I was not making up this stuff. Barnaby Wainfan (Technical Fellow for Aerodynamic Design and Analysis at Northrup Grumman, noted author, airplane designer and member here) presented on the topic at OSH a few years back. From his presentation, I got these takeaways:
- Putting a winglet on a wing does the same thing as adding span;
- Adding span is way more likely to give an efficiency boost than sticking on a winglet;
- Getting a net efficiency boost usually requires CFD and/or wind tunnel work and/or on-airplane development;
- Structural loads go up when you add a winglet, with potential structural weight increases.
Now maybe Barnaby will step in correct me on my memory of that day, but I was paying serious attention while he was talking for us.
Then along comes AutoReply (Good to have you back!) and tells us what the sailplane end of the spectrum has figured out.
The sailplane guys AutoReply is talking about are racers, and racers (of just about anything) will spend immense effort and expense to gain tiny advantages or not be left behind. We may be talking very modest differences yet the top guys plus the well moneyed also-rans in all racing feel a need to have the best speed mods.
The cited article is interesting. The author starts out telling us how many years of fooling around with winglets in sailplanes it took before they produced gain while thermaling slowly that did not cost you more while flying fast between thermals. Here is the first hint for us power guys: Lots of smart folks had trouble not making more drag in the operating area where the power guys cruise. here is the second hint: Then author also called the improvement possible for sailplanes from winglets "small but important". I talked about racers already.
The plots, while useful for sailplane folks, could use a bit of interpreting for us power plane design guys.
- Figure 5 - Sink speed vs airspeed at two weights with and without winglets. It appears that sink speed vs airspeed is not changed materially by going from a plain wing tip to a winglet. Again no info on more span vs winglets;
- Figure 6 and 7 - Increase in L/D is around 2.4% improvement over the base wing at 40 knots (where they thermal), and sinks away to about 1/4% at 120 knots. Yep, a well designed, developed, and executed winglet can be better than a plain wing tip even at 120 knots. There is no winglet vs span extension info on this plot, but I suppose somebody could paw through TOWS and tell us;
- Figures 8 and 9 - These are from simulations of cross country flights and displays increase in total trip speed and they compare span extension to winglets! How does this relate to power planes? Days with weak thermals mean a higher percent of the day flying slowly in thermals than on a strong lift day. On strong lift days, the pilots pare off a bunch of the time spent in thermals and thus spend a bigger fraction of the flight time doing what power planes do quite a bit of - cruise. So for power plane use, the high lift end of those curves is useful. Tip extensions on the ASW-22B are slightly better than winglets when light and about the same when ballasted, while the Discus comes in with a fraction of 1% improvement in cross country speed.
- Figure 10 - This show us how critical toe angle is to the gain or loss of cross country speed. For the ASW-22, +1 degree of toe out on the foil is better than -3 degrees or +5 degrees. In fact the other two angles are net losers in anything but light lift days.
Now we have some data points. What does the doc say to me? If we do a downtown job on designing a winglet, refine it in CFD, then, maybe some tunnel time (because sometimes CFD means Can't F'g Decide) and run a little designed experiment, so we can finally establish the design and go fly it to see how it does. Quite frankly that costs money, but in the competitive sailplane race environment that is the price of admission to the market. For us little homebuilt power plane guys, well, I can find you a lot more L/D and cross country speed by reducing baffle leaks, making it easy for air to get out of the cowling then start throttling down the inlet until CHT's get to 410 F on a 100 F day, fairing over protuberances, fixing the fit of fairings, cowlings, covers, and canopies, etc.
Oh, for perspective, if you do get 0.5% better speed on two hour trip, you save exactly 36 seconds. Maybe the racers would like that but the rest of us will play merry hell trying to even measure the difference. You could make your pattern tighter and your turns into and out of the pattern sooner and save more than that...
More later.
Billski