geodesics

Only in special cases.

One problem is that the solution can have more than one answer on a curved surface. SW doesn't like ambiguous.

snipped-for-privacy@hotmail.com wrote:

I'll give it an unqualified "I think" it can be done given the right set of circumstances. Perhaps you can be a little more specific? Your use of the word "goedesics" conjures up images, but.......

Read some of Einstein's works for a simple explanation.

On a sketch plane a line between two points is a geodesic.

On a sphere between two diametrically opposite points there are an infinite number of geodesics.

Jeff Howard wrote:

I have a complex surface (not a simple sphere) and two points on the surface. I'd like to connect them with a curve that lies on that surface so that the length of the curve is the shortest possible. A curve thru reference points simply draws a line from A to B but not on the surface. I know that projection may be needed but don't know how to assure myself its the shortest curve length. Although there can be an infinite number of solutions for some cases, I believe that in my cases there is only one. How/Can I find it without a lot of manipulation of the model?

Thanks MT

Back to my first response. Not with SW. If you have a mathematical definition of the surface you might find it.

I honestly don't know anything about the subject (just find it to be an interesting question). You might try doing a web search. I did a quick search (their server is very slow this morning and I'm a little pressed for time right now) of Rhino's site and found

looked like it was promising, but don't have time to follow up on it.

... the comment on mesh alogrithms was interesting.

Good luck with it.

For a sphere, if you create a sketch plane using the two points on the surface and the center of the sphere, the intersection curve of the sphere and sketch plane will be the shortest curve between those two points. This will not necessarily be true for a non-spherical surface.

What kind of precision would be required of your final curve?

Maybe I dont totally understand here.

If you have a noplanar surface with alot of topology and you are trying create a curve that follows the contour of the surface?

Heres what I would do

1. Create the intial surface
2. Create a new sketch with the location of the 2 points and draw a line connecting them
3. Extrude the line so it intersects the first surface
4. using your sketch tools, use the intersection curve to create the curve with the intersection of the 2 surfaces
5. You will have a 3d sketch that contains it.

If I am totally offbase here, disregard this message ;-)

FWIW,

Mike Lamora

Brian wrote:

Shortest distance between two points while staying on the surface. It is not necessarily a straight line projected.

What's wrong with the "sketch on surface"? If you draw a 2 point spline on a surface, it should connect the points with the shortest path. It seems like the simplest solution, or have I missed something?

matt

Mike Lamora wrote in news:3aj5c9F5rsa4iU1 @individual.net:

matt wrote in news:Xns96249B42B8DD1xxx@199.45.49.11:

Just to clarify, in 3D sketch there is a function called Spline on Surface. This is new with sw05.

It appears to create a curve that takes the shortest path sometimes. It is subject to singularities or forbbiden zones and when near them will act funny. I will try to post some of the pathology later.

Haven't seen the capability in Solidworks for geodesoics that say Catia has, in Catia you can even make a constant offset to a curve on a surface. as well as the shortest connection between 2 points on a surface.