---
jupyter:
description: 'Deformable shapes in planex2d (experimental): faces as dense polylines that follow displacements from
arrays or NGSolve GridFunctions, with holes opening and parts merging when the boundary comes too close.'
---
# Deformable shapes
- [Displacements and topology changes](displacements.md)
- [Displacements from NGSolve](ngsolve.md)
*Experimental: a proof of concept.*
A deformable shape represents a face as dense closed polylines whose vertices follow a displacement, given as an array
or as an NGSolve GridFunction. When parts of the boundary come closer than a minimum distance, the shape changes its
topology: a hole opens to the outside, and parts merge or split. `Shape()` returns the result as a planex2d shape at
any time, so netgen can mesh it again. This suits moving boundaries, such as shape optimization, erosion or growth,
where a mesh that is only deformed would degenerate.
In this section
