1. #1
    AndyJWest's Avatar Senior Member
    Join Date
    Jul 2009
    Posts
    5,656
    Some of you may have seen a bit of discussion on ground effect in this (stickied) thread: Patch 4.10 - Development Updates by Daidalos Team. Runyan99 posted this question:
    Is it possible to add Ground Effect? It's a big flaw in the flight model.
    After a bit of discussion about how significant it was, Viikate_ posted this:
    Actually the ground effect IS already in the sim. Try taking Ju-88 for example, trim it perfectly for level flight at 100m, then bring it as low as you can over flat surface (water or runway for example). Plane needs some down elevator to maintain level flight, because GE is increasing the lift.

    I removed the GE from code and tested some planes that I'm most familiar with. Totally different world! I wrecked big bombers like Ju-88 ALWAYS in landings Veryhappy. So that extra little lift that current GE gives is very important. When I put the GE back, I could do nice landings with Ju-88 again.

    However the current GE implementation could be improved. It's too weak for bigger planes and starts to kick in too low.
    This aroused my curiosity, so I did a little testing with my prototype autopilot setup (see http://forums.ubi.com/eve/foru...121016097#4121016097). Interestingly, this showed that on a given throttle setting (75% in this case), the Ju-88 was about 9 km/h TAS faster in level flight at 10m than it was at 100m - contrary to what one might expect from IL-2 Compare etc, which suggests TAS should increase with height:



    This is the Crimea map, with about 85% fuel load. I allowed the speed to stabilise before taking the screenshots, though there is a little fluctuation due to limitations in the autopilot.

    I can probably run some more tests, but before I do, has anyone got any comments?
    Share this post

  2. #2
    danjama's Avatar Banned
    Join Date
    Apr 2005
    Posts
    6,363
    That second screenshot, the plane is not level. Surely it needs to be level and stable for an accurate result.
    Share this post

  3. #3
    AndyJWest's Avatar Senior Member
    Join Date
    Jul 2009
    Posts
    5,656
    There is a little oscillation in pitch, roll and yaw with the autopilot engaged, but this effect is no more noticeable at 100m than 10m, and since I allowed the speed to stabilise, I doubt this is significant. If anyone can hold altitude sufficiently accurately to duplicate the test without the AP, I'd be interested to see if they got the same results, though I can't think of any reason why there should be significant differences - the AP only makes small control inputs.
    Share this post

  4. #4
    TinyTim's Avatar Banned
    Join Date
    Dec 2007
    Posts
    1,781
    What values for speed do you get at similar setup if you fly manually and simply engage "level autopilot"?

    I can't test ingame at the moment, but if memory serves me well I think Ju-88 should be faster than this at ground level. 390kph pops to mind from somewhere, but I'd need to test to be sure. 100% fuel, standard loadout, rads closed.
    Share this post

  5. #5
    AndyJWest's Avatar Senior Member
    Join Date
    Jul 2009
    Posts
    5,656
    Originally posted by TinyTim:
    What values for speed do you get at similar setup if you fly manually and simply engage "level autopilot"?

    I can't test ingame at the moment, but if memory serves me well I think Ju-88 should be faster than this at ground level. 390kph pops to mind from somewhere, but I'd need to test to be sure. 100% fuel, standard loadout, rads closed.
    To be honest, I've never really used the 'level autopilot' function, and I'm not sure how accurately it will work at this sort of altitude - I tried the Ju-88 down to 5m, but the props were getting a little close to the water.

    This test was at 75% throttle - I was more interested in the difference than the actual speeds. If this really is modelling ground effect, I'd expect the difference to be greater (in percentage terms) at low speeds than at full throttle - ground effect should reduce induced drag, rather than parasitic drag, and this will be more evident the slower you go.
    Share this post

  6. #6
    TinyTim's Avatar Banned
    Join Date
    Dec 2007
    Posts
    1,781
    Originally posted by AndyJWest:
    This test was at 75% throttle
    Ah, ok. I was under impression you are trying to measure the difference at max. speed.

    Slower speeds make perfect sense. One more idea (again without autopilot) - trim plane very accurately for level flight at a certain (low-ish) speed as close to the ground as possible, then take the plane up at, say, 30 meters, and see if the trim is still good. If ground effect is indeed modelled, it shouldn't be.
    Share this post

  7. #7
    AndyJWest's Avatar Senior Member
    Join Date
    Jul 2009
    Posts
    5,656
    I think that was more or less what Viikate was doing. The only reason I'm using my autopilot is that it is easier to hold steady(ish) conditions for long enough for the speed to stabilise, and for repeatability.

    I repeated the test at different throttle settings:

    110% rads open - 100m 376 Km/h, 10m 382 Km/h.
    110% rads closed - 100m 378 Km/h, 10m 392 Km/h.

    51% rads open - 100m 252 Km/h, 10m 267 Km/h.

    The 51% throttle test is rather slow for the AP to handle properly, and the aircraft rolls by about +- 4 degrees, together with lesser changes in pitch and yaw. I never attempted to cure this problem, as the intention was for the AP to be used at cruising speeds, rather than on the edge of stall - it tends to work better with single-engined aircraft too.

    EDIT ----
    Does the 'level autopilot' command actually do what it was intended to any more? It doesn't seem to...
    Share this post

  8. #8
    TinyTim's Avatar Banned
    Join Date
    Dec 2007
    Posts
    1,781
    Originally posted by AndyJWest:
    Does the 'level autopilot' command actually do what it was intended to any more? It doesn't seem to...
    If I recall correctly, there's a mess in a key configuration panel about the autopilot, level autopilot and autopilot automatization. They are named misleading or even plain wrong (switched).

    But yeah, level autopilot still works.
    Share this post

  9. #9
    the Ju-88 was about 9 km/h TAS faster in level flight at 10m
    Your speed will increase while in GE. Same power but with less drag equals more power available.

    It is not a "lift increase" but rather induced drag is dramatically reduced. Lift and drag are connected by design. The reduced in drag means the wing can lower its angle of attack to provide the amount of lift required. This increases the amount of available AoA to the aircraft. You can now use the same AoA at a lower velocity.

    Lift production stays the same and will only meet the amount of force required.
    Share this post

  10. #10
    AndyJWest's Avatar Senior Member
    Join Date
    Jul 2009
    Posts
    5,656
    Originally posted by Kettenhunde:
    <BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">the Ju-88 was about 9 km/h TAS faster in level flight at 10m
    Your speed will increase while in GE. Same power but with less drag equals more power available.

    It is not a "lift increase" but rather induced drag is dramatically reduced. Lift and drag are connected by design. The reduced in drag means the wing can lower its angle of attack to provide the amount of lift required. This increases the amount of available AoA to the aircraft. You can now use the same AoA at a lower velocity.

    Lift production stays the same and will only meet the amount of force required. </div></BLOCKQUOTE>

    That is what I was expecting. The results of trying this at different throttle settings seem to me to confirm this.

    I may try this again with a P-38. The contra-rotating props tend to result in reduced oscillations at low speeds, though I suspect there may be more to it than that.
    Share this post

Page 1 of 43 12311 ... Last ►►