1. ## Aerodynamic washout

While geometric washout seems to be relatively easy to understand (you twist the wing at the tip and it stalls later because of it's lower AoA - simple concept for the most part), aerodynamic twist seems to be more complex. Here is a statement by myself that was made in the "Fast Forward to the 1930s" thread started by ccg, and I have read statements by others that seem to say more or less the same thing:

My belief at present is that there needs to either be geometric washout or that the outer panel section needs to have a higher Clmax,...
I was suggesting that a higher Clmax would equate to a later stall at the tip. On reviewing some information in TOWS (Theory of Wing Sections, Abbott and Von Doenhoff), I see that this is not true. For example, with a root section NACA 4415, it's stall occurs at an AoA of 12°. If the tip section is the NACA 4418, it's stall will occur at an AoA of 14°, both according to the NACA lift curve charts at a Reynolds number of 3 million. Barring additional variables, this suggests that the 4415 root section will stall 2° before the 4418 tip section. Both sections have similarly gentle looking stall breaks, though the 4418 has a slightly more gentle one. The important thing here, with respect to the statement that I made above in ccg's thread, is that the NACA 4418 tip section actually has a slightly LOWER Clmax. "Stick and Rudder" constantly states that what is imprtant to how the airplane flies is the ANGLE OF ATTACK and not any other parameter. Regardless of the lift coefficient, the airfoil will stall at its stalling AoA. For washout, this section's stalling AoA needs to be higher than that of the root sections stalling AoA.

If the tip section is changed for the much thinner NACA 2412 section, the tip stall angle is now approximately 15°, but the stall break is much more abrupt due, I am guessing, to the thinner airfoils sharper leading edge radius. The symmetrical NACA 0012 also stalls at an AoA of 15° but it too has a very abrupt stall break.

Billksi has stated many times that thinner wing tip airfoils are not a good idea, and I'm just now beginning to grasp the reason why. There is way more to this, though, including the influence of the spanwise lift distribution. Orion has also talked about this many times.

I can't explain everything there is to know (or much of anything, actually) about aerodynamic washout, but its effect on handling and safety is obviously very important to anyone who will design and fly an aircraft.

It would be nice if there were a sticky with a good concise explanation of aerodynamic washout, considering the importance of this subject to safety.

2. ## Re: Aerodynamic washout

Originally Posted by Autodidact
While geometric washout seems to be relatively easy to understand (you twist the wing at the tip and it stalls later because of it's lower AoA - simple concept for the most part), aerodynamic twist seems to be more complex........
Assuming we aren't talking about a flying wing or FBW: Doesn't flap deployment of a portion of the span provide that function. So for instance if you wanted aerobatic performance like neutral handling and inverted flight capabilities but you want docile slow speed upright performance for takeoff and landing you could just require flaps to be at least at 10 degrees for anything below say 25% above stall. Say 50 or 60% span flaps that means 40-50% of the wing outboard is flying with minimum camber and a lower AOA than the rest. Or am I missing something?

The other curiosity I have about aerodynamic twist is Re. If the tips have a small Re in the area of the aileron and more resistance to stall (lower camber) that would make for docile handling right?

3. ## Re: Aerodynamic washout

Looking at the lift curve charts in TOWS, it does show that flaps reduce the stall AoA, so yeah, I think you're right about flaps creating effective washout. But I think that extending the ailerons towards the roots also helps with controlability at the stall (along with big radius leading edges), as well as increasing the roll rate. You'd want to have good roll control at stall
with the flaps up as well, as in a high speed stall?

Also looking at TOWS, it seems that a low Re, as with smaller chord tips, reduces the stall AoA in general. This is obvious for the 4415 in TOWS, but for the 4418 the stall AoA appears to stay pretty much the same. Greater camber doesn't necessarily mean a lower stall AoA; the NACA 2418 stalls at 14° just like the 4418 but has a lower Clmax (and camber). I think you just have to look at comparably generated data for the airfoils you wish to use and choose them according to what the data shows rather than categorical rules such as more camber, etc.

The other aspect of this that I think orion and Billski were alluding to is the stall break. It seems to be more abrupt for the thinner airfoils. Even though you have a tip that stalls after the root, it may stall very abruptly and after the peak end up with less lift than the root which may have peaked earlier but is still hanging on a little. The thicker sections tend to have larger radius leading edges and gentler stall breaks.

Ribblet seems to be saying that a moderately tapered wing with the same section root to tip can stall nicely since the Re is not that different and if the airfoil has a gentle break.

But I'm not trying to explain much here since I don't really know much and planform is really part of the other thread; I'm trying to correct a misperception that I have had and noticed that some others may (or may not) have had as well.

The bottom line is that the tip should stall at a higher AoA than the root however it is
accomplished and the relative angles of the zero lift lines are not necessarily important.

4. ## Re: Aerodynamic washout

The simple (perhaps over-simplified) approach would be to choose an airfoil at the tip that has a higher stall alpha than the root. As you reach the alpha at which the root stalls, the tip will still be flying. Of course, there are 3D effects to account for, but you can at least see the basic principle.

5. ## Re: Aerodynamic washout

There's several issues at play here so we can take a quick look for now, and expand on the details later. First off, the issue of tip stall is often not one to be too concerned about when landing since as Jay points out, the flaps are deployed at that point and as such, due to the steep geometric gradient of the root, it will be the flapped portion that will stall first. The major issue with adverse stall patterns has to do with flaps-up maneuvering - this could be relatively benign flight such as you'd have for pattern entry, or it could be during more abrupt maneuvers as you'd have with fun flying or with aerobatics. During these times a sudden, unpredictable stall could be anything from awkward to downright dangerous.

Simply said, washout is simply a form of flow control that assures a longer period of flow stability over one part of the wing as opposed to another. For the purpose of this discussion this means that you want to make sure that the outboard section of the wing has the flow attached longer than the inboard part and as such, you want to control said flow on the outboard section so that it maintains a stable boundary layer longer than that of the root. Keep in mind that there are magnitudes of stall and that the boundary layer separation does not necessarily happen all at once or at a sharply defined angle of attack. Even with sections whose data shows a sharp stall break will, on a three dimensional wing, have a generally gradual separation, albeit over a relatively narrow angle of attack range.

The latter statement is somewhat key since what you're trying to do is to control said separation in that narrow range. Don't think that you're going to somehow prevent the stall by something like five degrees aoa - most likely the difference will only be a degree or two. But that's enough to provide warning and control while the root separates. Now, how to do this. the best way to picture the goal here is to look at previous "fixes" and try to use them as a guideline for determining what works and what solution is questionable.

Back in the early seventies NASA undertook a general aviation safety study, evaluating several of the more popular trainers of the day. One of these was the Grumman Yankee (65-415 section), an airplane that was known for sharp stalls despite already having stall strips on the leading edge. For experienced pilots this was sufficient but not for students. To make a long story short, the fix to the Yankee was the incorporation of a drooped leading edge - the airfoil was identical except that a leading edge section was scarfed on over the span of the aileron. According to the flight tests the airplane became very predictable with a very benign stall pattern.

In designing an aerodynamically twisted wing then it is more beneficial to have a section at the tip that does a better job of attaching and controlling the boundary layer - this means a section with rounder nose (sections with round noses tend to be functional to slightly higher aoa's) however not necessarily one with a greater design lift coefficient. Depending on shape the design cl can be larger but it is important to keep in mind that some families the higher design cl will create a steeper gradient on the aft end of the section thus causing an earlier separation. All things in moderation.

It seems the better approach is to use a rounder front section and possibly a fatter section (for a tapered wing) at the tip, than one with a more substantial camber line. But keep in mind that this is only a guideline - there may be a few other approaches to this problem so this is by no means unique. For instance, Rn may be a significant issue so a section family may have to be chosen for the tip that has god stall performance at the slower Rn values.

6. ## Re: Aerodynamic washout

Originally Posted by orion
It seems the better approach is to use a rounder front section and possibly a fatter section (for a tapered wing) at the tip, than one with a more substantial camber line. But keep in mind that this is only a guideline - there may be a few other approaches to this problem so this is by no means unique. For instance, Rn may be a significant issue so a section family may have to be chosen for the tip that has god stall performance at the slower Rn values.
That's what I was trying to ask. I'll use one of my wings for example. The moving surfaces on the wing are about 25% chord. So I lined up the 75% chord line as the hinge line. 60% center section is not tapered. Tips are significantly tapered making a near elliptical compromise with slight sweep in the outer sections but not a straight trailing edge to keep the hinge line perpendicular to the flow. Now, the airfoil is symmetrical so the tips are flying at a much smaller Rn then the root. My thought was to have the max thickness of the tip airfoil farther forward than the root with the same overall airfoil function. The combination of slight sweep, lower Rn, and forward thickness should keeps the tips flying no matter the attitude or flap setting (sorry, I haven't run the Xfoil yet to give actual data across Rn and AoA). But that is the anchor in theory I am working with. Another thing to consider is the lift distribution. Keep the tips doing less work and therefore lifting less of the weight and they should stay happy at the onset of stall, is my guess.

7. ## Re: Aerodynamic washout

Originally Posted by Autodidact
But I think that extending the ailerons towards the roots also helps with controlability at the stall (along with big radius leading edges), as well as increasing the roll rate. You'd want to have good roll control at stall
with the flaps up as well, as in a high speed stall?
"Inner" parts of the ailerons contribute very little and going beyond 50% of your span or so does usually give pretty insignificant improvements.

Originally Posted by bmcj
The simple (perhaps over-simplified) approach would be to choose an airfoil at the tip that has a higher stall alpha than the root. As you reach the alpha at which the root stalls, the tip will still be flying. Of course, there are 3D effects to account for, but you can at least see the basic principle.
Well, solving it follows the same basic principles. Make a list/formula for the induced angle of attack as a function of the span, which is dependent on planform. Then you can tailor the different airfoils/twist/thickness for different scenarios (flaps up/down, ailerons actuated) and you're done.

A couple of remarks:
*Flow fences help a lot in preventing tip stall if mounted at the tip of the aileron. They also prevent (most of the) spanwise flow and are most effective between flaps and aileron in reducing drag, but also help quite a lot on the tip of the aileron and on the root of the flap. Typical roll rate just above stall improved by 50 (!) %:

*Don't extend the aileron all the way to the tip. This was told me by one of the designers of SH and apparantly helps a lot in preventing a less friendly tip stall.

http://js-dev.rockitg.com/content/bfW2xT45zcfbk.png

8. ## Re: Aerodynamic washout

Make a list/formula for the induced angle of attack as a function of the span, which is dependent on planform.
That sounds kind of fancy, is there a simple way to do this? The precentage of the spanwise distribution times the overall lift coefficient devided by the 3d lift curve slope, perhaps? But that's not right either, is it?

The combination of slight sweep, lower Rn, and forward thickness should keeps the tips flying no matter the attitude or flap setting
And doesn't sweep and low Re tend to exacerbate tip stall? I may be misunderstanding the above, Jay, but it sounds like you're implying that sweep and low Re is helping, the way you've worded that.

"Inner" parts of the ailerons contribute very little and going beyond 50% of your span or so does usually give pretty insignificant improvements.
Also, I'm "assuming" again about the full span aileron thing. I was assuming that the controllability at stall of something like an Extra 300 was partly due to the large ailerons. I am trying to take your signature, "Don't believe everything you think.", to heart.

9. ## Re: Aerodynamic washout

Originally Posted by Autodidact
And doesn't sweep and low Re tend to exacerbate tip stall? I may be misunderstanding the above, Jay, but it sounds like you're implying that sweep and low Re is helping, the way you've worded that.
I am asking the question about Rn as well. If you have a single symmetrical profile lets say, and the tip is very small compared to the root so much smaller Rn which stalls first due to just the size, I guess assuming the two profiles are flowing properly without a separation bubble on one and not the other?

When I say sweep I am talking about sweeping the tips like in multi-taper wings not sweeping the entire wing and only sweep in the outboard section to help the tip. But also, when I say sweep I am talking about mathematical sweep only between a straight leading edge and a straight trailing edge. My understanding of taper of that type as you head towards the tip is that if you line up the pressure numbers from profile to profile you can get favorable span-wise pressure distribution to ward off negative effects of span-wise flow that keeps the 3d wing losses as low as possible by keeping as much of the span working as late in stall as is possible. I also tend to visualize this as the shape of the vortex sheet in time which is a bit abstract but it is what helps me to understand it. I don't have a copy of Wil Schuemann's paper anymore on modifying tips in this way but it is always what I have gravitated to. Maybe I am all wet and I am worrying about details that don't yield quantifiable results.

10. ## Re: Aerodynamic washout

Originally Posted by Autodidact
That sounds kind of fancy, is there a simple way to do this? The precentage of the spanwise distribution times the overall lift coefficient devided by the 3d lift curve slope, perhaps? But that's not right either, is it?
When I tried grabbing my aero-book my friends were kind enough to put a bottle of Scotch in my hand... always great to have friends who support you in "difficult times"

The short answer; a numerical summation in Excel or a programming code (I use Java).
You need Lifting Line Theory to calculate circulation and toy with it a bit:
Lifting-line theory - Wikipedia, the free encyclopedia

Ain't exactly easy, but it does work pretty good with computers.

Here is an interesting presentation:
http://uav.ust.hk/resource/Fundament...0Chapter_5.ppt

31 and further discusses the "arbitrary planform".

Edit:
If I'm not mistaken, either XFLR5 or Xfoil has this built-in. (3d flow, including induced angle of attack)

And doesn't sweep and low Re tend to exacerbate tip stall? I may be misunderstanding the above, Jay, but it sounds like you're implying that sweep and low Re is helping, the way you've worded that.
I've neglected sweep (about 5 degrees in my outer panels), but lower Re (as in, below a critical value) usually decreases both Clmax and the aoa at stall. But yes, both sweep and lower chord make tip stall more likely to occur first.
Also, I'm "assuming" again about the full span aileron thing. I was assuming that the controllability at stall of something like an Extra 300 was partly due to the large ailerons.
Well, they went for everything they could get and while it does contribute a litte, even Extra said somewhere (can't recall where) that the effect was minimal.
I am trying to take your signature, "Don't believe everything you think.", to heart.
That one is in fact nicked from another user here

11. ## Re: Aerodynamic washout

From what I have read, a small amount of sweep doesn't make much difference and it is taper ratio that affects tip stall more than reynolds number. I think your main concern is the washout, which can be as complex as you make it to figure out as Autoreply indicated with the reference to induced angle of attack. Apparently it is a rather fancy calculation and if you understand vortex sheets and circulation theory you should be able to do it. But as he said, it depends on planform and that is what the other thread is about. I think orion answered your question but I was a little confused by the wording of it, hence my question.

My purpose was to correct a mistake I made earlier about the basic concept of aerodynamic washout - the optimization of local AoA to tailor the lift distribution is a little beyond what I was trying to do. Now I'm curious and am reading up on it but I doubt it will be simple enough that I will want to try and do the math unless I can find a very simple way to do it.

People apparently design airplanes without solving systems of differential equations but I guess you can do it with the complex math as well . I'm reading up on induced AoA in Airplane Performance, Stability and Control right now but I doubt if it is something I can calculate at all much less with difficulty.

Edit: I see Autoreply has addressed your questions. Thank goodness, I know I can't do justice to it!

12. ## Re: Aerodynamic washout

Originally Posted by Autodidact
"Stick and Rudder" constantly states that what is imprtant to how the airplane flies is the ANGLE OF ATTACK and not any other parameter. Regardless of the lift coefficient, the airfoil will stall at its stalling AoA. For washout, this section's stalling AoA needs to be higher than that of the root sections stalling AoA.
What you may have missed, or "Stick and Rudder" may have failed to mention, is the AoA where the airfoil starts lifting is just as important as the stall angle. Take two airfoils from the same family but different cambers and the one with more camber will have a wider range of operation. I've seen this type of wing design described in several books so I checked The incomplete guide to airfoil usage for examples. I found several drones from the '50s that used NACA 23012 at the root and NACA 4412 at the tip. The attached comparison polar shows the airfoils used on a trainer. When designing a wing this way you want the whole span to reach zero lift at the same time so you would twist the wing 3 degrees to archive zero aerodynamic washout. Since the slope of the lift curve is approximately the same for all airfoils the lift distribution stays the same at all AoAs below the stall angle of the root. This software is limited to 13 degrees AoA (if I need more I'll use XFLR5) so you can't see the actual stall but the 23015 shows a dip indicating that it's starting to stall but the 4412 is still climbing. Given the reputation of the 230 series I'd imagine that this wing stalls straight through and gives a good buffet.

13. ## Re: Aerodynamic washout

Thanks, Norman. I noticed while doing my original post that the 4415 and 4418 airfoils had the same zero lift angle, but I didn't know what, if any, significance to attach to this. I had only concerned myself with the fact that one had a higher stall aoa than the other.

So, I will suppose that if the zero lift lines are not lined up, then at high speed/low Cl, the inboard part will be creating negative lift and the out board positive, and a lot of drag?

14. ## Re: Aerodynamic washout

Originally Posted by Autodidact
, and a lot of drag?
That's why if you use more than one airfoil section the zero lift angle is the one to use when jigging the ribs not the chord line which is just a convenient reference line anyway and has no aerodynamic significance

15. ## Re: Aerodynamic washout

Jay, just my 2c, if I were you I would use the same section from root to tip, about an 0.7 taper ratio with a slightly thicker section at the tip, and I'd pick one that has a pretty round LE as well. In other words, in lieu of doing the complex analysis, I would follow the rule of thumb that makes for the simplest job to build combined with good handling. I would sacrifice a little top end speed for better stall handling which would mean a more turbulent airfoil with a lower Cmc/4. That's just me; if I were in college and going to make aero engineering a career (or just younger than I am), then I would certainly have to come to grips with the complicated side of things. But I'm not, so I just want to design something that I might have a chance to actually complete and get into the air. I love hot, fast and small aircraft, but I'm a little afraid of them.