Boundary surface tangency condition

Heh heh... I should know this, I was doing a search on this subject and I came up with old posts of mine advising people with a similar problem. I'm on 2001, trying to make a boundary surface with a tangency condition.

I have 5 sketched datum curves (1st dir), one after the other - each have a single spline. Their bottom end points lie on a datum plane (which is in edge view when I sketch them). The splines at these bottom end points have an angle dimension of 91.5 degrees to the edge of the datum plain. On this datum plane I sketch another curve (it becomes curve 1 in 2nd dir) which is another spline going through the

5 end points, plus I have 2 more to control the spline in other places. I use this curve in a subsequent sketch of a protrusion. I give this protrusion draft of 1.5 degrees on the surface which is made by the "use edge" of the sketched curve through the 5 end points.

You still with me? I make a boundary surface through the 5 curves in

1st dir and through 2 curves in the other; curve 1 being the sketched one through the 5 end points, curve 2 being a through points curve through the other 5 end points of the aforementioned curves.

So you guessed it, I am trying to give the surface tangency condition to the drafted surface along curve 1 2nd dir. I need it because the 5 angles on the 5 curves in 1st dir are not enough to hold it all the way along curve 1 2nd dir. It fails with the message"

"Bounding curve(s) are not tangent to tangent surface at highlighted point(s).."

It's referring to the first curve in the 1st dir, at the point where its aligned to curve 1 in the 2nd dir. Now I know the curve is tangent. It's 91.5 degrees and the draft is 1.5 and they're in planes that are normal... 91.5 - 90 = 1.5.

The end point of the curve also lies right on the vertex of the protrusion that is drafted. The draft is only on the one surface that I want tangent to the boundary surface though. There is no draft on the adjacent surface of the protrusion, the one that is planar to the surface that curve 1 is on. I have tried it with and without draft on this surface and it doesn't make any difference. I assume it's something to do with the fact that the curve in question is right on the edge of the drafted surface, but I have extended the surface, offset the protrusion, swept a surface instead of using the draft, but nothing works. My current workaround is to make the surface in 3 segments with the middle bit having tangency where I want it and the forward and backward bits without. This gets me to a quilt with 1.5 (more than in places)degrees along this edge but I would rather do it in one surface because I know this will complicate things later on.

Anyone got any ideas?

Reply to
Gra-gra
Loading thread data ...

A shot in the dark without seeing the file, but will it work better if you will first create the protrusion and draft, then come back and create your five curves (not necessarily planar) defining tangency with the surface or intersection curves laid on the surface? What I think might be happening is that the 1.5 deg draft isn't actually 1.5 degrees when measured in a plane that's normal to the edge at some location where your curves intersect it and that's the way the blend function is trying to look at it (just a guess, don't really know). Maybe playing with some of the blend options (sorry, I really don't know what's available in versions prior to WF) would help. Maybe try the blend skipping different first direction curves will indicate where the problem is? Create the blend without tangency, trim it back and create a second blend using just the edges and tangency or curvature definition? You can then go back and intersect the quilt to create new curves and create a new, single surface if desired. Five curves in one direction is a lotta curves for a single surface. Can you meet intent with fewer? Is the shape complex enough to warrant creating as multiple blended surfaces? Sorry, just ramblin' on...... 8~)

=================================

Reply to
Jeff Howard

Yep, you're probably right. I haven't done it that way but I might have to.

The other blend options are for making a more approximate version of the surface one way or another. I get what you mean about the surface angle though.

Tried that. I can't leave out the first one though and that's the one that has the problem.

Sounds messy, but I might give it a try.

I started with fewer and built it up, because it wasn't giving me the surface I wanted.

Is the shape complex enough to warrant creating as multiple blended

That was my workaround, and it's probably what I'll do. It's the best result I have achieved so far.

Ramble away, I appreciate your reply.

Reply to
Gra-gra

"Gra-gra" wrote in message news: snipped-for-privacy@posting.google.com... : Heh heh... I should know this, I was doing a search on this subject : and I came up with old posts of mine advising people with a similar : problem. I'm on 2001, trying to make a boundary surface with a : tangency condition. : : I have 5 sketched datum curves (1st dir), one after the other - each : have a single spline. Their bottom end points lie on a datum plane : (which is in edge view when I sketch them). The splines at these : bottom end points have an angle dimension of 91.5 degrees to the edge : of the datum plain. On this datum plane I sketch another curve (it : becomes curve 1 in 2nd dir) which is another spline going through the : 5 end points, plus I have 2 more to control the spline in other : places. I use this curve in a subsequent sketch of a protrusion. I : give this protrusion draft of 1.5 degrees on the surface which is made : by the "use edge" of the sketched curve through the 5 end points. : : You still with me? I make a boundary surface through the 5 curves in : 1st dir and through 2 curves in the other; curve 1 being the sketched : one through the 5 end points, curve 2 being a through points curve : through the other 5 end points of the aforementioned curves. : I read this through about 3 more times and think I finally understand what you're trying to do. A good night's sleep probably helped, too. See if this summary sounds right. You have five sketched, second direction curves of a boundary blend surface that you want to make tangent to a an underlying, drafted protrusion. Since the draft is 1.5 degrees and spline end is effectively the same, it should be possible to make them tangent.

My first question, assuming I've understood you correctly, is where does the protusion and draft appear in the model tree, before or after the the curves and surface. If before, you wouldn't have needed the 91.5 dimension in sketcher. You would have used a sketcher constraint to make the spline tangent to the drafted surface. Five curves constrained in this way would have made the tangency almost a given, even without going into the tangency option of the boundary blend interface. I also question if it's possible to create tangency to something that didn't exist when the curves were created, that is, that drafted protrusion. The thing that starts to occur to you, after working with enough boundary blend surfaces, is that their tangency conditions can't be any higher than the condition of their underlying curves. The biggest problem with sketched curves, as opposed to 'thru point' curves, is that they don't have to be tangent to anything. And even if they are tangent, the degree of tangency is still only C1 ~ simple tangency. You can't make a sketched curve C2 ~ curvature continuous ~ with another surface because of something Jeff mentioned: these curves are planar. So, as he was pointing out, these curves may not even be at the dimensioned angle, in relation to the drafted surface, because that drafted curvy surface is not planar and it's angle is not measured from the plane you sketched your curve on. 'Thru points' curves, on the other hand, have no such limitation and can be set to follow the curvature of an adjacent surface, even solid surfaces.

: So you guessed it, I am trying to give the surface tangency condition : to the drafted surface along curve 1 2nd dir. I need it because the 5 : angles on the 5 curves in 1st dir are not enough to hold it all the : way along curve 1 2nd dir.

If you like, I could suggest a technique that would accomplish what you are trying to do.

David Janes

Reply to
David Janes

There's lots here, and I'm not sure if I fully understand it, but there is a way to trouble-shoot your model.

Even though you have 5 curves in the 1st direction, test the model with a surface through just 2 curves in the 1st direction, and see if the tangency works. It doesn't even have to be the 2 outer curves. It could be any two curves (and then of course shorten your 2nd direction curves accordingly), and then test the tangency.

You can pick the 1st and 5th curve, 2nd and 4th curve, 3rd and 5th curve, 3rd and 1st curve, whichever combination. Just test each curve in a simple 2 curves in each direction surface, and test the tangency. When you know 2 curves are good, add a 3rd curve and see if it still works. Then add the 4th and see if it still works. Then add the 5th and see if it still works.

Most likely, one of your curves is tweaking the surface such that tangency conditions won't apply.

One other alternative:

When creating your surface, instead of using the tan-colored datum curves, use the edges of the protrusion or yellow surface edges instead. When you query select, datum curves usually highlight red, solid edges or surface edges highlight blue. Use the blue instead. Sometimes, these help tangency conditions work where it may fail with datum curves.

I discovered that one by accident when I got stuck on one project.

Hope this helps, and good luck.

Lee

Reply to
limacharliewhiskey

Should be....

5 curves, then curve in the other dir (through the other 5 endpoints), then protrusion (with a use edge to other curve), then draft, then surface. Can't easily order this any differently.

If before, you wouldn't have needed the 91.5 dimension in sketcher. You

I think that's it.

'Thru

Please do. This is a couple of days old now and I've struggled on, but maybe I can make what I've got better, or maybe I can use this info in the future.

Reply to
Gra-gra

: > If you like, I could suggest a technique that would accomplish what you are trying : > to do. : : Please do. This is a couple of days old now and I've struggled on, but : maybe I can make what I've got better, or maybe I can use this info in : the future. : The fundemental feature and the thing causing you the most trouble is that curve through the end points of the cross curve. It is unfortunately dependent on those curves and even makes a drafted solid dependent on them. But, it's clear from your description that you know where each of those end points is without creating the cross curves first. So, the best thing would be to create a series of datum points where you know those end points to be (offset distance of sketch plane, dimensioned distance of sketched spline end point). Create them on the datum plane the sketched splines were locked to. Starting this way has the advantage of allowing you to create your base curve independent of the cross curves, either as thru points or sketched, using datum points as sketcher references. With this curve created, you are now free to use it to create your solid and draft it.

These points can also be used for creating the cross curves, either as thru point (prefereable method because of highest tangency condition possible) or as sketched curves, using points for creating the sketch plane and as a sketcher reference for again anchoring the spline end points. Even in this case, you should be able to use the highlight edge of the solid for creating a tangent constraint to the drafted surface below. If it doesn't like the highlight edge for tangency, create first an axis thru the point and tangent to the surface. Construction geometry can also help with this. With tangency established on all of the curves, it should be possible to achieve it on the surface built with those curves. It bears repeating that the higher the quality of the curves, the better the quality of the surfaces.

I know this will seem like a lot more work and you may not be able to simply reorder things in your previous work to do it this way, but getting good at creating the datum rigging for boundary blend surfaces is about 90% of the battle. When it comes out right, without the backtracking and frustration, it won't seem like so much work.

Reply to
David Janes

Thanks for that. Yes, it comes down to having my features in a different order and using through points curves rather than sketched ones.

Reply to
Gra-gra

I always knew which one it was, the 1st curve in the 1st dir.

Thanks, I tried that one, but alas....

Reply to
Gra-gra

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.