elippse

Well Known Member
At many of the prop forums I've given at various fly-ins, I'm asked about the influence of Jack Norris' book on my propeller design. My friend Oscar Grassle offered 1/2 interest in his Lancair 235-320 while I was still building mine and I accepted. Because I had a passing interest in things aerodynamic, and had done a lot of reding about wings and airfoils, I felt that I could make a minor mod to the tip of his Great American prop to up its performance. Well, without changing the diameter and only a minor re-shaping of the LE and making the tip come to a sharp edge on the top of the blade, we increased speed from 215-216 mph to 218-219 mph! That 1.4% speed increase amounted to a 4.2% efficiency increase, just from this little change. Sometime after this there was a prop strike during taxi that required obtaining a new prop. Oscar called several prop makers and settled on one from whom we got a prop after telling them about our plane's performance. It was not as good, so we sent it back describing how it worked. We got it back and it still wasn't as good as what we had, so sent it back a second time. It still wasn't as good, so we gave up on that route. I had done design of electronics and physical processes, such as refraction and radar noise, and I knew that physical processes were very predictable, so I decided to write a program to design a propeller. After getting to know the program results, I designed a two-blade prop on 1998 August 11 to be able to turn up higher rpm for racing with Oscar's Lancair, as we wanted to participate in the EZ races. I gave the blade sections to Klaus of LSE, who was still making props, to make this prop for me. He tried to talk me out of it since it was unlike anything that had been made before, but I told him I needed it made in order to see what, if any, corrections needed to be made to the program if it didn't result in the predicted performance. I had predicted the performance of this prop at three different density altitudes at 100 rpm increments from 2400 to 3200. I wrote these predictions on a piece of paper and had Oscar sign it. I got the prop and flew it in January and February of 1999 and it performed exactly as predicted speed and rpm-wise.
After I completed my Lancair, I flew it with this prop initially, but my 125HP engine wouldn't turn it as fast as Oscar's 160 HP, so I designed a three-blade prop on 2002 July 25 and sent it off to Craig Catto to make for me. Again, I wrote down its predicted performance as before and had Oscar sign and date it, and it performed about 1-2 mph faster at the altitudes and rpm during testing at the end of 2002. I was getting really good cruise and climb efficiency, so I sent an e-mail to C.A.F.E to come to Santa Maria and check this thing out since they were supposedly interested in aircraft efficiency. They wrote back that they no longer did that, but they would pass this on to their prop expert. Several weeks later I got the first of many several-hour calls from Jack. He told me about Andy Bauer and him and their BGT concept, and asked me about my design concept. I told him that since Prandtl thought that the elliptical lift distribution on a wing was best, I felt that it was the best goal in my design equations. I didn't hear again from Jack until after the 2004 Reno air races when Tom Aberle's Phantom biplane, with the three-blade prop I designed for it, performed so outstandingly and garnered much publicity. From then on Jack would call or send me e-mails and mail print-outs of sections of his yet-to-be published book, whose first edition came on 2007 July. So as you can see from this tome, my original two-blade prop pre-dated any contact with Jack by over four years. I was delighted to see that Jack had included a comparative description on P. I-28 - Bk II of the BGT and elliptical designs and notes that the elliptical design, by moving the blade loading more inward, would "logically lose less off the tip, better than the B-G-T ideal" 'Nuff said!
 
Norris said even more

It's too much to type, so here is a PDF scan of the pages near where Paul is referring. Jack said of Paul: "Paul Lipps, of Arroyo Grande, California, a highly intelligent Radar Pro and Private Plane refiner, clearly deserves to become the 11th. benchmark achiever in the history of Propellers, for creatively providing us this path to ultimate efficiency, for the Second Century of Flight." The capitalization and boldface are Jack's.

Of course, if I were Paul, I'd be very proud of that. I agree that Jack is worth listening to, even when he disagrees with one. He's politely implied some rather unflattering things about some of my "work", so I know that feeling.


Read the PDF to get an understanding of what Jack is talking about. The handwritten notes are Jack's. To paraphrase an infamous Los Angelino, let's "all get along" while we seek clarity in discussing our disagreements. Jack goes on, in his lengthy book, to describe how the work of Betz, Goldstein and Theodorsen (BGT) (ending in 1948) can be applied via the "modern personal computer" (circa mid-eighties) to truly optimize the prop. Jack defines that optimization.

Paul and Jack do not completely agree, but they agree on a lot and Paul was clearly the first to discover some important truths about props and more importantly, to demonstrate them in reality. The thing that I, personally, think is worth noting is that Jack's design appears to accelerate extremely well from a standstill while Paul has at least implied that his design gives up roughly 10% on takeoff (I'm not sure what the standard of comparison is). That statement is based on Paul's own posting and Jack's prop's performance at Reno. This, to me, a mere layman, implies there are things yet to be learned and/or proven. I have no data whatsoever on which prop's performance is better at the top end or under what conditions of density altitude and RPM.

In closing, here's a thought experiment. Let's assume there is a prop that is truly 95% efficient, net of fuselage interaction. Let's further assume that the good prop increased a given airplane's top speed at a given condition by 10% which is 20 mph for an RV. Let's use the roughly accurate cube rule to discover that the thrust HP increase required to get 10% more speed is approximately the cube of 1.10 or 1.33. We just concluded that the first prop had a net efficiency of 0.95/1.33 = 71.4%. A modern sailplane with a 60:1 glide ratio has a L/D efficiency of 98.36%. It is a given that a rotating wing (prop) is not going to be that efficient and could not employ a similar aspect ratio. I don't know what the limit is on prop efficiency nor on net prop efficiency after taking into account fuselage interactions, but 95% seems beyond that limit to me.

I'm not saying anything about the results that Paul got, only making a mathematical deduction about the starting point. A C-152's net prop efficiency is in the range of 66.5% or less, mostly a little less. RV's are a lot cleaner; the same prop on an RV would perform better than on a C-152, net of fuselage interaction. the CAFE RV-6A's net prop efficiency was in the neighborhood of 80%. If an 80% prop improved by 33% it would produce 106.4%, thrust HP vs. BHP.

 
Is an elliptical distribution really the way forward...?

Obviously when running high alpha minimising tip-votices is a prudent thing to do, but is running a propellor blade at high alpha the most efficient thing to do? Clearly when on the beginning of a take-off roll, you have little choice, but in cruise and climb, surely you'd be needing each blade to be running at best L/D, in which case I would imagine tip vortices are a lot less significant compared to mach/compressibility loss issues?

To my feeble imagination, reducing the tip chord, which Paul Lipps does, reduces tip losses in a mach/compressibility sense more so than by achieving an improved lift distribution.

Just thinking and wondering out loud... :confused:

A
 
Last edited:
About L/D Max in a prop

Obviously when running high alpha minimising tip-votices is a prudent thing to do, but is running a propellor blade at high alpha the most efficient thing to do? Clearly when on the beginning of a take-off roll, you have little choice, but in cruise and climb, surely you'd be needing each blade to be running at best L/D, in which case I would imagine tip vortices are a lot less significant compared to mach/compressibility loss issues?

To my feeble imagination, reducing the tip chord, which Paul Lipps does, reduces tip losses in a mach/compressibility sense more so than by achieving an improved lift distribution.

Just thinking and wondering out loud... :confused:

A
It was also my first thought that L/D max would be the best. However, the HP minimum needed to support a given airplane weight happens at a lower speed, (higher alpha) the speed for best endurance rather than best range. If we translate that horizontal truth to the vertical, we get the same thrust (forward lift) at a slower speed (RPM) with less horsepower. In the case of the wing, the speed is approx. 76% of L/D max and the HP is lower by about 14%. But this is a simplistic analysis. Jack Norris points out there are many complexities which alter this. He uses a CL of 0.5 for best performance or 0.55 for almost as good with a smaller, lighter prop. No, I can't explain it, but his book does. www.propellersexplained.com . Notwithstanding that Paul's tips are different than Jack's, they are both trying to minimize tip vortices which is the consequence of elliptical distribution.

BTW: here is a graph of typical alpha vs. CL. You can see that at 0.5 CL, alpha is quite low. http://en.wikipedia.org/wiki/File:Lift_curve.svg You have to read the book to understand why this CL is used. I'm still not sure that I do.

 
Last edited: