fixed frame rate (Options → Application Preferences controls the playback speed) — useful for watching a field or a rotor sweep through a full electrical cycle without clicking through it by hand. All of the field plots, probes and region/shape results above read whichever step is currently selected, so scrub to the instant you actually want before reading a number off a region result. A time-harmonic run's snapshots are purely reconstructed instantaneous values (x(t) = Im{X·e^jωt}) walked with this same step control; there is no separate magnitude/phase (phasor) view of the complex solution.

Probes

Two kinds of probe, for two different questions:

• a field probe point, placed at a fixed x/y location, tracks one quantity's value over the whole run — the tool for "what does the field do at this exact spot over time";
• a contour probe (a line or arc, sampled at several points along its length) integrates or averages a quantity along a path instead of at a point.

The GUI's interactive contour-probe tool places a line or an arc by clicking, the same way as drawing ordinary geometry (see Geometry — drawing primitives). A full circular contour probe — the shape you would want to sweep a complete air-gap ring for a torque or flux check — is not available as a canvas tool; it exists only through the Python API / MCP (add_probe_contour_circle, giving a centre and radius directly), so build a circular probe from a script or an MCP call, or approximate one with two closing arcs from the canvas.

A fixed-position probe on a rotating machine reads whatever is physically under it at that instant, not a label that follows the rotor — keep that in mind if a probe's reading jumps as the rotor passes underneath it; see Troubleshooting — zero or unexpectedly small torque/force.

Region and shape results

Select a region and choose from its available results: Flux, Field Energy, Force, Joule Losses, Iron AC Losses (a thermal model instead offers only Average Temperature for a region). A selected shape (rather than a whole region) offers Normal Flux, Average B_n and Force (or, in thermal mode, Heat Flux and Average Temperature). A circuit element reports its own Flux Linkage, Voltage, Current and Losses, read from a separate list box next to the schematic once you select the element on the Circuit Editor canvas, rather than from the same region/shape results panel — select the coil or element first, then pick the quantity from its own list.

Forces and torque — the eggshell method

Force or torque on a region is computed with the eggshell method: rather than integrating stress on the region's own boundary (sensitive to exactly how the mesh sits on that boundary), Nabla integrates a Maxwell-stress-tensor quantity over a thin shell of elements just outside it, weighted so that it equals 1 on the region's own nodes and 0 away from it. Where two magnetic regions touch, the weight is shared between them so that the forces on each side of the contact still add up to the correct total force on the assembly — you can sum the force on a rotor's iron and its magnets separately and get the same answer as the whole rotor at once. For an axisymmetric model only the axial force is meaningful (the net radial force is zero by symmetry) so that is the one reported.

Read it from the region-results dropdown described above, after selecting the region whose force you want (a rotor's regions for torque, an armature's region for a linear pull). For a rotating machine, remember to let the run settle before reading torque — see Motion and Troubleshooting — non-convergence in a nonlinear solve.

Losses

Joule Losses comes straight from the solved current density and each region's conductivity. Iron AC Losses uses the Steinmetz coefficients from Regions & materials — iron-loss coefficients, evaluated against the region's own simulated flux waveform; fitting those coefficients from a material's measured loss data, like reading the loss result for a region that already has coefficients set, is free on every tier. The Steinmetz model is a per-cycle average and tends to under-predict loss for a waveform with many small reversals (minor loops) rather than one clean sinusoid — keep that in mind comparing against a measured value on a heavily-loaded machine.

Demagnetization risk

Demagnetization Risk (%) is a field quantity (and, at a probe point, a per-point result) showing how close each point of a permanent magnet came to its own material's demagnetizing knee during the run: 0% means the operating point never approached the knee, 100% means it reached or passed the intrinsic coercivity where the magnet is expected to lose strength permanently. It uses the demag-knee table stored on the magnet's own material (see Regions & materials — B-H curves) at the temperature you solved at, so a magnet material with no demag table recorded reads 0% everywhere rather than a wrong number.

Exporting results

Model → Export Field (PNG) renders whichever field is currently selected in Postprocessing to an image file, framed to the whole model and coloured to that field's own value range (not whatever zoom or manual colour range you have set on screen). It needs both a field selected and a solved result, in memory or on disk, to be enabled. Field-on-shape samples, coil time series and probe time series each have their own CSV/TXT export next to the tool that produced them, for taking a result into a spreadsheet or another analysis tool.

Machines Model → Create Report generates a full PDF report for a solved PMSM or induction-machine model — see the machines chapter for what it contains.

Next steps

• Regions & materials — where iron loss and demag data come from.
• Troubleshooting.
• Back to the contents.