loft function bug?

[charlieuk]

I finaly thought I had worked out how to loft wings last week however I cam back to try again today and now the same method gives me a error.

trying to find the fault I have ended up going back and using the same file I made last week. I resaved this file, closed the body I lofted last week and tried to do the exact same loft again and this time it has the error

Warning: Loft to point would have non-smooth neighbourhood of point

I have been trying to figure out this loft for 9mths now and I thought I finally figured it out but now this!.

the original file that works is attached bellow however try re saving it to see that it now will not loft.

edit ( I have realized the originally loft was done with the original imported sketch of the airfoil so it is just the fact it does not work with the new spline sketch I did to smooth out a few bumps the original had. )

many thanks

[laughingcreek]

Here i split your profile in half, and did the loft half at a time (top and bottom).  When you do it this way you can see why the loft fails.

On the second loft in the time line, the first rail is that leading edge of the wing surface from the first loft.  When that rail is set to free, the loft works, but if it is set to tangent or smooth, you get an intersecting surface. 

Trying to do the loft in a single profile without splitting it would have the same result as whet you see here, but with that rail set to smooth.  thus it fails.

[laughingcreek]

ok, that explanation may have been totally bogus.  But i do think it had to do with the trailing edge rail causing the surfaces to be so thin that they end up intersecting.  Here I created 2 edges off the trailing edge to use as rails, and then patched in the small flat section on the trailing edge.  You way have to zoom way in to see the additional rails. 

[charlieuk]

thanks for taking a look

I have been playing around trying to re create it, I got half way there but then got stuck on a simple thing of trying to rotate the anhedral sketch around the tip point to create the new ones?

[laughingcreek]

You still need your original curve in the sketch named "anhedral" to create the project lines in sketch "3d le", so don't move that one.

I just created additional lines above and below it for the tailing edge, used to create sketch "3d te" and "sketch 2"  (Note: I forgot to activate component 2, so "sketch 2" is in in the master level.  you can just move it down to where it should be by grabbing it and dropping in "component 2")

I just free handed the two additional curves added to sketch "anhedral"  with the point close to the body constrained to a project end of the short line at the trailing end, and the point at the tip constrained to the original curve end.  You could gain some control over the geometry by throwing in some construction lines and dimensions (might want to "fix" the original curve so it doesn't move around)

[charlieuk]

ok many thanks yes I was going to leave the origanel there its just I was having this rotate problem on another thing so was trying to crack that along the way.

after you said about splitting the foil into parts and lofting separately I was playing around and noticed if you split the airfoil vertically in the center it comes up with the front part being a open sketch however it is one continuous spline from the trailing edge bottom to trailing edge  top. could this be something that is causing problems? it seams a bit strange.

I will carry on trying to recreate your loft tonight. many thanks

[laughingcreek]

It really is about lofting to that point, and the trailing edge rail being a hard edge.  it would work the way you had it originally if the rails weren't curved so much.  at some point the loft just gets to thin and fails.

look at the attached study. (all free form with no thought to dims).

body 1 and body 2 start with different profiles, but the rest is the same (rails, loft to point)

take the point at the wing tip and use the move command to move it up about 1/4" or so.  (you don't need to go to the sketch, just select the point, right click "move", and pull up in the z direction)

 loft for body 1 fails at this extreme,  but the loft for body 2 doesn't. 

[charlieuk]

i have had a go at recreating your first drawing but still getting a problem im not sure were i went wrong?

http://a360.co/2vwmiSt

i also tried splitting the airfoil in 3 bits a top bottom and center but no luck with that idea ether

[charlieuk]

going back and looking at the original loft using the profile i imported before i cleaned it up i guess you can see where its having the problem right on the curve near the trailing edge towards the tip. I thought i was going to be able to shift the thickness down using the direction function and it would solve that but using the original loft again it doesn't appear that way.

maybe i need to go back to trying to use a second profile some were near the apex of the curve.

 I really wish i could go take lessons in this some were!

[laughingcreek]

The first design (wing 1) was just to illustrate why it wouldn't work with out more guide curves.

The second (wing2) I think is right on as far as shape goes.  adding the 2 rails at the tailing edge is going to be necessary past a certain amount of curvature at the tip.  Only change I would make to that one is to put some defining dimensions in to make sure the lines are clean and not wavy.

pulling the tips up (down?) is changing the cross-section of the wing, so if that foil cross-section is calculated from some where, you may need to add a couple more to maintain the proper shape.  Don't know how important that is to you.

[laughingcreek]

here's a screen cast of lofting from the last file.  I think you just have to be careful to get the right rails at the trailing side selected.