2003-march.txt

acis说明文档

Hi all,

how can I get the intersection curve of two bodies? api_slice does not
work for us (as I posted 2 days ago). Is there another way to get the
intersection curve?

thanks a lot in advance,
Hannes

From [e-mail]  Fri Mar 14 17:32:36 2003
From: [e-mail] (Ames, Arlo)
Date: Fri, 14 Mar 2003 10:32:36 -0700
Subject: [Acis-alliance] intersection curve
Message-ID: <[e-mail]>

I tried an approximation of what your code does in the test harness and it
works.  Have you tried to do what you're attempting in either the test
harness or the scheme interface?  I can imagine a wide variety of geometric
or programming effects that can result in you getting no answer (a lot of
ACIS programming problems involve issues like forgetting api_start_modeller;
api_slice doesn't change models if they don't actually overlap...).  Your
code looks OK but I haven't the time to compile it to check.

As to other approaches, first verify that things are really broken.  Any
other approach to getting an intersection curve relies on the same
underlying technology (surface/surface intersectors, etc), so api_imprint,
api_slice, api_boolean, merely give you variations on the same answer, and
if one is broken they probably all are.

Arlo Ames
[e-mail]


-----Original Message-----
From: Hannes Kaufmann [mailto:[e-mail]]
Sent: Friday, March 14, 2003 10:05 AM
To: ACIS List
Subject: [Acis-alliance] intersection curve


Hi all,

how can I get the intersection curve of two bodies? api_slice does not
work for us (as I posted 2 days ago). Is there another way to get the
intersection curve?

thanks a lot in advance,
Hannes
_______________________________________________
Acis-alliance mailing list
[e-mail]
http://lists.cs.columbia.edu/mailman/listinfo/acis-alliance


From [e-mail]  Mon Mar 17 14:50:35 2003
From: [e-mail] (Marius Verkauskas)
Date: Mon, 17 Mar 2003 16:50:35 +0200
Subject: [Acis-alliance] Autocad 2002 ACIS
Message-ID: <000c01c2ec94$912967c0$[e-mail]>

Hello,

Can someone tell me which version of ACIS is compatible with AutoCad
2002?


From [e-mail]  Wed Mar 19 09:39:03 2003
From: [e-mail] (Hannes Kaufmann)
Date: Wed, 19 Mar 2003 10:39:03 +0100
Subject: [Acis-alliance] "clipping" an intersection graph
Message-ID: <[e-mail]>

Dear all,

we have a problem with the intersection graph of 2 surfaces. The
intersection graph always consists of the full curves e.g. when
intersecting a box with a sphere we are getting 6 full circles instead
of the circle arcs that the box "cuts out" of the sphere. What we want
is to "clip" the intersection graph with the box. 
Is there an api function to do this clipping, to split and cut
curves/intersection graphs with BODYs.

Thank you very much in advance,
Hannes

-- 
    Hannes Kaufmann | Vienna University of Technology
         Researcher | Interactive Media Systems Group
T [phone] | F [phone]
           kaufmann @ ims.tuwien.ac.at

From [e-mail]  Wed Mar 19 16:17:43 2003
From: [e-mail] (Ames, Arlo)
Date: Wed, 19 Mar 2003 09:17:43 -0700
Subject: [Acis-alliance] "clipping" an intersection graph
Message-ID: <[e-mail]>

Hannes,
How are you computing the intersection graph?  Generally api_slice or
api_imprint produce trimmed curves, not the complete curves you describe.
Directly calling the surface/surface intersectors can produce full,
unlimited curves, but there aren't too many other ways of doing that.  It IS
possible to get unwanted full curves if your geometry exhibits large
inaccuracy (i.e. the edges are sufficiently far from the surface that they
don't intersect the imprint curves).

If you can't find a way to correct the construction process (which I would
generally prefer), you can probably try to imprint/slice/whatever the wire
body of the intersection graph with the box's body.  I suspect your answer
won't necessarily change with such an approach, though, as the same
underlying intersectors will be called unless you did something strange in
construction.

And now I have another question:  is the answer you're getting really wrong?
I'd like to see your geometry, but there are a lot of sphere/box
intersections where the full circles are absolutely correct.

In test harness (forgive me for being old fashioned), I did the following:
block a w 10 d 10 h 10
sphere b1 r 5
sphere b2 r 6
sphere b3 r 7.0710678118654752440084436210485
sphere b4 r 8
sphere b5 r 10
slice a with b as c
draw a

the slice of a and b1 is a grazing case -- the imprint consists of
degenerate vertex loops.  Each degenerate loop points to a single coedge,
which points to an edge which has NO curve, and starts/ends at the
degenerate vertex.

The slice of a with b2 gives you 6 complete circles, and is correct.

The slice of a with b3 has the sphere just grazing each edge of the block.
You get more coedges, as each complete circle intersects 4 others, so you
expect 24 coedges.

The slice of a with b4 produces 6 triangles of circular arcs.

The slice of a with b5 could theoretically produce a set of degenerate
vertex loops on the sphere, but instead reports no intersection -- an answer
somewhat lacking in symmetry, but nonetheless reasonable.  ACIS performs
intersections starting with surfaces, moving to edges, and limiting with
vertices.  Vertex/surface interactions are not generally considered
important for imprinting, as they don't produce anything interesting, by
themselves, in boolean terms.  Since the slicing functionality was derived
from boolean functionality, this isn't at all surprising.  I generally check
vertices myself if they're important.



-----Original Message-----
From: Hannes Kaufmann [mailto:[e-mail]]
Sent: Wednesday, March 19, 2003 2:39 AM
To: ACIS List
Subject: [Acis-alliance] "clipping" an intersection graph


Dear all,

we have a problem with the intersection graph of 2 surfaces. The
intersection graph always consists of the full curves e.g. when
intersecting a box with a sphere we are getting 6 full circles instead
of the circle arcs that the box "cuts out" of the sphere. What we want
is to "clip" the intersection graph with the box. 
Is there an api function to do this clipping, to split and cut
curves/intersection graphs with BODYs.

Thank you very much in advance,
Hannes

-- 
    Hannes Kaufmann | Vienna University of Technology
         Researcher | Interactive Media Systems Group
T [phone] | F [phone]
           kaufmann @ ims.tuwien.ac.at
_______________________________________________
Acis-alliance mailing list
[e-mail]
http://lists.cs.columbia.edu/mailman/listinfo/acis-alliance


From [e-mail]  Tue Mar 25 15:09:13 2003
From: [e-mail] (Alexander Muthler)
Date: Tue, 25 Mar 2003 16:09:13 +0100
Subject: [Acis-alliance] thicken a parametric free-form sheet
Message-ID: <[e-mail]>

Dear Alliance,

I want to thicken a sheet body with a specific thickness. 
The sheet body is of type spline. Do you know a method to do that quickly? 

Your help would be appreciated.

Best Regards,

Alex

From [e-mail]  Tue Mar 25 15:48:56 2003
From: [e-mail] (Harigopal Polisetty)
Date: Tue, 25 Mar 2003 10:48:56 -0500
Subject: [Acis-alliance] thicken a parametric free-form sheet
In-Reply-To: <[e-mail]>
Message-ID: <l03130300baa62a5ea019@[192.168.1.141]>

api_sheet_thicken()


At 10:09 AM -0500 3/25/03, Alexander Muthler wrote:
>Dear Alliance,
>
>I want to thicken a sheet body with a specific thickness.
>The sheet body is of type spline. Do you know a method to do that quickly?
>
>Your help would be appreciated.
>
>Best Regards,
>
>Alex
>_______________________________________________
>Acis-alliance mailing list
>[e-mail]
>http://lists.cs.columbia.edu/mailman/listinfo/acis-alliance




From [e-mail]  Thu Mar 27 11:14:26 2003
From: [e-mail] (lzguo)
Date: Thu, 27 Mar 2003 19:14:26 +0800
Subject: [Acis-alliance] How to save *.sat to *.igs with InterOp
Message-ID: <001801c2f452$0781adf0$af501e0a@t6>

This is a multi-part message in MIME format.

------=_NextPart_000_0015_01C2F495.158F9130
Content-Type: text/plain;
	charset="gb2312"
Content-Transfer-Encoding: base64
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------=_NextPart_000_0015_01C2F495.158F9130
Content-Type: text/html