boundary file interpolation looks off on GEOM terrain

[ericvmueller]

Run the LS_ellipse_0ms_30deg.fds case and pull up the IGNITION TIME boundary file in smokeview. Clip the upper bound to something close to T_END (125 s).

It looks like the interpolation to the nodes is flipped in each patch or something

[mcgratta]

Is this a Smokeview or FDS issue. Can you assign it.

[ericvmueller]

Ah sorry, Glenn asked me to post it here so I will assign him for now.

Edit: or maybe I wont... I don't seem to have that power on this repo

[gforney]

I assigned myself

[gforney]

marcos, are geometry boundary files node centered or cell centered?

[gforney]

smokeview is drawing geometry boundary files as if they are "cell" centered - that is coloring each triangle according to the value passed for that triangle

[ericvmueller]

Is that the expected behavior, then? I'm not so familiar with GEOM boundary outputs ...and equally likely the appearance of an inverted color gradient with the cells is just my bad eyesight

[gforney]

smokeview is drawing the same color for each vertex in a triangle. I assume because fds is passing data to smokeview by triangle, one value for each triangle. marcos is there an option to pass values to smokeview by vertex?

[marcosvanella]

cell_centered is set to true always in DUMP_BNDF. Not clear why as we have the option to do VERTEX writing in DUMP_SLICE_GEOM_DATA. Take a look in there in the code section:

! these two routines need to be moved and called only once
      CALL GET_GEOMINFO(SLICETYPE,I1,I2,J1,J2,K1,K2,NVERTS,NVERTS_CUTCELLS,NFACES,NFACES_CUTCELLS,VERTS,FACES,LOCATIONS)
      CALL MAKE_UNIQUE_VERT_ARRAY(VERTS, VERT_UNIQUE, NVERTS)
      CALL AVERAGE_FACE_VALUES(VERT_UNIQUE, VERT_VALS, NVERTS, FACES, VALS, NFACES)

[gforney]

@ericvmueller is this still and issue?

[ericvmueller]

@marcosvanella and I were discussing whether we should allow CELL_CENTERED=F for GEOM boundary files. If we want to implement that maybe lets leave this open until its sorted out.