SCOREC / core

parallel finite element unstructured meshes
Other
179 stars 63 forks source link

chef: geombc boundary element 'face type' string shouldn't be in the interior block tag #381

Open cwsmith opened 1 year ago

cwsmith commented 1 year ago

From @KennethEJansen in slack:

I had intended to only add face type to boundary elements (wedges and pyramids appearing on boundaries should have this in general) but it looks like my development leaked into interior elements who do not need this. Basically the difference is adding the word quadface. I don’t think paraview would see any difference in these inputs. PHASTA probably would not either.

grep -a ' : < ' good_case/4-procs_case/geombc.dat.1 > /tmp/goodGrep
grep -a ' : < ' run_case/4-procs_case/geombc.dat.1 > /tmp/testGrep
(base) kjansen@viz003: /users/cwsmith/toKenFeb1_BL_query $ diff /tmp/goodGrep /tmp/testGrep 
18c18
< connectivity interior linear pyramid  : < 121 > 6 5 1 5 3 3 5 
---
> connectivity interior linear pyramid quadface : < 121 > 6 5 1 5 3 3 5 

However, the other parts show a similar issue for interior wedges AND my intended/necessary development for boundary elements

(base) kjansen@viz003: /users/cwsmith/toKenFeb1_BL_query $ for i in {1..4}; do grep -a ' : < ' good_case/4-procs_case/geombc.dat.$i > /tmp/good.$i; done 
(base) kjansen@viz003: /users/cwsmith/toKenFeb1_BL_query $ for i in {1..4}; do grep -a ' : < ' run_case/4-procs_case/geombc.dat.$i > /tmp/test.$i; done 
(base) kjansen@viz003: /users/cwsmith/toKenFeb1_BL_query $ for i in {1..4}; do diff /tmp/good.$i /tmp/test.$i ; done

yields

18c18
< connectivity interior linear pyramid  : < 121 > 6 5 1 5 3 3 5 
---
> connectivity interior linear pyramid quadface : < 121 > 6 5 1 5 3 3 5 
18,19c18,19
< connectivity interior linear pyramid  : < 21 > 1 5 1 5 3 3 5 
< connectivity interior linear wedge  : < 3649 > 152 6 1 6 3 3 3 
---
> connectivity interior linear pyramid quadface : < 21 > 1 5 1 5 3 3 5 
> connectivity interior linear wedge triface : < 3649 > 152 6 1 6 3 3 3 
27,29c27,29
< connectivity boundary linear wedge  : < 1249 > 52 6 1 6 3 3 3 6 
< nbc codes linear wedge  : < 417 > 52 6 1 6 3 3 3 6 
< nbc values linear wedge  : < 2497 > 52 6 1 6 3 3 3 6 
---
> connectivity boundary linear wedge triface : < 1249 > 52 6 1 6 3 3 3 6 
> nbc codes linear wedge triface : < 417 > 52 6 1 6 3 3 3 6 
> nbc values linear wedge triface : < 2497 > 52 6 1 6 3 3 3 6 
18c18
< connectivity interior linear wedge  : < 3073 > 128 6 1 6 3 3 3 
---
> connectivity interior linear wedge triface : < 3073 > 128 6 1 6 3 3 3 
26,28c26,28
< connectivity boundary linear wedge  : < 1057 > 44 6 1 6 3 3 3 6 
< nbc codes linear wedge  : < 353 > 44 6 1 6 3 3 3 6 
< nbc values linear wedge  : < 2113 > 44 6 1 6 3 3 3 6 
---
> connectivity boundary linear wedge triface : < 1057 > 44 6 1 6 3 3 3 6 
> nbc codes linear wedge triface : < 353 > 44 6 1 6 3 3 3 6 
> nbc values linear wedge triface : < 2113 > 44 6 1 6 3 3 3 6 
18c18
< connectivity interior linear pyramid  : < 181 > 9 5 1 5 3 3 5 
---
> connectivity interior linear pyramid quadface : < 181 > 9 5 1 5 3 3 5 

To be clear, PHASTA tolerates the new inputs but, if I recall correctly, I lacked the C++ skills to keep the new information that I wanted to be associated with boundary elements ( lines with the word boundary AND the lines with nbc codes and nbc values) without also getting that same adjustment for lines for interior element headers (lines with interior).