I talked to one engineer who helped design the wing tips on the new 747-400. He said with all the tests they ran the best wing tip ended up not sticking straight up but flat with the wing,

I.E. longer wingspan. Same goes for flying wings. As he told me wingtips are a quick fix for a poor wing design.

From : Don Stackhouse

This is basically true. Burt Rutan's comment on the subject was that a 3' high winglet has about the same effect on wing efficiency as a 2' increase in span, but it has the same effect on wing root bending moment as a 1' increase in span. Theoretically this means that some savings in the weight of the wing spar is possible, but this will only be realized if the wing spar design is optimized perfectly to begin with, which also requires that the mission be so well defined that the load spectrum the wing spar sees can be PRECISELY determined. In my experience, for the vast majority of model designs these requirements are almost never met.

However, the one place where winglets do seem to make sense is where the induced losses can be expected to be fairly high (so that there is enough energy to be harvested that you stand a reasonable chance of gaining more than the winglets themselves are costing in drag), AND where you have to put some vertical surfaces anyway, AND where the longitudinal moment arm of the vertical surfaces will be about the same whether they are used for winglets or not. As you can see, this is a fairly restrictive set of requirements, and fits well with your friend at Boeing's comments. Yes, winglets are not generally a good idea for most applications. However, I think that the specifics of Neil's application fit the above guidelines reasonably well.

In Neil's case, we're dealing with a very low aspect ratio design with zero taper, probably a wonderful example of what your friend calls "a poor wing design". Before we get too harsh with Neil's design, however, let's remember your other comment:

Also, you are flying a 1/8th scale model in FULL scale air and the rules change from full size aircraft.

What you're referring to is the effects of Reynolds number ("Re"). On very slow and small aircraft (such as Neil's application) it becomes one of the dominant factors in the design. For example, I looked at one of the more popular R/C sailplane airfoils a while back. Between Re = 120K and Re = 80K, the drag coefficient doubled. It doubled again between Re = 80K and Re = 60K! As size and speed goes down, you either have to reduce aspect ratio in a desperate attempt to keep Re up to something reasonable, or else you have to be VERY good at low-Re airfoil design. That second one usually requires a lifetime or so of study in that specific field to be any good at it, since most of what's in the textbooks becomes wrong at Re's below about 80K - 100K (guess how I know!). Most of the software becomes VERY subject to interpretation in that realm as well, to the point that your experience with your particular software and your ability to interpret its results at low Re becomes much more important than the raw accuracy of the software itself. In Neil's case, his extremely low aspect ratio and lack of taper helps keep the Re's up to a reasonable level. Whether or not his particular wing dimensions are an overcompensation for the low Re problem is open for discussion. Personally I don't feel we have enough data and analysis of this specific case at this point to form a valid opinion one way or the other. There are a lot of factors to consider, and many of them don't have anything to do with aerodynamics. It's very dangerous to jump to any conclusions about someone else's design, unless you thoroughly understand WHY the designer made each decision in that design.

We are just playing with model airplanes and my fix would be to put 1 rudder on it and fly it. Any ways 1 rudder in the middle has advantages, LESS BUILDING, easier to mount, takes less time to cover, less weight at wing tips for better roll control, HARDER to knock off, and you probably will not know the difference.

Good points, to which I'd add that a single fin will have a better span and/or better Re's than dividing up that same area between two separate surfaces. As far as easier to mount, that depends on the details of the structure at the center section and the tips, that issue could fall either way. However, in Neil's case, the relatively low speed and high span loading says that there is a good chance of harvesting enough energy from the tip vortices to at least break even on the drag of his vertical surfaces. A central fin will add the fin's drag to the induced losses.

Whether we say it's because of a "poor wing design" or not (and that's subject to debate, we have not analysed all the facets of Neil's design in sufficient depth at this point to for a valid opinion one way or the other at this time), the fact is that this particular case may be one of the the exceptions where winglets are an advantage. Note, if Neil was using a conventional tail on a conventional moment arm (like most of the designs your friend at Boeing is probably used to working on), I would probably feel differently. It's just that in this case, as long as Neil has to install some vertical surfaces anyway, he might as well use them to help with the induced drag problems as well.

If you are worried about which will be better just build 2 models with 2 different rudder configurations.

Very good point. If I was doing this design myself, I'd probably at least analyse both in-depth, and if there was still any question in my mind I would probably build and test both. If you visit the "Ask Joe and Don" section of our website, you can read an article in the "Design" category on the merits of kits vs. scratch building. The bottom line is that if you want something to fly, build a good kit. If you want the experience of developing something, where experiencing the development process for its own sake is your goal, then scratch build an original design. Just don't expect the first prototype to meet all your design goals. That usually requires going through several airframes and a bunch of mods, if your standards are very high at all. In my case, I just happen to be interested in the design and development process for its own sake, and I've been lucky enough to develop a full-time job for myself that involves doing exactly that. Trust me though, it's not easy, and it's not for everybody.

BTW, I am designing a new airplane and I need to know what is a good airfoil.

This question is impossible to answer. It's like asking what the best shoe size is, without supplying any context for that question. It all depends on the size and shape of your foot ... er ... your airplane and its mission profile. Yes, there are shoes that don't fit anyone's feet well, just as there are airfoils that don't work well on any airplane. There are shoes that are good for playing basketball that would not be good for playing baseball or for getting married in. The shoe/airfoil must be matched to the mission and to the requirements of the foot/wing they're to be used on.

However, just as Nike and Adidas both make good basketball shoes, in all probability there are a number of airfoils that can be made to do reasonably well on your design. It is also almost inevitable that the best airfoil(s) for one part of the operating envelope will not be the same as the best ones for a different portion, and that the best airfoil(s) for the root will not be the same as in the midsection of the wing or at the tip. If you want to do it right, the airfoils are just a few of the parameters in the overall wing design, and usually not the controlling ones. When I design a new wing, the final airfoil designs are usually among the LAST things to be determined. Only in the context of the local airflow and lift requirements along the wing and the global interactions between the different wing segments over the entire mission profile can you determine the appropriate airfoil requirements. Just about anything can be made to fly, but if you want to consistently be able to fly WELL, you'd better be prepaed to do your homework!

Don Stackhouse
DJ Aerotech