You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Instead of specifying geometries "as fabricated", it should be doable to "emulate" process fabrication. This is not as good as physical process simulation, but can be appealing.
substrate
GMSH box
size defines process domain in x, y, z
masked conformal growth/deposition of thickness t
create a new entity by occ.fuse all existing entities
masking: occ.cut vertically across the domain where the mask covers the substrate
apply occ.dilate on the entity by the desired thickness (isotropically a=b=c=t or directionally e.g. a=b=0, c=t; other types of growth at other angles also possible)
if isotropic, occ.fillet the dilated entity's curves by radii = t; this will maintain a distance of t from the previous surface everywhere
occ.cut by the original occ.fuse: this also allows tagging of the new grown layer
masked etching
full overetch hack: occ.cut the tag of a grown layer by an infinitely tall box (or other shape if isotropic or non-vertical anisotropy) with x-y dimensions matching mask opening
otherwise can occ.cut with a grown gas layer (complement of current fuse with process domain) where exposed
alternate depositions and etches as needed
cleanup
occ.cut outside the process domain in x,y
The text was updated successfully, but these errors were encountered:
Instead of specifying geometries "as fabricated", it should be doable to "emulate" process fabrication. This is not as good as physical process simulation, but can be appealing.
occ.fuse
all existing entitiesocc.cut
vertically across the domain where the mask covers the substrateocc.dilate
on the entity by the desired thickness (isotropicallya=b=c=t
or directionally e.g.a=b=0
,c=t
; other types of growth at other angles also possible)occ.fillet
the dilated entity's curves byradii = t
; this will maintain a distance of t from the previous surface everywhereocc.cut
by the originalocc.fuse
: this also allows tagging of the new grown layerocc.cut
the tag of a grown layer by an infinitely tall box (or other shape if isotropic or non-vertical anisotropy) with x-y dimensions matching mask openingocc.cut
with a grown gas layer (complement of current fuse with process domain) where exposedocc.cut
outside the process domain in x,yThe text was updated successfully, but these errors were encountered: