Thank you very much Don for your in-depth reply to my Whitcomb Winglet question.

I am getting afraid to ask any more questions, as I am sure you have much better things to do than educate me via email.

From : Don Stackhouse

You're welcome, no problem at all, I've found that the best way to make sure you really understand something is to try to explain it to someone else! However, I must admit that I am getting behind on several projects (including some slow flyer kit designs), so I may have to slow down my response time on this thread a bit.

How does this sound for experiments - fixed fin at the wing tip TE, with a hinged rudder extending rearwards? Then I can experiment with various rudder-out angles, and glue them when an optimum is found. I could test fly for duration easily. Would you use a constant percentage chord rudder, or make it wider at the top?

Sounds like a good approach to me. One other point: as you add camber and incidence to your winglets through rudder deflections, I expect you will find that the performance gains at the optimum operating point will increase (up to a point of course), but that the range of operating points where the winglets are a benefit wil get marrower.

On winglets you're not quite as concerned with tip stall of the winglet itself. I think I'd go with constant % chord.

Despite calling my initial design a park flyer, and being on the slow fly list, I don't think this will be a particularly slow model, and changing the induced drag by +/- 5% may not be discernible...

The key parameter here is lift coefficient, which is very dependant on airspeed. If you're going to fly fast enough, the induced drag won't be an issue.

With regard to designing very efficient ailerons:

Easier typed than done! Given that the wing will be a flat 20" * 10" sheet of 3mm Depron (with a carbon rod LE) I really don't have that many choices regarding the elevons...I don't see that I have much opportunity to make them more or less efficient.

Probably the biggest thing you can control here is chord. At low Re, it usually works better is to move the hinge line further forward, with much less angular deflection. At very low Re's, the flow tends to be laminar (even if you turbulate it, it will go back to laminar), which means that the kinetic energy the boundary layer starts with at the leading edge is all the energy it has to make it to the trailing edge without separating. It isn't getting fresh infusions of kinetic energy from the layers above it the way a turbulent boundary layer does (this is why turbulent boundary layers can do a better job of staying attached). Meanwhile, it is continuously losing kinetic energy through skin friction. As a result, the further back on the airfoil you get, the less energy is left in the boundary layer airflow. It has less ability to tolerate discontinuities, like hinge lines, and the increased adverse pressure gradients from control surfaces with too much angular deflection, the closer you get to the trailing edge. You need to treat the flow gently and with respect, and keep any discontinuities well forward where the flow is best able to deal with them. If you don't, the air molecules will get overworked, frustrated and mad at you, and refuse to follow the surface of your airfoil.

Several times Don you have talked about optimising the design (or airfoil) for the plane's mission. But while the goals are clear for HLGs (and full size), and an HLG can be tested to see how well it meets the design goals, this is *very* different for sports models.

Not really. You have a list of things you want the model to do, and the goal is to accomplish those tasks as efficiently as possible. The payoff is better climb rates, better max and min speeds, longer flight times, better control response and turn rates, etc.. These all make the airplane more fun to fly, which is the ultimate goal of a sport model. You can define fairly clear goals for any model, including a sport model, if you think about what it is you are really trying to accomplish. For example:

I suppose my design goals for this model are: a) Slow/agile enough to fly in the meadow close to my house, but fast and stable enough to cope with up to 10mph winds.

So does this mean you want a max cruising speed of say, twice the wind speed (20 mph in this case)?

Strong enough to withstand my poor flying.

A bit nebulous, but I think we all know what you mean. Weight figures into this as well, a lighter model doesn't hit as hard!

Aerobatic enough for me to improve my rolls and inverted flight, and be fun to fly.

So we need a decent roll and pitch rate, but good damping about all three control axes ("twitchiness" is partly the result of poor damping, not just too much control throw; you can tolerate a lot of control authority if the damping is sufficient).

Quick to build.

Everybody's in a hurry these days! Maybe we should start putting Ritalin in the water supply. We put flouride in the water to fight tooth decay; why not something to fight TV-induced decay of our attention spans? ;-)

At least 5 minutes duration when flying aerobatics. f) Ascertain if the IFO type model's smoothness and stability is a function of its swept/tapered tips.

So we're talking about a controlled experiment. In that case, you need to make at least one of these that's as identical as possible to the IFO except for the taper. You need to make another that's identical to the IFO including its sweep (as measured at 25% chord), but with constant chord.

It looks to me like you're already well on your way to a fairly clear set of design goals.

Even without goal (f) I don't know of any analytic way of determining if a given design will meet the goals that wouldn't take *much* longer than making the model.

So how does one go about optimising the design of a sports model?

Depends on the tools available. If you don't have a bunch of the right software AND a lot of experience using it, you're probably right. Also, you've locked in enough aerodynamic parameters already just because of your choices of materials and construction techniques, that there isn't a lot of room left for optimization within those artificial limits.

I think we've just answered this question. The same way you optimize ANY model. First you get a clear definition of what it is you expect it to do, and then you use all the tools at your disposal to the best of your ability to make the model achieve or exceed those goals as much as possible. BTW, your initial choice of goals is the foundation that the rest of the design will be built upon, and in that sense is probably the most critical and important step. Most of the time, when I've seen designs fall short of the mark, it's because they did a poor job of defining exactly what the goals should be.

In this specific design, I think the real key is your choice of speed envelope. You need to have enough top speed to penetrate (as we've already discussed), but you also need a slow enough minimum speed so that you can keep things under control going downwind in the same wind conditions. You also need enough turn rate (time) and turning radius (distance, including wind drift during the turn) to keep the model within the field boundaries. Ultimately what will set all of these is your mental workload while flying it. The fun you have flying it will depend a great deal on how much stress the model imposes on your brain cells until it's safely back on the ground. This is probably one of the biggest criteria for a good sport model.

You need to measure your field, then go out and mentally fly a few flights in it. Find out how many seconds you need for a crossing of the field in each direction (this determines the maximum number of seconds between each maneuver) in order for your brain to keep up with the model. How much room do you have to make a turn, and how fast a turn can your brain keep up with before you start to overshoot the ending point of the turn? Bring a stopwatch, visualize the imaginary flight, and time the various segments. Visualize typical "unusual attitude" situations, such as a stall or an upset from tree-induced turbulence, and how you would recover from those situations. It may help to bring a toy airplane along (or you can even do this sitting in a chair at home) and "fly" it with your hands through an imaginary flight. I know it sounds a bit silly, and your "significant other" may head for the phone to call those nice men in the white coats to come get you, but in fact the top full-scale aerobatics pilots and military combat instructors use exactly this technique to help visualize their routines before flying them.

Once you have a clear definition of time and speed properties, you can then figure bank angles, turn rates, min and max speeds, stall characteristics and other handling qualities, etc..

As you can see, if you really want to understand the details of what's going on and improve the odds of designing a successful model, one that meets your mission requirements, it's possible to get some very quantitative specifications even with a simple "sport model". Yes, it is possible to design a model like this with zero math and a lot of trial and error. The down side of that method is that there's a good chance you might run into troubles with one of your goals, and without a clear goal and the tools to understand why your model isn't meeting it, you may have no clues about how to fix the problem, or even what the problem is. It will be like hunting birds in the dark with a shotgun. You will blast away with attempt after attempt, possibly creating a lot of mahem in the process, and quite possibly never hit anything useful. There's also the chance (especially if you didn't start with clear goals to begin with) that you will hit your target, and not even realize it, continuing to blast away with additional but pointless attempts.

A great deal of the rewards in the whole "design experience" lie in the setting of a specific set of goals, working to achieve them, actually reaching those goals, and then KNOWING that you accomplished what you set out to do! Don't go through all this effort and then cut yourself out of half the fun!

Don Stackhouse
DJ Aerotech