# Unpaired faces in mesh

**URL:** <https://pyfr.discourse.group/t/unpaired-faces-in-mesh/216>\
**Category:** General\
**Created:** [25 May 2018 17:51 UTC](https://pyfr.discourse.group/t/unpaired-faces-in-mesh/216 "2018-05-25T17:51:04Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![Andrew\_Hedrick](https://avatars.discourse-cdn.com/v4/letter/a/5f8ce5/32.png) [@Andrew\_Hedrick](https://pyfr.discourse.group/u/Andrew_Hedrick)\
**Post date:** [25 May 2018 17:51 UTC](https://pyfr.discourse.group/t/unpaired-faces-in-mesh/216/1 "2018-05-25T17:51:04Z")

</div>

The attached wedge\_2.geo file produces the error

> File “/usr/lib/python3.4/site-packages/pyfr/readers/base.py”, line 194, in get\_connectivity  
> raise ValueError(‘Unpaired faces in mesh’)

when attempting to run ‘pyfr import wedge\_2.msh wedge.pyfrm’.

If I uncomment the line ‘//Physical Line(“wedge”) = {5, 6, 7};’ then the mesh will import, but running py produces the error

> File “/usr/lib64/python3.4/configparser.py”, line 1115, in \_unify\_values  
> sectiondict = self.\_sections[section]  
> KeyError: ‘soln-bcs-wedge’

Which makes sense because I haven’t defined a soln-bsc-wedge in my ini file.

Any help correcting the geo file is much appreciated.  
As a possible recommendation, would the PyFR team consider including the .geo files in their examples?  
I understand they are not necessary with the mesh files currently being provided, however having the .geo files available would be helpful when attempting to create a project from scratch.

[wedge\_2.geo](https://pyfr.discourse.group/uploads/short-url/pBQnfORVHYJQ3Bgh3gaKH96PpUg.geo) (738 Bytes)

[wedge\_2.msh](https://pyfr.discourse.group/uploads/short-url/8NvTJJLamRm6CM5hf7bFxrPJART.msh) (119 KB)

---

<div class="post-metadata">

**Author:** ![nnunn](https://avatars.discourse-cdn.com/v4/letter/n/c89c15/32.png) [@nnunn](https://pyfr.discourse.group/u/nnunn)\
**Post date:** [26 May 2018 02:03 UTC](https://pyfr.discourse.group/t/unpaired-faces-in-mesh/216/2 "2018-05-26T02:03:03Z")

</div>

Hi Andrew,

As you said, you need to handle both sets of wall type faces (wedge and channel) in your .ini input file.

If you can post your .ini file, we might be able to spot the problem, and adjust the .geo file.  
Nigel

---

<div class="post-metadata">

**Author:** ![Andrew\_Hedrick](https://avatars.discourse-cdn.com/v4/letter/a/5f8ce5/32.png) [@Andrew\_Hedrick](https://pyfr.discourse.group/u/Andrew_Hedrick)\
**Post date:** [30 May 2018 14:00 UTC](https://pyfr.discourse.group/t/unpaired-faces-in-mesh/216/3 "2018-05-30T14:00:38Z")

</div>

Attached is the ini file as requested. I am sure that the flow equations and parameters are not well optimized for the problem I am attempting. My approach was it iteratively improve, the first step of which was to get a working mesh.

[wedge\_2d.ini](https://pyfr.discourse.group/uploads/short-url/zhg8Toc9lIKgBPTakd3xOa5rl1J.ini) (1.8 KB)

---

<div class="post-metadata">

**Author:** ![nnunn](https://avatars.discourse-cdn.com/v4/letter/n/c89c15/32.png) [@nnunn](https://pyfr.discourse.group/u/nnunn)\
**Post date:** [30 May 2018 20:17 UTC](https://pyfr.discourse.group/t/unpaired-faces-in-mesh/216/4 "2018-05-30T20:17:28Z")

</div>

Hi Andrew,

To get started, try adding the 3 wedge faces to the set of “wall” boundary faces:

// Physical Line(“wall”) = {1, 3};  
// Physical Line(“wedge”) = {5, 6, 7};

Physical Line(“wall”) = {1, 3, 5, 6, 7};

In your wedge2d.ini file, all these faces will get handled by the section:

[soln-bcs-wall]  
type = no-slp-wall

If you want to apply a different BC to the wedge faces, you’ll need to add a section in the .ini file, e.g.

[soln-bcs-wedge]  
type = hedrick-wall

Nigel

---

<div class="post-metadata">

**Author:** ![Andrew\_Hedrick](https://avatars.discourse-cdn.com/v4/letter/a/5f8ce5/32.png) [@Andrew\_Hedrick](https://pyfr.discourse.group/u/Andrew_Hedrick)\
**Post date:** [6 June 2018 16:51 UTC](https://pyfr.discourse.group/t/unpaired-faces-in-mesh/216/5 "2018-06-06T16:51:45Z")

</div>

Thanks for your time Nigel,

You’re correction makes sense now that I see it. Implementing your change allows the mesh to import. PyFr runs with a new error that I haven’t triggered before, so I’ll work on that for a bit.
