Offsetting a spline limitations, wildfire 3.0 1

[marcpbcup]

Hello,

I am trying to create a normal direction offset to a 2-D spline created in a sketch. I followed the technique in the following thread: thread554-159791 ,
i.e. Create a sketch of the spline to be offset. Create a second sketch that uses the first sketch as a reference and use the "offset edge" command to create a spline that is offset in the normal direction to the spline.

The trouble with this method is that there seems to be some limitations on how much offset is allowed. I get the following when trying to offset by -0.5:
"Range is -1.4473e-01 - 1.0000e+09. Please re-enter:"

Does anyone know why it won't let me offset by my required amount? My spline is half of an airfoil (unclosed, no sharp geometry changes). Are there any good workarounds for this?

Much appreciated,

Marc

 

[antran7]

is it a really complicated spline with a lot of nodes? Whenever I use splines, I don't use the interpolation-point method. Rather I use the control points, and there's only 2 control points (and 2 end points) in each spline.

I suspect there might be a small radius somewhere in your spline, hence the offset limits.

 

[marcpbcup]

antran,

Thanks for the reply. Yes, it is a spline (probably technically a B-spline) with 160 closely spaced nodes. I tried to convert it to the control point method but Pro-E thinks for about 5 minutes before neglecting to preform the command without any warning or error messages. Is there a max limit on # of nodes that can be converted to control polygon method?

Another note- I think your small radius guess is correct. I tried the offset with a simpler shape of airfoil, still 160 nodes, and it worked perfectly, suggesting it is a geometry problem.

 

[antran7]

wow... 160 nodes... does it have to be that many? i'm assuming you've got 160 points in space that you have to go through, but could you approximate some?

if you want or need the clean single entity of a spline, another thing that might work is to build a curve using simplier geometry: like arcs, and simple 2-point splines. then do an approximate copy of that curve, which will splinify it into a single sketch entity. it won't absolutely go through your 160 nodes, but will be close.

and another trick i've done (coincidentally on an airfoil), is to offset the points, and then make the spline through those offset points. again, there's some approximation because you have to offset the points normal to undefined spline going through the original points. you could use your 160-node spline as the entity to be normal to. then generate another 160 nodes. pop a spline through that, and voila...

that might also help you understand why the original offset curve method didn't work.

 

[marcpbcup]

To close this thread-

I finally got it to work. Antran- thanks for all of your insight.

I got inspiration from the method you suggested, but instead of offsetting all of the points normal, I divided the spline into several sections (Divide an Entity command), some larger ones based on sections that have very small geometric changes and the offset points method near the leading and trailing edges. There was a lot of dynamic trimming, normal constraint using and connection issues, but doing it by hand allowed me to see why Pro/E couldn't do it either. The leading edge was a mess.

In order to extrude it, I ended up creating another sketch over it and putting a single 120-pt spline back around the entire foil. I don't know if there's a numerical way to check how much my new, shrunk foil is off, but it's probably not more than my machining tolerance.

P.S. Femap and ABAQUS/CAE also ran into the same trouble that Pro/E did. I'd be interested to see if Catia could do it. It's my opinion that Catia is better at complex surfacing.