Question about swept wing airfoils

[SteveC68]

What airfoil shape matters in a tapered swept wing? Is the shape that is parallel to airflow important or is it the airfoil that is perpendicular to the leading edge that is more important? I ask this because it is sometimes easier to cut foam wing cores as just a tapered wing and then come back and cut the sweep into the root. See the image below.

Thanks,
Steve

[JetPlaneFlyer]

The airfoil shape parrallel to the airflow is what matters. There is still nothing stopping you cutting the wings as you describe, just thicken the chosen airfoil by by dividing it's thickness by the cosine of the sweep angle.

i.e. if your airfoil is 10% thickness and your wing sweep is 30 degrees:
10% / Cos 30 = 11.5%

So for cutting if you thicken the airfoil to 11.5% then your final 'swept' airfoil will be the 10% thickness you were aiming at.

This method works perfectly for symmetrical airfoils but on a cambered airfoil it will slightly reduce the effective camber unless you also increase the camber by the same proportion.

[MarkusN]

Actually it's a bit more complicated than that. As wing sweep affects flow, the wing section that's effective on the airflow is curved, so the airfoil you cut is not the airfoil that's seen by the flow. Flow over a swept wing is three-dimensional, so strictly speaking the theory normally applied to calculate behavior of wings does not apply.

But for sweep in the order of magnitude that you have drawn it's insignificant. Just don't expect a scale Sabre or Lightning to fly the way you calculated with the standard theory.

If you want to apply JPF's correction (nothing wrong with that) just scale all the airfoils ordinates with 1/cos(sweepback); that will correct thickness as well as camber.

[markdrela]

Actually it's a bit more complicated than that. As wing sweep affects flow, the wing section that's effective on the airflow is curved, so the airfoil you cut is not the airfoil that's seen by the flow. Correct. For high aspect ratios, the relevant airfoil shape is perpendicular to the leading edge, not parallel to the flow. The lift also depends only on the speed component perpendicular to the leading edge, not on the flight speed. In the aero business this is called "infinite swept wing theory", and it's been around since the 1930's.

For low aspect ratio wings, this sweep theory strictly doesn't apply. But here the concept of a locally-2D airfoil flow is also dubious. In any case, the "streamwise" cross sectional shapes on a wing are mostly irrelevant.

The infinite swept wing theory doesn't just apply to streamlined wings, BTW. Consider a long circular cylinder (e.g. a wire) in a flow at two orientations:
A) 0 degree sweep
B) 80 degree sweep
Each case will have the same massive separation behind the cylinder, and the same drag coefficient relative to the LE-normal velocity. The fact that case B has long slender ellipses for its "streamwise" cross-sectional shapes is irrelevant, and doesn't help to reduce the amount of separation.

[SteveC68]

So the component of the airflow that is perpendicular to the leading edge is what we should design to? If that is the case then cutting the wing using an unmodified airfoil (not divided by cos of sweep angle) would be fine as that will ultimately be the shape that is perpendicular to the leading edge. Is that correct?

Thanks,
Steve

[nmasters]

The infinite swept wing theory doesn't just apply to streamlined wings, BTW. Consider a long circular cylinder (e.g. a wire) in a flow at two orientations:
A) 0 degree sweep
B) 80 degree sweep
Each case will have the same massive separation behind the cylinder, and the same drag coefficient relative to the LE-normal velocity. The fact that case B has long slender ellipses for its "streamwise" cross-sectional shapes is irrelevant, and doesn't help to reduce the amount of separation.
Well that’s just weird :) I wish I’d known that 20 years ago.

[markdrela]

So the component of the airflow that is perpendicular to the leading edge is what we should design to? If that is the case then cutting the wing using an unmodified airfoil (not divided by cos of sweep angle) would be fine as that will ultimately be the shape that is perpendicular to the leading edge. Is that correct? Yep.

[JetPlaneFlyer]

Yep.

I guess you live and learn!

[SteveC68]

I did a little reading on the infinite swept wing theory and came up with a pretty good explanation about the cylinder model. Imagine the cylinder of infinite length placed in a wind tunnel. Now pull the cylinder through. Put the two velocity vectors (air speed and speed of the cylinder) together and solve for the angle. The faster the cylinder is being pulled through the more the relative sweep.

[macboffin]

Note that (Abbott and Doenhoeff, "Theory of wing sections") that if sweep angle exceeds 27 degrees,(measured at quarter chord point spanwise) the actual section becomes of little importance except as regards form drag, since CL is dependant primarily on angle of attack.

[nmasters]

For high aspect ratios, the relevant airfoil shape is perpendicular to the leading edge, not parallel to the flow.
Like this? :)
http://www.standardcirrus.org/120702-0-002.jpg
http://www.standardcirrus.org/120702-1-005.jpg

[miniphase]

wow, great fotos Norm, very effective demonstartion of putting a trip on, what's the fin/tip from?

[HerkS]

Correct. For high aspect ratios, the relevant airfoil shape is perpendicular to the leading edge, not parallel to the flow. The lift also depends only on the speed component perpendicular to the leading edge, not on the flight speed. In the aero business this is called "infinite swept wing theory", and it's been around since the 1930's.

For low aspect ratio wings, this sweep theory strictly doesn't apply. But here the concept of a locally-2D airfoil flow is also dubious. In any case, the "streamwise" cross sectional shapes on a wing are mostly irrelevant.

This is a very interesting thread. Thanks for getting it going.

I have a technical question.

Given a relatively high aspect ratio (~10) model wing with 45 degrees of sweep, a constant chord, perpendicular to the leading edge, equal to ten inches, and flying at a free stream speed of 30 miles per hour.

What is the (sea level) RN that the wing is experiencing? Is it about 230K or is it about 330K. Seems like the difference should be important.

[nmasters]

wow, great fotos Norm, very effective demonstartion of putting a trip on, what's the fin/tip from?
Looks like a truck :D Seriously though they're from the “Standard Cirrus (http://www.standardcirrus.org/12-07-02oiltests.html) ”web site. I thought they would be relevant to this discussion because they show the boundary layer flow is perpendicular to the leading edge. This is to illustrate the point made earlier that the cross section that the air sees is perpendicular to the LE and not parallel to the free stream. However I believe that that's not always the case, sometimes it's much worse. I've seen flows that follow an ogive like in this old drawing. I think the difference is related to CL but I don't really know. (Bet there's an NACA tech report or something that explains sweep effect. Wish someone who dose understand this stuff would recommend reading material)

--Norm

[biber]

What did you expect?

I'd expect a vectorial superposition of a common speed distribution around the airfoil (perpendicular to the LE) and a constant component along the LE.
If you do that superposition of the two vectors/components you will get that S-shape bent streamlines close to the surface.
Or am I mistaken there?

biber

[HerkS]

Hi Norm
Great pictures - Really shows a separation bubble and the effectiveness of a turbulator.

My experience with oil for flow visualization is that it's affected by gravity so don't take the vertical trend too seriously. Also the flow field around a Whitcomb winglet is complex because of the tip vortex that it takes advantage of. Comparing the way a winglet works to the function of a wing panel, the photos would be showing the equivalent of the lower surface.

No special point here - just thinking about the photos.

[HerkS]

(Bet there's an NACA tech report or something that explains sweep effect. Wish someone who dose understand this stuff would recommend reading material)

--Norm

Norm, Here's one that's kind of old, but has a lot of graphical data and is reasonably easy to get into.

[nmasters]

WooHoo gotta love this technology :cool: I just downloaded a scientific document during breakfast that would have taken weeks to get here by interlibrary loan in the old days. Not to mention that librarys in rural towns like mine don't usually have the resources to look up the tn numbers of NACA reports.

Thanks bunches, Herk.

--Norm

[nmasters]

My experience with oil for flow visualization is that it's affected by gravity so don't take the vertical trend too seriously.
Yeah, it would be more convincing if he'd used something that doesn't run downhill (spray paint?) or if the winglet were horizontal. I had noticed the problems WRT gravity, the boundary layer of the truck and the fact that there's no wing to generate the tip flow. All that may have skewd his results but if you hold a piece of paper up to the screen with one edge on the leading edge you'll see that the streak lines near the TE really are perpendicular to the LE. I don't think that's just coincidence but I'm often wrong.

[HerkS]

NASA Technical Reports Server.

http://ntrs.nasa.gov/search.jsp?N=287

Have a ball Norm.

[Troy]

I wanted to open up this discussion again to ask a quick question. If you are cutting cores of a highly swept wing (say 30 degrees) with multi tapered sections along the leading edge, how do you address where the cores meet if you cut the foil parallel to the LE. I am imagining that the airfoils are not going to line up perfectly from one core to the next. The tip here as an example....

[nmasters]

Whether the templates are parallel to the free stream flow or perpendicular to the LE isn't going to make a big difference in performance. It's more a mater of do you want to cut the foam blanks into trapezoids and then hotwire or would you prefer to hotwire the blanks square and then cut the cores into trapezoids. If you want to try it with the airfoils perpendicular to the LE lay it out like this and then cut of the red triangles.

--Norm

The end of the second segment will have to be cut of like the tip to get the ends to match up well and even then you'll have to sand or fill the joint a bit before you glass it otherwise there will be a little step

[Troy]

Thanks for the illustration, that's about how I was thinking it would work. I have always gone to the trouble to make my core blanks with the proper sweep prior to cutting. My biggest concern was to keep an accurate airfoil in thickness and camber with regard to flow over the chord.

[Joel K. Scholz]

I built highly swept wing about a year ago with the help of many peoples input from e-zone. If you take the time to read thru the post you will learn a great deal about flying wings


http://www.rcgroups.com/forums/showthread.php?t=782118

[nmasters]

Here's a newly posted old video of a boundary layer visualization test of a swept wing being rotated through it's AoA range.
http://www.youtube.com/watch?v=v2yipTXBP-Y

[BMatthews]

Now THAT was an amazing bit of video. You can clearly see the initial separation bubble effect out near the tip and how it grows as the angle of attack increases. I'd heard about the spanwise flow issues and understood them in principle but it's another thing to see this effect so clearly in a real life setting. Note how it even has mid span vortices near the total stall point.

It's also highly interesting to note that with a swept wing that the tips suffer first in all stages. A strong support for the use of higher than normal washout angles on highly swept wings. It also explains the need for almost rediculously high washout angles on some of the swept flying wing designs I've seen such as the CO series.

[nmasters]

Here are two more swept wing flow studies http://www.youtube.com/watch?v=hutx8ivD_yc&feature=channel_page
http://www.youtube.com/watch?v=w8SOuH9sNYc&feature=channel_page