| Author |
Message |
Bullman
Guest
|
Posted:
Fri Nov 25, 2005 9:10 am Post subject:
Filleting 3 line 3D sketch problem |
|
|
Hello
I have been trying to find the right technique for achieving the
following.
You have two random straight line ends in 3D space. The end points are
fixed and can not (should not) be moved. You want to link the two line
ends up using the following:
- two tangent arc elements (one for each line end)
- one straight line (to join the free ends of the two tangent arcs
together and complete the smooth tangential chain)
This can not be done using 3D fillet command as the end points of the
lines will be shortened and the original location for the line end
points will now become the location of the "virtual sharps".
I am finding that manipulating the 3D tangent arcs to be difficult.
They have three degrees of freedom.
1. Rotation angle about line end
2. Radius
3. Arc length (or arc angle)
SW is not able to automatically solve the sketch if you try and rely on
adding relations only (even without any dimensions being defined).
Perhaps you need to more closely geometrically approximate the answer or
perhaps there are just too many variables that it can't solve them all
at once.
Is there a fail safe efficient technique for ensuring a solution?
Bullman
--
Internal Virus Database is out-of-date.
Checked by AVG Free Edition.
Version: 7.1.361 / Virus Database: 266.0.0/162 - Release Date: 5/11/2005
|
|
| Back to top |
|
 |
LinSD
Guest
|
Posted:
Fri Nov 25, 2005 1:10 pm Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
It's impossible to have a 3D fillet feature, I think you can connect
this kind of 2 space line segment by 3D spline. after draw the spline
on any plane, you can add tangent relation to found a "3D fillet".
You need to use other spline tools to fully define it, to make it looks
like a fillet. |
|
| Back to top |
|
 |
That70sTick
Guest
|
Posted:
Fri Nov 25, 2005 9:10 pm Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
Got an idea. Workng on it. Be back in a day or two.
|
|
| Back to top |
|
 |
matt
Guest
|
Posted:
Sat Nov 26, 2005 8:27 am Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
What I would do would be to turn the original lines into construction
geometry, draw new lines and arcs in space disconnected from the
original lines, and then once the tangent relations are working and the
shape is approximately correct, drag the new lines/arcs onto the old
construction lines.
Or, just draw the line connecting the endpoints, make 2 reference points
at random locations, then key in coordinates to put them at the
endpoints of the lines, then fillet the corners and drag the moved
endpoints back to the reference points.
Another thing you could try would be to leave the original lines alone,
draw the 3rd line, build sketch planes using the 2 endpoints of the 3rd
line and the end point of an original line that the arc is to hook up
to, and then you have planes for the arcs. Of course the new sketch
plane function in 2006 is pretty fragile, and pretty likely to fail.
Or you could just say "screw it" and put a couple tangent splines in
instead of arcs. You'll solve one set of problems and create another
set.
There's not much in a 3D sketch that is what you call "fail safe", and
because you're usually working with workarounds, nothing's really very
efficient either.
good luck,
matt
In article <43868314$0$12444$5a62ac22@per-qv1-newsreader-
01.iinet.net.au>, "Bullman" <bullman<NOSPAM>@iinet.net.au> says...
| Quote: | Hello
I have been trying to find the right technique for achieving the
following.
You have two random straight line ends in 3D space. The end points are
fixed and can not (should not) be moved. You want to link the two line
ends up using the following:
- two tangent arc elements (one for each line end)
- one straight line (to join the free ends of the two tangent arcs
together and complete the smooth tangential chain)
This can not be done using 3D fillet command as the end points of the
lines will be shortened and the original location for the line end
points will now become the location of the "virtual sharps".
I am finding that manipulating the 3D tangent arcs to be difficult.
They have three degrees of freedom.
1. Rotation angle about line end
2. Radius
3. Arc length (or arc angle)
SW is not able to automatically solve the sketch if you try and rely on
adding relations only (even without any dimensions being defined).
Perhaps you need to more closely geometrically approximate the answer or
perhaps there are just too many variables that it can't solve them all
at once.
Is there a fail safe efficient technique for ensuring a solution?
Bullman
--
Internal Virus Database is out-of-date.
Checked by AVG Free Edition.
Version: 7.1.361 / Virus Database: 266.0.0/162 - Release Date: 5/11/2005
|
|
|
| Back to top |
|
 |
TOP
Guest
|
Posted:
Sat Nov 26, 2005 9:10 am Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
| Yes, if you don't care too much about the fillet radii. |
|
| Back to top |
|
 |
TOP
Guest
|
Posted:
Sat Nov 26, 2005 9:10 am Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
| Yes, but don't always count on the fillets haveing the same radius. |
|
| Back to top |
|
 |
TOP
Guest
|
Posted:
Sat Nov 26, 2005 9:10 pm Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
Yes, but......you can't always get the fillet radii you want.....it
doesn't work when geometrically impossible.....and SW gets a bit quirky
even when it is geometrically possible. |
|
| Back to top |
|
 |
TOP
Guest
|
Posted:
Sat Nov 26, 2005 9:10 pm Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
Yes, at least from a geometry standpoint it works except for two
special cases one of which you probably will never want to try.
However, SW gets quirky when implementing this.
Bullman wrote:
| Quote: |
Is there a fail safe efficient technique for ensuring a solution?
Bullman |
|
|
| Back to top |
|
 |
That70sTick
Guest
|
Posted:
Sun Nov 27, 2005 1:10 am Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
<http://www.esoxrepublic.com/models/>
Look at model "SkewFillet.SLDPRT"
First sketch is two skew (non-intersecting) lines.
Second sketch connects skew lines w/ fillets and line in middle.
Key elements: Sketch points coincident with original lines and middle
line keep middle line intersecting the two skew lines so a fillet can
be applied. Fillets were created by creating temporary construction
lines collinear with original lines. |
|
| Back to top |
|
 |
matt
Guest
|
Posted:
Sun Nov 27, 2005 1:10 am Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
In article <1133036342.028653.239730@g44g2000cwa.googlegroups.com>,
kellnerp@cbd.net says...
| Quote: | Yes, but......you can't always get the fillet radii you want.....it
doesn't work when geometrically impossible.....and SW gets a bit quirky
even when it is geometrically possible.
|
Paul,
Was that "turkey" or "Wild Turkey" that you were getting into? I see
that you're responding, but what are you responding to?
Matt |
|
| Back to top |
|
 |
TOP
Guest
|
Posted:
Sun Nov 27, 2005 1:10 am Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
I tried to respond when the post first came out. Nothing happened so
posted again. Then today they all appeared at once after That 70s Tick
responded. I was responding to Bullman's post. |
|
| Back to top |
|
 |
TOP
Guest
|
Posted:
Sun Nov 27, 2005 8:04 am Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
Similar to the construction I had. But.....just like mine it doesn't
like having the first sketch changed. I doesn't want to solve. I used
2D planes in 3Dsketch to hold down the 3 point arcs. You can put all
the geometry in one sketch. It still doesn't update very well. |
|
| Back to top |
|
 |
That70sTick
Guest
|
Posted:
Sun Nov 27, 2005 9:10 am Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
Suppose I should post the file location:
<http://www.esoxrepublic.com/models/>
SkewFillet2.SLDPRT |
|
| Back to top |
|
 |
TOP
Guest
|
Posted:
Sun Nov 27, 2005 9:10 am Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
One thing is apparent. The constraint solver in 3D Sketch doesn't work
very well, even in 2006. I have used the second method when reverse
engineering imported CAD geometry. Sticking with planes and 2D sketches
is pretty robust. |
|
| Back to top |
|
 |
That70sTick
Guest
|
Posted:
Sun Nov 27, 2005 9:10 am Post subject:
Re: Filleting 3 line 3D sketch problem |
|
|
Another stab. More complex, yet more robust, yet more complex. Pairs
of dimensions in "Midline" sketch need to match to produce dependent
fillet of equal length. |
|
| Back to top |
|
 |
|
|
|
|