# pvtu with polyhedra issue
**URL:** https://discourse.paraview.org/t/pvtu-with-polyhedra-issue/13858
**Category:** Development
**Tags:** vtk
**Created:** [February 3, 2024, 3:31pm UTC](https://discourse.paraview.org/t/pvtu-with-polyhedra-issue/13858 "2024-02-03T15:31:12Z")
**Posts on this page:** 3
**Page:** 1
### Author:  [@kirill-terekhov](https://discourse.paraview.org/u/kirill-terekhov)
#### Post date: [February 3, 2024, 3:31pm UTC](https://discourse.paraview.org/t/pvtu-with-polyhedra-issue/13858/1 "2024-02-03T15:31:12Z")
The display of polyhedra in Paraview depends on the order of pieces in the pvtu file.
In the attached example [parallel\_polyhedron.pvtu](https://discourse.paraview.org/uploads/short-url/h8LUAzD8ZT6Yt2gAAek5PEKoNHH.pvtu) (754 Bytes), there are 3 hexahedra. In the first piece [parallel\_polyhedron\_0.vtu](https://discourse.paraview.org/uploads/short-url/5EB9kkCS8lxINP4GO4kOpLQ8yab.vtu) (1.0 KB), there is the usual hexahedron of type 12. In the following two [parallel\_polyhedron\_1.vtu](https://discourse.paraview.org/uploads/short-url/szvMHf8KFFa1G9G14yjDjEcnwJA.vtu) (1.1 KB) and [parallel\_polyhedron\_2.vtu](https://discourse.paraview.org/uploads/short-url/9iF5AGmvGQQrrAeq0BxAEYJFkeE.vtu) (1.1 KB), there are two hexes stored as polyhedra of type 42 with additional “faces” and “faceoffsets” arrays.
In the original order of the example, Paraview displays a gap in the middle of the mesh, whereas there should be three tight hexahedra. If the order of the pieces in the pvtu file is changed with parallel\_polyhedron\_0.vtu going last, the mesh is displayed correctly.
It appears that Paraview renders the mesh correctly if the pieces with polyhedra are followed by the pieces without polyhedra. It skips rendering the polyhedra if the piece without polyhedra is followed by the piece with polyhedra.
Is it an issue in pvtu file or a bug in Paraview?
Encountered with Paraview 5.11.1 on windows and tested with ParaView-5.12.0-RC2-Windows-Python3.10-msvc2017-AMD64.msi.
The next issue with the attached example is that if you apply a “Clip” filter, it will give a warning “A cell with a non-manifold triangulation has been encountered. This cell cannot be contoured.”. Only the top cell is rendered in this case. However, if you check “Crinkle clip”, it will show the cell in the middle and wouldn’t render the cell at the bottom.
This works with Paraview 5.11.1 on windows. ParaView-5.12.0-RC2-Windows-Python3.10-msvc2017-AMD64.msi crashes at “Clip” filter.
---
### Author:  [@mwestphal](https://discourse.paraview.org/u/mwestphal)
#### Post date: [February 5, 2024, 8:24am UTC](https://discourse.paraview.org/t/pvtu-with-polyhedra-issue/13858/2 "2024-02-05T08:24:09Z")
There is definitely a bug there. Could you open an issue?
[https://gitlab.kitware.com/paraview/paraview/-/issues](https://gitlab.kitware.com/paraview/paraview/-/issues)
---
### Author:  [@kirill-terekhov](https://discourse.paraview.org/u/kirill-terekhov)
#### Post date: [February 5, 2024, 2:34pm UTC](https://discourse.paraview.org/t/pvtu-with-polyhedra-issue/13858/3 "2024-02-05T14:34:34Z")
Done: [https://gitlab.kitware.com/paraview/paraview/-/issues/22471](https://gitlab.kitware.com/paraview/paraview/-/issues/22471)
Thank you, Mathieu, for your time.