[elippse]

Here's a rather simple formula to determine your plane's best L/D speed. This is not best ROC speed, since both L/D AND prop efficiency figure into it! It requires you to know the plane's weight, W, aspect ratio, AR, wing area, Aw, wingtip-shape effect on the wing's Oswald Efficiency factor, OE, and a good estimate of the plane's equivalent parasite drag area, Ap. What this will give is CAS in mph.
CAS, mph = 15/22[4 X W^2 / (rho^2 X Ap X Aw X pi X AR X OE)]^1/4
where rho = 2.377E-3 and OE for a slashed tip is 0.8 to 0.82, and a rounded tip is o.7 to 0.75.
For an RV-6A with 110 sq ft, 1440 lb, OE 0.81, AR, 4.8, and Ap 2.3 as described in the Sport Aviation article, CAS is 100.7 mph.

 

[Kevin Horton]

The CAFE foundation test for the RV-6A matches this reasonably well, as the formula predicts a best L/D speed of 107.4 mph CAS at 1650 lb, and they found 106 mph CAS via flight test. The difference is well within the range of error expected in a flight test.

 

[hevansrv7a]

Finding The Factors

Without knowing either the MinimumSink speed or the BestL/D speed, how would one estimate the parasite drag of, for instance, a -7A? Wouldn't a more direct approach begin with the MinimumSink speed? If you already know the parasite drag at either MinSink or BestL/D then at BestL/D, the induced drag is the same, yada yada. I have no issue with the formula, but how does someone with only an airplane and it's own instruments use it?

 

[Kevin Horton]

Without knowing either the MinimumSink speed or the BestL/D speed, how would one estimate the profile (i.e. parasite) drag of, for instance, a -7A?

If you have a type-certificated prop and engine, you could measure top speed, then use published power charts and prop efficiency maps to determine the total drag. Use the classical induced drag theory (e.g. based on wing area, aspect ratio, Oswald span efficiency, etc) to estimate the induced drag, and subtract that from the total drag to get the profile drag.

Or, you could make an estimate of the profile drag based on the CAFE data for the RV-6A. I suspect the RV-7A flat plate drag area is only a bit more than the RV-6A. The fuselage, landing gear and empennage drag should be similar. There would be a bit more drag from the extra wing span, as a SWAG, say the flat plate drag area is maybe 2-3% more than the RV-7A. You could refine this number by comparing Van's published max speed data - the induced drag is so low at the max speed point that any difference in top speed at the same power is largely due to differences in flat plate drag area. I note that Van claims the exact same top speed for a 180 hp RV-6A as for a 180 hp RV-7A, which would suggest the flat plate drag area was very similar. Maybe the decreased drag from newer wheel pants makes up for the profile drag from the longer wing span

Wouldn't a more direct approach begin with the MinimumSink speed? If you already know the parasite drag at either MinSink or BestL/D then at BestL/D, the induced drag is the same, yada yada. I have no issue with the formula, but how does someone with only an airplane and it's own instruments use it?

If you measure minimum sink speed with engine turning, you can't separate thrust or drag from the prop from airframe drag. If you are prepared to shut the engine down, and slow to stop the prop you might get useable data. I plan on eventually doing such tests, right overhead a suitable airfield, and after I am well practiced in forced landing patterns.

Whatever test approach you use, gather data from several flights. The data from any one flight could be suspicious, and lead to to bad conclusions. If you do the same test several different times, you can see whether you have a repeatable question or not.

 

[hevansrv7a]

...

Or, you could make an estimate of the profile drag based on the CAFE data for the RV-6A. I suspect the RV-7A flat plate drag area is only a bit more than the RV-6A. ...


If you measure minimum sink speed with engine turning, you can't separate thrust or drag from the prop from airframe drag. If you are prepared to shut the engine down, and slow to stop the prop you might get useable data. I plan on eventually doing such tests, right overhead a suitable airfield, and after I am well practiced in forced landing patterns.

Whatever test approach you use, gather data from several flights. The data from any one flight could be suspicious, and lead to to bad conclusions. If you do the same test several different times, you can see whether you have a repeatable question or not.

I think these excellent and well intentioned contributions go to support my point. First, if the -7A in question has a different cowl, plenum and wheel fairings, then the CAFE report is off to an unknown degree. Second, we are just running around the other side of the circle. The formula works if you already know what you are trying to find out but it doesn't help you find out what you want to know if you don't already know it.

I agree about gliding tests with engine turning. However, I think we agreed earlier that the lowest power setting that will maintain level flight is at the speed that equals minimum sink (correcting my earlier thought that it was best L/D). So that method can find minimum sink reasonably accurately and then best L/D is just minimum sink times 1.316. When I get a little better weather (Michigan is much like Ontario) I will try to demonstrate this with engine off and prop stopped (above an airport!).

I suggest that this is emprically sound and that the formula can then be used to derive the various factors with reasonable accuracy.

 

[Kevin Horton]

I agree that the speed for minimum power required in level flight equals the speed for minimum sink rate in a glide. It won't be easy to accurately determine this speed via flight test though, as there will be a range of speeds where the power required will vary very little. Also, if you get too slow, you'll be on the back side of the power curve, which makes it much harder to fly a stabilized test point. It is worth doing the test - I'm just not sure how accurately you can determine the speed for minimum power required.

I do agree that the speed for best L/D (with prop stopped) should equal 1.316 times the speed for minimum power required in level flight.

If the intent is to validate your assumed induced drag model, I think you will also need to nail down a high speed point so you can determine the profile drag. Then you can play around with the flat plate drag area and Oswald's span efficiency factor to find the combination of those two that best fits your data.

 

[hevansrv7a]

Follow Up for Kevin

I agree that the speed for minimum power required in level flight equals the speed for minimum sink rate in a glide. It won't be easy to accurately determine this speed via flight test though, as there will be a range of speeds where the power required will vary very little. Also, if you get too slow, you'll be on the back side of the power curve, which makes it much harder to fly a stabilized test point. It is worth doing the test - I'm just not sure how accurately you can determine the speed for minimum power required.

I do agree that the speed for best L/D (with prop stopped) should equal 1.316 times the speed for minimum power required in level flight.

If the intent is to validate your assumed induced drag model, I think you will also need to nail down a high speed point so you can determine the profile drag. Then you can play around with the flat plate drag area and Oswald's span efficiency factor to find the combination of those two that best fits your data.

Yes, it is a little tricky to get to least power level flight, but I've got it over several trials within plus or minus 1.5 kts or better, I think. I also think it improved as much as 2 kts when I realigned the wheel fairing - as would be expected. Using the GRT with %Power indicated helps a lot and I don't think that it matters how valid the reading is so long as the changes can be detected easily. Fuel flow also supports this if mixture is not changed. Validating the airspeed thus determined would be (will be in my case) aided by an AOA instrument and a VSI. Once the best guess is determined, the engine-off test can be performed from a high altitude starting with the AOA found during power-on testing and bracketing the speeds one kt/mph at a time while watching the VSI for minimum reading. This can perhaps be done before the air changes density enough to mess up the findings.

I think that finding BestL/D directly through flight testing is at least as tricky, requiring many more iterations (and calculations) through changing atmosphere conditions and still not solving the problem of the contribution (negative or positive) of the prop. At least with the MinSink method you have a chance to get a stable reading over a minute or more in the same atmostphere.

Here's the question: Isn't the parasite drag equal to the induced drag at BestL/D? And don't the drag polar curves usually show only those two kinds of drag? And don't both those curves vary as the square of the velocity, one up, the other down? Where in the drag polar does the profile drag fit? How is it determined from top speed if power is unknown?

As Vinnie Barbarino said, "I'm so confused". And now he's a very accomplished pilot, so there's hope for me.

Thanks.

 

[Kevin Horton]

Here's the question: Isn't the parasite drag equal to the induced drag at BestL/D? And don't the drag polar curves usually show only those two kinds of drag? And don't both those curves vary as the square of the velocity, one up, the other down? Where in the drag polar does the profile drag fit? How is it determined from top speed if power is unknown?

Profile drag is just another name for parasite drag.

Yes, both induced drag and parasite drag vary as you described, as long as the mach number is low enough. Mach number should not be a problem for RVs below 10,000 ft.

If we don't know the power, and the prop efficiency, the only thing we have left is engine off, prop stopped glide testing. The force of gravity provides the power required to maintain the speed. The weight times vertical speed (with appropriate unit conversions) will tell us the amount of power required at the speed of interest.

 

[Bevan]

Don't we have to measure it?

The weight times vertical speed (with appropriate unit conversions) will tell us the amount of power required at the speed of interest.

I'm no expert, but it seems to me that the complex formulas are methods for calculating things such as theoretical Best L/D based on previous tests. To actually measure something, don't you just have to go out and measure it, as Kevin descibes above, to validate the theory?

Very interesting thread.

Bevan
RV7A wiring