Performance evaluation proposal


Here is a proposal for a comparative performance evaluation.  While it could be done for many types of aircraft, I’m primarily picturing this event taking place at the Zenith Aircraft factory’s annual Homecoming in September.  It is focused on highlighting the aircraft created by builders, and gathering data from them to assist current and future builders in making choices which tailor their projects to better suit their particular needs.


In recent years, STOL contests have become very popular, and Zenith has run them at the Homecoming as far back as eight years ago. In national and international STOL contests, regular Zenith STOL planes have consistently demonstrated their performance and value, even against purpose built, trailered in, aircraft costing many times more.  This is good, but STOL contests only provide one facet of a planes performance, and most pilots don’t fly that way on a typical day. This performance evaluation is aimed at gathering comparative data over a much broader flight envelope, one that represents more typical use of the designs.


Although I am an engine guru, I have grown weary of nearly every opportunity for builders to learn, being distorted into a marketing opportunity for somebody. I have spoken at many Zenith gatherings in the last 15 years, and other than the engine panel discussions, I make a point of not using the word “Corvair”,  when doing so. Marketing has a place, but it should never be allowed to displace builder camaraderie or learning.  While the very nature of the evaluation will provide accurate, useful data for comparing power plants, my goal is to provide a much broader overview of the performance potential of the designs, their utility and to specifically highlight the achievement of individual builders and encourage those currently building.


Above, Phil Maxson’s 601XL over the Florida coast at Ponce Inlet, 2006.


Here are the steps a builder takes , participating in the evaluation: 


A ) After being warmed up, the plane’s fuel tanks are topped off to the bottom of the filler necks, a level that can be visually replicated at the completion of the flight.


B ) The plane is placed on electronic scales and weighed in, without the pilot.


C ) Plane heads to the designated runway and takes off.  At he 800′ mark there is a very light tape spaning the runway, 6′ off the ground. Any Zenith model will clear this without difficulty, even with full tanks, but we have it to preclude anyone using a very high prop pitch setting which would skew the cross country speed evaluation.


D ) Plane proceeds to fly a designated 3 leg course, approximately 100 miles.  Two very visible land marks are used as pylon turn points. The flight is done at a set altitude, perhaps 2,500′. There are observers at the turns, pilots make a radio call when approaching. The visual is just a back up, the data can be collected for any aircraft with a GPS with a system like this:


E ) The three legs are flown, the last one over flying the point of origin at 2,000′. The time to fly the course is halted there. Plane comes back to land and proceeds immediately to the fuel pumps.


F ) The plane is re-weighed for fuel burn.  The plane is topped off, refilled to the exact same location on the filler necks, and re-weighed to confirm the fuel consumption.


G ) Plane goes back out. It must use the exact same configuration as the cross country course, particularly the prop pitch setting.  Pilot climbs to 2,000′ and flies at a stable, level speed of 75mph for one minute. From there, he climbs at any airspeed he chooses, directly to 4,000′.  He calls on the radio, but again the performance is measured electronically.


H ) Builders can fly the cross country portion at any pace they like, but they are encouraged to demonstrate flat out performance. I  would also like at least some of the planes to do a second run at the course completing A – F again, but at a typical cruise power setting for comparative purposes. This could be run at a single model appropriate speed, such as running the 601/650’s at 115 mph, as it would give a comparison of fuel burn rates at casual cruise setting.




What do we get from this? Real data. If you have been in the EAA for 30 years, you can remember the CAFE foundation’s performance evaluations, published for many aircraft, in long articles in Sport Aviation. This was information I treasured. The EAA was pressured by some kit manufactures to not publish information that contradicted their marketing departments. That was a transition point of the organization being asked to serve the manufacturer and not the member. In the internet age, things have deteriorated, and quality information is even harder to find, and the hidden compensated relationships between the ‘evaluator’ and the product being harder, not easier to see. We can lament a bygone time, or we can take actions to provide our own real data. These projects have lasting value. Ask any Pietenpol builder, the W&B work done a decade ago by Ryan Meuller and myself has improved a generation of Piet’s by by giving their builders real data to plan their builds and making the operation of the design safer. It is hard to overstate the lasting effect of real, accurate information.


Who does this serve? Builders. I speak to people potential builders who mistakenly think they must have a STOL model to fly off a 2,200′ airstrip; there are also people who float on the choice between a STOL and a Cruiser, have questions about typical builders useful loads vs factory design prototypes. The 100 mile cross countries will provide a very good look at the general potential utility of each design. Comparisons of a sustained 2,000′ climb under known conditions will answer many questions builders have. This type of data will be much more useful than pictures of glass cockpits with a single set of information.  The better data builders have going into the process, the much more likely they are to finish their plane, and they are vastly more likely to be happy with their creation.  The current data available is ok, but a lot of it is really marketing and not an evaluation. This can be corrected in one weekend, and it will have the additional benefit of being a deterrent to future excessive marketing claims.


This will be a builder focused event. I don’t want to include data from aircraft that belong to, or were built by engine guys as demonstrators….  Between parts from myself and Dan from SPA, I can assemble a 3,550cc Corvair, and I have nearly 40 years of experience with nitrous oxide installations on all kinds of vehicles. But a $22K Corvair festooned with Fogger nozzles isn’t what I teach builders to assemble nor what we sell. It would be useful if this was a ‘contest’ but it is not, it’s an evaluation.


Additionally, the factory planes will be left out also, because, like planes coming out of professional shops, they are not representative of current builders creations. The evaluated planes will be in the stock airframe configuration with only minor mods like fairings.

Additionally, each plane must complete all the steps A-G. If the weight of the plane is a ‘secret’ or a plane with a very high pitched prop and poor climb performance doesn’t want section G recorded, it will not give a complete data set, and that would only serve someones ego or marketing plan, not builders.


I value everyone’s constructive input here. Please use the comments section to add any thoughts which might improve this concept, or explain how you would find this data useful.  Thanks in advance.








19 Replies to “Performance evaluation proposal”

  1. That is an amazing concept. I have become very frustrated and somewhat jaded after years of being wowed by the claims of many that never pan out. It is true of trucks, greenhouses, and of course home built aircraft. I am suspicious it may be difficult to get a large enough and broad enough cross section to be as useful as this could be, but WOW, what a database it would be. I think of all the guys that have invested in hangers full of parts that will never work and I have seen so many planes finished at great sacrifice only to find out their particular conversion/ PSRU doesn’t actually work. They end up leaving home building and often aviation in general. Thank you for being our champion.

    I really enjoyed the comparison flight of the rv 10 vs a Cirrus. Very enlightening and encouraging to see the rv-10 perform as advertised. It would be interesting to mix some certified GA planes into this analysis and comparison. I also like the idea of the side by side comparison of a flight video because it is really easy to track the differences visually and is easier for non- statistic people to get the message. Perhaps some of these evaluations could be safely flown two at a time?

    1. OK, Who said I was going to be comparing 601s to 701’s? Why would you assume That the Corvair guy would bring a junk pile to an evaluation? I said I would not evaluate aircraft done by professionals, so why are you suggesting I’m going to have “factory-installed and factory tuned engines” in this? Why are you suggesting that we have ” full factory support is implemented,” when I have said it is about builders aircraft in half a dozen places in the story?

      I’m a bit concerned that other people are misreading the story also, and deciding based on their incorrect reading that I’m going to get an ” erroneous outcome.” Honestly, get a coffee and actually read the story and apply a bit of common sense. I have a long track record of running good educational events, and valid testing. Making statements about the proposal without reading it is ridiculous.

  2. Hey William, I would like to see this kind of testing with Bearhawk LSAs flying Corvairs. There are probably not enough of them flying Corvairs to get a good data set but its not because I dont push the engine with every chance I get, Im a loyal disciple of the Corvair gospel!!!

  3. Maybe another point to be recorded is the distance from start of roll to the point of lift oft.

  4. I would find real data like this very useful, and if my plane is done, I would participate.
    I too read the CAFE reports with great interest, and wished they had done more, especially of models that were typical of ‘rank and file’ builders and pilots.
    One possible addition, to complete the landing performance picture, would be to make one landing over the hypothetical “50 ft obstacle” as typical builders could expect to achieve. Riding with Roger on my demo was a treat, but he could make any Zenith dance and sing whatever tune he wanted. To confirm that mere mortals could match it in real life would be valuable data I would use for sure.

  5. Good luck with that, William!
    Once you get that rolling, how about a similar program to get reliable data on the weight of alternative engines. You’d think it would be easy, (just set it on a scale) but it’s not. Next, horsepower—harder, but possible with your dynamometer setup, as you proved in ’04 with your test of the O-200. Of course, an airplane doesn’t fly on horsepower, it flies on thrust. It would be even harder to determine thrust, but a sophisticated computer program could do it from propeller data and carefully measured flight data. (Static thrust won’t cut it.) Climbing at 65 knots., a “65 hp” VW turning a 56” prop at 3,300 rpm has far less thrust than a “65” Continental turning a 72” at 2300, but there’s no way to measure it. Everybody would have to use the same computer program, of course.
    The bottom line is, I don’t think most people really want to know—especially vendors. They prefer unicorns to ponies!
    The plural of anecdote is not data.
    Dave Gingerich

    1. Hi Dave:

      Why do you say “Static thrust won’t cut it” ? Seems to me that test gives important data as long as it’s interpreted correctly.

      I think a ‘Static Thrust Shootout’ would be interesting.

      Done after the flight performance test William proposes, seems to me it would show which planes are optimised for a point of the envelope as opposed to optimized over all flight regimes.

      The results of the real world flight performance test might suggest a “correction factor”

      as mentioned at ,

      for comparing apples to apples in the static thrust test.

      Or crunch the numbers in the opposite direction: the static thrust measurement might suggest a correction factor for the flight test.

      The ‘correction factor’ might work like a sailboat racing handicap where they measure sail area, boat length, etc., for example. I think WW’s thrust testing accounts for differences in setup:

      “When this propeller was replaced with the 72″ Warp Drive, a prop appropriate for a plane with an 85-100mph cruise speed, the thrust shot up to 470 pounds.”

      – from

      So… static thrust, prop pitch and diameter, fuel burn, speed, time to climb, engine RPM, prop RPM, weight, and cubic inches displacement ? What else would be relevant?

      1. Static thrust is easy to measure and might be useful for comparing different props. What we really want is thrust at climb airspeed with WOT and redline rpm for a climb prop, or maybe at cruise airspeed with a cruise prop. Neither of these can be measured directly. Both are lower than static thrust.

    2. Dave,
      By definition, you would never have “climb airspeed and WOT and redline rpm” at the same time in a real flight condition, unless you use a prop set to over-speed in level flight at the same throttle setting.
      The timed climbs between specific altitudes give us an idea of actual climb thrust, and with the ability to measure prop setting on the ground, and knowing what the climb speeds were, we should be able to control for those variables.
      By repeating the course at typical cruise, as proposed, we will have some of that info as well.

  6. What a wonderful idea.

    Making it an annual event would give builders an opportunity to show self improvement, as well as provide a chance for people for whom September is a tight month. Plus it would generate more data as different airplanes would attend who otherwise only can every few years.

    I bet individual builders will kind of be in competition with themselves, meaning that an individual builder will want to do better in Year 2 than they did in Year 1. That means improving their Pilot Skills.

    A worthwhile endeavor.

    1. Pat, are you thinking about flying in for this if I can get it approved? I think Sebastien is looking for a cross section of builders to say they are in favor of it before he green lights anything. William

      1. Yes, I would like to. I have plans to be in Scotland then, but that whole Coronavirus thing might impact that. If it’s just me I’d fly the Zodiac to KMYJ, but if Mary comes we’d probably fly the Grumman. Either way it sounds like it’ll be a lot of fun.

  7. William,

    If you still have the gauge setup you used for the O-200 and Rotax static thrust tests, a ‘Static Thrust Shootout’ would be interesting.

    If you do it after the flight performance test you proposed, you could weed out people bringing in a ringer, a ‘one trick pony’ optimised for takeoff as opposed to a plane set up for normal use.

    Seems pretty useful to get reliable data as opposed to marketing based brochureware.

    It worked for tractors:

  8. Love the idea. I was unhappy when EAA cancelled the annual LBF 500 but it had kind of run it’s course when the AJ special showed up as a single purpose design to take all the prizes.
    My only suggestion is Keep It Simple. Specific and simple. Might be interesting to see how my RV-9A with a Subaru would score. Push the pilot to practice more for sure!

  9. I didn’t know a 3,550cc Corvair was possible. Wat kinda HP would that put out?

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: