[Courtesy of Michael S. Selig UIUC Applied Aerodynamics Group/UIUC LSATs Department of Aeronautical and Astronautical Engineering University of Illinois at Urbana-Champaign, Urbana, Illinois 61801 - posted to RCSE Dec 1997]
RC Soaring Exchange:
During my holiday browsing of my backlog of email, I started reading the RC Soaring Exchange and came across some references to the similarities between the SD7080 and SA7035. Everyone is right---these airfoils are very similar in shape. As stated in the article on the new series [ref 1 below], the SD7080 was the benchmark for the lower camber version of the series, so it can only be expected that the 7035 would be very similar to the 7080.
A word about the design process: These airfoils were designed by first "redesigning" the SD7037, that is, we created a new input file for the PROFOIL airfoil design code [ref 2]. This new input file differs from the original one because back in 1987 when John Donovan and I designed the SD7037, we used the Eppler code, while today I have my own (it was the topic of my PhD thesis). So in designing the new series Ashok and I needed to get the old Eppler-code 7037 input file into the PROFOIL input file format. This "redesigned" SD7037 is not out and about because it is the same as the original (same coordinates). From that input file we then changed the operating lift range, up for the 7038 and down for the 7036 and 7035. Of course, other things we tweaked too. We did not scale the camber, but did things aerodynamically through the design parameters. For anyone interested in the details, we incremented the alpha*'s to increase the lift range (7038) and decremented them to decrease the lift (7036 and 7035). What are alpha*'s? See the PROFOIL-WWW page [ref 2].
Since the SD7080 and the SA7035 are similar in shape, the performance is also similar. But from the predicted performance, the trends [ref 1] show that the 7035 is a little bit better in penetration than the 7080. And yes, one swipe with some sand paper in the wrong place and this better performance of the 7035 will not be realized. The performance of the SA7035 is as good as we could get it. It may not seem possible that there are these differences in performance since the shapes are so much alike, but it is for real and a flyer can notice it. (A flyer can notice as little as a 3-4% change in drag.) In the SoarTech books , there are plenty of cases that show significant changes in performance due to small deviations in airfoil shape. This sensitivity of airfoil performance to shape changes is one thing that has always intrigued me about airfoils at low Reynolds numbers. I continue to appreciate it more and more ... with each new series of tests.
Here's one example, we were testing an airfoil for AeroVironment and by adding one piece of tape (0.002 inch thick) to the surface of the 12 inch chord airfoil (a turbulator), we measured a more than 50% drag reduction. We have seen the same thing on the Miley airfoil. (0.002 in is less that the thickness of a sheet of paper.) These are extreme cases, but the point is that 'yes' performance is really sensitive to the airfoil shape. It is all tied up w/ the laminar separation bubble, a good picture of which is show here: http://ae.uiuc.edu/m-selig/uiuc_lsat.html
The upshot of this discussion related to those comments about the 7035 and 7080 similarities is that to get the most out of an airfoil, it should match the true airfoil as designed. And if one is very serious about flying the best, fly only wings coming out of CNC cut molds .... and as a check have the airfoil digitized and compared with the true airfoil.
Keep in mind, the SA* series was not designed to test the limits of one's building expertise, but they were instead designed to give all of us a family of airfoils similar to the SD7037 with incremental changes in lift range (speed range) - in either direction. I saw what I thought to be some stagnation in terms of design choices, and that's what provided the motivation.
I also saw that someone pondered where my funding might be coming from for the low Reynolds number airfoil testing. It started out 4 years ago as Univ of Illinois funding (through my start-up offer), then it evolved to be mostly funding from modelers with some help from the UIUC Aero/Astro Dept. Support from modelers is still our bread and butter, but lately we have benefited (and in turn modelers as well!) from 1) our work with the DOE National Renewable Energy Laboratory (NREL) on small wind turbine airfoil design, and 2) our work with AeroVironment for their high-flying solar-electric airplanes under NASA ERAST funding (ERAST = Environmental Research and Aircraft Sensor Technology - atmospheric science work). Out of the NREL work, we have tested several airfoil that would not have been otherwise tested. Also, we did several airfoils that were aimed at high L/D, and the outcome are several airfoils that achieve L/D's higher than anything else that we have tested. From the AeroVironment work, we were able to buy more instrumentation (to add some redundancy) and also paid for part of a new x-y traverse system for moving to probes around to measure drag. This will save us a lot of time in the future, or alternatively make it possible to take a lot more data. I like the latter view...but the test team tries to coach me in the other direction. :)
Recently, we put together another newsletter to let everyone know that we are still in business and still need their support. The last two years have been busy ones (and good ones) for me, and I have not been beating the drum too hard on the modeling front, mostly because the situation was stable and healthy. Now I am looking again for more help!
Ref 2 http://www.profoil.org
Regards, Michael Selig m-selig "at" uiuc.edu
For actual simulations and comparisons you can go to: