3D Modeling

[sawarsi]

I have a couple of questions regarding spur EDGEs and non-
manifold BODYs. When modeling a fluid domain, some of my
customers want to embed baffle surfaces into an otherwise
manifold BODY--in order to simplify modeling and analysis.
Per the ACIS documentation on FACE sidedness, sense, and
containment, these FACEs would be "double-sided, both-
inside". This is all clear to me. However, I'm a little
unclear about how these FACEs might interact with other
FACEs in the BODY. In the first figure, the bottom surface
represents one of the bounding FACEs of a hypothetical solid
BODY. The smaller FACE is a double-sided face embedded in
the BODY's interior. It terminates exactly on the interior
of the larger FACE. My understanding is that the EDGE at the
intersection would be used 3 times in the BODY. The first
usage is (via a COEDGE) by the smaller FACE. The second and
third are by the larger FACE. In particular, within this
FACE, the EDGE is a spur EDGE with 2 COEDGEs (running in
opposite directions) defining a hole LOOP with no area. Is
my understanding correct? Could such configurations cause
any trouble for ACIS modeling operations, Interop output,
etc? Note that this EDGE sharing is particularly important
to me because it helps my mesher ensure continuity between
the meshes on the 2 FACEs (see second figure).

I can also envision situations in which the user may wish to
embed spur EDGEs in the interior of FACEs bounding a solid
BODY--such that the EDGEs don't touch any other LOOPs. This
essentially amounts to cutting 0-area holes in the FACEs.
This might be so that, during mesh generation, the mesh
points on the EDGE can be captured in the FACE mesh. Cutting
such holes may also be useful if the user knows that,
somewhere down the road, he might wish to sweep some baffles
from them. In general, is it a bad idea to cut 0-area holes
in FACEs? Is there an api or simple procedure (in C++) that
would allow me to, given a FACE and an EDGE(s) in its
interior, cut a 0-area hole(s) in the FACE. I know how to
manually construct the COEDGEs and LOOPs, but I'm hoping for
something to take care of these details for me as the actual
cases may be arbitrarily complex. Also, might any ACIS
modeling ops automatically remove such "holes", even against
my wishes?

Any insight will be appreciated.

[ybiyani]

Hello Saif,

Such bodies are valid configuration in ACIS and work well. The apis that will allow you to create scribe edges on a face are api_embed_wire_in_faces and api_scribe (this a new interface). http://doc.spatial.com/qref/ACIS/html/group__BOOLADVEMBED.html#g87f6f5d260cb55a6c5e7e3fb9adb4ab3.  ACIS will not create a face for such scribe edges (zero area face ).

I hope this helps.

Regards,

Yogesh

[sawarsi]

Hi Yogesh,

I'm still looking for a little clarification on one point
of this issue. In the earlier example, I showed a face (the
one on the bottom) with the following loops:

1) A simple periphery loop.
2) A hole loop consisting of 2 coedges (running in opposite
directions) built upon an edge. In other words, this loop
consists of a single spur edge that is not connected to any
other edge on the face. It is contained fully within the
bounds of the face and does not touch the periphery loop in
any way. Even though the loop cuts a hole in the face, the
hole has no area.

Is this a valid construct in ACIS? I ask this question for
2 reasons:

1) In our system, WIRES do not quite do what we want since
they can be associated only with SHELLs or BODYs. At the
time at which a user wants to embed an edge into a face, there
may not even be any SHELLs or BODYs. As we already know
whether or not the edge lies inside the face, we only need a
mechanism that allows us to define a parent-child relationship
between the face and the embedded edge. Since FACEs have no
pointer back to a WIRE, we need to do something else.

2) Constructing a zero-area hole loop gives us the information
we need, as it allows us to treat such embedded edges like any
other spur or prop edge during mesh generation and other
operations. I've actually created such cases in ACIS. When
faceting such faces or doing boolean ops on solids constructed
from them, ACIS seems to have no trouble whatsoever.

In your earlier response, it was not clear to me whether it
is OK to cut a zero-area hole in a FACE. I'd like to continue
with this approach, but I'm wondering whether this is going
to bite me in some other way.

Thanks for any insight,