CAD-friendly plotting

A generic mesh viewer draws the triangulation: with show_edges=True you see every facet edge, an artifact of the tessellation tolerance that means nothing to a CAD user. What a CAD user wants is the model’s topological edges (the B-rep feature curves) over smoothly shaded faces, exactly like a CAD application’s viewport.

pyvista-cad reproduces that pipeline:

• pyvista_cad.topods_to_multiblock() tessellates each topological face independently and attaches analytic surface normals (a coarse cylinder still shades perfectly round), plus a sibling edges block of the real B-rep curves recovered from the same mesh pass.

• The plotter.cad component and the .cad.plot() accessor render faces smooth with their triangle edges hidden and the topological edges overlaid in the theme edge colour.

import numpy as np
import pyvista as pv

import pyvista_cad
from pyvista_cad import examples

Read a STEP bracket. It comes back as a pyvista.MultiBlock with the originating TopoDS cached on each block, so the CAD-view helpers can recover exact topology.

mb = pyvista_cad.read_step(examples.bracket_step_path())

The generic view: triangle mesh with facet edges. The filleted bend looks polygonal and the edges are tessellation noise, not features.

combined = mb.combine().extract_surface()
combined.plot(show_edges=True, color='lightblue')

Static Scene

Interactive Scene

/home/runner/work/pyvista-cad/pyvista-cad/examples/06_advanced/cad_friendly_plot.py:40: PyVistaFutureWarning: The default value of `algorithm` for the filter
`UnstructuredGrid.extract_surface` will change in the future. It currently defaults to
`'dataset_surface'`, but will change to `None`. Explicitly set the `algorithm` keyword to
silence this warning.
  combined = mb.combine().extract_surface()

The CAD-friendly view: smooth analytic shading, only the topological edges drawn. Same data, but it reads like a CAD viewport.

mb.cad.plot()

Static Scene

Interactive Scene

The data behind the render. topods_to_multiblock keeps per-face identity and classifies every edge curve.

.cad.cad_view() resolves the read result (recovering each block’s cached B-rep) into a {'faces', 'edges'} MultiBlock.

cad = mb.cad.cad_view(linear_deflection=0.2)
edges = cad['edges']
legend = edges.field_data['cad.edge_kind_legend'][0]
kinds, counts = np.unique(edges.cell_data['cad.edge_kind'], return_counts=True)
n_faces = sum(1 for _ in cad['faces'].cad.walk())
print(f'topological faces : {n_faces}')
print(f'topological edges : {edges.n_cells}')
print(f'edge-kind legend  : {legend}')
print(f'edge-kind counts  : {dict(zip(kinds.tolist(), counts.tolist()))}')

topological faces : 1
topological edges : 19
edge-kind legend  : 0=line;1=circle;2=ellipse;3=hyperbola;4=parabola;5=bezier;6=bspline;7=offset;8=other
edge-kind counts  : {0: 17, 1: 2}

A silhouette-only outline, handy for drawings and thumbnails.

pl = pv.Plotter()
pl.cad.add(cad, color='white', silhouette=True)
pl.show()

Static Scene

Interactive Scene