firemodels / smv

Smokeview
Other
44 stars 172 forks source link

Boundary values not being computed correctly #1928

Open rmcdermo opened 2 months ago

rmcdermo commented 2 months ago

If you run the attached case, you will see that the _devc values of HRRPUA are being written. But when you look at the case in Smokeview, the boundary values report 0.

Archive.zip

Note that @johodges reported this worked correctly in an older version of Smokeview, circa 6.8.0

When I go in and reset the bounds, things look reasonable, see below. But notice that the computed bounds (see background terminal) are 0.

Screenshot 2024-06-06 at 9 37 24 AM

gforney commented 2 months ago

I'll take a look

gforney commented 2 months ago

I reproduced the problem. FDS is creating the .bnd file ok. the .bnd file has 0.0 and 500.0 as min and max before smokeview opens the case. smokeview overwrites the file for some reason - with the wrong bounds.

gforney commented 2 months ago

should be fixed. there is a bug in my scripts that upload test bundles to github .
for now linux and osx test bundles are here https://github.com/firemodels/test_bundles/releases/tag/SMOKEVIEW_TEST windows test bundle is in the right place at https://github.com/gforney/test_bundles/releases/tag/SMOKEVIEW_TEST

after the bug is fixed all the smokeview test bundles will go here https://github.com/gforney/test_bundles/releases/tag/SMOKEVIEW_TEST

drjfloyd commented 1 week ago

Now it is not showing X and Z, but two X. There is a change in mesh resolution which seems to have smokeview consider this as two different slice files. I think the menu should only show one entry. For cases with multiple mesh resolutions it is a bit of a pain to have to do multiple slice loads. I think we should just group these all by the global slice index and for the location I'd suggest the location from the smallest mesh.

image

There is also a gap in the slice file. I am not sure if this is an FDS issue or Smokeview. The coarse mesh should be the first row of grid celss above the finer meshes as it is just on &SLCF input.

smvtest_0003

&HEAD CHID='smvtest'/ &TIME T_END=0.1/ &RADI RADIATION=F/

&MESH ID='4-in', IJK=48,72,32, XB=-7.7216,-2.8448,-10.9728,-3.6576,-2.2352,1.016, MULT_ID='M1'/ &MESH ID='Top-8', IJK=68,46,20, XB=-11.7856,2.032,-15.0368,-5.6896,10.7696,14.8336, MULT_ID='M15'/

&MULT ID='M1', DX=4.8768, I_UPPER=1, DY=7.3152, J_UPPER=2, DZ=3.2512, K_UPPER=3/ &MULT ID='M15', DX=13.8176, I_UPPER=0, DY=9.3472, J_UPPER=2, DZ=4.0640, K_UPPER=0/

&SLCF QUANTITY='VELOCITY',CELL_CENTERED=T,XB=1.1684,1.1684,-10.9728,18.2880,-2.2352,10.78,ID='V CHR'/

gforney commented 1 week ago

update your smv repo and rebuild smokeview - this should fix the menu issue. we need to pay attention to slice menus in other cases to make sure a game of whackamole hasn't started

I'm investigating the gap issue

gforney commented 1 week ago

think the K indices for the 1-D slice output to the slice file should be 0 0 not 1 1 . smokeview starts numbering at 0 not 1. that is 64 64 0 46 0 0

not

64 64 0 46 1 1

gforney commented 1 week ago

update your smv repo and try again

gforney commented 1 week ago

is this resolved?