PDA

View Full Version : Ground Effect?



Pages : [1] 2

AndyJWest
04-18-2010, 03:15 PM
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 (http://forums.ubi.com/eve/forums/a/tpc/f/23110283/m/5121031528/p/10). 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 (http://forums.ubi.com/eve/forums/a/tpc/f/23110283/m/4121016097?r=4121016097#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:
http://i958.photobucket.com/albums/ae65/ajv00987k/Ju-88-GE-10m.jpg
http://i958.photobucket.com/albums/ae65/ajv00987k/Ju-88-GE-100m.jpg

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?

danjama
04-18-2010, 03:31 PM
That second screenshot, the plane is not level. Surely it needs to be level and stable for an accurate result.

AndyJWest
04-18-2010, 03:41 PM
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.

TinyTim
04-18-2010, 03:51 PM
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.

AndyJWest
04-18-2010, 04:16 PM
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.

TinyTim
04-18-2010, 04:36 PM
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.

AndyJWest
04-18-2010, 05:15 PM
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...

TinyTim
04-18-2010, 05:33 PM
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.

Kettenhunde
04-18-2010, 05:34 PM
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.

AndyJWest
04-18-2010, 05:44 PM
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.

M_Gunz
04-18-2010, 05:57 PM
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>

This part:
You can now use the same AoA at a lower velocity.

And these:
It is not a "lift increase" but rather induced drag is dramatically reduced.
Lift production stays the same

Are they not contradictory?
I can see that with less drag you can go faster and get more lift for same AOA or same speed with more AOA
or same speed on less power but how to get same lift with same AOA at lower velocity without "lift increase"?

Kettenhunde
04-18-2010, 06:03 PM
Are they not contradictory?


No they are not. In fact this is a key concept to the correct understanding of how airplanes work.

Look at the force of lift over Coefficient of Lift and AoA:


We can also see in the figure that the lift force curve is perfectly constant, which is reassuring,

http://www.vanhaesendonck.com/...s.html#fig-force-ias (http://www.vanhaesendonck.com/de_wouw/how/htm/4forces.html#fig-force-ias)

GE allows our wing to move to a lower Coefficient of Lift at a lower angle of attack at a slower speed to produce the same amount of lift force.

AndyJWest
04-18-2010, 06:08 PM
In a steady state, lift = weight. The point is that in ground effect you can get the same lift with less drag, or in the (simpler to demonstrate) case I give with constant power, same lift with lowered induced drag gives excess power, and hence increases speed. It is easier to measure speed accurately in IL-2 than induced drag, though the results I get make sense in this context.

It might just be possible to measure the change in AOA (though not the absolute values, as I don't know the angle of incidence relative to the aircraft datum) - this will need averaging a decent sized sample of values at each of the different speeds and altitudes.

Kettenhunde
04-18-2010, 06:55 PM
Good explanation, AndyJWest.

WTE_Galway
04-18-2010, 07:44 PM
Most obvious example of ground effect in real life is these things ...

http://www.youtube.com/watch?v...PzDM&feature=related (http://www.youtube.com/watch?v=TYFEFekPzDM&feature=related)

JtD
04-18-2010, 10:13 PM
Try switching off "wind and turbulence" and the extra km/h at low altitude are gone.

JtD
04-19-2010, 12:29 AM
The whole thing got me curious, so I did a little test, where I'd fly the P-38 at 2-5 and 40-50 meters altitude at various low speeds. This would give some AoA's and some cl's. The graph below sums it up, while I wouldn't want to bet at the exact characteristics (I would need to keep altitude exactly constant for that), it is obvious that there is some extra lift at low altitude in game and the ground effect is modeled. Please note that wind and turbulence have been turned off for this test, you leave it on you get other results.

I'd suppose the ground effect goes up to something like 5m only, which is why the red line is steeper than the other - I started at 5m and kept descending down to 2-3m. So it might be parallel to the green at a higher level. Like I said, not exact, just proof of concept.

http://mitglied.lycos.de/jaytdee/testgraph/groundeffect.JPG

AndyJWest
04-19-2010, 04:54 AM
Interesting stuff, JtD. The effect of turning off "wind and turbulence" needs further investigation - I'll repeat my tests with this.

I've also had a brainwave - I may be able to run a test at constant altitude above sea level, running over one of the mountain-top runways on the relevant map, just to confirm that it is proximity to the ground that is causing the effect. More to follow...

thefruitbat
04-19-2010, 06:28 AM
one of the patches a while back, forget which one, had really prenounced ground effect, and then it got toned down with future patches.

JtD
04-19-2010, 09:27 AM
Originally posted by AndyJWest:
I've also had a brainwave - I may be able to run a test at constant altitude above sea level, running over one of the mountain-top runways on the relevant map, just to confirm that it is proximity to the ground that is causing the effect. More to follow...

It might be easier to fly on a map like Burma, where you have rivers at altitudes higher than 0. More flat terrain than just a runway.

runyan99
04-19-2010, 09:51 AM
If ground effect is working, then I stand corrected. I'll go back to blaming my botched landings on myself.

Now if I could just get static campaign designers to start throwing in some random wind components...

AndyJWest
04-19-2010, 01:42 PM
As JtD suggested, turning off 'wind and turbulence' seems to remove the increase in speed at very low level I'd measured earlier. I double checked my results, and also flew a reciprocal course to confirm it wasn't an effect of wind - I'd set the weather to 'clear'

One thing I can't understand is how, if the ground effect phenomenon disappears with 'wind and turbulence' off, he is getting the results he showed in his graph, though to be honest, I'm not entirely sure I understand what he is measuring.

This finding has some interesting consequences: if you select a 'no-wind' weather setting, you will find takeoffs and landings easier with 'wind and turbulence' on!

JtD
04-19-2010, 01:59 PM
It's not the ground effect that disappears, it's the speed boost. That's not related. The ground effect gives you more lift at the same angle of attack just above the ground (which is also what I was measuring).

Viikate said he edited the code to turn it off, so there's no difficulty switch to do so, I guess.

AndyJWest
04-19-2010, 02:05 PM
I'd assumed that the speed boost was due to ground effect - reduced induced drag giving excess power to increase speed at a given throttle setting. If the speed boost isn't down to ground effect, then what is causing this?

Off-topic, I must remember my autopilot wasn't designed to do 180 degree turns at 10m. It's just as well the max bank angle for bombers is set at 30 degrees. http://forums.ubi.com/groupee_common/emoticons/icon_eek.gif http://i958.photobucket.com/albums/ae65/ajv00987k/il2fb2010-04-1920-05-38-60.jpg

EDIT ----
I have now confirmed I get the same results for my Ju-88 test using the standard 'level stabilizer' autopilot function as I do using my AP.

K_Freddie
04-19-2010, 05:05 PM
http://www.av8n.com/how/img48/3v.png
This pic is 'wrong' with regard to bernoulli's principle, very wrong!!!
http://forums.ubi.com/groupee_common/emoticons/icon_wink.gif

AndyJWest
04-19-2010, 05:20 PM
The pic might well be wrong, but I'm not sure I see the relevance?

If you are saying 'it is wrong because the airflow over the top must meet up with the same airflow at the bottom when it meets the trailing edge', or something like that, I think this is a misinterpretation anyway.

TS_Sancho
04-19-2010, 05:24 PM
Sorry to break it to you Freddie, but applying bernoulli's principle and equal transit time to describe how an airfoil generates lift is a fallacy. http://forums.ubi.com/groupee_common/emoticons/icon_razz.gif


A false explanation for lift has been put forward in mainstream books, and even in scientific exhibitions. Known as the "equal transit-time" explanation, it states that the parcels of air which are divided by an airfoil must rejoin again; because of the greater curvature (and hence longer path) of the upper surface of an aerofoil, the air going over the top must go faster in order to "catch up" with the air flowing around the bottom. Therefore, because of its higher speed the pressure of the air above the airfoil must be lower. Despite the fact that this "explanation" is probably the most common of all, it is false.

Link (http://www.aviationexplorer.com/fixed_wing_aircraft.htm)

http://upload.wikimedia.org/wikipedia/commons/9/99/Karman_trefftz.gif

Kettenhunde
04-19-2010, 05:35 PM
JtD says:
That's not related.

IRL, it is very much related....


JtD says:
The ground effect gives you more lift at the same angle of attack just above the ground

Negative Ghost Rider.

You do not get "more lift", that is not how it works at all. That is fundamentally not correct. The wing only produces the amount of force required.

In a climb or descent, the wing will produce the amount of force equal to weight offset by the angle of climb or descent. The forces are in balance.

In a steady altitude coordinated turn, the wing will only produce the centripetal force required and the weight offset by a component of thrust.

The effect of an aircraft entering GE is the wing will move to a lower angle of attack at the same velocity. You must reduce power to maintain the same velocity.

Real pilots who do not understand this can get in trouble and kill themselves on both take offs and landings.


Ground effect reduces induced drag and the airplane is able to reach a speed where it can stagger off. As altitude is gained, induced drag increases as the effect of the ground effect diminishes. Twenty or thirty feet up, ground effect vanishes, the wing encounters the full effect of induced drag and the struggling airplane which got off the ground on the ragged edge of a stall becomes fully stalled and drops to earth.

http://www.pilotfriend.com/tra...raining/aft_perf.htm (http://www.pilotfriend.com/training/flight_training/aft_perf.htm)


Drag is reduced and at a constant power speed will increase.

julian265
04-19-2010, 07:57 PM
Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> JtD says:
The ground effect gives you more lift at the same angle of attack just above the ground

snip

The effect of an aircraft entering GE is the wing will move to a lower angle of attack at the same velocity.

snip </div></BLOCKQUOTE>

If you disagree with JtD, then why are you saying that the lower AoA is required?

AndyJWest
04-19-2010, 08:32 PM
I think we are getting into an unnecessary question of semantics here. For an aircraft to remain at a constant altitude, the upward force - 'lift' has to equal the downward force - 'weight'. Since ground effect gives the same lift at a lower AoA, to maintain this balance, the AoA needs to be reduced. I think my experiments seem to suggest this is occurring, though I haven't measured it directly, and JtD seems to be observing a similar effect. There seems to be a little uncertainty about measurement techniques, but so far I've seen no evidence that either (a) IL-2 isn't modelling the effect, or (b) it is giving implausible results for the scale of the effects. If anyone can come up with quantitative figures for what these effects should be in real life, we will be able to make a comparison. Otherwise, arguing about abstract questions will tell us little. Has anyone got any real data on predicting ground effect?

Kettenhunde
04-19-2010, 08:32 PM
why are you saying that the lower AoA is required?

I am just telling you how a real airplane behaves. You can see this when you establish a stabilized approach to make a wheel landing in a tail dragger.

JtD's statement is factually as well as conceptually incorrect in the part where he advances the concept of the wing creating "more lift". The wing does not produce "more lift" even at a higher coefficient of lift or changing angle of attack.

Kettenhunde
04-19-2010, 08:40 PM
Has anyone got any real data on predicting ground effect?


Yeah, it is generally expressed as a standard rate of percentage reduction in Coefficient of drag over the ratio of wingspan to height.

WTE_Galway
04-19-2010, 08:41 PM
Ground effect improves the lift to drag ratio by reducing drag.

This may be mistaken for an increase in lift but its not.

Viikate_
04-19-2010, 09:28 PM
Nobody mentioned yet span dominated ground effect & chord dominated ground effect. First one reduces induced drag and latter actually does generate lift at alt < chord length.

I think X-plane's approach is to take in count these two components.

AndyJWest
04-19-2010, 09:36 PM
Any chance of some actual real-life figures, Viikate? Even approximations have got to be better than the assumptions we are making at the moment.

Since you have access to the IL-2 FM, perhaps you could also tell us what factors it takes into account when modelling ground effect?

I suspect the whole subject is more complex than we have been assuming, and I'd like to see something a bit more definitive.

WTE_Galway
04-19-2010, 09:45 PM
well chord dominated occurs very very close to the ground .. not the 5 meters discussed here

what would be really interesting is to know the effect with carrier landings, whether ground effect suddenly kicks in as you cross the threshold of the deck

AndyJWest
04-19-2010, 10:10 PM
I've not actually tried to make measurements below 10m above sea/ground level.

I suspect that modelling the airflow around a moving carrier (or at least, one moving relative to the air) is beyond what one could reasonably expect of IL-2.

I'll see if I can come up with some results from my 'high altitude runway test' tomorrow, if I can get my AP to do what I want - I may have to tweak it a bit to allow more precise altitude settings.

If I get the chance, I'll also see if I can get a better idea of where GA is beginning to kick in, though if it is properly modelled, it should depend on aircraft configuration, and on airspeed...

JtD
04-19-2010, 10:29 PM
Originally posted by AndyJWest:
I'd assumed that the speed boost was due to ground effect - reduced induced drag giving excess power to increase speed at a given throttle setting. If the speed boost isn't down to ground effect, then what is causing this?

Let me be more precise - the "wind and turbulence" speed boost is not because of the ground effect. The ground effect appears to give you a some speed advantage, at higher AoA's. I haven't measured it exactly, but in the data from the above test I was eventually using lower throttle setting when I was 2m above the ground then when I was 50m above the ground to maintain the same low speed.
I know from earlier testing (couple of years back) that the top speed at 2m is lower than at 10m, so here you have a disadvantage.

JtD
04-19-2010, 10:37 PM
Originally posted by WTE_Galway:
Ground effect improves the lift to drag ratio by reducing drag.

This may be mistaken for an increase in lift but its not.

If the ground effect only reduces drag and does not produce extra lift, then why have I never seen a hovercraft flying at 100m altitude? Or in fact other dedicated ground effect vehicles that simply cannot not go higher than the air cushion created by the ground effect reaches? How come those nifty racing boats don't fly at all altitudes?

Kettenhunde
04-19-2010, 10:56 PM
Nobody mentioned yet span dominated ground effect & chord dominated ground effect. First one reduces induced drag and latter actually does generate lift at alt < chord length.

Once again, no additional lift force is created. Lift force only meets the amount required.

Cdo also changes in ground effect....is that going to be very useful information at this level for your game?

http://img526.imageshack.us/img526/9492/groundeffect.jpg (http://img526.imageshack.us/i/groundeffect.jpg/)

Kettenhunde
04-19-2010, 11:03 PM
If the ground effect only reduces drag and does not produce extra lift, then why have I never seen a hovercraft flying at 100m altitude?


There is no extra lift and what the heck does a hovercraft have to do with this conversation??

A hovercraft is generating a downward thrust vector and has absolutely nothing to do with ground effects on an airplane.

Perhaps you mean a WIG and they do not fly at 100M because they do not have the wing area required at the dynamic pressures they generate.

If you study the thrust required over velocity in knots graphic in the bottom right hand corner, you will see that speed will increase at a constant power.

http://img526.imageshack.us/img526/9492/groundeffect.jpg (http://img526.imageshack.us/i/groundeffect.jpg/)

AndyJWest
04-19-2010, 11:06 PM
As I suggested earlier, ground effect seems to be a complex issue. We probably need to deal with 'Il-2 ground effect' separately from 'real life ground effect' at least until we have some way to compare one with another. We probably also need to discuss the issue in terms of reproducible results, so anything I've come up with using my AP needs independent verification - but so, unfortunately, does 'earlier testing' and assertions about what 'must happen'.

I started this thread hoping to learn a little more about the complexities of aerodynamics. I'm no sure how far I've got...

WTE_Galway
04-20-2010, 12:18 AM
Originally posted by AndyJWest:
As I suggested earlier, ground effect seems to be a complex issue. We probably need to deal with 'Il-2 ground effect' separately from 'real life ground effect' at least until we have some way to compare one with another. We probably also need to discuss the issue in terms of reproducible results, so anything I've come up with using my AP needs independent verification - but so, unfortunately, does 'earlier testing' and assertions about what 'must happen'.

I started this thread hoping to learn a little more about the complexities of aerodynamics. I'm no sure how far I've got...

These threads always deteriorate http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif generally because what actually is happening in aviation is often different to the "seat of the pants" feel and what you think is happening.

I recall one thread that went for 20 pages with numerous people insisting (against all logic) that applying flaps resulted in an aircraft flying "nose high" with poorer visibility basically because that is what they perceived to be happening in the game when they applied flaps.

I don't have the patience for it these days http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif

julian265
04-20-2010, 07:58 AM
Originally posted by Kettenhunde:
JtD's statement is factually as well as conceptually incorrect in the part where he advances the concept of the wing creating "more lift". The wing does not produce "more lift" even at a higher coefficient of lift or changing angle of attack.


Originally posted by Kettenhunde:
The effect of an aircraft entering GE is the wing will move to a lower angle of attack at the same velocity.

If the plane is moving at the same speed, and wind isn't blowing downward, then what makes the AoA reduce?

JtD
04-20-2010, 09:02 AM
Originally posted by AndyJWest:
As I suggested earlier, ground effect seems to be a complex issue. We probably need to deal with 'Il-2 ground effect' separately from 'real life ground effect' at least until we have some way to compare one with another.

I don't actually think that's necessary, because there is a drag loss, and there is a lift gain in game, both fairly consistent with what you can expect from real life, too. What I'm most curious about is the drag loss with wind and turbulence disabled, I don't think it exists. Which would lead to the conclusion that the drag loss is covered in the wind and turbulence switch, quite reasonably since the drag loss is connected to different turbulence characteristics, while the lift gain isn't, as the air cushion is independent from wind and turbulence.

I'd like to do a bit more testing with this, but I'm currently busy with other things. Unless you've found a way to clarify all open points, I'll get back to this in a week or so.



We probably also need to discuss the issue in terms of reproducible results, so anything I've come up with using my AP needs independent verification - but so, unfortunately, does 'earlier testing' and assertions about what 'must happen'.

What exactly do you think needs to be verified? I could make a short list about what I'm curious about, in case you run out of ideas.

Anything on your high altitude airfield test yet?

AndyJWest
04-20-2010, 09:35 AM
Anything on your high altitude airfield test yet?

As always, setting it up proved more complex than I expected - should have something to show later, if it works...

Kettenhunde
04-20-2010, 10:19 AM
If the plane is moving at the same speed, and wind isn't blowing downward, then what makes the AoA reduce?


From the first page:



Crumpp says:

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. OR you can fly at a lower angle of attack for the same velocity.

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



http://forums.ubi.com/eve/foru...031092458#8031092458 (http://forums.ubi.com/eve/forums/a/tpc/f/23110283/m/3121062458?r=8031092458#8031092458)

I added the results of flying at the same velocity in bold to hopefully make things clearer for you. You can also see this on the lower left hand corner of figure 6.9 which I posted above on this page.

You got it now, Julian? It is not an intuitive concept and more than one real world pilot has killed themselves because they did not understand Ground Effect.

It is just one more thing that dramatically illustrates how an hour in the library cannot give you an education in aerodynamics.

M_Gunz
04-20-2010, 12:23 PM
It seems that JtD is right given the proviso of "at same speed and AOA" and that Crumpp has implied pilot
or other action that actually lowers AOA.

I still don't see how change in drag alone "at same speed and AOA" enhances lift but then I'm running a fever
and not up to understanding this new mystery. You either have the same lift curve or you don't? Oh no, there's
more of course!

I expect that since it is induced drag that is lowered to see the big increase at speeds where induced drag is
much higher than parasitic drag, ie low speed and low power not 100% where induced is a few percent at most.

AndyJWest
04-20-2010, 12:42 PM
I expect that since it is induced drag that is lowered to see the big increase at speeds where induced drag is
much higher than parasitic drag, ie low speed and low power not 100% where induced is a few percent at most.

That was the result I got with my Ju-88 tests, M_G:

110% throttle, rads open - 100m 376 Km/h, 10m 382 Km/h. Difference 6 Km/h, 1.6%
75% throttle, rads open - 100m 345 Km/h, 10m 354 Km/h. Difference 9 Km/h, 2.6%
51% throttle, rads open - 100m 252 Km/h, 10m 267 Km/h. Difference 15 Km/h 6.0%

My 'high altitude ground effect' test isn't producing the effects I expected, but it may be due to not all planes modelling ground effect the same - I was using a P-38, but I need to check this at sea level first.

Kettenhunde
04-20-2010, 02:32 PM
that Crumpp has implied pilot
or other action that actually lowers AOA.

Nowhere have I said anything about pilot input.

The physics changes the AoA and not the pilot.

JtD is wrong in the silly notion of "increased lift". In his defense, it is a common misconception. It just does not work that way however.

M_Gunz
04-20-2010, 02:35 PM
And yet same lift at lower AOA or more lift at same AOA but not increased lift..... I see word puzzles here.

Kettenhunde
04-20-2010, 02:41 PM
And yet same lift at lower AOA or more lift at same AOA but not increased lift..... I see word puzzles here.


Don't know what to tell you.

How hard is it to understand that the lift force does not change?

A coefficient is not the lift force nor is an Angle of Attack.

If you establish a stabilized approach or in other words, a steady state descent, the lift force required will be exactly the same at 5 feet as it is at 5000 feet.

M_Gunz
04-20-2010, 02:41 PM
Originally posted by AndyJWest:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">I expect that since it is induced drag that is lowered to see the big increase at speeds where induced drag is
much higher than parasitic drag, ie low speed and low power not 100% where induced is a few percent at most.

That was the result I got with my Ju-88 tests, M_G:

110% throttle, rads open - 100m 376 Km/h, 10m 382 Km/h. Difference 6 Km/h, 1.6%
75% throttle, rads open - 100m 345 Km/h, 10m 354 Km/h. Difference 9 Km/h, 2.6%
51% throttle, rads open - 100m 252 Km/h, 10m 267 Km/h. Difference 15 Km/h 6.0%

My 'high altitude ground effect' test isn't producing the effects I expected, but it may be due to not all planes modelling ground effect the same - I was using a P-38, but I need to check this at sea level first. </div></BLOCKQUOTE>

I think that GE should happen 1/2 wing span or lower, not a fixed distance. P-38 is not as wide as Ju-88 IIRC.
Also to test near landing speed or at least below 220kph. 6% is not very much at all but 252kph is over 150mph.

Insuber
04-20-2010, 03:22 PM
It's basic physics and common sense, it suffices to look at the 4 forces of flight: if the plane flies leveled, the lift equals the weight. If it flies faster at the same thrust, then drag is lower.
Drag = constant x medium density x frontal surface x speed squared/2. Given that the drag coefficient is a constant, that speed is increasing, air density is about the same (actually slightly higher) at 5 m and 100m, the only factor which can be lower is the frontal surface. Since the plane is not shrinking, the only way to change the frontal surface is to lower the AoA. So same lift force, at lower AoA.




Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">And yet same lift at lower AOA or more lift at same AOA but not increased lift..... I see word puzzles here.


Don't know what to tell you.

How hard is it to understand that the lift force does not change?

A coefficient is not the lift force nor is an Angle of Attack.

If you establish a stabilized approach or in other words, a steady state descent, the lift force required will be exactly the same at 5 feet as it is at 5000 feet. </div></BLOCKQUOTE>

AndyJWest
04-20-2010, 03:40 PM
Given that the drag coefficient is a constant

But what everything I've found so far via Google (admittedly not a lot) seems to say is that ground effect reduces induced drag - it isn't constant.

Kettenhunde
04-20-2010, 03:52 PM
And yet same lift at lower AOA or more lift at same AOA but not increased lift


M_Gunz,

It might be helpful to understand the change in angle of attack is in the induced angle of attack and not the aerodynamic angle of attack.

I think I have explained induced angle of attack to you already. Do you remember?

Insuber
04-20-2010, 04:08 PM
Originally posted by AndyJWest:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Given that the drag coefficient is a constant

But what everything I've found so far via Google (admittedly not a lot) seems to say is that ground effect reduces induced drag - it isn't constant. </div></BLOCKQUOTE>

You probably read about the induced drag coefficient, which actually varies with the lift.

AndyJWest
04-20-2010, 05:05 PM
I think the point is that ground effect alters the entire flow pattern around the aircraft, so pretty well everything is different than in free air. Whether the plane changes it's pitch angle or not may have little direct relationship to lift or drag - there may also be stability changes due to the flow over the tailplane altering - no doubt this will depend on aircraft configuration to some extent. Ground effect is expected to reduce the strength of wing-tip vortices though, which as I understand it are a major factor in induced drag. All rather complex anyway...

I tried my 'high altitude ground effect' test with a Ju-88, and it didn't seem to be there either. Possibly a bug, but if it is, it is possibly a minor one, unless you are actually flying on the mountain map (online4summer) - I'd need to try the same thing over a more normal airfield to check further.

WTE_Galway
04-20-2010, 05:51 PM
Originally posted by JtD:
Which would lead to the conclusion that the drag loss is covered in the wind and turbulence switch, quite reasonably since the drag loss is connected to different turbulence characteristics, while the lift gain isn't, as the air cushion is independent from wind and turbulence.


If i understand correctly this second effect occurs at some fraction of the wing chord, in other words within a meter or so of the ground for a fighter and maybe a few meters for a bomber at the most.

Which is why the second effect is noticeable in GA flying a Warrior but not a 172 for example.



Originally posted by AndyJWest:
I tried my 'high altitude ground effect' test with a Ju-88, and it didn't seem to be there either. Possibly a bug, but if it is, it is possibly a minor one, unless you are actually flying on the mountain map (online4summer) - I'd need to try the same thing over a more normal airfield to check further.

High altitude summer and maximum load should be optimal.

Kettenhunde
04-20-2010, 06:26 PM
Which is why the second effect is noticeable in GA flying a Warrior but not a 172 for example.

There is no "lift gain". That line of thinking will get you in trouble as a pilot.

AndyJWest
04-20-2010, 07:30 PM
There is no "lift gain"
One of the graphs you posted earlier seems to imply otherwise, Kettenhunde:
http://i958.photobucket.com/albums/ae65/ajv00987k/GEgraph.jpg

Isn't a higher CL at a given AoA a gain? Of course, as I said earlier, the whole airflow is going to be affected by ground effect, so any simple formula is likely to be misleading.

Actually, trying to measure the effect at altitudes below wing chord would be decidedly difficult I think, so I'm inclined to think it isn't that relevant to IL-2, or at least something I can investigate.

Perhaps more to the point, my results seem to indicate that if anything the drag reduction is kicking in far too high, but I need to look into this further.

Kettenhunde
04-20-2010, 07:52 PM
One of the graphs you posted earlier seems to imply otherwise, Kettenhunde:

Coefficient of lift is NOT lift force....

Force vs Airspeed:


http://img442.imageshack.us/img442/2124/forceias.png (http://img442.imageshack.us/i/forceias.png/)

Coefficients vs Airspeed:

http://img59.imageshack.us/img59/3744/coeffias.png (http://img59.imageshack.us/i/coeffias.png/)

In GE, if the airplane is held in a constant angle of attack, the coefficient of lift will increase.

To meet this new coefficient of lift, the airspeed and rate of descent changes to the dynamic pressure that provides the same amount of lift force.

AndyJWest
04-20-2010, 08:12 PM
In GE, if the airplane is held in a constant angle of attack, the coefficient of lift will increase.

To meet this new coefficient of lift, the airspeed and rate of descent changes to the dynamic pressure that provides the same amount of lift force.


...which is another way of saying that there is a lift increase, which is compensated for by alterations to the airspeed and/or rate of descent - implying an acceleration (or deceleration) which need a force applied to occur.

julian265
04-20-2010, 08:23 PM
Originally posted by Kettenhunde:
It might be helpful to understand the change in angle of attack is in the induced angle of attack and not the aerodynamic angle of attack.

"might be helpful"? Actually 'induced' was a necessary term, as what you were saying was simply wrong for 'normal' AoA.

Kettenhunde
04-20-2010, 08:31 PM
...which is another way of saying that there is a lift increase,

No there is not an increase in lift force!

There is increase in coefficient of lift if the aircraft is held at a constant angle of attack.

Coefficient of lift is a dimensionless ratio of lifting pressure to dynamic pressure.

Airspeed will reduce, the rate of descent will reduce and most importantly our thrust required will be reduced.

If you do not reduce power, you will experience a floating sensation.

Lift force has not changed at all. LIFT FORCE REMAINS THE SAME.

Not understanding the difference between a coefficient and a force is a fundamental misunderstanding. Simply put you will not understand what is going on with the aircraft if you keep the idea of "increased lift" in your head.

Here is a nice explanation for you. Read it and study the earlier graphics showing the relationship of coefficients and force.

http://img707.imageshack.us/img707/3885/groundeffectexplainatio.jpg (http://img707.imageshack.us/i/groundeffectexplainatio.jpg/)

http://img9.imageshack.us/img9/3885/groundeffectexplainatio.jpg (http://img9.imageshack.us/i/groundeffectexplainatio.jpg/)

Earlier graphics:

http://forums.ubi.com/eve/foru...981010558#9981010558 (http://forums.ubi.com/eve/forums/a/tpc/f/23110283/m/3121062458?r=9981010558#9981010558)

If the light bulb does not come on after this, I can't help you! http://forums.ubi.com/groupee_common/emoticons/icon_frown.gif

julian265
04-20-2010, 08:34 PM
Originally posted by AndyJWest:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">In GE, if the airplane is held in a constant angle of attack, the coefficient of lift will increase.

To meet this new coefficient of lift, the airspeed and rate of descent changes to the dynamic pressure that provides the same amount of lift force.


...which is another way of saying that there is a lift increase, which is compensated for by alterations to the airspeed and/or rate of descent - implying an acceleration (or deceleration) which need a force applied to occur. </div></BLOCKQUOTE>

Kettenhunde stated that *induced* AoA reduces when in ground effect - which may explain why, for a higher CL, you have the same lift.

julian265
04-20-2010, 08:37 PM
http://forums.ubi.com/groupee_common/emoticons/icon_rolleyes.gif

Originally posted by Kettenhunde:
WTF are you even talking about??


Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">And yet same lift at lower AOA or more lift at same AOA but not increased lift


M_Gunz,

It might be helpful to understand the change in angle of attack is in the induced angle of attack and not the aerodynamic angle of attack.

I think I have explained induced angle of attack to you already. Do you remember? </div></BLOCKQUOTE>

Kettenhunde
04-20-2010, 08:42 PM
"might be helpful"? Actually 'induced' was a necessary term, as what you were saying was simply wrong for 'normal' AoA.

Thanks for the clarification. It took me a while to realize what you were talking about. I have explained some of this to M_Gunz on how a wings angle of attack is derived.

What I said was absolutely correct. The wings angle of attack is comprised of both the section angle and the induced angle.

Wing AoA = Section AoA + Induced AoA

Section AoA is the airfoils 2D aerodynamic angle of attack to reach the required coefficient of lift.

What is a "normal" angle of attack?

julian265
04-20-2010, 08:50 PM
What a lot of people regard AoA as - the angle difference between wing chord line and incident air. Maybe I should have said 'common' AoA. Either way, it's the root of a few people's, including my own misinterpretation of your posts in this thread.

Kettenhunde
04-20-2010, 09:11 PM
What a lot of people regard AoA as - the angle difference between wing chord line and incident air. Maybe I should have said 'common' AoA.

FWIW, there is no such thing as a "common angle of attack" or a "normal" angle of attack.

There is just angle of attack for a wing, angle of attack for an airfoil, and induced angle of attack.

Angle of attack for an airfoil is the angle between the chord line and the average relative wind. Airfoils are also called "2D data" or an infinite wing. It is just a two dimensional shape.

Induced angle of attack is the angle between the average relative wind and the remote relative wind. This represents the size as well as movement of the upwash and downwash relative to the aerodynamic center. It is the additional angle the three dimensional wing must achieve to reach the same coefficients as the 2D airfoil.

The angle of attack for a three dimensional wing is the sum of those two angles and represents the angle between the chord line and the remote relative wind.

AndyJWest
04-20-2010, 09:46 PM
Airspeed will reduce, the rate of descent will reduce and most importantly our thrust required will be reduced.

If you do not reduce power, you will experience a floating sensation.

Lift force has not changed at all. LIFT FORCE REMAINS THE SAME.

Um,how can you 'experience a floating sensation' unless a force is applied?

A practical example: a model glider (no pilot, no power input, but same aerodynamic principles) is carefully trimmed so when launched at the appropriate speed and angle, it continues to descend at the same, in free air. Ignoring changes in air density with altitude (lets not complicate things too much http://forums.ubi.com/groupee_common/emoticons/icon_rolleyes.gif) this is an equilibrium state. Said glider then approaches deck, experiencing ground effect. What happens?

(a) Nothing at all, it continues at the same airspeed and angle - no 'ground effect'.

(b) Its speed remains the same, but its rate of descent reduces.

(c) Its speed remains the same, but its rate of descent increases.

(d) Its rate of descent remains the same, but it speeds up.

(d) Its rate of descent remains the same, but it slows down.

If I had to guess, I'd say it was probably a bit of (b), and a bit of (d), but regardless, for anything other than (a) to occur, a force must have been applied to the glider. About the only possible exception to this I can thing of is a combination of (c) and (d), but that doesn't sound like 'floating' to me - more like diving into the deck.

Kettenhunde
04-20-2010, 10:15 PM
Um,how can you 'experience a floating sensation' unless a force is applied?

Because when you are descending at one rate and it changes....it feels like you are floating.

What you feel in the airplane has nothing to do with the airplane is actually doing.

http://www.youtube.com/watch?v=uw2qPLEgKdQ

AndyJWest
04-20-2010, 10:19 PM
Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Um,how can you 'experience a floating sensation' unless a force is applied?

Because when you are descending at one rate and it changes....it feels like you are floating.

What you feel in the airplane has nothing to do with the airplane is actually doing.

http://www.youtube.com/watch?v=uw2qPLEgKdQ </div></BLOCKQUOTE>

If the rate changes, the aeroplane is doing something...

Kettenhunde
04-20-2010, 10:20 PM
Said glider then approaches deck, experiencing ground effect. What happens?


At a constant airspeed, the angle of attack will lower and the rate descent will remain the same.

Read the explanation I posted from Aerodynamics for Naval Aviators. It tells you the exact same thing I told you on page one.

Kettenhunde
04-20-2010, 10:20 PM
the aeroplane is doing something...


Sure, the drag and thrust required is changing but that is not the force of lift.

The force of lift is not changing.

JtD
04-20-2010, 10:28 PM
Originally posted by WTE_Galway:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by JtD:
Which would lead to the conclusion that the drag loss is covered in the wind and turbulence switch, quite reasonably since the drag loss is connected to different turbulence characteristics, while the lift gain isn't, as the air cushion is independent from wind and turbulence.

If i understand correctly this second effect occurs at some fraction of the wing chord, in other words within a meter or so of the ground for a fighter and maybe a few meters for a bomber at the most. </div></BLOCKQUOTE>

My in game test was done at 5-2 m altitude, where 2 m meant that the vertical stabs and the props of the P-38 nearly touched the water. Like I said in my initial evaluation, I suppose that 5m is the maximum altitude for the effect, however at 2m a distinct difference was there. The maximum chord of the P-38 is about 3m. I don't know if the effect in game is related to wing geometry, that would be one thing I'd like to test.

JtD
04-20-2010, 10:33 PM
Originally posted by AndyJWest:

110% throttle, rads open - 100m 376 Km/h, 10m 382 Km/h. Difference 6 Km/h, 1.6%
75% throttle, rads open - 100m 345 Km/h, 10m 354 Km/h. Difference 9 Km/h, 2.6%
51% throttle, rads open - 100m 252 Km/h, 10m 267 Km/h. Difference 15 Km/h 6.0%

That's pretty neat. How accurate do you think you are?

AndyJWest
04-20-2010, 10:33 PM
Maybe I'm misunderstanding, but you seem to be contradicting yourself, Kettenhunde:


If you do not reduce power, you will experience a floating sensation
This implies an acceleration.


At a constant airspeed, the angle of attack will lower and the rate descent will remain the same.
This implies no acceleration.

Actually, I'm not particularly concerned about what 'will' occur, as what I'm trying to determine how the IL-2 simulation of what occurs compares with what actually happens in practice, and it seems to me that the only way one can verify any of this is through observation and measurement. Aerodynamic theory might be right, wrong or irrelevant, but it is completely off-topic if one is trying to compare the 'real world' with a simulation. Scientific theories are an attempt to describe the world, not a set of rules defining how it actually works.

M_Gunz
04-20-2010, 10:38 PM
So as you enter GE the nose drops without stick movement. And this is due to trim and change in flight condition.
So your total lift stays the same.

JtD
04-20-2010, 10:39 PM
An aircraft descending into the ground effect at a constant AoA will go faster and reduce the rate of descent, because it generates less drag and more lift at the same AoA.

AndyJWest
04-20-2010, 10:43 PM
Originally posted by JtD:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by AndyJWest:

110% throttle, rads open - 100m 376 Km/h, 10m 382 Km/h. Difference 6 Km/h, 1.6%
75% throttle, rads open - 100m 345 Km/h, 10m 354 Km/h. Difference 9 Km/h, 2.6%
51% throttle, rads open - 100m 252 Km/h, 10m 267 Km/h. Difference 15 Km/h 6.0%

That's pretty neat. How accurate do you think you are? </div></BLOCKQUOTE>

It will depend to some extent on fuel load, and even in 'clear' weather I seem to get a difference of 1-2 Km/h on a reciprocal course. Also, I am having to estimate when TAS has stabilised, and given a slight variation in TAS, estimate the 'average' (though it usually varies by only 1 Km/h or so). I am also relying on my autopilot not causing any extra anomalies, tough I can see no reason why it should. In any case, I was more interested in speed difference than absolute values, so many of the factors involved should cancel out.

There is a slight problem in that I'm assuming constant throttle setting = constant power, which clearly isn't true over a large altitude difference.

Ultimately, I'd not say I was producing definitive results, but something for others to look into if they are interested - the more input we get, the more reliable the measurements should become.

JtD
04-20-2010, 10:52 PM
All in all it sounds alright. Engine power difference over 90 m altitude isn't that much, like 3 hp or so, translating into insignificant %.

If you speed is stable within a km/h, then I guess your results are as reliable as they can be.

WTE_Galway
04-20-2010, 11:06 PM
Originally posted by AndyJWest:

If I had to guess, I'd say it was probably a bit of (b), and a bit of (d), but regardless, for anything other than (a) to occur, a force must have been applied to the glider.

Applied or removed. Vectors don't care.

Oversimplifying a bit but a reduction in drag is equivalent to applying a force in the direction of movement. The effects are identical.

AndyJWest
04-20-2010, 11:32 PM
Originally posted by JtD:
All in all it sounds alright. Engine power difference over 90 m altitude isn't that much, like 3 hp or so, translating into insignificant %.

If you speed is stable within a km/h, then I guess your results are as reliable as they can be.
Well, I could run the tests again, but to be honest, I'd be happier to have independent verification of the trends than worry about the exact figures. Reproducing my results using the 'level stabilizer' control rather than my autopilot might be a bit awkward, but it should be possible. In any case , this is one particular aircraft, at one throttle setting, and from what I've done since I'm beginning to suspect that the IL-2 model doesn't conform with 'real world' expectations that well - I need to do more tests though.

M_Gunz
04-20-2010, 11:36 PM
Hmmmmm. I remember reading some on the subject not terribly long ago.
Please keep in mind that many things go on and there are big pictures discussed that only slowly the details are told
so no recriminations please since part of one picture does not always apply to part of another.

av8n.com -- author is as much an expert as anyone I know of (http://www.av8n.com/how/htm/airfoils.html#sec-circulation-vortices)


3.12.5 Bound Vortex

Let’s not forget about the bound vortex, which runs spanwise from wingtip to wingtip, as shown in figure 3.27.

When you are flying in ground effect, you are influenced by the mirror image of your bound vortex. Specifically, the flow circulating around the mirror-image bound vortex will reduce the airflow over your wing. I call this a pseudo-tailwind.16

Operationally, this means that for any given angle of attack, you need a higher true airspeed to support the weight of the airplane.


As a less-precise way of saying things, you could say that to compensate for ground effect, at any given true airspeed, you need more coefficient of lift. This explains why all airplanes – some more so than others – exhibit “squirrely” behavior when flying near the ground, including:

* Immediately after liftoff, the airplane may seem to leap up a few feet, as you climb out of the pseudo-tailwind. This is generally a good thing, because when you become airborne you generally want to stay airborne.
* Conversely, on landing, the airplane may seem to drop suddenly, as the pseudo-tailwind takes effect. This is unhelpful, but it’s not really a big problem once you learn to anticipate it. It does mean that practicing flaring at altitude (as discussed in section 12.11.3) will never entirely prepare you for real landings.
* The wing and the tail will be influenced by ground effect to different degrees. (This is particularly pronounced if your airplane has a low wing and a high T-tail, but no airplane is entirely immune.) That means that when you enter or exit ground effect, there will be squirrely pitch-trim changes ... in addition to the effects mentioned in the previous items. Just to rub salt in the wound, the behavior will be different from flight to flight, depending on how the aircraft is loaded, i.e. depending on whether the center of mass is near the forward limit or the aft limit.

During landing, ground effect is a lose/lose/lose proposition. You regret greater speed, you regret lesser drag, and you regret squirrely handling.


7.7.3 Skimming in Ground Effect

Here is trick for saving a little bit of energy. I hope you never get into a situation where you need to use this trick — but it might save your bacon if the situation arises.

Suppose no engine power is available, and the aircraft is too low and/or too far from the desired landing place. Using our energy-management logic, we see that the only real way to stretch the glide is to find a low-drag mode of operation. The solution is sort of the reverse of a soft-field takeoff (section 13.4) — you should make use of ground effect.

Specifically, the procedure is to maintain best-glide speed right down into ground effect, even if this means that you enter ground effect over the swamp a tenth of a mile short of the intended landing place. Once you are in ground effect, start pulling back on the yoke. Because there is very little induced drag in ground effect (as discussed in connection with soft-field takeoffs in section 13.4), the airplane can fly at very low airspeeds with remarkably little drag. You can then fly all the way to the landing area in ground effect. It is like a prolonged flare; you keep pulling back gradually to cash in airspeed and pay for drag. This technique will not solve all the world’s problems, but it is guaranteed to work better than trying to stretch the glide by pulling back before entering ground effect.

Conversely: if you are approaching a short runway and have a few knots of excess airspeed on short final, you should pull back on the yoke and get rid of the excess airspeed before entering ground effect. If you think you can’t get rid of it on short final, remember it will only be harder to get rid of in ground effect. A timely go-around might be wise.

If you want to practice skimming in ground effect, find a long, long, long runway to practice on, and be careful not to run off the far end.


A LOT of different things happening at once, not in sequence. Add what the pilot does.. must do, or not changes the picture
of what is being said.

Ever watch "The Usual Suspects"? First time around the plot keeps changing but once you know it is a different movie.

Kettenhunde
04-21-2010, 05:23 AM
An aircraft descending into the ground effect at a constant AoA will go faster and reduce the rate of descent, because it generates less drag and more lift at the same AoA.

Wrong.

The wing does not generate "more lift force".

You claim to know how to do the math.

What is going to change our lift force required at a constant Angle of Attack?

The answer is nothing.....the aircraft's slows rate of descent at the same lift force.

You see this effect every time you land a real airplane.

If you want to get technical, if you enter GE at a constant Coefficient of lift and allow angle of attack to change, lift force experiences a small reduction in a amount of force required.

Why do you think that occurs?

Kettenhunde
04-21-2010, 05:28 AM
Oversimplifying a bit but a reduction in drag is equivalent to applying a force in the direction of movement. The effects are identical.


Exactly.

M_Gunz
04-21-2010, 05:49 AM
If you increase power to a plane in flight, it will climb. That's how I check throttle/prop pitch combinations.

AndyJWest
04-21-2010, 09:18 AM
More interesting test results

One of the options with my autopilot setup is 'pitch hold' which keeps the pitch axis (nose up/down) constant - this tends to be more stable at high AoA than the alternative 'altitude hold' mode, but it also enables me to measure the altitude at which a given pitch/throttle setting will maintain height.

With A Ju-88 on the Crimea map, with fuel load at 50%, and 'unlimited fuel' selected to keep the fuel weight constant, pitch set to +6 degrees, radiators open, I get the following results:

50.0% throttle, level flight at 70m
47.3% throttle, level flight at 22m
44.0% throttle, level flight at 3m

The throttle settings are as reported by DeviceLink, avoiding rounding problems with the in-game display.

I tried 42.9% power, and it looked like settling at about 2m, but a slight AP induced roll caused a prop to catch the water.

http://i958.photobucket.com/albums/ae65/ajv00987k/il2fb2010-04-2115-04-39-37.jpg
http://i958.photobucket.com/albums/ae65/ajv00987k/il2fb2010-04-2115-04-58-51.jpg

JtD
04-21-2010, 10:02 AM
Definitely interesting, a good idea for a test setup, too.

Was the speed 221ish at all three altitudes?

Was wind and turbulence on or off?

AndyJWest
04-21-2010, 10:28 AM
I forgot to note TAS/IAS but I don't think it changed much. I'll be rerunning the test anyway, at different pitch angles, though +6 is probably near the maximum my AP will hold a ju-88 steady at - higher angles at low airspeed tend to cause excessive rolling, and height/airspeed instability.

'Wind and turbulence' was on. I'll see if turning it off makes a difference, though previous results imply it should.

Kettenhunde
04-21-2010, 10:51 AM
M_Gunz says:
If you increase power to a plane in flight, it will climb.


Exactly!

Let's talk about our lift force picture in a climb or descent compared to level flight. This is key to understanding ground effect.

Do we require more lift force in a climb or less left force compared to level flight?

How about a descent? Does our wing need to generate more lift force to descend compared to level flight or less?


Answer these questions correctly and the lack of understanding in the following quote becomes obvious:


An aircraft descending into the ground effect at a constant AoA will go faster and reduce the rate of descent, because it generates less drag and more lift at the same AoA.

AndyJWest
04-21-2010, 11:24 AM
lack of understanding

Kettenhunde, for any of us to understand ground effect, we need a reliable source of information. As I've already stated, there is some information on the net, as found via Google, but it is patchy and not entirely consistent. Perhaps it would be more helpful to give us guidance as to where we can find 'understanding', rather than just telling us we don't have it. From what little I've learned so far, none of it is 'obvious', or if it is, it is probably wrong.

JtD
04-21-2010, 11:30 AM
Originally posted by AndyJWest:
I forgot to note TAS/IAS but I don't think it changed much. I'll be rerunning the test anyway, at different pitch angles, though +6 is probably near the maximum my AP will hold a ju-88 steady at - higher angles at low airspeed tend to cause excessive rolling, and height/airspeed instability.

'Wind and turbulence' was on. I'll see if turning it off makes a difference, though previous results imply it should.

Thanks for the infos.

Regarding the AP, you can adjust the sensitivity of the controls in the autopilot.ini file. For level flight, you can reduce the aileron sensitivity to a tenth of the standard values and use really long filter times. It takes a while to get you level, but once you are you're in a very stable condition.

Unless you want to stick to the Ju-88, I'd recommend the P-38 because the counter rotating props mean no slip. This way you can turn off the slip=0 command which makes everything more stable.

AndyJWest
04-21-2010, 11:39 AM
Regarding the AP, you can adjust the sensitivity of the controls in the autopilot.ini file. For level flight, you can reduce the aileron sensitivity to a tenth of the standard values and use really long filter times. It takes a while to get you level, but once you are you're in a very stable condition.

Unless you want to stick to the Ju-88, I'd recommend the P-38 because the counter rotating props mean no slip. This way you can turn off the slip=0 command which makes everything more stable.

I'm using my own prototype autopilot program. Which AP would this be referring to?

JtD
04-21-2010, 11:44 AM
Originally posted by Kettenhunde:

Answer these questions correctly and the lack of understanding in the following quote becomes obvious:

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> An aircraft descending into the ground effect at a constant AoA will go faster and reduce the rate of descent, because it generates less drag and more lift at the same AoA. </div></BLOCKQUOTE>

Look, if a planes load factor changes there's a change of lift. Like when a plane descents into the ground effect, and then stops descending. That's 1g in the descent, more than 1g when it stops descending, and 1g again after it stopped descending and goes level.

Rambling for 5 pages about the silliness and wrongness of others while not even understanding what they are talking about. That's brilliant. At least when it comes to displaying oversized egos...

JtD
04-21-2010, 11:45 AM
Originally posted by AndyJWest:

I'm using my own prototype autopilot program. Which AP would this be referring to?

Oh, interesting. I was referring to LesniHu's autopilot from ages ago.

AndyJWest
04-21-2010, 12:23 PM
I tried out the LesniHu AP some time back, but couldn't seem to get it working properly - possibly me misunderstanding how it works, rather than anything wrong with the program. At that stage my own AP was working anyway, so I didn't pursue it further - mine has the advantage that you can set parameters in flight, which I don't think you can do with LesniHu.

Edit:
In case you missed the earlier link, some info on my AP here: http://forums.ubi.com/eve/foru...121016097#4121016097 (http://forums.ubi.com/eve/forums/a/tpc/f/23110283/m/4121016097?r=4121016097#4121016097)

JtD
04-21-2010, 01:03 PM
I might actually have missed that. Or I might have had a glance, and decided that I'm very familiar with LesniHu's AP now and don't want to bother with it. In that case, my bad.

Might take a look next week.

AndyJWest
04-21-2010, 02:24 PM
My autopilot is very much an unfinished prototype, and has never been released - it has a nasty habit of turning itself off, which can be tricky in situations like this!

I repeated my last test, noting speeds this time. Same conditions - Ju-88 on the Crimea map, with fuel load at 50%, and 'unlimited fuel' selected to keep the fuel weight constant, pitch set to +6 degrees, radiators open:

50.0% throttle, level flight at 69m, groundspeed 240 Km/h, indicated airspeed 230 Km/h
47.3% throttle, level flight at 20m, groundspeed 233 Km/h, indicated airspeed 224 Km/h
44.0% throttle, level flight at 3m, groundspeed 222 Km/h, indicated airspeed 213 Km/h

Groundspeed is from 'wonder woman view' instrument, IAS is from DeviceLink.

This was on a heading of 180 degrees. on a 0 degree heading the 50% throttle speeds were 238 Km/h and 230 Km/h, suggesting there was about a 1 Km/h tailwind, even in 'clear' conditions. The discrepancy between groundspeed and indicated airspeed must therefore mostly be due to either (simulated) instrument position error, or the conditions on the map not being the 'standard' ones assumed for calibration.

I also tried the test with 'wind and turbulence' disabled, and as I suspected, the 'ground effect' no longer occurred - at 50% throttle, the plane descended to sea level at about 0.3 m/s

Kettenhunde
04-21-2010, 02:55 PM
we need a reliable source of information.

It says the exact same thing I have already posted but from another textbook. I would steer clear of most internet sites. There is quite a bit of flaky information on most of them.

John Denker's site is one of the better ones on the net.

http://www.av8n.com/how/#contents

http://img231.imageshack.us/img231/6519/groundeffectagain.jpg (http://img231.imageshack.us/i/groundeffectagain.jpg/)

http://img46.imageshack.us/img46/6653/groundeffectagain2.jpg (http://img46.imageshack.us/i/groundeffectagain2.jpg/)

http://img130.imageshack.us/img130/1102/groundeffectagain3.jpg (http://img130.imageshack.us/i/groundeffectagain3.jpg/)


Another common phenomenon that is misunderstood is that of ground effect. That is the increased efficiency of a wing when flying within a wing length of the ground. A low-wing airplane will experience a reduction in drag by 50% just before it touches down. There is a great deal of confusion about ground effect. Many pilots (and the FAA VFR Exam-O-Gram No. 47) mistakenly believe that ground effect is the result of air being compressed between the wing and the ground.

To understand ground effect it is necessary to have an understanding of upwash. For the pressures involved in low speed flight, air is considered to be non-compressible. When the air is accelerated over the top of the wing and down, it must be replaced. So some air must shift around the wing (below and forward, and then up) to compensate, similar to the flow of water around a canoe paddle when rowing. This is the cause of upwash.

As stated earlier, upwash is accelerating air in the wrong direction for lift. Thus a greater amount of downwash is necessary to compensate for the upwash as well as to provide the necessary lift. Thus more work is done and more power required. Near the ground the upwash is reduced because the ground inhibits the circulation of the air under the wing. So less downwash is necessary to provide the lift. The angle of attack is reduced and so is the induced power, making the wing more efficient.

Earlier, we estimated that a Cessna 172 flying at 110 knots must divert about 2.5 ton/sec to provide lift. In our calculations we neglected the upwash. From the magnitude of ground effect, it is clear that the amount of air diverted is probably more like 5 ton/sec.

http://www.allstar.fiu.edu/aero/airflylvl3.htm

GE is more like the effect of a tailwind than any silly notion of "more lift" being created.

If you held a constant angle of attack and airspeed additional lift is never created. The wing can generate more energy and that energy is what changes the rate of descent.

If you held a constant rate of descent, then the energy is what changes the velocity. Lift force remains constant.

The only thing that changes lift force is another force. Dynamic pressure is the same in GE as it is out of GE.

This goes back to my questions to M_Gunz:


Let's talk about our lift force picture in a climb or descent compared to level flight. This is key to understanding ground effect.

Do we require more lift force in a climb or less left force compared to level flight?

How about a descent? Does our wing need to generate more lift force to descend compared to level flight or less?


Answer these questions correctly and the lack of understanding in the following quote becomes obvious:

http://forums.ubi.com/eve/foru...111012558#3111012558 (http://forums.ubi.com/eve/forums/a/tpc/f/23110283/m/3121062458?r=3111012558#3111012558)

AndyJWest
04-21-2010, 03:49 PM
Thanks for that, Kettenhunde. Very useful. http://forums.ubi.com/images/smilies/25.gif

I've done a bit more experimenting, and confirmed that in the sort of situation I described - holding a constant pitch angle of +6 degrees at 50% or so throttle, the Ju-88 tends to follow the terrain hight, rather than sea level altitude, though obviously it will only climb relatively slowly, so this will be of little use in mountain avoidance! I'd hope to gather some data for a graph demonstrating this, but my AP threw one of its periodic hissy fits and cut out at the critical moment. http://forums.ubi.com/images/smilies/compsmash.gif

I'll probably have another go at this later, and will report back...

Kettenhunde
04-21-2010, 08:14 PM
Thanks for that, Kettenhunde. Very useful.


You are welcome. I hope it is useful to you in determining if your game exhibits the correct effects.

The change in slope of the lift polar appears correct to me.

AndyJWest
04-21-2010, 08:44 PM
...if your game exhibits the correct effects
I think the answer to that has to provisionally at least to be no - it seems to model the drag reduction due to ground effect, and the change in CL, but these are evident at much greater heights above ground than I'd expect. From my measurements, at highish AoA (no way to determine the angle actually modelled, since I don't know the angle of incidence relative to the aircraft datum) there is strong evidence for ground effect at 70m, with an aircraft with a 20m wingspan. I'd almost suggest that there was a misplaced decimal point involved, though there may be more to it than that. Perhaps Team Daedalos can be asked to look into this for a later patch, but for now at least I've shown that IL-2 can model it, and that its presence can be verified. I should probably do some testing with other aircraft types, to see if that throws any further light on what is going on. Meanwhile, I've learned a little more about aerodynamics...

Kettenhunde
04-21-2010, 08:46 PM
The wing can generate more energy and that energy is what changes the rate of descent.

That should read convert and not generate. That is how gliders work, btw.


I've done a bit more experimenting, and confirmed that in the sort of situation I described - holding a constant pitch angle of +6 degrees at 50% or so throttle, the Ju-88 tends to follow the terrain hight, rather than sea level altitude, though obviously it will only climb relatively slowly, so this will be of little use in mountain avoidance! I'd hope to gather some data for a graph demonstrating this, but my AP threw one of its periodic hissy fits and cut out at the critical moment

Will your game airplane hold altitude out of ground effect at those settings?

AndyJWest
04-21-2010, 08:56 PM
Will your game airplane hold altitude out of ground effect at those settings?

A good question. If it does, then something is clearly wrong, though I'd have thought I'd have noticed it if it did. As a said in my last post, the effect seems to occur at a greater height off the ground than I'd expect, but I haven't as yet tested enough to see how much it varies - I'll look into this. It may not be that easy to actually measure though, because as the aircraft moves further out of ground effect, the changed flight characteristics diminish, and finding the 'level flight' altitude for a given throttle setting becomes harder (and less meaningful in any real world context too - I'm effectively testing with a constant-mass aircraft - no fuel burn - which is a little unrealistic).

M_Gunz
04-21-2010, 10:09 PM
Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">M_Gunz says:
If you increase power to a plane in flight, it will climb.


Exactly!

Let's talk about our lift force picture in a climb or descent compared to level flight. This is key to understanding ground effect.

Do we require more lift force in a climb or less left force compared to level flight? </div></BLOCKQUOTE>

Actually it is less lift in a -steady- climb of descent when the vectors are shown, but more drag. The rise of the
plane is a matter of power, not lift at all as long as it is steady. Nose up tilts the lift vector back, it is thrust that
raises the plane and lift actually decreases as the path no longer crosses gravity at 90 degrees ++ in a steady climb or
descent ++. However it did take me seeing diagrams and a few re-reads to understand that and I don't show them here. For
one, I don't know where those sites are now that I went to back before 2000.

What most people do is add power then keep pushing the nose down to gain speed but if you don't then you just climb.
You're not going faster. You're not getting more lift. ** The climb is about power alone. **

Throttle controls height, pitch controls speed as long as you stay on the front side of the power curve. Back side of the
power curve.. uh-uh.

M_Gunz
04-21-2010, 10:33 PM
For one thing though, IL2 is Oleg Maddox and Maddox Game's 'game' and not 'my game'.
I just use it, I don't own it and I sure didn't make it.
A more proper reference is 'IL2' than 'your game'.
It's a matter of respect to the maker as well as for clarity and correctness.

JtD
04-21-2010, 10:34 PM
Originally posted by AndyJWest:

50.0% throttle, level flight at 69m, groundspeed 240 Km/h, indicated airspeed 230 Km/h
47.3% throttle, level flight at 20m, groundspeed 233 Km/h, indicated airspeed 224 Km/h
44.0% throttle, level flight at 3m, groundspeed 222 Km/h, indicated airspeed 213 Km/h

Groundspeed is from 'wonder woman view' instrument, IAS is from DeviceLink.

Very good again, also a good confirmation that your testing setup works, a 2m variance is good. Thanks for posting these. This confirms both drag and lift change.

Regarding the difference between ground speed and indicated air speed, the indicated air speed is off. On our so called "standard atmosphere" map, I'm using a cubic correction formula to get a good relation between indicated and total air speed over altitude. At sea level, my factor is 1.04. It is my impression that the IAS isn't even the one determining the flight physics, if you test the cl/AoA relation you can get the programmed number fairly accurately if you use TAS for EAS at sea level, but IAS is off.

Your findings with w&t off contradict mine, that's something that should be looked into.

Interesting info on the terrain!

I don't consider the w&t effect the ground effect because it goes up that high. That's why I recommended you to switch it off early on. But then you might be right that it is supposed to be it, or part of it, but just goes up too high. It's also not depending on wingspan.

Thanks for taking the time to do this testing!

Kettenhunde
04-21-2010, 11:39 PM
A more proper reference is 'IL2' than 'your game'.


Noted...not trying to offend anyone, just separate the game from reality.

http://forums.ubi.com/images/smilies/16x16_smiley-wink.gif

M_Gunz
04-22-2010, 12:45 AM
Originally posted by JtD:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by AndyJWest:

50.0% throttle, level flight at 69m, groundspeed 240 Km/h, indicated airspeed 230 Km/h
47.3% throttle, level flight at 20m, groundspeed 233 Km/h, indicated airspeed 224 Km/h
44.0% throttle, level flight at 3m, groundspeed 222 Km/h, indicated airspeed 213 Km/h

Groundspeed is from 'wonder woman view' instrument, IAS is from DeviceLink.

Very good again, also a good confirmation that your testing setup works, a 2m variance is good. Thanks for posting these. This confirms both drag and lift change. </div></BLOCKQUOTE>

In level flight isn't the lift = the weight?

I think that what it shows is the ability to maintain lift with less power due to GE in IL2 rather than differences in lift.

BillSwagger
04-22-2010, 02:49 AM
http://www.youtube.com/watch?v...bTjA&feature=related (http://www.youtube.com/watch?v=aY_o_cLbTjA&feature=related)


Bill

deepo_HP
04-22-2010, 02:57 AM
Originally posted by JtD:
Regarding the difference between ground speed and indicated air speed, the indicated air speed is off. On our so called "standard atmosphere" map, I'm using a cubic correction formula to get a good relation between indicated and total air speed over altitude. At sea level, my factor is 1.04. as far as i know, devicelink doesn't read out any TAS... at least haven't yet seen any key for TAS in the 'Devicelink.txt'.
i regularly use 'UDPGraph' for flightpath-visualisations and distance-measuring and the TAS-calculation is described as IAS/3.6*(1 + ALT/15000), with IAS in km/h and TAS in m/s. on the Crimea-map this compares well to any table-resolved TAS/IAS-relation with an error of less than 0.1% up to 150m (i just have never logged at higher altitudes so far).
it is not surprising that the game is all around IAS (or maybe CAS), at least simulationwise. very little errors in a calculated TAS are probably considered nothing to bother for the player.

have i missed anything about a TAS-readout by the game? why would a simulation want to include TAS anyway?
i got 'UDPGraph' and 'UDPTrack' from AVC-site in 2004, but i haven't found anything in the actual 'Devicelink.txt', which indicates additional reads or writes for TAS. but, as i said, i use the tools just for logging recorded tracks and displaying the flightpath then.
however, an error as high as 1.04 on the Crimea-map seems quite a lot. it is obvious that the calculated TAS, which i get, has to be nore or less according the IAS... so i wonder where you get the TAS from and why the IAS should be off?

WTE_Galway
04-22-2010, 03:10 AM
Originally posted by M_Gunz:


In level flight isn't the lift = the weight?



Only kinda more or less approximately sort of ... http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif


For starters at low airspeed the high angle of attack means the Thrust vector acquires a vertical component and hence Lift + the vertical component of thrust are equal in magnitude to weight.

AndyJWest
04-22-2010, 05:05 AM
JtD writes:

I don't consider the w&t effect the ground effect because it goes up that high. That's why I recommended you to switch it off early on.
With 'wind and turbulence' off I don't see any evidence for ground effect at all, from the tests I've done. Given that my 'terrain following' test seemed to indicate it was related to height above ground, rather than height above sea level, I can't think what else to describe it as than ground effect, possibly miss-scaled.

I need to investigate this further.


Deepo_HP writes:

as far as i know, devicelink doesn't read out any TAS

Quite true, it doesn't. The only way you can derive this is by using the 'wonder woman' view dial, which gives groundspeed, and then subtracting windspeed - though as I said, running a reciprocal course at 70m on the Crimea map with 'clear' weather suggests that this 'wind' is only around 1Km/h. (Actually, you can measure windspeed at ground level by looking at the devicelink readings for a stationary aircraft - not realistic, but useful).

In any case, arguing about which is 'correct' is rather missing the point. This is a simulation, so there are no true figures anyway. All you can do is compare what you have, and check for internal consistency, and for a match with what you know about real-world aircraft behaviour.

Kettenhunde
04-22-2010, 05:44 AM
In any case, arguing about which is 'correct' is rather missing the point. This is a simulation, so there are no true figures anyway. All you can do is compare what you have, and check for internal consistency, and for a match with what you know about real-world aircraft behaviour.

From what I can tell based on the slope change of the coefficient of lift JtD posted, the aircrafts behaviors in GE are well simulated.

If you are looking for something to correct, get the taildragger physics corrected. You have to put the wrong control input on many of the games airplanes.

I noticed that the few times I have played.

deepo_HP
04-22-2010, 05:55 AM
hi andyjwest,

it wasn't my intention to take a guess on 'correct' speeds, my post was only related to the correction factor which jtd mentioned.

but since devicelink is pretty much reflecting the data, which the game uses, i would think that the TAS, as it is seen in the no cockpit gauges, is the same calculated from IAS, as 'UDPGraph' does.
i am not at all that much into the origin of the Devicelink-data, but the TAS calculated by 'UDPGraph' is the same as shown in the gauges of wonderwoman. since the data which are missing in Devicelink can be calculated from those on readout, it seems very probable to me that the TAS is indeed derived from IAS in game models (as to be seen in 'wonderwoman' and as logged by simple formula in 'UDPGraph').

you mentioned that windspeed can be determined by readout the data for a stationary aircraft. this is, in my opinion, very much supporting what i was assuming: the game has only one speed, which is IAS (or CAS, i don't know).
which makes good sense for the flightmodel, but would result in erratic TAS (for wind), and not wrong IAS (which was what i was wondering about).

basically, my point is that for any attempt to check for internal consistency, i would consider IAS to be the 'correct' speed to log from game.

AndyJWest
04-22-2010, 06:41 AM
I've not used UDPGraph, so can't really comment on that, but as I see it, the 'wonder woman' instrument is giving groundspeed, which only equates to TAS with no wind, and I don't see how UDPGraph can make an allowance for windspeed.

Ultimately the only 'speed' we can really definitively talk about is that of the simulated aircraft over the simulated map, and given that we know the map scale (10 Km for each numbered/lettered square), we can determine speed relative to that. As far as I'm aware, this speed coincides with the 'wonder woman' readout.

As for what determines the 'physics' of the sim plane behaviour, I'd assumed that an 'equivalent altitude' is calculated from the map weather conditions and the plane altitude, the TAS is determined by subtracting the wind from the groundspeed, and the flight model works from this height and TAS to determine behaviour - you can't really do this with IAS alone because you need TAS for questions of inertia. Most likely, the IAS shown in-game and via DeviceLink is derived by a 'conversion factor' from TAS and equivalent altitude.

One would really need access to the source code to see exactly how this is all done, but there isn't much chance of that...

My 'terrain following' test is proving a little awkward, and so far I've merely demonstrated that a plane reacting to ground effect from rising terrain by climbing at 0.1 m/s is unlikely to get very far if the terrain rises at 2 m/s!

M_Gunz
04-22-2010, 08:18 AM
Devicelink was made to provide instrument readings to game-external instruments or emulators. Devicelink IAS should give
the same value as you see in cockpit.
Crimea summer map should be very close to standard atmosphere where close to seal level IAS=TAS AFAIK.

thefruitbat
04-22-2010, 08:20 AM
Originally posted by M_Gunz:
Devicelink was made to provide instrument readings to game-external instruments or emulators. Devicelink IAS should give
the same value as you see in cockpit.
Crimea summer map should be very close to standard atmosphere where close to seal level IAS=TAS AFAIK.

i always thought that, but it transpires that all the il2 compare data was done on Smolensk, and is actually closer than crimea.

JtD
04-22-2010, 09:14 AM
Originally posted by M_Gunz:

In level flight isn't the lift = the weight?

I think that what it shows is the ability to maintain lift with less power due to GE in IL2 rather than differences in lift.

It shows that you can produce the same lift at the same AoA at a lower EAS, and therefore have a higher lift coefficient at the same AoA.

JtD
04-22-2010, 09:22 AM
Originally posted by deepo_HP:
have i missed anything about a TAS-readout by the game? why would a simulation want to include TAS anyway?
i got 'UDPGraph' and 'UDPTrack' from AVC-site in 2004, but i haven't found anything in the actual 'Devicelink.txt', which indicates additional reads or writes for TAS. but, as i said, i use the tools just for logging recorded tracks and displaying the flightpath then.
however, an error as high as 1.04 on the Crimea-map seems quite a lot. it is obvious that the calculated TAS, which i get, has to be nore or less according the IAS... so i wonder where you get the TAS from and why the IAS should be off?

No, you haven't missed a TAS readout. Unless I have missed it, too. The TAS I am referring to is from the yellow cockpit off gauge. I've checked it against the map distances once and it is as correct as I could determine. I tried to do a constant say 400 km/h, and after an hour I had traveled like 400.02 km across the map. So the yellow gauge is my reference. I've done a few calibration flights, noting the speed on the yellow gauge, checking against the IAS from devicelink and came up with my correction formula.

If you check Andys numbers, you can find the 1.04 factor at sea level, say 233 / 224 = 1.0402.

Note: Just checked my correction and it's 1.0285 for me, not 1.04. Sorry for the memory lapse. At any rate, it's not insignificant.

Edit: I see Andy's told you most of it by now. Nevermind.

JtD
04-22-2010, 09:27 AM
My 'terrain following' test is proving a little awkward, and so far I've merely demonstrated that a plane reacting to ground effect from rising terrain by climbing at 0.1 m/s is unlikely to get very far if the terrain rises at 2 m/s!

http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif

M_Gunz
04-22-2010, 11:23 AM
Originally posted by JtD:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by M_Gunz:

In level flight isn't the lift = the weight?

I think that what it shows is the ability to maintain lift with less power due to GE in IL2 rather than differences in lift.

It shows that you can produce the same lift at the same AoA at a lower EAS, and therefore have a higher lift coefficient at the same AoA. </div></BLOCKQUOTE>

Oh, yah. I see. Or at least at the same pitch. And from John Denker there is less upwash so the actual AOA is even... less?

K_Freddie
04-22-2010, 02:17 PM
Originally posted by TS_Sancho:
http://upload.wikimedia.org/wikipedia/commons/9/99/Karman_trefftz.gif
I'll be curious to know whether these results were done on a wind tunnel (air forced over a wing), or actual tests done on an aircraft flying through an air volume - there is a big difference.
http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif

M_Gunz
04-22-2010, 02:35 PM
Originally posted by K_Freddie:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by TS_Sancho:
http://upload.wikimedia.org/wikipedia/commons/9/99/Karman_trefftz.gif
I'll be curious to know whether these results were done on a wind tunnel (air forced over a wing), or actual tests done on an aircraft flying through an air volume - there is a big difference.
http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif </div></BLOCKQUOTE>

There's not a big difference between the two, the software models are based on what happens in real wind tunnels.

AndyJWest
04-22-2010, 08:55 PM
The 'terrain following' tests are producing some very odd results, though they need more analysis before I can come up with anything definitive. For now, I'll just report that 'ground effect' seems to be much more pronounced over land than over water! Having set my trusty Ju-88 into its 6 degree pitch/50% power mode, and having it settle at about 70m above sea level, it seems to want to climb to more like 200m as soon as it encounters land - I need to analyse my latest tests, and do a few more, before I can be more definitive, but this seems distinctly odd. I'll post links for a couple of .ntrk files tomorrow, though whether they will be of any use to anyone else I don't know - you'd really need DeviceLink to understand what is going on.

AndyJWest
04-22-2010, 09:07 PM
Originally posted by K_Freddie:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by TS_Sancho:
http://upload.wikimedia.org/wikipedia/commons/9/99/Karman_trefftz.gif
I'll be curious to know whether these results were done on a wind tunnel (air forced over a wing), or actual tests done on an aircraft flying through an air volume - there is a big difference.
http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif </div></BLOCKQUOTE>

From the point of view of physics, air 'being forced' over a wing is no different from a wing 'being forced' through the air, I'd have thought, though the relative size of the wing and the tunnel is clearly relevant if one wants to avoid effects caused by interference between the tunnel walls and the airflow over the wing, and careful design is needed to avoid turbulence in the airstream. On the other hand, a wind tunnel does at least give controlled conditions for comparative tests. Once again, our old friends Orville and Wilbur seem to have been ahead of their time - before they built the first practical powered heavier-than-air aircraft they built the first wind tunnel:
http://i958.photobucket.com/albums/ae65/ajv00987k/WrightWindTunnel.jpg
http://en.wikipedia.org/wiki/Wright_brothers

deepo_HP
04-22-2010, 09:08 PM
Originally posted by JtD:
I've done a few calibration flights, noting the speed on the yellow gauge, checking against the IAS from devicelink and came up with my correction formula.

hi jtd and andyjwest,

i guess that i failed to express myself... was a bit late (or early).

i was relating to the speed which is used for data-collection and if they need to be corrected. since devicelink is used, it seems important to know which data devicelink can read and perhaps also conclude, which data are used for flightmodeling.
devicelink has only one entry for speed, which is IAS. in my opinion the IAS is the only speed used for the game's tables and that there is no seperately processed value for the TAS shown in instruments... this makes sense for a simulation and also any other speeds can be calculated on that base (pls someone correct me, if this wrong for a computed model). the point here is that in this case any IAS-correction - based on TAS, which is calculated from IAS - would be the wrong way.
the question remains, how correct the TAS for the gauges is calculated.

i did a quick speed check with a P-38 L on Crimea map.
i used 'UDPGraph' to read the IAS and altitudes (5m, 30m, 50m, 100m, 200m). TAS was noted by gauges in 'no cockpit', temperature (constant 25°C) by instruments on spawn.
before the actual flights, i tested wind on the map, which seemed to be 1.25 km/h from N at airfield. the tests were done southwards then and the wind added to IAS. speed at the point of measure were between 510-540 km/h IAS.
for the calculation of TAS by IAS i used this calculator (http://www.luizmonteiro.com/Altimetry.aspx#TrueAirspeed).

to shorten this off-topic post, i will just call the resulting numbers (if someone is interested in the screenshots, anytime)...
according to other findings posted, there was at all altitudes a difference between TAS calculated from IAS and TAS as shown on instruments, however only in the range of 0.8-0.9% (or 4-5 km/h).
i did 3 more snaps at 30m, 50m and 400m, but at ~210-280 km/h IAS. the differences between theoretical and displayed TAS were this time around 1.5-2.2% - but still the same 4-5 km/h in absolut terms.

in summary, the 'Devicelink' has a readout for IAS, but none for TAS. nevertheless TAS is shown as gauge in 'no cockpit view', but the value is different from what could be expected from IAS. the difference is small, and doesn't seem to be expressed as percentage, but in absolut terms of 4-5 km/h.
since my theoretical background is near to none, i very likely had just no clue how to use the conversion tables - any hints are very welcome.

nevertheless, my main interest is in frequent logging of speed and distance... which makes TAS on gauge useless to me. in this regard:
1. the IAS is the only speed-readout possible by Devicelink. i (still) assumed, that all other speeds can be derived from it - which still works very well, if IAS is used as EAS in tables, but then
2. the TAS-display on 'no cockpit' gauges is too high... in my opinion about a constant value of 4-5 km/h, not a percentage and by far not in the range of 3%! i had the impression of some added corrective, perhaps map-specific (just hypothetical)?

since you both use the displayed TAS for testing and even change the IAS-readout accordingly, i am honestly interested which your the reasons are. true, the speeds don't seem to fit as they should (which i haven't noticed yet, so thx), but i don't see the conclusion why the gauge should be considered the right speed instead of IAS as the only available speed-data (and the only speed which can be observed in all other views)?
btw, there isn't any direct distance-read in Devicelink as well and neither would i think that map-distances are reliable. which is my most concern, as an important reason for my own logging is to determine the length of courses.

sry for the threadnapping, since this is not exactly on topic of 'ground effect'. but nevertheless i general thoughts on measuring ingame data still have a relevance here perhaps. any comments are appreaciated - also if i have shown silly ignorance in the matter of velocities.

AndyJWest
04-22-2010, 09:33 PM
I wouldn't call this 'threadnapping, Deepo, you raise some relevant points. There is little point in trying to measure speeds etc unless you are reasonably sure the figures are meaningful.

Having said that though, I have to say that I'd stick by my assertion that the 'wonder woman' groundspeed is the nearest thing we have to a definitive measure, and that IAS is less reliable. Try running a few tests in the Gulf of Finland winter map - Airspeed at a given altitude is definitely lower than on the Crimea map, say, and indeed IAS is higher than TAS at very low altitudes. This makes no sense at all unless you assume that 'high air pressure' and possibly lower temperatures are responsible for 'instrument error'. As for map-distances not being reliable, I can't see how you can model a 3-D environment without having a consistent scale, and representing it accurately on a map should be simpler than applying IAS fudge-factors.

As I suggested earlier, this is all hypothetical, and we'd need access to IL-2 source code to really understand what is going on.

JtD
04-22-2010, 10:39 PM
Like I said already, I don't think that the FM is depending on the IAS devicelink reads (this being equal to the number in the red HUD or on the cockpit gauges). The P-38 has a function of cl = 0.17+0.092*pitch, which at 12.5° gives you 1.32. That is from the flight model file, so game data. Now if you look at the graph I posted on the first page, you can see that at 12.5° pitch I have a cl of 1.320 (no ground effect, green line), this cl is based on the TAS, with my correction applied. If I were to use the IAS as read from devicelink, I'd end up with a cl of 1.397 for the same pitch. Obviously, the TAS gets a much better match and therefore I think that the IAS from devicelink is nearly irrelevant for FM purposes.

I'd say map distances use the same scale as all other objects in game, if I give a plane a wingspan of 10m, it's the same reference system as stacking two poles 10m apart on a map. If I say it goes 500 km/h, it needs to do 500km on a map within an hour. Why would you use different references?

I think you can trust me on that, I did extensive testing some time ago to find the "correct" speed and a good relation between IAS and "correct" TAS. I found this formula to be fairly exact (maybe 1-2 km/h off at most) from sea level up to 10000m.

ch('TAS'):=ch('v')*1.0285+ch('v')*ch('alt')/21200+ch('v')*ch('alt')*ch('alt')/400000000

WTE_Galway
04-22-2010, 10:40 PM
Originally posted by AndyJWest:
I wouldn't call this 'threadnapping, Deepo, you raise some relevant points. There is little point in trying to measure speeds etc unless you are reasonably sure the figures are meaningful.

Having said that though, I have to say that I'd stick by my assertion that the 'wonder woman' groundspeed is the nearest thing we have to a definitive measure, and that IAS is less reliable. Try running a few tests in the Gulf of Finland winter map - Airspeed at a given altitude is definitely lower than on the Crimea map, say, and indeed IAS is higher than TAS at very low altitudes. This makes no sense at all unless you assume that 'high air pressure' and possibly lower temperatures are responsible for 'instrument error'. As for map-distances not being reliable, I can't see how you can model a 3-D environment without having a consistent scale, and representing it accurately on a map should be simpler than applying IAS fudge-factors.

As I suggested earlier, this is all hypothetical, and we'd need access to IL-2 source code to really understand what is going on.

Just to make things even more interesting, in the case of Real Life aircraft the positioning of pitot tubes and static ports on some aircraft can be such that the GE directly effects the pressures on the ports giving an instrumentation error with IAS reading lower than it should normally.

Fortunately this is not the sort of detail likely to be modeled in a flight simulator.

deepo_HP
04-22-2010, 11:03 PM
hi andyjwest,

you can model any map and give it any size. that does not at all mean it has any scaled dimesion - a 'scale' in virtual environments can only relate to itself. the degree of proper 'scale' can therfore only be found in the detail and consistency of the modeled parts.
i think we have very few, if any near to 1:1 'scaled' maps modeled and those who refer to real scapes are often twisted, like the sicily map for example. in no case you can tell how exact the 'scale', if not by the ingame relation of time and distance. which leads to the problem (actually mine), that i can't test speed by travelling over a map - which is in all it's dimensions only relative to the plane's speeds!

sure it is hypothetical what we talk, but i think relevant for both of us.
so, which problem do you see in TAS being lower than IAS? my first guess would be, that temperatures are lower than ISA - just by the physics of altitude and speed, nothing to do with instruments. but as i said, all these abbreviations are not my favourite playground.
in any case, map-wise different IAS-TAS relations are to be expected, i think. the very matter of all the pressures and temperatures...

i agree, that code-knowledge would be nice.
but i don't have access or knowledge for code-reading and -talking. however, the tools which i have found (and even those were foreign enough to me) show just one single speed available and nothing more would be needed for the game. so happy was i with my li'l logs, until i read here about this gauge being different, which i had never noticed. and since it indeed is different, i am curious about:
- why the gamewise most unrealistic display of TAS in 'no cokpit' gauge not consistent with calculations?
- why is there only IAS as the only speed possible to read, the same which shows on the 'normal' cockpit and which is also used for so many other factors of the flight-model? and
- which one is reflecting the 'real' thing... IAS-data or TAS-display?

if the displayed TAS were correct (in the sense of flight-simulation), it could have been used easily for data-layout (or IAS corrected from the start). since not, it must have a reason... like to express hard to calculate effects, like perhaps ground-effect. in which case it would make sense, if the displayed TAS is representing the 'real' speed in the boundaries of ground-effect - not as part of the basic equations but as an added value or factor. but it doesn't make sense to use it outside the effect (see above).
this is not about fudge-factors which it would be anyway and for both speeds, not to say top of fudgyness to dive with IAS and travel on a boosted TAS-display. either TAS is displayed wrong from a certain altitude on, where IAS comes to play then, or a confusing double-speed-warp will scratch my belief in a beloved simulation.
to some extent, the IAS data and the displayed TAS and IAS seem not according each other (as confirmed and documented), but does this mean that only one is the 'correct' speed, or both are used the same for issues with flightmodels, or one speed is only 'added' to represent a limited effect?

anyways, here is the one solution of the problem which came to my mind - and i am glad things still come there:
not knowing of the trustworthyness of landscale, it would be needed to compare the times for flying the exact same path (waypoint to waypoint) at the exact same TAS on display - but in different altitudes... one in ground-effect and one outside.
if the travelled distance is the same by TAS in both cases, the game uses different speed-tables depending on purpose (heartbreaking then).
if the distances show different, the game has the displayed TAS only for the (assumed) reason of ground-effect, and just keeps it on display but makes further use of a properly calculated TAS by IAS (however strange)... this might have been a workaround perhaps, for a feature added in a patch?

if the distances show other results, this one solution which came to my mind could still be used as a script for a historical comedy. in which case i am still glad for things to come and i would love to behappied by mindful explanations.

TheGrunch
04-22-2010, 11:10 PM
Map scale isn't an issue because the scale of the map squares doesn't change, it's always 10km IIRC.

AndyJWest
04-22-2010, 11:34 PM
not knowing of the trustworthyness of landscale, it would be needed to compare the times for flying the exact same path (waypoint to waypoint) at the exact same TAS on display - but in different altitudes... one in ground-effect and one outside.
if the travelled distance is the same by TAS in both cases, the game uses different speed-tables depending on purpose (heartbreaking then).
if the distances show different, the game has the displayed TAS only for the (assumed) reason of ground-effect, and just keeps it on display but makes further use of a properly calculated TAS by IAS (however strange)... this might have been a workaround perhaps, for a feature added in a patch?

I think you are making the whole thing unnecessarily complex. The simulation must be working at some scale factor, converting form native units (probably floating point) to map units (most likely metres, or centimetres). In order to determine where an aircraft will be at a given point in the future, you need to know (a) it's position, and (b) its velocity. You cannot use any other units than the assumed map scale to determine these. If you say the map is 'wrong' then there isn't anything that is 'right' because nothing has a position at all except in relation to the map. Objects in the sim have to move relative to the defined origin - basic Cartesian geometry. If the groundspeed (not TAS - these are different things) as indicated in the 'wonder woman' view doesn't reflect this, then I can't see how you can make any statement about velocity at all. Nothing is moving, except in relation to something else, and we know that IAS isn't a measure of relative velocity except in identical atmospheric conditions - if I pursue you at a constant distance behind you, at a constant height above you, my IAS will be lower than yours, but our velocity over the ground will be the same.

JtD
04-22-2010, 11:42 PM
deepo, devicelink is a tool that allows you to put the readings of the in game cockpit instruments onto your external devices. For instance, you built yourself an instrument panel, you can use devicelink to make the work. Devicelink is not a tool to read exact flight model workings, so if Oleg has decided to simulate some kind of instrument error, the in game cockpit gauges will show them just the same way devicelink will show them. It's not an accurate reading and it's not intended as such. It needs to be calibrated.


Andy: If the groundspeed ... as indicated in the 'wonder woman' view"

WW does not show ground speed. Try going up vertically, you'll still get a reading, with zero ground speed. It's some sort of total speed, and if you switch wind and turbulence off, it's exactly TAS.

Hannibales
04-23-2010, 03:58 AM
Originally posted by AndyJWest:
Once again, our old friends Orville and Wilbur seem to have been ahead of their time - before they built the first practical powered heavier-than-air aircraft they built the first wind tunnel


Afraid that honour goes to this chap : Frank H. Wenham. In 1871 he raised funds through the Aeronautical Society of Great Britain to build a wind tunnel in Greenwich. http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif


And now back to Ground Effect.......

M_Gunz
04-23-2010, 05:08 AM
Using the FMB or just editing the mission file you can be sure of location of objects and places in game scale.
Sicily as mapped is much, much, much too small but 1m in game is 1m regardless of object or terrain sizes.

Flying long distance low over water from concrete bases in Crimea map I have gotten that IAS is very close to
TAS and no change for compression between low and full speed either. This is minutes long at over 500kph and
only decimals precision difference just looking for clues of what the gauge is telling.

If you want to bad enough then you can tell the position of your plane within maybe 10m on X & Y but you have
to set up ground objects as references and some non-trivial trig to do so. Maybe 10m, or maybe a longer smear
of possibles representing device/data delays it is hard to say and that is the point where I dropped checking
it out.

deepo_HP
04-23-2010, 05:19 AM
hi jtd,

i úse devicelink since i found the 'UDPTools' in 2003, which are just grafical loggers for the data.
devicelink is just a networkclient who receives and sends data from and to the game as host. since UDP is the used protocol, the mentioned tools were named like that.

i am also quite eloborate in landscape-modelling - lately in relation with 'UDPGraph' trying to combine 3D-flightoaths of racecourses with landscapes generated from the unwrapped SFS-files.
3D-flights are very useful to analyse different techniques, like in turns, including all possible energystates during turns.
so much only to explain my practical (if one can say so) interest in the, for me new and intersting observation of divergence between displayed and calculated TAS - which i have never noticed, since i don't like crappy full screens when flying.

so please be assured, that i don't consider it trivial.

however, my interest may be less in the aerodynamical effects of flight-modeling (like yours in 'ground-effect'), but is mainly in the raw data which relate to it. that's why i said sry for the interference. still i hope that i may pose my questions on these data... even if for you the data are probably more useful for prognosis than for their value.


i agree that TAS on display is not as expected as calculated from IAS, at least in a range of altitude.
on a (short) check of my own, i found this difference in altitudes up to 400m and on several maps. i couldn't find any factored difference, in my opinion it looks like an added value - but this could very sure be in the roughness of my speedchecks.

(i'd like to adress andjwest from here on as well... several points are related to his reply) so... hi, andyjwest!


i don't want to talk you into your and andyjwest's way to test 'ground-effect'. the results will be interesting and are awaited for by me as well, but until then i consider too detailed for me.


i said it before, too, and by the risk of repetition (but obviously i failed miserably to communicate this)...
my point of interest and reason for posting in this thread is, that i would hope to come to a decision for myself about which value represents the game's TAS. therfore here another try:

1. my understanding of IAS/CAS/EAS is it's meaning for plane performance (no need to explain the difference to groundspeed/TAS here, but thx) - in my opinion quite important for a simulation. i also think, it is ok to say, that in a tabled simulation it any of the speeds can be used for calculating all others. nevretheless, i consider IAS as pretty much the base speed for modeling, not TAS.

2. i think, there are exact ways to calculate the basic speed-frames if all atmospheric data are known. 'UDPGraph' offers at least 3 different ways to calculate the TAS from 'Devicelink', one of them with the same values as i got by using scientific-calculators, another simplier but so far good for my purpose. for my own, short speed tests i 'calculated' TAS from the IAS-readout by helpful websites full of fantastic atmospheric parameters and compared them to the displayed TAS-value. i am sure, that jtd's formula is the in the same way exact, but i am not good in these kind of equations.
nevertheless, i don't see any reason, why data from devicelink would be considered just as for 'the in game cockpit gauges will show them'. quite a far fetched statement! i always get all detail of engine-management and my positional data from devicelink calculations exactly i have recorded them online or played... without any need for 'calibration'! for very sure, devicelink delivers the data which are actually used as such - perhaps with exceptions (like the speed-derivations we are discussing). any detail in engine-management and of course devicelink does not 'read out exact flight model', nor have i said so.
(to show my use of devicelink, this is a 3-minute-racetrack (http://www.dadatainment.info/wb/alpine_run.gif), plotted from a record by 'UDPGraph'. the coursemarkers are a simple excel-diagram of the mission-file, which fit by +/-2m to the data from track)


considering the above said...
3. i wonder in first order why the IAS data ingame are not reflected in the TAS displayed in 'no cockpit gauge'?
this is a general wondering... since it does not make any sense. it would be a great relief if there was a simple explanation like: 'it is not IAS, but ...', or 'it is not TAS in wonderwoman, but overground speed' (my favourite), or 'TAS in wonderwoman is only relevant for certain effects, like GE - otherwise it is just there on display'.
if there is no such an easy answer, the following question for me is of course, which one is the 'real' TAS?

without any means to have a test-backup here, but only by desperate conclusions:
3.a) the 'real' TAS is calculated from IAS! my name is deepo and i approve to this conclusion.
i haven't yet found a reasonable idea, why IAS would be the only available raw data from the game, but delivering wrong values. furthermore, the full-cockpit instruments show the same as IAS, so they would be wrong as well. the speedbar would be wrong, too.

3.b) the TAS displayed in 'no cockpit' view is the real one!
without evaluating 'no cockpit' view, but why would a regarded simulation offer the only 'real' TrueAirSpeed only in the most un-simulationlike view? the same gauges which display TAS in no cockpit view, also display the very unrealistic 'Altitude over Ground', far from reality.
for all that matters in dogfight and combat, the TAS doesn't need to be displayed for the player. it is an important factor, but the same as it was of no concern in fighter's reality, it is none of the gamer to see. the no cockpit gauge is pure luxury. those who like the hyperreal thing and navigate by map, won't bother as well.

i guess my opinion is easy to recognise, but nevertheless is my attitude not to despite the existence of that display. it is somehow different and somehow it is not easy to decide which is 'real', in my opinion. so, although with a biased view, but not as a conclusion... some possible thoughts:

4. there must be a reason for the slight difference in TASs, so there is probably a reason for the existence. for the 'usual' altitudes i would be surprised if there were a need for correction of IAS calculated TAS. for anything hard to include in the code, it might make sense... like near ground effects. very much to my concern, as most race-maps are very goundloving. other reasons for special use of a corrected IAS would be a feature included in a patch. maybe jet-starts/landings, maybe taildragger taxiing, and last maybe weather-features in different altitudes (like snowflakes in a storm).
in any case, it is my very interest to know where which TAS is taken. i still doubt very much, that we have an overall TAS different from the aerodynamical expected - except perhaps for the purpose of some illusion. still i wouldn't exclude the possibility it is there to camouflage some minor flaws in the model.


that's it. just had to made this post since i was under the impression that i might have been considered either a crittery dork or very simple-minded. both would darken my day.
sorry to those, who accidentally read this and are traumatised by boredom now.



(short addition, for andyjwest's eyes only)

hi andyjwest,

i know how to scale a 3d-landscape. what i tried to express is, that for example the scale of a printed map has it's reference outside the medium - 1cm on map is 1 km in the reference and so it is.
a virtual landscape can reflect real matter as well, but never can the scale be referenced to anything outside. the scale in virtual environments is self-referencing and can only be confirmed by interaction with virtual objects and by consistency. i have no doubts that grids in the game often reflect the distances of the real thing (like okinawa or the newer maps), but for the lack of physical existence it can only be proven by the planes in our case.
the point is that the distances on maps should be absolutely easy to adjust to the desired travel-times and would also be most easy to check. therefor it can not be an option to check the time to travel map-distances to achieve correct speeds, because that is exactly the very thing a developer will have adjusted the maps for.
in that regard, map-travelling can be used to find out which speed they are based on - but quite as well can the speed have been 'adjusted' to make maps smaller. for the question of 'real' TAS this is a main point, since a second TAS makes much sense for being used just as Ground-Speed for map-scaling and traveling.

btw, the Wright's apparatus looks like a salesman advertising a 'coffin-accelerator' in a gothic sci-fi novel. 'buy it in time and your heirs won't be troubled with your remains and love you till the end.'

AndyJWest
04-23-2010, 05:58 AM
I think we'll have to agree to disagree on this one, deepo. In any case, I try to give any speed measurements from my results in both IAS and WW-View groundspeed (or more accurately, WW-View velocity relative to whatever coordinate frame it is using as a datum, since as has been pointed out, it is still there in a vertical climb). If you want to analyse my results using a calibration of the WW-Velocity back to IAS, or indeed to make your own measurements any way that suits your methodology there is nothing to prevent you.

M_Gunz
04-23-2010, 06:09 AM
TAS calculation for the WW view could be via shortcut, or less precise floating-point or even fast integer methods.
How many places are the constants used stored to?

What was pointed out about a constant 4-5kph regardless of speed or alt -- does that say rounding error?
Modeling speeds... are we not looking at m/s or similar? 1 m/s is 3.6kph. Add roundoff to IAS you get.

deepo_HP
04-23-2010, 06:13 AM
hi andyjwest,

we are not talking about the same at all. i guess, due to my failure in comprehensible writing instead of carpet-texting, sry.

anyways, good luck.

AndyJWest
04-23-2010, 06:20 AM
TAS calculation for the WW view could be via shortcut, or less precise floating-point or even fast integer methods.
How many places are the constants used stored to?

'Could be'? True, but you can say that about any data from IL-2. If I'm right and it is velocity relative to game-coordinate-space, then there are no constants to apply at all, other than a simple scaling multiplier.

deepo_HP
04-23-2010, 06:40 AM
hi m_gunz,

i have no clue how any of the values reflect the actual calculation-depth.

i agree though with the general rounding-issue... if it makes sense to give speeds in decimals, might be questionable for a realo-practical view.
i for my part am interested, if and why there are possibly different speed-models - and also in regard to using the game as a racing-ground. there are many issues to handle when turning around pylons, which make it very obvious that the game has not been intended for bulk-flying nearest to obstacles (still the best i ever could expect).
the topics of 'ground-effect' and energy-/speed-relations are of much interest to me, even if my theoretical approach hardly goes further data-sampling - the more i would like to know how near the game is in reflecting real aerodynamic relations.


on a slightly different note... it might be worth to consider the online-aspect of the game. server-client communication is not in high frequency, and it could perhaps be helpful for the prerendering of positions to have some 'corrected' data at hand, which have less relevance to single-player action.

JtD
04-23-2010, 07:15 AM
I think it might have gotten lost, but at sea level, WW-TAS equals EAS, while IAS does not. Also WW-TAS agrees with the coordinate systems of the maps, both horizontally and vertically, while IAS does not.

I'd also say that TAS is more important for flight modeling, since most flight physics relate to that. There are just a few aerodynamic forces calculated from IAS/EAS, and for these you don't even need to calculate it.

I don't know where UDP gets the IAS from, but the devicelink IAS output will not give you exact coordinates (with a standard 0 correction at sea level on Crimea map).

Edit: I think I should explicitly note that nearly all my experiences regarding in game testing come from the Crimea map, 12:00 o'clock noon. Of course, other maps are completely different.

Kettenhunde
04-23-2010, 09:18 AM
I don't know about IL2 but most home computer FM's use EAS.

At least all the ones I have worked on and built FM's for....

That is what the airplane feels and reacts too. TAS is required for modeling mach effects.

If you use EAS, you basically just have to adjust to power at altitude and can disregard all the density effects as they are already accounted for in the EAS velocity adjustment.

It is not very hard to convert EAS to TAS.

Saves a lot of computing power.

AndyJWest
04-23-2010, 10:20 AM
I think you will need to use EAS to calculate aerodynamic forces, but you need to work with TAS and any allowance for wind to calculate updated positions, relative to whatever coordinate frame is used. If both are required, I can't see why the 'wonder-woman' instrumentation can't be giving the correct velocity relative to the map coordinate frame. For now, I'm going to assume it does, unless someone can provide actual evidence that it doesn't. This is all getting rather off-topic anyway. I was trying to investigate if and how IL-2 models ground effect, and I'm now more confused than I was when I started...

Kettenhunde
04-23-2010, 10:28 AM
If both are required,

Both are required. Pure TAS methods require a considerable amount of measured data for accuracy.

M_Gunz
04-23-2010, 02:25 PM
I don't know the form of internal number representation used but I do know what runs faster and might give the
kinds of differences reported so I mention the possibilities as possibly relevant.

For sure they have to position the models in 3D constantly. Ntrk data contains position updates as well. And
they do have to deal with air density -and- supply the instruments with constant data which as far as I have
seen the IAS we get needs no correction for CAS, probably a calculation too many on the framerate hit-list.

Working in meters or feet per second.. that's how it gets down at most aero sites. Does IL2? I dunno but it
seems that native coding in units-per-hour would require a lot of unneeded divides. Perhaps units in the 10ths
or 100ths of a second would make more sense depending on how far apart the flight calculations (as opposed to
the graphics tweens) are made and at that point I'd be looking at integers to do the actual math in since they
are a whole lot quicker in general. 32 bit integers even as millimeters allows +/- 2 million km.
But what they use... I dunno!

Who uses EAS to calculate the rectilinear equations in a zoom? There is much better done either TAS or EAS so
why not they use the best/quickest to calculate each force used in the physics-based model? It is forces and
outcome, not pre-determined results from control actions aka-table-driven modeling used in the receding past.
Even EAW is beyond that as the file structure reveals.

deepo_HP
04-23-2010, 04:17 PM
hi jtd,


Originally posted by JtD:
I think it might have gotten lost, but at sea level, WW-TAS equals EAS, while IAS does not. not sure if i got it...
the EAS calculated from 'WW'-TAS certainly won't match the IAS, if the 'WW'-TAS already is different from the TAS calculated from IAS! that is the whole point here, not?
do you relate to other performance analysis you have done, from which you concluded back on EAS?


Originally posted by JtD:
I don't know where UDP gets the IAS from, but the devicelink IAS output will not give you exact coordinates i might suffer from a springrelated slowdown, but you lost me again...
'UDPGraph' uses devicelink, so it gets IAS via devicelink, key 30. coordinates are not included in returns on speed.

for grid-coordiantes in general... at least i don't know of any possibility to gather absolut positional coordinates, but i haven't been in need of them so far. for distances, the relative positions can be derived from devicelink readouts.
that is pretty much why 'UDPGraph' is that valuable tool to me: it doesn't just read the data, but offers trigonometrical functions, basic energy-equations, integration of custom formulae - and all the comfy logging and timestamping.
the logged data are displayed as an editable 3D-visualisation, and can be called from there stepwise.
for my purpose, i especially like that i get distances which are not grid-based, but actually related to the flight-path itself.


Originally posted by JtD:
I think I should explicitly note that nearly all my experiences regarding in game testing come from the Crimea map, 12:00 o'clock noon. Of course, other maps are completely different. yep, that was understood. i'm not much in testing, but i took it that crimea is considered to have a standardised atmosphere for that purpose? anyway, that's why i did my quick-checks there according to your actual testing.
and sure, other maps are different - i mentioned them only in reagard to andyjwest's comment on speeds at low temperatures.

WTE_Galway
04-23-2010, 09:29 PM
Originally posted by AndyJWest:
I think we'll have to agree to disagree on this one, deepo. In any case, I try to give any speed measurements from my results in both IAS and WW-View groundspeed (or more accurately, WW-View velocity relative to whatever coordinate frame it is using as a datum, since as has been pointed out, it is still there in a vertical climb). If you want to analyse my results using a calibration of the WW-Velocity back to IAS, or indeed to make your own measurements any way that suits your methodology there is nothing to prevent you.

AFAIK TAS is a vector with vertical and horizontal components not a ground speed so the WW velocity may well be estimating TAS.

AndyJWest
04-23-2010, 10:01 PM
Isn't insomnia wonderful. Still awake, and browsing the forums...
(Me I mean, not you. It may be an entirely sensible time of day to be online where you are)

WW Velocity appears to be a vector, as you say, but won't match TAS if there is any wind involved.

Back on the original topic, I've come to the conclusion that at least in the case I'm testing (Crimea map in 'clear' weather at noon or thereabouts), there are two different 'ground effect' scales - one over water, and another over land. In fact if you experiment with a low and slow aircraft, you can demonstrate this without resorting to autopilots, DeviceLink, or anything else. As you cross from sea to land, you get a nose-down pitching movement, followed by a tendency to climb, whereas crossing from land to sea you go nose-up and then descend. Odd, but definitely there. Even apparent as you cross a river inland. What I haven't yet determined is the degree to which ground effect over land is actually related to terrain height. I also need to do tests on other maps, and with other aircraft.

JtD
04-23-2010, 11:28 PM
not sure if i got it...
the EAS calculated from 'WW'-TAS certainly won't match the IAS, if the 'WW'-TAS already is different from the TAS calculated from IAS! that is the whole point here, not?
do you relate to other performance analysis you have done, from which you concluded back on EAS?

I'll sum it up again. It's only my opinion, but I'm fairly certain I'm correct.

1. WW-Speed = total speed in the il-2 world
2. WW-Speed = TAS if there is no wind (WW-TAS)
3. Measured lift coefficient over angle of attack is wrong if you use devicelink-IAS at sea level.
4. Measured lift coefficient over angle of attack is correct if you use WW-TAS at sea level.
5. For the determination of these characteristics, you'd have to use a correct EAS.
6. Both, devicelink-IAS and WW-TAS should equal EAS at sea level (in a standard atmosphere).

Conclusion:
1. devicelink - IAS does not equal EAS at sea level, it can not be used directly for the calculation of the aerodynamic forces.
2. WW-TAS equals EAS at sea level and can be used for the calculation of the aerodynamic forces.

Eventually, what I'm trying to say is that the IAS you can see in the cockpits or read with devicelink, seems to matter very little in any regard to the simulation, while the WW-TAS seems to be a very solid base for several things.

---

I don't know UDPthings. That's the essence of my statement. You're saying that you can very exactly reconstruct a race / race track through it, which you wouldn't be able to if you were just using devicelink-IAS without correction at sea level, imho.

JtD
04-23-2010, 11:35 PM
Originally posted by AndyJWest:

Back on the original topic, I've come to the conclusion that at least in the case I'm testing (Crimea map in 'clear' weather at noon or thereabouts), there are two different 'ground effect' scales - one over water, and another over land. In fact if you experiment with a low and slow aircraft, you can demonstrate this without resorting to autopilots, DeviceLink, or anything else. As you cross from sea to land, you get a nose-down pitching movement, followed by a tendency to climb, whereas crossing from land to sea you go nose-up and then descend. Odd, but definitely there. Even apparent as you cross a river inland. What I haven't yet determined is the degree to which ground effect over land is actually related to terrain height. I also need to do tests on other maps, and with other aircraft.

On hot and sunny summer days on the coast, you'll be getting an upwind over land and a downwind over sea. The sun can heat the land much better than the sea, so the air above the land gets hotter than over the sea, so it rises. Also, since all the hot land air is rising, there's local low pressure at low altitude over land, while there is local high pressure at higher altitude over land, the sea air will flow towards the land at low altitude, while the land air will float towards the sea at higher altitude. This should not exist over large land masses to that scale, though.

I wonder if that's got something to do with the land-sea effect you describe.

AndyJWest
04-24-2010, 08:00 AM
On hot and sunny summer days on the coast, you'll be getting an upwind over land and a downwind over sea. The sun can heat the land much better than the sea, so the air above the land gets hotter than over the sea, so it rises. Also, since all the hot land air is rising, there's local low pressure at low altitude over land, while there is local high pressure at higher altitude over land, the sea air will flow towards the land at low altitude, while the land air will float towards the sea at higher altitude. This should not exist over large land masses to that scale, though.

I wonder if that's got something to do with the land-sea effect you describe.

Possibly this effect is what IL-2 is trying to model JtD, though I suspect that if it does it is on a very simplistic basis. I'd like to check comparative outside air temperature over land and sea, too. This isn't given in DeviceLink data, but I think some aircraft may have a dial - probably too small to read accurately though. I might get a clue by looking at engine temperatures. Actually, in the real world the effects of convection currents on a coastline would be more complex than a simple updraft - the air rising has to be replaced, so at ground level the vertical component is at its lowest, and horizontal landward flow is greatest, you certainly wouldn't consistently get the sudden transitions that IL-2 gives.

M_Gunz
04-24-2010, 08:12 AM
Isn't there an OAT gauge on the wing of the Bf-110? My XP is down, I can't even check but....

AndyJWest
04-24-2010, 08:35 AM
Originally posted by M_Gunz:
Isn't there an OAT gauge on the wing of the Bf-110? My XP is down, I can't even check but....
I'd have thought it was more likely to be an engine temp gauge there. I'm sure some aircraft have them though, the question is (a) do they work, and (b) are they going to register a small variation?

M_Gunz
04-24-2010, 09:47 AM
I never really looked, only knew there was one on one plane that is maybe the 110 and that after discussion
thread pointed it out.

In the morning and at night the water may be warmer than the ground, you get shifting winds in shore regions
depending on time of day and they can be quite strong too. I've fought such winds cycling before, would much
rather have the tailwind esp when going uphill from Waialua back to post on afternoons. The freaking backroad
hill is steep enough as it is.

AndyJWest
04-24-2010, 10:00 AM
I've just done a quick search on Corsican Corsair's 'Cockpit Arrangement Guide' (http://www.mission4today.com/i...file=details&id=3265 (http://www.mission4today.com/index.php?name=Downloads&file=details&id=3265)) and the following are shown as having OAT instruments:

Do 335
Go 229
Bf 110 (instrument panel top right)
Mosquito FB VI
F4F
SM 79

Yet another thing to investigate...

Edit: corrected link...

Kettenhunde
04-24-2010, 03:16 PM
IAS does not equal EAS at sea level,

Why is this surprising??

It is not going to without a specific installation PEC.

Kettenhunde
04-24-2010, 03:41 PM
the following are shown as having OAT instruments:

All airplanes should have an OAT. It is necessary to avoid icing and calculate speeds.

deepo_HP
04-24-2010, 04:40 PM
hi jtd,

rgr, i haven't read carefully enough and didn't get that 'WW-TAS' relates to 'no wind'-setting.
and thx for the explanation how you calculate EAS. that was what i wondered: which way another reference could be derived.

just out of interest (so no need to answer if considered off-topic):
how big would you estimate the error to be in the calculation? or better: in the data used for it - like 'pitch' for AoA?
sry, if it was told already... will read from start again later.

so i get it right that in conclusion, i would need 'no wind' in difficulty for testing-purpose. only then TAS is displayed correctly in WW and would equal 'Ground Speed' - while IAS will always read unusable data.
and the displayed WW-speed does not fit with 'wind' either.


in any way, i can live with these findings, as long as the differences between loggable and displayed data are not necessarily reasoned by random 'correction factors'. which is some relief http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif
very interesting all and i definitely have to read some more for my background. i always just thought of performance in relation to IAS and V-speeds... which made it so confusing to me that the game would have unusable data for it.

AndyJWest
04-24-2010, 04:43 PM
Originally posted by Kettenhunde:
All airplanes should have an OAT. It is necessary to avoid icing and calculate speeds.
I'm listing the ones that have them in IL-2 - how well they work I don't know. The sim doesn't simulate icing. I believe Oleg Maddox said that this will be modelled in SoW:BoB though. It is definitely modelled in the later MSFS versions, and needs attention - carb icing results in quite a power drop.

The reason air temperature seemed relevant to this thread was that there seems to be a performance difference between flying over land and sea - at least at low level. I think I'm right in saying that higher air temperatures tend to result in lower performance though, so the 'land warming faster than sea' explanation doesn't make sense unless an updraft is also modelled, and that ought to be a local phenomenon.

AndyJWest
04-24-2010, 04:50 PM
just out of interest (so no need to answer if considered off-topic):
how big would you estimate the error to be in the calculation? or better: in the data used for it - like 'pitch' for AoA?

I don't think you can use pitch axis measurements for AoA, given that it is probably measured relative to the aircraft datum, and wings are often at a positive angle of incidence to this - wings may also have a twist from root to tip too - The Spitfire is a good example of this. You should also remember that even if the datum is aligned with the wing chord line, AoA would only equal pitch angle in level flight.

deepo_HP
04-24-2010, 05:52 PM
hi andyjwest,


Originally posted by AndyJWest:
I don't think you can use pitch axis measurements for AoA yes, that's why i asked. i just assumed 'pitch'-data were taken...
but just an example i could think of.
my interest is only in a rough estimation of error in the mentioned plot.



Originally posted by AndyJWest:
The reason air temperature seemed relevant to this thread was that there seems to be a performance difference between flying over land and sea - at least at low level. wouldn't it be sufficient to read the data on water- or oil-temperatures immediately after spawn at the required altitudes?
in my opinion, these reflect the outside air of the map - at least they are always the same in any craft, with or without a cockpit hood and show a constant 25°C by instruments on 'Crimea' (up to 50m at airfields and over sea)

about the finding of pitch/lift changes when passing from water to land.
i would be surprised if atmospheric details like local pressure-differences were modeled - while wind-settings inflict with speed readings.
perhaps the differences are caused by the way how water and land are generally defined in their dimensions and their physical attributes.

for example, water is in my opinion for most a simple plane with waves as a bump-map. the landscape is achieved from a very low-resolution greyshade only, but still the rendered polygons need to be high enough (and the same for all) for convincing collision-detection. so i assume that it is not a fully cached model, but rendered related to the players visibility.
besides the details, near ground physics over water have a dimension less to consider and are probably not as demanding.
the noticeable transition effect here could perhaps be caused by the animation of bordeline (isn't that the only animation in maps anyway?). the foam might be mapped on a seperate layer with a changing z-coordinate - which then causes possibly a transition effect.


although crimea is a common standard for tests, wouldn't it be worth to look for a test-map with sufficiently flat terrain (and water) for the purpose of investigating the possible differences in the aerodynamical model over water/land? if they will be observed for sure than, they could be standardised by referencing the water data to crimea-map later.

AndyJWest
04-24-2010, 06:15 PM
The temperature-difference explanation for the effects observed was only one possible factor to consider.

I obviously need to look at other maps too if I'm going to come to a general understanding of IL-2s 'ground effect' modelling, though there are enough flattish areas to do some of the tests required.

Kettenhunde
04-24-2010, 07:58 PM
The temperature-difference explanation for the effects observed was only one possible factor to consider.


Is it non-standard conditions?

What is the speeds and the temperature?

Do you guys know how to convert? It is pretty easy to do.

Kettenhunde
04-24-2010, 07:59 PM
The reason air temperature seemed relevant to this thread was that there seems to be a performance difference between flying over land and sea - at least at low level.

There should be....

It will reverse at night.

AndyJWest
04-24-2010, 08:13 PM
What is the speeds and the temperature?
Unfortunately there is no direct way to obtain OAT from the sim - it isn't one of the parameters provided by DeviceLink.

As you say, a reversal of performance figures at night would be indicative that this was what was being modelled - I'll try that out tomorrow. I also need to see the extent to which this land-sea effect varies with altitude, though this gets problematic as it appears not to be linear, so one is dealing with very small effects at height anyway.

Kettenhunde
04-24-2010, 09:53 PM
Unfortunately there is no direct way to obtain OAT from the sim

Can you read the pressure anywhere? Most altimeters have it.

Can you determine true altitude vs indicated altitude? You could get the temperature from that too.

If you get the pressure then you can figure the temperature.

AndyJWest
04-24-2010, 10:10 PM
Can you read the pressure anywhere? Most altimeters have it.

Again. not given directly by DeviceLink. Actually, since WWII altimeters were little more than a barometer with an alternative scale marked on the dial from what I understand, you can only determine altitude if you know sea-level pressure and vice versa. In fact, don't airports provide air-pressure reports to incoming aircraft even today?



Can you determine true altitude vs indicated altitude? You could get the temperature from that too.

In this case, I think that both DeviceLink and in-game altimeters read 'true'.

We are dealing with a sim that nobody has ever claimed models the 'real world' 100%, but what strikes me both in these discussions and in my earlier efforts to develop an autopilot is how much you can grasp of what is going on by assuming, at least until you get evidence to the contrary, that the IL-2 world conforms to real-world physics. I think with this investigation of ground effect we are maybe finding the limits of the IL-2 model but I still find assuming parallels with 'real world' phenomena are a good starting point for explanations.

deepo_HP
04-25-2010, 06:11 AM
spawning a Corsair on Sarah Toga (17.90m deck-altitude) in open sea, the temepratur of either 'engine.oil in', 'engine.oil out' or 'engine.cyl' reads as 24.90°C. spawning a Corsair without Sarah (0.20m) in open sea: 25.02°C
the same keys deliver for a A6M2-N on a shore (1.80m): 24.10°C, a runway (7.32m) in near distance to the sea: 24.96°C, a spot in midst the forgotten crimean peninsula (86.2m): 24.43°C and in the open sea (1.80m): 24.10°C
and last (accidentally)... after attaching a I-16 to a TB-3, the start of the mothership's engines leads to a sudden rise of baby's temperature to 40°C and during their trip at medium altitude (70.20m), the readouts drop in steps until they stay at 24.00°C over any place.

i have left the numbers with 2 decimals, because that is what the data reads. however meaningless this generally is, it shows the proper change in temperature by 0.0065 K/m - as could be expected from barometric altitude in the standard atmosphere of Crimea.
but... not always!
sweet polikarpov, attached to big mama, has a temperature too low. this seems to reflect a cooling by wind - would it be an expression of real-world instrumental issues, the temperature should have fallen by altitude and not over time.
the stranded floatboat was well in the barometrical temperature, but too cool in the float. water can't be the cause since the Corsair was good on Sarah and even better when it drowned at optimal 25°. so far it just leaves that the bird cooled down by floating the wind (which was on). and again it is less likely a technical effect of the real-world, since the faster Corsair had nothing like that (which raises the question though, why cooling does not happen on Sarah's deck).

in my opinion, this little collection shows that the game relates well to real-world physics, but has also it's limits herein - my interpretational thoughts on the data are the following...
besides windy weather and shaking, one would think that moving air has an effect on cooling which could then be expected as a function of the relative speed, however:
the Corsair reads the standard barometric temperature and the speed of Sarah, the A6M has a significant drop in temperature by just floating, and the attachment has just about the same cooling although on a fast trip.
it has to be assumed, that there is no wind in game.
the observation that engine-temperature sometimes changes abrupt or in discrete steps, can possibly mean that it is not part of the environment as well.

i'd see this as some more evidence that atmosphere is not physically modelled, at least not in temperature and air. related effects are not simulations of the dynamics between plane and surrounding medium, but already fully present in the flightmodel, triggered by positional data and with parameters set by mapproperties.
this is not news... propwash and mach-effects are frequent topics. in my opinion, the preferences have been set from flightmodel over damage and ballistic to atmosphere - which i find reasonable, satisfying and very well done.
in relation to the topic of the thread, i would be surprised, if temperature or any atmosphere had a parameter in the x-y-dimension to represent land-water-distribution.
nevertheless, this was just a short setup (never knew how to get attached before http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif) and not been reproduced yet. please point out what i have forgotten to take into consideration, wrongly concluded or left out as alternative. this is quite exctiting.

AndyJWest
04-25-2010, 07:02 AM
Nice idea with the I-16 and TB-3, deepo. Strange results though.

It is a bit misleading to expect 'cooling by wind'. So-called 'wind-chill' only occurs with an object significantly warmer than the airflow, so in this case all it would do (in the real world) would be help the gauge read the correct temperature more quickly.

As you say, there seems to be some evidence that the sim isn't really modelling the real world that well here, but it would be nice to find out what is going on.

Kettenhunde
04-25-2010, 07:38 AM
: 24.10°C

IS Crimea supposed to be standard conditions? 15C is standard temp at sea level.

JtD
04-25-2010, 08:30 AM
Originally posted by AndyJWest:

I don't think you can use pitch axis measurements for AoA, given that it is probably measured relative to the aircraft datum, and wings are often at a positive angle of incidence to this - wings may also have a twist from root to tip too - The Spitfire is a good example of this. You should also remember that even if the datum is aligned with the wing chord line, AoA would only equal pitch angle in level flight.

Pitch equals AoA in level flight in game. There's also no variable incidence for the wing, and the wings incidence equals pitch. The game's fm isn't that complex in this regard.

thefruitbat
04-25-2010, 08:40 AM
Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">: 24.10°C

IS Crimea supposed to be standard conditions? 15C is standard temp at sea level. </div></BLOCKQUOTE>

no, there is no standard unless you use mods, and use the flight test map. Smolensk is closest in the stock game.

Kettenhunde
04-25-2010, 08:42 AM
There's also no variable incidence for the wing, and the wings incidence equals pitch.

Very confusing statement due to misuse of terms.

"wings incidence equals pitch" means the wings are attached at the angle of pitch.....

???? Is that what you are trying to say??

I think you mean to say there is no angle of incidence or aerodynamic twist modeled just as bet there is no PEC correction either.

Kettenhunde
04-25-2010, 08:54 AM
there is no standard

OK!

It would be nice to get a pressure reading. Somebody look on the altimeters and give me the settings if you can.

Assuming standard pressure, we already up to a 1.03156 density ratio.

MMMMMMMMMMMMM

I am glad you can be an expert at this stuff in an hour at the library!


JtD says:
you can find the 1.04 factor at sea level,

Probably would not be floundering wondering about this mystery "correction factor".

It is the density ratio corrected for actual conditions.

thefruitbat
04-25-2010, 09:09 AM
from the load.ini of the crimea map,

MONTH=7
DECLIN = 43
PRESSURE = 745
TEMPERATURE = 25

Smolensk,

MONTH=7
DECLIN = 52
PRESSURE = 745
TEMPERATURE = 17

from the flight test map,

MONTH=6
DECLIN = 30
PRESSURE = 760
TEMPERATURE = 15

JtD
04-25-2010, 09:39 AM
Did you hack the game to see this or can I find that, too? Would be nice to know more.

So we're dealing with a too high temperature, too low pressure (at sea level only?) and what's the declin?

Kettenhunde
04-25-2010, 09:40 AM
Assuming 745 is mmHG....

The density ratio at sea level for the conditions posted is 1.0554.

That is close enough. That 1.04 is just the density correction to bring the data to standard conditions.

AndyJWest
04-25-2010, 09:44 AM
Thanks, fruitbat, useful to know.

At the moment my AP test setup isn't working with mods, though there is no reason in principle why it shouldn't.

Out of curiosity, do you have the figures for the Finland winter map? I'd always thought that from performance, it was 'high pressure', hence the ASI reading being higher than TAS (or whatever WW-Velocity gives) at low altitude.

JtD
04-25-2010, 09:56 AM
Could also be low temperature. http://forums.ubi.com/images/smilies/winky.gif

thefruitbat
04-25-2010, 10:10 AM
JTD, check PM's,

Andy, here is the data for the Finland winter map,

MONTH=11
DECLIN = 53
PRESSURE = 750
TEMPERATURE = -20

Kettenhunde
04-25-2010, 10:20 AM
Could also be low temperature.


It is a math formula and at the conditions fruitbat gave, that is the result assuming 745 is in mmHG. Density is pressure corrected for temperature.

That means the temperature is one of the numbers you crunch....

You know everything, why don't you continue to discover and wonder at mysterious corrections?

Once more, I have waited a whole week for you to refute the lift force claim.

Simple basics....

I purposely wrote that weight is the sin of the angle of climb. The weight vector is the cosine and thrust is the sine.

Any climb triangle will show you that.

Then I led you right to the answer by talking about force vectors.

It would take about 2 minutes with some simple high school algebra and a little knowledge of the basics to prove that in some conditions lift force does increase in GE.

M_Gunz
04-25-2010, 10:31 AM
Originally posted by JtD:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by AndyJWest:

I don't think you can use pitch axis measurements for AoA, given that it is probably measured relative to the aircraft datum, and wings are often at a positive angle of incidence to this - wings may also have a twist from root to tip too - The Spitfire is a good example of this. You should also remember that even if the datum is aligned with the wing chord line, AoA would only equal pitch angle in level flight.

Pitch equals AoA in level flight in game. There's also no variable incidence for the wing, and the wings incidence equals pitch. The game's fm isn't that complex in this regard. </div></BLOCKQUOTE>

That or they assume one AOI for all planes. What's the Cl at zero pitch?

M_Gunz
04-25-2010, 10:37 AM
Originally posted by thefruitbat:
from the flight test map,

MONTH=6
DECLIN = 30
PRESSURE = 760
TEMPERATURE = 15

Flight Test Map: temperature, check,....

Declin as in declination might be latitude?

M_Gunz
04-25-2010, 10:39 AM
Originally posted by Kettenhunde:
It would take about 2 minutes with some simple high school algebra and a little knowledge of the basics to prove that in some conditions lift force does increase in GE.

Assuming you are not on ignore and your posts are even seen.

JtD
04-25-2010, 10:45 AM
With the infos given by fruitbat regarding the conditions of the crimea map, my theories stating that the devicelink IAS does not equal EAS do not really work out. My test based 1.0285 correction factor would be calculated 1.0274 with the numbers given, so devicelink IAS would be EAS exactly.

But then I wonder why the cl's don't match the IAS. Guess I'm off to more testing. Or we could just ask Team Daidalos. But where's the fun in that. http://forums.ubi.com/groupee_common/emoticons/icon_biggrin.gif

Kettenhunde
04-25-2010, 11:08 AM
Assuming you are not on ignore and your posts are even seen.


My post are seen, he was answering the points made in the discussion.



There is nothing to argue either. It just some simple math that is what it is....

There is no need for drama or any lengthy reply or even a requirement to address anyone specifically.

AndyJWest
04-25-2010, 11:12 AM
Just to throw a bit more confusion into the mix, I've just now a bit of testing on the Crimea map at midnight, and the 'ground effect' I was seeing over water seems to have disappeared!

I'll have to look into this further - I may have been doing something daft - lack of sleep is catching up with me, and my brain says do something less taxing...

JtD
04-25-2010, 11:52 AM
So it might be an effect related to thermals, not ground effect.

Might be a good idea to do a temperature check the same way deepo did it.

JtD
04-25-2010, 11:54 AM
Originally posted by M_Gunz:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by JtD:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by AndyJWest:

I don't think you can use pitch axis measurements for AoA, given that it is probably measured relative to the aircraft datum, and wings are often at a positive angle of incidence to this - wings may also have a twist from root to tip too - The Spitfire is a good example of this. You should also remember that even if the datum is aligned with the wing chord line, AoA would only equal pitch angle in level flight.

Pitch equals AoA in level flight in game. There's also no variable incidence for the wing, and the wings incidence equals pitch. The game's fm isn't that complex in this regard. </div></BLOCKQUOTE>

That or they assume one AOI for all planes. What's the Cl at zero pitch? </div></BLOCKQUOTE>

According to my testing, it's the same number you find in the fm files for zero AoA. For instance, 0.17 for the P-38.

Kettenhunde
04-25-2010, 12:53 PM
According to my testing, it's the same number you find in the fm files for zero AoA. For instance, 0.17 for the P-38.


All the wings properties such as twist, incidence, and airfoil selection are accounted for in this number.

Basic stuff.

AndyJWest
04-25-2010, 04:14 PM
The outside air temperature gauge on the Bf 110 seems to work, and give the same results as the map temp settings at ground level. The gauge is top right on the panel, and reads +-40 degrees C. Reading the gauge is possibly complicated slightly by parallax error.

Crimea, noon, 41m (map temp 25 C):
http://i958.photobucket.com/albums/ae65/ajv00987k/Bf110-Crimea-noon-41m.jpg

Smolensk, noon, 138m (map temp 17 C):
http://i958.photobucket.com/albums/ae65/ajv00987k/Bf110-Smolensk-noon-138m.jpg

Finland, noon, 2m (map temp -20 C):
http://i958.photobucket.com/albums/ae65/ajv00987k/Bf110-Finland-noon-2m.jpg

There didn't seem to be any temperature difference at midnight.

The temperature reduces with altitude, as would be expected.
Crimea, noon, 4000m (gauge reading about -2 C?):
http://i958.photobucket.com/albums/ae65/ajv00987k/Bf110-Crimea-noon-4000m.jpg

Crimea, noon, 8000m (gauge reading about -28 C?):
http://i958.photobucket.com/albums/ae65/ajv00987k/Bf110-Crimea-noon-8000m.jpg

M_Gunz
04-25-2010, 07:37 PM
Originally posted by JtD:
According to my testing, it's the same number you find in the fm files for zero AoA. For instance, 0.17 for the P-38.

Unless the actual P-38 Cl at zero AOA is 0.17 then might I suggest that Cl is tied to pitch-into-path as a way to save storing
an extra value and repeatedly fetching and adding that value during FM calculations? As in "at zero pitch, this is the Cl"?

I would think that someone with blueprints for a P-38 would be able to tell the AOI very closely.

I do wonder if any of these warbirds has a fast enough top end to need to lower the nose below zero to stay level at top speed.
It seems logical to me since they should run pitch=zero at combat speeds, IAS of course.

JtD
04-25-2010, 07:38 PM
Andy, your altitude tests indicate standard temperature drop rate. Thanks for posting your findings.

JtD
04-25-2010, 07:44 PM
M-Gunz, that's what I'm saying. While the game calls it AoA, it is actually aircraft pitch. And it's the same pitch you read from the devicelink output.

There are some planes that have a less than 0 pitch in high speed level flight.

DrHerb
04-25-2010, 07:52 PM
Apparently, Ground Effect should kick in approximately at the height of the wingspan of the aircraft. Just to add to the conversation.

Source (http://www.pilotfriend.com/training/flight_training/aft_perf.htm)


ground effect

Every pilot has encountered the term ground effect. What exactly is it?

The total drag of an airplane is divided into two components, parasite drag arid induced drag. Induced drag is the result of the wing's work in sustaining the airplane. The wing lifts the airplane simply by accelerating a mass of air downward. It is perfectly true that reduced pressure on top of an airfoil is essential to lift, but still that is but one of the things that contribute to the overall effect of rushing an air mass downward. The amount of downwash is directly related to the work of the wing in pushing the mass of air down and therefore to the amount of induced drag produced. At high angles of attack, induced drag is high. As this corresponds to lower airspeeds in actual flight, it can be said that induced drag predominates at low speed.

When a wing is flown very near the ground, there is a substantial reduction in the induced drag. Downwash is significantly reduced; the air flowing from the trailing edge of the wing is forced to parallel the ground. The wing tip vortices that also contribute to Induced drag are substantially reduced; the ground interferes with the formation of a large vortex.

Many pilots think that ground effect is caused by air being compressed between the wing and the ground. This is not so. Ground effect is caused by the reduction of induced drag when an airplane is flown at slow speed very near the surface.

Ground effect exerts an influence only when the airplane is flown at an altitude no greater than its wing span, which for most light airplanes is fairly low. A typical light airplane has a wing span of perhaps 35 feet and will experience the effect of ground effect only when it is flown at or below 35 feet above the surface (ground or water). A low wing airplane is generally more affected by ground effect than a high wing airplane because the wing is closer to the ground. High wing airplanes are, however, also influenced by this phenomenon.

Pilots get into trouble because of ground effect when they precipitate take-off before the airplane has reached flying speed. Take the scenario of a pilot trying a take-off from a poor field. He uses full power and holds the airplane in a nose high position. Ground effect reduces induced drag and the airplane is able to reach a speed where it can stagger off. As altitude is gained, induced drag increases as the effect of the ground effect diminishes. Twenty or thirty feet up, ground effect vanishes, the wing encounters the full effect of induced drag and the struggling airplane which got off the ground on the ragged edge of a stall becomes fully stalled and drops to earth.

Ground effect is also influential in landing. As the airplane flies down from free air into ground effect, the reduction of induced drag as it nears the runway comes into, effect to make the airplane float past the point of intended touchdown. In the common case of an airplane coming in with excessive speed, the usable portion of the runway may slip by with the airplane refusing to settle down to land. A go around will probably be necessary. On short fields, approach as slowly as is consistent with safety.

An airplane also tends to, be more longitudinally stable in ground effect. It is slightly nose heavy. The downwash from the wing normally passes over the tail at an angle that produces a download on the tail. Ground effect deflects the path of the downwash and causes it to pass over the tailplane at a decreased angle. The tailplane produces more lift than usual and the nose of the airplane tends to drop. To counteract this tendency, more up elevator is required near the ground. During take-off as the airplane climbs out of ground effect, the download on the tailplane increases and the nose tends to pitch up.

AndyJWest
04-25-2010, 07:58 PM
There are some planes that have a less than 0 pitch in high speed level flight

It is possible to get lift from an asymmetrical aerofoil at an AoA of zero, or even a negative AoA, depending on the camber, so I don't think you can really state definitively one way or another how angle of incidence relates to the datum.

JtD, I've probably seen the figure somewhere, but what is the 'standard temperature drop rate'?

I think I'll do some tests at 6 am and 6 pm on the Crimea map, after repeating my midnight test just to check that I'd not accidentally turned 'wind & turbulence' off - this would account for my strange findings in the last run.

I'll also do some trials on the Finland winter map. Does it regard sea-ice as water?

Edit ---
Off-topic, but does anyone (Kettenhunde?) know how the 'Höhe über Grund' meter worked? I've seen it described as a radio (radar?) altimeter, but I've a vague recollection of reading that the Germans also used either ultra-sound or possibly infra-red for this. It is actually quite a useful instrument in the Bf 110 - make a pass over a runway before you land, and you can determine its height above sea level! (this only works because of the infallible IL-2 sim altimeter of course.) I'd also be quite interested to learn what blind-landing and autopilot equipment (if any) it had.

Kettenhunde
04-25-2010, 08:16 PM
'standard temperature drop rate'?


He means lapse rate. There is no such thing as "drop rate".


Standard Pressure, Temperature, and Lapse Rate
Sea level standard pressure = 29.92" hg
Standard lapse rate = -1" hg. for each 1000' increase in altitude
Sea level standard temperature = 15°C / 59°F
Standard Lapse Rate = -2°C / -3.5°F for each 1000’ increase in altitude

http://www.swaviator.com/html/...JF03/Basics1203.html (http://www.swaviator.com/html/issueJF03/Basics1203.html)

Kettenhunde
04-25-2010, 08:17 PM
I don't think you can really state definitively one way or another how angle of incidence relates to the datum.

Look at the airfoil characteristics....

It is easy to tell.

AndyJWest
04-25-2010, 08:25 PM
Apparently, Ground Effect should kick in approximately at the height of the wingspan of the aircraft

That seems to be the consensus of most of the sources I've seen, but I'm observing effects in the sim at much higher altitudes, which made me wonder whether it is just 'ground effect' that is being modelled. As I suggested earlier, figures I was getting over water might just be down to a misplaced decimal point in the modelling, but this doesn't explain the land-sea differences.

I think if I've shown nothing else, I've at least demonstrated that measuring aircraft performance in IL-2 isn't as easy as it might seem, even when we can control conditions. 'Real-world' testing must have been much more difficult, even ignoring the problems inherent in testing in a war zone, so how anyone can expect WWII performance figures to be as accurate as some who post on this forum expect, is beyond me.

AndyJWest
04-25-2010, 08:30 PM
Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> I don't think you can really state definitively one way or another how angle of incidence relates to the datum.

Look at the airfoil characteristics....

It is easy to tell. </div></BLOCKQUOTE>

True, but since we are dealing with the IL-2 sim, any 'characteristics' are those of the modelling, and may not comply well with real-world expectations. I've no idea if IL-2 models things like aerofoil camber, or just uses standard assumptions.

EDIT ---
Thanks for the standard atmosphere figures, Kettenhunde. http://forums.ubi.com/images/smilies/25.gif

JtD
04-25-2010, 10:13 PM
In the standard atmosphere, the temperature drops with 6.5 K per 1000m of altitude. So you should be at -1°C at 4000 and -27°C at 8000m on Crimea, very close to your readings.

I think the "Höhe über Grund" was nothing but a standard altimeter with a different zero, basically you'd set it to zero on the runway and could have an easier time landing in poor visibility, provided the weather did not change. It wouldn't give you a reading of how much altitude there was between your plane and that mountain en route, though.

Kettenhunde
04-25-2010, 10:19 PM
I've no idea if IL-2 models things like aerofoil camber, or just uses standard assumptions.

Of course it does, that is part of the airfoil section and is included.

You can look at the airfoil section polar of the P-38's wing and tell they were integrated in your game too.

M_Gunz
04-26-2010, 04:01 AM
They fold as many pre-calculations into one as possible while still modeling the desired characteristics to
cut down on framerate hit. Airfoil camber is something you want to calculate for once and that only during
development, not during load or run time.

The labels used in the data which was never intended for users to see are going to be misleading to anyone
who takes them for pure explanations.

M_Gunz
04-26-2010, 04:06 AM
Originally posted by AndyJWest:
I've a vague recollection of reading that the Germans also used either ultra-sound or possibly infra-red for this.

If they had working IR that good and compact then the night fighters would not have been the way they were.

AndyJWest
04-26-2010, 08:31 AM
You are probably right about the "Höhe über Grund" gauge not being infra-red, M_G. It definitely works like a radar altimeter in-game, though, so I think that JtD's suggestion that it is just another altimeter set to zero at airfield height seems unlikely, unless the sim has it wrong. It only reads to 750m too, which makes this suggestion seem less likely.

Kettenhunde
04-26-2010, 09:39 AM
You can look at the airfoil section polar of the P-38's wing and tell they were integrated in your game too.


Just to be clear, the root airfoil and tip airfoil sections are obviously integrated in your game in order to produce a lift coefficient of .17 at zero angle of attack.

The tip produces a CL of ~.4 at zero angle of attack and the root ~.1.

All the info I have shows there is no aerodynamic twist in the P-38's wing design. The AOI is 2 degrees but that has no bearing on the wings Angle of Attack but is important for Body Angle determination.

M_Gunz
04-26-2010, 12:21 PM
Originally posted by AndyJWest:
You are probably right about the "Höhe über Grund" gauge not being infra-red, M_G. It definitely works like a radar altimeter in-game, though, so I think that JtD's suggestion that it is just another altimeter set to zero at airfield height seems unlikely, unless the sim has it wrong. It only reads to 750m too, which makes this suggestion seem less likely.

It would be very interesting to know just how it worked and the accuracy too.

With old radar ranging you emit a signal and on the return the time interval gives you distance. The thing is that the
length of the emitted signal defines your minimum range at light speed and even the 50's radar unit I worked with had
a minimum range of 250m.

A high-frequency sound bounce using the same technique would have a much shorter minimum range, and maximum too.

They could have used return signal strength though. The proximity shells worked that way, probably still do. That
could have worked with IR emitter and a cadmium cell with light filter available at the time. I would think about
how the texture of the ground would affect the return though, same with a return-strength based sound unit.

If you had an emitter at one end of the plane and a receiver at the other that could determine what angle the
return came in then you could tell by geometry up to the resolution of your angle detection. Perhaps phase of
the return would give that.

I'm sure there's other ways. If they had the ground height radar then would they have used that as a bombing aid
at least for night bombing when the sight would not (?) be more accurate?

So I think that it would be very interesting to know and as far as bets, I think that JtD has the answer.

AndyJWest
04-26-2010, 12:31 PM
More results from testing - I'm comparing airspeed at noon and midnight:
http://i958.photobucket.com/albums/ae65/ajv00987k/Ju-88-ASI-Alt-Throttle-test-results.jpg

Very strange - much more difference at 100m than at 10m. http://forums.ubi.com/images/smilies/blink.gif The more I looks, the confuseder I gets...

---------------

Regarding radio altimeters, the technology seems to have been developed by WWII, though at what time the Germans adopted it I don't know:

In 1924, American engineer Lloyd Espenschied invented the radio altimeter. However, it took 14 years before Bell Labs was able to put Espenschied's device in a form that was adaptable for aircraft use.
http://en.wikipedia.org/wiki/Radar_altimeter

M_Gunz
04-26-2010, 12:49 PM
Excellent!

deepo_HP
04-26-2010, 12:54 PM
Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Standard Pressure, Temperature, and Lapse Rate
Sea level standard pressure = 29.92" hg
Standard lapse rate = -1" hg. for each 1000' increase in altitude
Sea level standard temperature = 15°C / 59°F
Standard Lapse Rate = -2°C / -3.5°F for each 1000’ increase in altitude </div></BLOCKQUOTE>hi kettenhunde,

i took for my comparison a value according to the one you posted:
0.0065 K/m.

(edit: i see, jtd made a note on it.)

the temperatures i gathered, did reflect that value - except for aircrafts which were moving (floatplane, attached I-16). i intentionally read 2 decimals, not because i think it would make atually sense, but to show that the temperatures are indeed an exact calculation by barometric formula in the very small range of altitude between 0-70m.

my guess for higher altitudes is, that temperatures are exactly calculated by lapse rate. the gauges on a Bf-110 should show Indicated Temperatures, not Outside Temps... the difference can be calculated by TAS. comparison between temperature by lapse rate and the temperature on gauges would show, if the game does make this difference.

and you are very right, that 25°C are not representing the actual standard atmosphere. i thought this might refelect the different/other 'standards' of the era...


andyjwest's finding of a temeperature change at night is very interesting here!
it should be worth to find out, how it changes over daytime - if it changes all together at a certain time, or if it changes gradually from noon to midnight.
this could give a further hint on the overall exactness of atmosphere representation.

AndyJWest
04-26-2010, 01:01 PM
Deepo_HP writes:

andyjwest's finding of a temeperature change at night is very interesting here!

Actually, I haven't found any direct evidence for temperature change at night, from looking at the Bf 110 OAT gauge. I'm beginning to wonder if that really gives a 'true' reading, or is just based on the nominal map temperature, minus the factor for altitude, rather than another figure used for calculating aircraft performance.

deepo_HP
04-26-2010, 01:31 PM
hi andyjwest,

ok, but your speed-tests indicate something changes at night.

if the gauges were truly exact, they won't show the OAT, but IAT... with a temperature difference over TAS. which would be nice to see considered, but probably not.
(btw, i think that you have measured 'headwind', not 'wind'... a difference of 1 km/h TAS in winddirection would mean about 1.3 km/h wind. true wind is probably better to determine by turning the airplane on ground to different directions and read the speed indicated)

still your speed-tests reveal differences between day and night, for which i might consider temperature being the main reason. it should be possible to find the according temperatures by the speeds you gathered?
very surprising though that the speeds are increased!

is it possible to switch to various daytimes during one run of your autopilot? this should give instant changes in the speeds - and show if they are gradually or just day/night-like.
it also would confirm, if 'altitude' is 'true altitude'.
it maybe not related directly to the topic, but could give further hints on the structure of the game's atmosphere.

AndyJWest
04-26-2010, 04:34 PM
Further confirmation that performance varies between sea and land:

Another test. Crimea map, running north to south over the Kirch peninsula at 100m above sea level.
Throttle 75.9%, Fuel 50%/unlimited, rads open.

Speed over water 338 Km/h IAS - 351 Km/h TAS
Speed over land 346 Km/h IAS - 358 Km/h TAS

Reciprocal S to N

Speed over water 338 Km/h IAS - 348 Km/h TAS
Speed over land 346 Km/h IAS - 357 Km/h TAS

We are therefore about 8 Km/h (2%) faster over land!

This confirms wind N to S at about 1 Km/h. Height AGL seemed to make no difference to IAS - the
terrain rises to about 60m max.

The test should probably be repeated at different altitudes.

deepo_HP
04-26-2010, 08:35 PM
Originally posted by M_Gunz:
The labels used in the data which was never intended for users to see hi m_gunz,

the data dicussed here are for most gathered via devicelink. devicelink was always a given tool and so the data were always likely to be seen and openly available.



Originally posted by AndyJWest:
I'll also do some trials on the Finland winter map. Does it regard sea-ice as water? hi andyjwest,

to find out if ice has a thickness:
i spawned a A6M2-N in water, where it showed 1.80m altitude.
then i spawned an A6m2-N on a runway (8.32m) and a F4u on runway (8.14m), with a difference of 0.18m.
on finland-map, i spawned the same planes on ice...
A6M2-N was 2.32m, the F4U 2.15m, so showing about the same difference.
this leads to a thickness of ice on finland map 2.32m - 1.80m = ~0.50m.

on a side-note: the temperature readout started for both planes at 0°C, then gradually lowering to 19.99°C - at which point it stayed constant for the F4U, but jumped back to 19.1°C for A6M2-N. this might be another hint, that map-temperature is individually corrected in (some) plane-models.


related to your water/land-speedtests...
have you logged throttle ('power') and 'altitude' from devicelink to check if they stay as you have set them?
perhaps it is useful to include 'manifold' and 'prop pitch' as well in the logging of devicelink-data.

AndyJWest
04-26-2010, 08:49 PM
have you logged throttle ('power') and 'altitude' from devicelink to check if they stay as you have set them?
perhaps it is useful to include 'manifold' and 'prop pitch' as well in the logging of devicelink-data.

I get throttle and altitude continually updated from DeviceLink. I'm not making any changes to prop pitch, so that should stay constant - they do (within reasonable limits, altitude tends to fluctuate slightly - no more than +- 1m). Manifold pressure is unlikely to change much, at a constant altitude over 8 Km/h speed difference I'd have thought.

In principle I can log everything from DeviceLink, but I'd rather stick to measuring direct performance effects first, and then try to figure out the cause.

deepo_HP
04-26-2010, 10:03 PM
hi andyjwest,

i don't know, how your autopilot works and by which means it achieves to keep the set values constant. it was just a thought to check on any possible change of variables.

in this regard, it might be worth to test, how throttle reads if speed is set constant... or if altitude will read as 100m, if throttle and speed are preset.
the idea was to see if the autopilot can perhaps handle some demands better than others (like a tool which i once used, but never investigated further)

Kettenhunde
04-26-2010, 10:52 PM
and you are very right, that 25°C are not representing the actual standard atmosphere.

Great! Just convert those non-standard condition data to standard conditions. Things should line up then.

BTW, I would not get wrapped around the axle on wind speeds. The airplane does not feel a steady wind. IAS, CAS, EAS, and TAS are not effected by wind.

A steady winds primary effect is in relation to the ground.

AndyJWest
04-27-2010, 04:44 AM
BTW, I would not get wrapped around the axle on wind speeds. The airplane does not feel a steady wind. IAS, CAS, EAS, and TAS are not effected by wind.
True, but the problem is that we only have two measures of speed in the sim: IAS (from DeviceLink) and the 'Wonder Woman' view gauge, which seems to register groundspeed in level flight. If you allow for wind (determined by runs on reciprocal courses), and pressure/altitude, the results seem to correlate quite well, at least at low altitudes.

Deepo, I've already given results for some constant throttle/constant pitch axis/altitude tests. All my autopilot is doing is reading parameters from DeviceLink, and then moving the controls (again via DeviceLink) to reach and hold the parameters I set it to. It should be possible to verify most of my results by other means. Actually, it might be useful if other people did - I seem to be the only person in this thread that is actually measuring anything.

Kettenhunde
04-27-2010, 04:55 AM
which seems to register groundspeed in level flight.


That is certainly possible that the it reads ground speed instead of TAS.

In real aircraft testing, we use ground speed to construct the position error curve and determine all the airplanes speeds. Might provide some insight...

Do you have a way to measure distance?

AndyJWest
04-27-2010, 05:27 AM
Do you have a way to measure distance?
The maps seem to have a fixed scale in relation to objects on them, which again correlates with the speed values if you assume that the grid is 10 Km per numbered/lettered square. It is also possible to check the internal coordinate values for objects placed on maps etc. There isn't a 'GPS' type readout for position though, but depending on difficulty settings, an indication of the the aircraft's position on may be shown the in-flight map. I added a feature to my autopilot to do continuous 'dead reckoning' to calculate coordinates relative to the starting point, which worked reasonably well, if there was no wind - a bit of a gimmick though really.

JtD
04-27-2010, 01:56 PM
Regarding Höhe über Grund in the Bf-110, it's a Funkhöhenmesser (radio altimeter).

Good overview over German cockpit insturments. (http://www.cockpitinstrumente.de/instrumente/katalog/hohenmesser/hohenmesser.htm)

Some info on Funkhöhenmesser FuG101. (http://www.dornier24.com/pages/equipment/FuG101.html)

---

Interesting night / day info.

Re 100m / 10m effect, the proximity to the water might keep the conditions at 10m more constant than at 100m.

Not sure why the plane would go this much faster at night, though.

And before we all go nuts about temperatures and pressures, we can't be sure that density in game actually depends on them like it does in real life. And the EAS / TAS depends on density, nothing else.

You can find your position by making an event that's recorded in the log file, like wingtip smoke, going to a gunner seat, crashing or starting the plane.

M_Gunz
04-27-2010, 03:00 PM
Originally posted by deepo_HP:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by M_Gunz:
The labels used in the data which was never intended for users to see hi m_gunz,

the data dicussed here are for most gathered via devicelink. devicelink was always a given tool and so the data were always likely to be seen and openly available. </div></BLOCKQUOTE>

And for those it's okay to go by the labels knowing they are game data.

But there have been more than two posts referring to what I wrote about, otherwise I wouldn't have posted that.
Thank you for your concern.

M_Gunz
04-27-2010, 03:03 PM
There's a map called Flight Test Map with Std Atm and perhaps still weather?

AndyJWest
04-27-2010, 03:31 PM
JtD, thanks for the info on the Funkhöhenmesser. http://forums.ubi.com/images/smilies/25.gif

I hadn't thought at looking at log files - useful to know.

M_G, I know about the mod Flight Test Map. I'll do a few tests on it at some point, though my AP isn't working with mods currently - I can do basic speed runs though.

M_Gunz
04-27-2010, 05:06 PM
A mod... that's why I never saw it.

deepo_HP
04-27-2010, 09:46 PM
Originally posted by AndyJWest:
Actually, it might be useful if other people did - I seem to be the only person in this thread that is actually measuring anything. as you wish...

i placed 2 'objects 61' with a distance of 150km of each other, over sea in the most western part of Crimea.
the test was setup as a mission, with a Ta-152 at unlimited fuel:

<pre class="ip-ubbcode-code-pre">[g0122]
Class air.TA_152H1
Fuel 100
[g0122_Way]
NORMFLY 2900.00 2000.00 150.00 340.00 &0
NORMFLY 2900.00 8000.00 150.00 340.00 &0
NORMFLY 2900.00 165000.00 150.00 340.00 &0
[Buildings]
0_bld House$IndustrialFactoryChimney1 1 2900.00 10000.00 360.00
1_bld House$IndustrialFactoryChimney1 1 2900.00 160000.00 360.00</pre>

tests were recorded and 'IAS', 'altitude' (=TA) were read from devicelink, the speed displayed in 'no cockpit' view (=WWS) was noted during the (boring) run. the times (=T) when the plane passed over the chimneys were taken by stopwatch.
i did two runs with Heading 0 and two runs Heading 180 (with and without wind each).

general results:
the differences for all runs were very small and well inside the manual use of stopwatch! therefor i measured windspeed at 3 points on the course by a Corsair placed on a stationary CV... it showed, that the speeds were varying between 0.1, 0.2 and 0.9 m/s. turning the CV revealed even different -0.1, -0.1 and -0.8 m/s. it seems, that wind is not any given local setting, but randomly created by the plane (as i have assumed before).

Originally posted by Kettenhunde:
Just convert those non-standard condition data to standard conditions. Things should line up then. thx for the hint! the issue of non-standard temperature has been accounted for so far, but i think pressure still is considered standard. i remembered the basic configuration for 'UDPGraph', in which TAS was calculated by IAS at 745 mmHg and 25°C. so i used these values and indeed they lined up http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif (at least in hope so)

the following shows the <span class="ev_code_YELLOW">measured results</span> and the <span class="ev_code_GREEN">calculated numbers</span>, considering a <span class="ev_code_RED">29.33 InHg and 25°C at sea-level</span>.
the flight heading 180 gave all the same results, so i just noted the slightly different times and GroundSpeeds (=GS)

Originally posted as text:

http://www.dadatainment.info/wb/test.png

my conclusions:
Crimea is as much as 'standard atmosphere' map as any other map... the atmosphere at sea-level is given as 745 mmHg and 25°C. the calculations for TAS from the IAS devicelink-reading just confirm this.
so IAS from devicelink appears to be exactly what it says: Indicated Airspeed in the game! the TAS, calculated from IAS, fits well to what could be expected by Speed over Ground, which was measured seperately. the displayed speed in 'no cockpit' matches GroundSpeed exactly (rounded, since the gauge doesn't display decimals).
and last - as already said above - wind has to be considered just a random factor, which seems to kick-in at various times and is just limited in range by the weather-settings.

the results include some error, mainly in the times taken by stopwatch. also the 'WWS' is only displayed without decimals, whereas devicelink-readouts give two. furthermore, the tests are so far just for one altitude and don't explain some of andyjwests findings.
however, the intention was to explain why the WWS (WW-TAS) does not agree with the game's IAS, if a pressure of 760 mmHg is assumed. at sea-level, the temperature is 25°C and the pressure 745 mmHg (29.33 InHg).


(edit:
sry for the messed up table... this is probably the very first forum-software ever created - without the slightest way for custom postlayouts. besides one million ridiculous smileys
i have replaced it with an image
edit2:
better image)

AndyJWest
04-27-2010, 10:12 PM
I can see your logic, deepo. <STRIKE>However, if there is a direct relationship between IAS and TAS/WWS one cannot say which one is calculated from the other. One therefore has to look to external evidence, and I think it has been shown that objects are placed on the map at positions defined by a known scaling factor, and that other objects (aircraft) move relative to these objects at a measurable rate - a velocity. This is how the game must work. Nothing can exist unless it has a position, and for moving objects this varies over time. This variation over time is a velocity, by definition. I may be labouring the point, but I can't see how to express it more concisely. IAS isn't a measure of velocity - it can't be, given what we know about how IAS/TAS ratios vary with pressure density. In most situations, this makes little practical difference if one applies the appropriate scaling factors for pressure altitude, but why complicate things? If one assumes that the sim operates in a uniform coordinate-space relative to a fixed 'zero' point, the maths works. Why would anyone wish to program a sim using any other frame of reference?</STRIKE>

Edit----
I'd misread deepo's last post http://forums.ubi.com/images/smilies/blush.gif
Ignore. Original text left, but with strikethrough to remind me that being able to write is no indication of being able to read. http://forums.ubi.com/groupee_common/emoticons/icon_rolleyes.gif

deepo_HP
04-27-2010, 10:25 PM
sry for the missing data, will be replaced soon.


hi andyjwest,

have you read the testsetting at all?
it says exactly what you are suggesting...

again:
i measured groundspeed by map!
i logged data by devicelink and as displayed in 'no cockpit'!
all speeds agree to each other - if calculated by an atmospheric standard, which is different than had been assumed before.

i can not see the point in your last post...

AndyJWest
04-27-2010, 10:51 PM
Sorry deepo, by the look of it I misunderstood your last post - I should have gone to bed hours ago. We seem to be in agreement and I'd misread something. http://forums.ubi.com/images/smilies/blush.gif

I'll have another look at what you wrote tomorrow, when my brain should be less fuddled...

M_Gunz
04-27-2010, 11:01 PM
Originally posted by AndyJWest:
However, if there is a direct relationship between IAS and TAS/WWS one cannot say which one is calculated from the other.

Devicelink IAS is probably the last EAS used in the flight calculations. At least I never was able to find any
correction for PE or high speed and so far never saw a post here or the ORR showing any going back to late 2001.

As far as position there must be a datum (zero point) since online packets do contain at least occasional absolute
position data, as do ntrk files.

deepo_HP
04-27-2010, 11:13 PM
hi andyjwest,

not a problem... i was in rage myself when i replied - the silly 'eve'-board cooked me!

deepo_HP
04-27-2010, 11:26 PM
hi m_gunz,

at least on the 'KhalkinGol'-map the 'Zero'-point is clearly to be seen. start at Tamsaq and head just a bit SWS - and you will see the white mark like an inlay to the ground.
looks a bit like an alien airport. well, for small aliens...

Kettenhunde
04-28-2010, 05:53 AM
thx for the hint! the issue of non-standard temperature has been accounted for so far, but i think pressure still is considered standard.


You are welcome!

I can outline the formula's and steps to convert from non-standard conditions.

deepo_HP
04-28-2010, 10:32 PM
hi kettenhunde,

thx for the offer.

for the test, i took the (true) altitude from devicelink and used the conversion-tables at http://www.luizmonteiro.com/Altimetry.aspx to get the corresponding pressure altitude (for 745/25 at sea-level).
the true temperature was calculated by a lapse rate of 6.5 K/km.

with these values i get a true airspeed of 1 km/h less than indicated in game.
this difference can can have several reason, most of all some data show hyperexact with 2 decimals, whereas others are rounded in integers... so i consider the result sufficiently well.

i don't know which equations are used in the mentioned link above, so i don't know how good the conversions by it's tables are. but at leat it looked very complex.
i always get a bit of shiver when these formulae come up, so was glad to find the site.
what is your opinion on it?

Kettenhunde
04-29-2010, 06:52 PM
what is your opinion on it?


If I understand you correctly, the games TAS lines up with what it should be under those non-standard conditions.

If that is so then I would have to say the game is well modeled.

I wouldn't worry about the rounding differences. Significant digits is good enough.

AndyJWest
04-29-2010, 07:18 PM
I've done a little more testing myself, and have confirmed the over land/over water anomaly occurs at night as well, except this time the difference is the other way round.

Another test. Crimea map, running north to south over the Kirch peninsula at 100m above sea level.
Throttle 75.9%, Fuel 50%/unlimited, rads open, Wind and Turbulence on, conditions clear. Midnight.

Speed over water 354 Km/h IAS - 366 Km/h TAS
Speed over land 346 Km/h IAS - 359 Km/h TAS

About 8 km/h (2%) faster over water than land.
Run N to S A reciprocal course revealed a N to S wind of approx 1 Km/h - ('TAS' is actually WW-Velocity, not corrected for this)

Compare with my earlier test - same conditions at noon:

Speed over water 338 Km/h IAS - 351 Km/h TAS
Speed over land 346 Km/h IAS - 358 Km/h TAS

About 8 Km/h slower over water than land. Speed over land the same.

A test at noon with wind & turbulence off gave 337 Km/h IAS - 348 Km/h TAS.

This is all very strange. I may see if I can obtain indirect relative temperature measurements, from engine temps, to see if there is a land/water difference - the OAT gauge results obtained earlier might be suspect?

Edit ---
Added 'Midnight' to first test results, for clarity.

Kettenhunde
04-29-2010, 07:45 PM
Temperature/pressure change over water and land will have an effect on aircraft performance.

You should see the faster speeds occur in higher temperatures / lower pressures.

The effect is you are changing altitudes. A friend of mine used to race cassut's in the midget class.

http://cassutt.lornet.com/

He won by taking the time to climb to a higher altitude than the rest of the competition. He would use the lower pressure to gain speed. His typical race profile left him above and behind the pack after the climb. He would then work his way to the front and on the last lap, dive for the finish pulling ahead for the win.

AndyJWest
04-29-2010, 08:06 PM
That would make sense, though if it is correct, I think IL-2 has got something backwards - as I understand it, air temperatures should be more stable over sea than land.

It may be complicated by engine power modelling though - as I understand it higher air temperatures can result in reduced power?

I'll have to do some temperature testing first though - this may involve altering my AP program, and I haven't looked at it in months - I hope I can remember how it all works. http://forums.ubi.com/groupee_common/emoticons/icon_rolleyes.gif

Kettenhunde
04-29-2010, 08:54 PM
Yes the higher the temperature, the less power the engine can develop.

The water is also relatively stable in temperature compared to land.

Your game is relatively stable over land.

The effect appears backwards. The over ocean speeds should be very close for both night and day. The land speeds should have the largest change.

http://www.srh.noaa.gov/jetstr...ocean/seabreezes.htm (http://www.srh.noaa.gov/jetstream/ocean/seabreezes.htm)

JtD
04-29-2010, 10:29 PM
Interesting again, Andy. I'd suppose the OAT is about right, though, and I don't think it's all down to a temperature change. You'd see a difference in the IAS/TAS relation if it was, and what you show is marginal. So it can't be that important.

Could you record the pitch of the aircraft? Maybe wind and turbulence does indeed model up- and downdrafts, which you could find that way. A couple of days back you said when you went from sea to land at a constant pitch, you kept climbing, iIrc. So I don't think it's something static, like temperature, but more likely something dynamic, like updraft.

M_Gunz
04-30-2010, 04:33 AM
From sea to land, the land under you rises.
From land to sea the surface under you is going down a slope and then onto level.

You set to fly at constant sea level height (?) but your height over surface and
the slope below you changes. Yet Andy had noted that with ground rising 60m IIRC....

Perhaps I am just confused but the whole thing seems to be confusing and going into
more and more complex explanations for what has in the past been shown to be a pretty
simple and somewhat limited simulation by comparison.

Kettenhunde
04-30-2010, 05:14 AM
Are you 100% sure the game is showing you airspeed?

True Airspeed = Groundspeed in the absence of wind.

AndyJWest
04-30-2010, 06:50 AM
Originally posted by M_Gunz:
From sea to land, the land under you rises.
From land to sea the surface under you is going down a slope and then onto level.

You set to fly at constant sea level height (?) but your height over surface and
the slope below you changes. Yet Andy had noted that with ground rising 60m IIRC....

Perhaps I am just confused but the whole thing seems to be confusing and going into
more and more complex explanations for what has in the past been shown to be a pretty
simple and somewhat limited simulation by comparison.

The speed run tests were at a constant height above sea level, at least according to the altimeter readout - I was running at constant altitude and throttle setting, and measuring the airspeed and groundspeed. In no case did reciprocal runs suggest a wind of more than 1 km/h or so.

It IS confusing. As I've already noted, there seems to be evidence that the 'increased airspeed at very low altitude due to ground effect' modelling is incorrect in that it starts to occur at far to great a height over ground, so it is entirely possible that this land/sea difference is flawed too. At the moment, I'd guess that there is one 'atmosphere' modelled over land, and another over water, but how they differ I can't really say yet.


Are you 100% sure the game is showing you airspeed?
Well, the way I see it, the only 'true' speed in a simulation is that represented by coordinate changes over time. I THINK that this is what the 'Wonder Woman view' instrument gives (as a vector length, but it should give 'groundspeed' in level flight). So is the airspeed 'true' or at least consistent during the tests? I can't say for sure, but I'm getting differences of the same scale in both groundspeed and airspeed, so any error must be slight.

JtD
04-30-2010, 08:48 AM
If you're testing wind speed, can't you simply spawn vertically (Lerche, for instance) and check if there's a vertical wind over land and sea?

15% extra power (which is about the difference between your day and night speeds over the sea) are a bit too much for a slight change in the air. On the other hand, 1m/s up-/downdraft is enough.

AndyJWest
04-30-2010, 09:17 AM
Interesting idea with the Lerche, JtD - I'll try it.

Edit----

It looks as though JtD may be on to something here. I get 'airspeeds' of -3 to -4 Km/h with wind & turbulence on, and 0 km/h with it off, with a stationary (vertical) Lerche either on land or on a carrier - day or night. This suggests that there may be a vertical component to the windspeed - an upcurrent. I'm only measuring IAS to the nearest Km/h, so I can't make meaningful comparisons between the land/sea figures.

Of course it is impossible in reality to get a uniform upward airflow at groundlevel, with no horizontal component - the sim must be wrong here.

Off topic - placing a Luftwaffe plane on a RN carrier for a test isn't very sensible, unless you are interested in testing the carrier AA capabilities: http://forums.ubi.com/groupee_common/emoticons/icon_rolleyes.gif
http://i958.photobucket.com/albums/ae65/ajv00987k/il2fb2010-04-3017-33-23-45.jpg

deepo_HP
04-30-2010, 11:19 AM
hi andyjwest,

can't you assign the carrier and lerche to the same country-colour?
(the carrier shouldn't move also. your pic seems to show a speeding carrier)

about the 'lerche' wind... have you crosstested it by placing another aircraft next to the lerche and checking the wind for it?