investigating FB.VI deck speed is ~15 mph too slow

[Racoon]

Just run some tests yesterday on the actual Open beta build. Tried to fly with opened and closed radiators.

Screenshots and the track included.

Mossie Max speed.trk

[Bozon]

@Racoon and @Yo-Yo

Thanks for posting the above recording, it helped me find the source of the speed difference.

There is an issue with the slip/turn indicator. In the recording above, the rudder trim is pointed at the "R" of the "TRIM" label. At that state the slip needle is NOT CENTERED but the mosquito reaches 340 mph - faster than my tests. Rotating the rudder trimmer to the neutral position (white triangle) also centers the slip needle - this is how I flew in my tests. My no-slip trim state should have been faster, however it is the other way around!

To confirm, I tested the Mosquito with the un-centered slip needle and indeed was able to obtain 296 kts (340 mph). Then trimmed the rudder to center the slip needle and lost 7-8 mph to get the familiar 332-333 mph result.

Edited June 28, 2023 by Bozon

[Art-J]

Ha! That was it then. I was also making sure to cancel slip by trimming and watching the needle... as that's what one expects to do to minimize drag in any airplane.

Quite puzzled to see the sideslipping (?) Mossie being faster than trimmed one, what's up with that?

Or could it be aileron trim modelling issue as well? Shouldn't be, but anything's possible at this point. I'm asking because I can see on Racoons's screens (thanks for these BTW) aileron trim zeroed, while I have to keep two full notches to the right all the time in most flight phases to fly hands off without Mossie wanting to roll left. It's been that way since module release.

Edited June 29, 2023 by Art-J

[Yo-Yo]

[Yo-Yo]

Ok, I tested it carefully. As you can see both rudder trim position give the same energy and, thus, max speed. Screenshots show needle position for both cases.
Vvert equiv on the chart is Vvet equiv 2

 

[Hallsy]

This is phenomenal data, everyone. Really appreciated from my end. I am doing some tests again myself. Thank you very much!

[Art-J]

Thanks to @Racoon and @Yo-Yo for all the data. I watched Racoon's track and I still don't know how you guys manage to get up to and maintain 340 for any longer period of time. Tested the plane again with ~20 000 lbs weight and the results were puzzling (replay track attached):

a) If I fly off-trim on purpose, like in Racoon's screenshots, I can indeed get up to 340 for a short time;

b) If I want to trim the plane for coordinated, hands-off flight, I need to use ~2.2 div down for elevator, ~1.6 div right for aileron and about 0 for rudder. That, however, will inevitably cause the speed to drop to ~335 at best, maybe a bit higher shortly when I dive a little.

Must be some form of a trim drag then. Not a big problem, just curious what's going on.

Mossie speed test 2.trk

[Yo-Yo]

I watched your track and I can say that the problem is not in the FM but more in the way you are trying to measure the speed. Even with ideal trim the plane was not hold  at the same altitude for a reasonable time to find the max speed. Then, as I made the same E-record for you track, I noticed that there is a significant difference between my test and yours: as you can see, the points that are filtered for g=1 with small tolerance are very dispersed for TRIMMED flight and, vice versa, more condensed for UNTRIMMED.  It is an evidence that your flight was performed with constant g changes, small, but bleeding enough energy, For untrimmed flight the points are more solid, more organized, and the speed is exactly as in my test.

I can suggest that it could be a joystick issue, so for these precise tests the curve must be set like this.

 

null

null

[Art-J]

Thanks for taking a look at the track. It's true I was experimenting with different trim settings every couple of minutes to try to find "the best" combination based on the screenshots you guys provided.

I use Warthog sick with 30 cm extension, which gives me a pretty good precision when held in some off-center position (maybe that's why "untrimmed" points less dispersed), but it's obviously not convenient in the long run. That's why I prefer the "trimmed-for-hands-off" approach to avoid touching the joystick at all. But since I use keyboard for trimming (to get shorter input pulses compared to hat switch) it's not easy to keep the variometer locked at zero indeed, even with short taps of elevator trim keys. So maybe that was the source of the problem indeed. I'll try harder next time ;).

[Ala13_ManOWar]

27 minutes ago, Art-J said:

Thanks for taking a look at the track. It's true I was experimenting with different trim settings every couple of minutes to try to find "the best" combination based on the screenshots you guys provided.

I use Warthog sick with 30 cm extension, which gives me a pretty good precision when held in some off-center position (maybe that's why "untrimmed" points less dispersed), but it's obviously not convenient in the long run. That's why I prefer the "trimmed-for-hands-off" approach to avoid touching the joystick at all. But since I use keyboard for trimming (to get shorter input pulses compared to hat switch) it's not easy to keep the variometer locked at zero indeed, even with short taps of elevator trim keys. So maybe that was the source of the problem indeed. I'll try harder next time ;).

Are you aware that variometers never, ever, read instantly, right? Well, IIRC some really old ones do, and they weren't very useful jumping all around constantly. Usually vario has a delay, a few seconds until settled, so if you chase the needle you won't get a stable reading at all. Maybe that's one source of your problems. Watch the altimeter instead if you wanna check stable flying and trim with that.

Besides, a couple of minutes, even if you managed to really set it absolutely stable, isn't enough to achieve top speed anyhow. It's usually a really stable long run thing more than a few moments to reach it. That's why it's commonly really hard to get that top speed.

[Yo-Yo]

37 minutes ago, Ala13_ManOWar said:

Are you aware that variometers never, ever, read instantly, right? Well, IIRC some really old ones do, and they weren't very useful jumping all around constantly. Usually vario has a delay, a few seconds until settled, so if you chase the needle you won't get a stable reading at all. Maybe that's one source of your problems. Watch the altimeter instead if you wanna check stable flying and trim with that.

Besides, a couple of minutes, even if you managed to really set it absolutely stable, isn't enough to achieve top speed anyhow. It's usually a really stable long run thing more than a few moments to reach it. That's why it's commonly really hard to get that top speed.

You are absolutely right. The best way to use variometer is to combine its readings with aircraft pitch (horisont line, clouds, etc) to have steady pitch (steady g-load=1) for a certain time. It's not a big mistake, for example, to have -0.5 m/s for 10 seconds, then +0.5 for another 10 seconds and take the average from two TAS readings.
The most accurate way to obtain it is to use, of course, energy curve. If the flight was steady for a while, the real max speed can not be even reached a bit, but extrapolation wil give the exact value.

[Bozon]

This is great, but just remember that the historical tests we compare to were performed by a human flying the real plane, including the minor control inputs and imperfect coordination of flight.

I still get a perfectly straight slip needle when rudder trim is exactly neutral (knob on the white triangle), and that gives a slower speed than a rudder trim offset to the right and a small left-leaning slip needle.

I am away from home for the next two weeks - after that I will map the top speed vs. rudder trim (and resulting slip reading), so we’ll be able to see where the max speed is obtained.

[Yo-Yo]

I can not agree, that in RL is harder to obtain steady flight because in RL a pilot has instant acceleration feedback.