Sweep with guide curve fails to make wing

I'm trying to create a wing with quater-ellipsoid taper.

I have two paths on the front plane for leading and trailing edge of
the wind,
and a profile on the top plane for the root profile.

Lofting to a smaller profile at the wing tip doesn't work because the
Loft profile is not maintained exactly between profile sketches.

A sweep with path and 1st guidecurve ought to do it, but it comes out
weird.

(I know about creating the profile after the path and guid, so it's
not that...)

See http://img147.imageshack.us/img147/5429/33907542cb5.jpg

SolidWorks 2007SP5

Any thoughts?
/gh

 

I would use the surface loft with guide curves. If you draw your profiles
centered on 2 parallel planes with guide curves the profile taper is pretty
consistent. See image

http://img185.imageshack.us/img185/1642/wingts7.jpg

weevoe

 

Well, that sort of works, except that "pretty consistent" is not
really what you want when you spent all that effort in developing the
profile.

Seems this is not a task for SW.

I had it done in Rhino3D using "Sweep along two rails". Doesn't need
all the intermediate profiles, and is fully consistent rather than
just "pretty".

/g

 

What intermediate profiles? There are two, the others you see are generated
by the two with the guide curves. I don't know what you mean by "fully
consistent" so I used the term "pretty". Your term "fully consistent" is
ambiguous at best. It would depend on how you measure/evaluate the surface
profile, where, how many places & to how many decimal places. I am not going
to take the time to do surface profile analysis...that would be on you. I
was just showing you how it could be done since obviously you do not know
SolidWorks.

 

My gosh! It's so funny with people who use anonymous forums to vent
their frustration...

Using you method SW generates a surface which is a "pretty" good
approximation of a sweep.

It is not 100% correct. If you make a cut of the resulting body the
profile is NOT an exact scaling of the original profile.
It is slightly different. A swept wing should have each section be
exactly the same profile, that's what I mean by fully consistent. Not
some ambiguos at all I would say...

Make your guides curve a bit more and you will start to see
significant deviation in intermediate wind sections.

And I have many years of SW experience going back to when parts could
only contain one solid, thank you very much, so stop trying to
intimidate your fellow forum members and show some manners please!

 

I smell a bit of confusion & perhaps language problems.
Each "profile" would provide more options for control,
IOW you can do more (if you want to).
"fully consistent" seems unclear.

So does the later "exact scaling", "swept wing should have each section be
exactly the same profile" , "fully consistent" & "intermediate wind sections".

IF you curve your guide curves what would you expect? The "scale" is
different along each curve (as they have different lengths, curves, etc.,
right?

 

Sorry, typo: "Wind sections" should read "wing sections".

The scale is different, yes. Of course. But any section of the wing at
any location along the should still be an ortogonally scaled copy of
the original profile. This is what I mean with "fully consistent". A
consistent (scaled) profile as you move along the sweep.

What happens in SW when you use a Loft here, is that some kind of
surface approximation math comes in to play which creates a surface
that looks sort of OK, but makes no effort to keep the wing profile
consistent. If you section the wing in the middle the profile will be
(from an aerodynamic point of view) very different from the root and
tip profile.

I made an image to show the problem: http://img525.imageshack.us/img525/4840/wingsgm4.jpg

(If the guide curves are straight lines the profile is correct all the
way, though, but that's not a very good wing...)

 

If you look at real aircraft ....

IF you want such linear scaling you'd probably best first
scale the control points for each section curve from the
original scale and then properly align the sections.
Then your sections *parallel* to the section curves
would perhaps be what you want (assuming that the sections
were parallel in the first place).
Don't know about the SW "loft" function but should work
for most section-controlled surfaces without other constraints.

I doubt your guide curves are producing a linear scaling in any case
& have no idea what the other section curves are like.

 

That does not sound like linear scaling to me.
BTW, IIRC Most modern wing shapes have a
"twist" to them in the "as designed free state"
(no loads or gravity).

I don't think I want to fly anymore.