PDA

View Full Version : Unpowered dive acceleration



FritzGryphon
10-10-2005, 02:09 AM
I was curious to see how well aircraft fall, so I tested some planes, and compared it vs. an airless environment.

Next thing you know, I have a spreadsheet and a graph. Enjoy.

http://members.shaw.ca/evilgryphon3/dive.jpg

Fun fact: Reducing the fuel load makes the plane fall slower. Adding a bomb makes it fall faster.

danjama
10-10-2005, 03:42 AM
Your graph shows that you started at 0kmh from 4000m. Is that correct? How did u start at 0kmh? Are these actual graphs or made up numbers, caus thats what the 0kmh indicates!

FritzGryphon
10-10-2005, 03:45 AM
How dare you suggest it's made up :P

If you set a takeoff waypoint, then change it to a normfly waypoint, you can get 0 km/h.

I'd upload a track, were this not so easy to replicate.

F16_Neo
10-10-2005, 04:48 AM
Interesting, a drag measuring method (edit: tho not perfect, weight contributes of course)!
Tried the P-51 (laminar flow wings & "magic" radiator)?

Blottogg
10-10-2005, 05:47 AM
Thanks FritzGryphon, that shows drag and weight modeling that I didn't think was so completely modeled. I'm glad to be proven wrong yet again.

Another question I've got is whether prop drag is modeled (I think it is.) Did you do these tests with the prop set to high or low RPM, or auto for the German fighters?

DaimonSyrius
10-10-2005, 08:17 AM
Originally posted by FritzGryphon:
I was curious to see how well aircraft fall, so I tested some planes, and compared it vs. an airless environment.

Next thing you know, I have a spreadsheet and a graph. Enjoy.

For the sake of empirical correctness, the first set of datasamples (at time=0) could be just discarded, since it's attributable to the experimental setup, I mean test conditions. Besides that, since the purpose of the test is to observe what the sim does at t>0, after the initial instant, I guess how did we manage to get the sim plane to t=0 doesn't matter much.

In this sense, Tagert is correct, if we are just 'dialing in' V=0 at t=0, and we're interested in what happens after that in the system, the first set of data points are not truly experimental results, they're just a statement of initial experiment conditions.

Another way to do this might be, for instance, to set (or fly) the plane at, say, 3000 m and 400 km/h TAS, then doing a vertical climb (90 deg) until all or most speed has decayed (or momentum invested in potential E) and the plane falls on itself, and then using only the samples taken after getting the plane on a 90 deg downwards attitude. I would expect that to take several seconds too. I guess the 'preparatory flying' part wouldn't be easy, since of course each aircraft would zoom up to a different point, so an additional requirement might be that the sampling frequency be high enough to provide enough measurements at comparable altitude points for all planes. It would be more cumbersome to do the tests, I would imagine.

Anyway, looking at the general shape of the plot by FritzGryphon, IMO it would be a reasonable prediction that nothing really interesting actually happens in the segment from t=0 to t=5 s, where the magnitudes we're looking at are not deviating very much from the calculated 'ideal' vacuum conditions (in other words, are not deviating from the behaviour of a Galilean brick, let's say http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif)

So either way, it's nice to see that plot, thanks FritzGryphon http://forums.ubi.com/images/smilies/25.gif

S.

Jetbuff
10-10-2005, 08:54 AM
Might I suggest you redo the tests for the planes with CSP using prop-pitch set to 0%? With closed throttle flat blades (trying to maintain 100% rpm) would act like giant disc-brakes. Oh, and don't argue; try it first and compare the results. I know because I could get to ~870kph in the P-47 in only 3000m of free-fall.

MEGILE
10-10-2005, 08:56 AM
The anton has some bada$$ acelleration http://forums.ubi.com/groupee_common/emoticons/icon_eek.gif

Diablo310th
10-10-2005, 09:02 AM
Hmmm a FWA5 outdives a D27 Jug...interesting. That explains why i can't outrun a FW in a dive...not even accounting for initial acceleration. I find it strange that a FW which is much lower in weight in an unpowered dive is outrunning a much heavier Jug.

LEBillfish
10-10-2005, 09:18 AM
What Jetbuff suggests sounds like the way to see how weight/drag are affected. Button it up and let it fall straight down (though might I suggest from a common point of say 300 k/h IAS)...Simply as it hits that mark roll it over and dive OR place your way points 1 over the other with altitudes differing (though there is a minimum speed for norm. fly).

Trouble with dive speeds is "many" do not utilize prop pitch well IMLTHO, and then wonder why a P47 at 40PP in a vertical dive beats one at say 90PP......I rarely play with pitch on climb or level, yet always on all non-auto/LW planes to good effect. Naturally as well trim has a huge effect.

DaimonSyrius
10-10-2005, 09:28 AM
Originally posted by Diablo310th:
Hmmm a FWA5 outdives a D27 Jug...interesting. That explains why i can't outrun a FW in a dive...not even accounting for initial acceleration.
If I understand correctly what has been said here by others, you're not just looking at the effects of weight alone, but also to those of drag. A combination of both. So maybe the plot is indicating a better drag performance by the FW over the P-47 in the sim model? Just suggesting.

Anyway I wouldn't know if that's how it is in the real aircraft. Is it?

I find it strange that a FW which is much lower in weight in an unpowered dive is outrunning a much heavier Jug.

Regarding effects of weight alone on a free fall (with no drag), if you would drop (that is, from V=0) a Jumbo jet and a FW and a penny at the same time and altitude on the Moon (or on an airless Earth) they would "land" at the same time, AFAIK. It's Galileo again.

Someone will correct me if I'm wrong http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

S.

StellarRat
10-10-2005, 01:14 PM
Weight does not matter, but drag does. It appears that Oleg thinks the A5 has a cleaner airframe then the 47. I tend to agree with this (much as it pains me), however the 47 will usually win a full power diving contest with any plane because it falls apart last and has excellent elevators to pull out of the dive. Also, due to the size and weight of the 47 it should retain speed out of a dive longer than the A5.

Stackhouse25th
10-10-2005, 01:19 PM
in

faustnik
10-10-2005, 01:29 PM
Nice test FritzGryphon! Would you have time to do a full power dive test?

This is not only an interesting test but, very practical as well. It's nice to know what you can dive away from and what you can't in the sim. http://forums.ubi.com/images/smilies/25.gif

IL2-chuter
10-10-2005, 05:04 PM
Throttle off prop drag would definately be an issue. The German fighters should have, as in RL, full feathering props and American constant speed props (hydraulic and, I believe, electric) won't bump up off the low pitch stop until about 50% rpm due to governor operation limitations.

As far as the 0/4000 starting point not being a test "result" . . . maybe it's not, but it's the most consistant starting point you can come up with and is the direction from which your next point comes from so I would say it's a valid point on the chart. Just my (the science major part of me, that is) opinion.

FritzGryphon
10-10-2005, 06:43 PM
It appears that Oleg thinks the A5 has a cleaner airframe then the 47


One other thing that is usually overlooked, is the actual size of the planes.

FW-190 was tiny for a 5 ton plane. I think even smaller, length and spanwise, than a 109. And the wings, very thin.

P-47 is 7 tons, but also very large (largest single engine fighter of all?). Frontal area would be great, and wings, more thick.

----

In this case, propellors were left 100% or auto on all. I agree that the propellor may alter the results somewhat.

In fact, when you turn off the engines, you get bizarre results. Some planes will actually slow down at some points during the dive. Try to dive an I-16 straight down with the motor off; it will alternately accelerate, then decelerate.

IL2-chuter
10-11-2005, 02:58 AM
Concerning the I-16 . . . having actually shut off a motor in flight then having trouble getting it started again (no starter . . . http://forums.ubi.com/images/smilies/16x16_smiley-surprised.gif . . . it was a long time ago . . . 1978 . . . ), I would guess that when the prop stops rotating the aircraft speeds up and then the prop starts rotating and the aircraft slows down till the prop stops again.


PS. I got the engine (A65, wood prop) going by diving at 135mph (129mph Vne, 7AC) from about 2000ft. There's some sort of lesson there . . .

Diablo310th
10-11-2005, 06:09 AM
Originally posted by faustnik:
Nice test FritzGryphon! Would you have time to do a full power dive test?

This is not only an interesting test but, very practical as well. It's nice to know what you can dive away from and what you can't in the sim. http://forums.ubi.com/images/smilies/25.gif

exactly faustnik

WOLFMondo
10-11-2005, 06:18 AM
Originally posted by FritzGryphon:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">It appears that Oleg thinks the A5 has a cleaner airframe then the 47


One other thing that is usually overlooked, is the actual size of the planes.

FW-190 was tiny for a 5 ton plane. I think even smaller, length and spanwise, than a 109. And the wings, very thin.

P-47 is 7 tons, but also very large (largest single engine fighter of all?). Frontal area would be great, and wings, more thick.
</div></BLOCKQUOTE>


Also the P47 has 8 guns which protrude from the wings which would interupt airflow, the FW190 only has 4 or 2 depending on the version. Its annular radiator doesn't interfier with the airflow as much as the flaps on the P47's would also.

WholeHawg
10-11-2005, 06:59 AM
That is interesting considering the 47 was historicaly known for its ability to dive.


Has anyone done the same thing with 100% power on?

Aaron_GT
10-11-2005, 07:32 AM
That is interesting considering the 47 was historicaly known for its ability to dive.

This was an unpowered test. Most dives done in combat are powered dives and the engine in the P47 retains its power very well at high altitude. The power loading at high altitude changes the picture quite a bit. Also drag at higher mach also isn't brought out fully in the tests posted at the head of this thread.

IL2-chuter
10-11-2005, 11:42 AM
P-47 pilots were told they COULD NOT power-on split-esse below 10000ft, not enough altitude. I do it routinely (in the game, not RL http://forums.ubi.com/images/smilies/16x16_smiley-indifferent.gif) below 2000ft. WOOHOO. http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif


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

FritzGryphon
10-11-2005, 03:20 PM
COULD NOT power-on split-esse below 10,000ft
Who said? A 747 can split-S in 3km of altitude.

Probably one of those massive safety margins designed to keeps noobs from lawn darting.


Power on test
Not as simple as a power-off test. The engine doesn't reach max RPM until the test is 1/3 over, and each engine spools up differently. Have to figure out some other way to do it.

Takata_
10-11-2005, 03:46 PM
Originally posted by FritzGryphon:
If you set a takeoff waypoint, then change it to a normfly waypoint, you can get 0 km/h.

I did this test, setting 0% fuel at 0 km/h. I used P-47D-27 and Fw-190A-8 and I set a vertical target to dive in at 90?. (propeller pitch set to manual and zero). To do it, just edit the mission file and set values to zero.

The most difficult was to trim the plane right as the speed change. I made a lot of dive, starting at 10 km, 9 km, 8 km, etc.

Without trim, both dived the same, no real difference, or outside my error margin.

Takata.

IL2-chuter
10-11-2005, 08:03 PM
In P-47 power-on (full power) dives the Vne of Mach 0.80 was reached (and easily exceeded) rather quickly and, although the aircraft held together quite well, going throttle-off and waiting to slow down enough to pull out could take quite a lot of altitude - or so says a few training films. (I imagine that if the tail fell of in the process dive brakes would have been fitted.) The aircraft was deemed unsuitable for dive bombing because of this . . . feature.

Your results ay vary. http://forums.ubi.com/images/smilies/16x16_smiley-indifferent.gif

JR_Greenhorn
10-11-2005, 10:25 PM
The results on the graph in the original post aren't that surprizing if one considers the relative wingloading values of each plane.

If the induced drag of the wings is the major factor in the test, and the effects of propellor and fuselage drag are assumed to have no significant difference among the planes tested, the results are almost as expected.

I calculated quick-&-dirty wingloading values for these planes from a plane data quick reference sheet I have. If somebody has better figures, by all means, post them.


<pre class="ip-ubbcode-code-pre">Plane Wingloading
===========================
P-47D 5136 psi
FW 190A-8 5080 psi
Bf 109G-6 4562 psi
F6F-5 3906 psi
A6M2 2208 psi
===========================</pre>

The real surprize here is that the P-47 has the highest wingloading, but it doesn't win the dive test. Still, the wingloadings are very close, and if one notes the significant difference in frontal area between the two planes, it becomes apparent that the FW 190 will start to win the free-fall dive.

Another factor is propellor drag. While I'm not prepared to speculate about contribution of propellor drag, it would stand to reason that the larger the propellor diameter, the larger the drag contribution of the propellor. I have the P-47 listed with a 12-foot-diameter prop. Does anyone know what the FW 190's is?



To further investigate propellor drag contribution as modelled in game, it would be interesting to see the results of the twin-engined P-38 and Bf 110 compared as well.

IL2-chuter
10-12-2005, 12:40 AM
"If the induced drag of the wings is the major factor in the test, and the effects of propellor and fuselage drag are assumed to have no significant difference among the planes tested, the results are almost as expected."


Actually, in a 90% dive I would think induced drag would not be a factor, as in shouldn't exist, and total form drag (Drag coefficient times front plate area) and prop drag would be the only factors. You're not generating any lift going straight down, are you? You should have the plane nosed over to a neutral lift AOA.

Here's a wrinkle. Try your time to climbs inverted. I've not been able to tell much difference between climb rates right-side-up or inverted. A bit of a difference in longitudinal AOA, attributable, in my mind, to wing incidence.

I'm ever so happy to contribute. http://forums.ubi.com/images/smilies/16x16_smiley-happy.gif


A side note. The more aft your CG, the less downforce (less induced drag) produced by the stabilizer. This means less wing lift (less induced drag) to carry the stabilizer downforce. I suppose we can't aft load any of these planes for performance tests, though. This is what makes the stalls in the game a little goofy, when the wing stalls the nose should drop as the center of lift and aerodynamic center are both behind the CG, more or less, not the plane go into a sort of squat-roll-thingy . . . sorry . . . tangent-thingy. My bad. http://forums.ubi.com/images/smilies/16x16_smiley-sad.gif


Carry on. http://forums.ubi.com/images/smilies/16x16_smiley-indifferent.gif

Aaron_GT
10-12-2005, 03:37 AM
Actually, in a 90% dive I would think induced drag would not be a factor, as in shouldn't exist,

Just because the plane is diving near vertically it doesn't mean that the wing isn't generating lift. It is. Induced drag is still there. Depending on COG it will either move the plane horizontally across the map when pointed down or tend to pull the plane out (hence the need to trim). When trimming the trim needed to fight the tendency for the lift to pull out will also add drag.

IL2-chuter
10-12-2005, 02:35 PM
The center of gravity is always ahead of the center of lift. This is stability. The CoL moves forward with an increase in AoA. At stall the CoL has moved to its maximum forward position. The aft CG limit is some distance forward from this at a point determined by flight testing. (On the Mustang they THEN added the 85gal fuselage tank which, when full, moved the CG to a point where, when the AoA was increased to about 2G accelleration, it matched the CoL and got stick reversal.)


Anyhow . . . you can nose the plane over enough to zero out lift, just like you can nose it over further to create negative lift (like when you fly upside down). So, your line of sight forward (direction of flight) when going straight down is at an angle above the normal line of sight for level flight but not as extreme as inverted flight. So, when you view the aircraft externally, the nose would be angled quite a bit below (pilot's orientation) the vertical.

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


Way too much info . . . I'm a ramblin' man http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif.

Aaron_GT
10-12-2005, 03:21 PM
The center of gravity is always ahead of the center of lift.

Wasn't one of the problems aflicting the P39 was that the CoL was pretty much at the CoG in a stall (or at least much closer than most aircraft of the period), hence the spin problems.

Grey_Mouser67
10-12-2005, 08:02 PM
Odd that you posted this..since the roll rate posts came out, it got me wondering about dive modelling so I did the exact same thing, but a differing methodology...I wanted to understand how planes dove without power...that is how drag and mass were modelled...the plane with the most mass in combination with the least drag, should fall the fastest.

Now all I needed was a method to control dive angle so this is what I did...QMB, Pacfic Islands map at 3000 meters...start with engine at 0, all settings auto/default/full fuel/CEM on and I picked an end of the moon shaped island and dove till i crashed and recorded the speed...here goes
D9 45 761
D9 44 755
A9 741
A8 737
A6 735
A4 706

P51D5 705
P51B 695
P47D27 685
P47D10 689
P38J 685
P38L 644
P38Llt 643

Spit mk ixc 637
Hellcat -3 646
Corsair 1A 673
La 7 x3 613
Yak 9D 650
P-40E 651
Spit VIII 638
Wildcat -4 632
A6M2 551
A6M5 562
Ki 43 Ic 521
Ki 84 1a 649

Bf 109G2 662
G6late 687
G6A/S 692
G10 690
K4 687

I repeated at a lower altitude and under full power and the results changed slightly and I'm not sure why.

I'm not even entirely sure what this is telling me, but one thing is for sure...the Fw has the best weight/E/drag combination...I would have expected the Mustang to be tops, followed by Jug, then Lightning, Fw, Spit and 109 close but distance from the above group...the Hellcat and Corsair aught to be right in there with the Jug but, the Hellcat is porked based on this test.

What I think this is telling me is really about E retention...I have to think about it in reverse to get there, but from a physics standpoint, your propelling forces are engine power(thrust would be a better description) and gravity and your resisting forces are drag which is created by the airframe, propeller and the lift created by the wings...I'd expect heavy aircraft, with low drag and high wingloading to be the best and lighter aircraft with high drag and low wingloading to be worst in this test....it sort of worked out that way but both the 109 and Fw were too high from a relative standpoint as the Fw had good power, high wingloading but was medium weight...Mustang had best drag, high weight, and great thrust...Jug/Hellcat/Lightning had all the weight you wanted but more drag and the Hellcat in particular had low wingloading.

Anyways, this shows me that there are some real issues in dive modelling but I have a nagging suspicion this really shows up in zoom climb and e retention based on ingame experiences. The weight and thrust should be additive in the dive since gravity is working for you which is opposite of climbing where thrust to weight ratio in combination with wingloading determined climb angle/angle of attack and climb rates.

I would love to hear what Oleg has to say...I have only a very rudimentary knowledge of the physics involved with flight, diving and climbing so I'm sure I'm missing something, but I don't think I'm wrong in that these results don't model what would have happened in real life...again, under power, things change quite a bit...all planes get closer together including the La7 which gets up to 772 vs a Dora at 771 and a Mustang at 774...this was the same testing but starting at 2000 meters diving at full power plus wep/boost. Both the Mustang and Dora were sporting similar thrust to the La, so why would it be so good with power on...the ratio should have been similar...thrust to weight should not matter in the dive because gravity makes mass/weight and additive property rather than a resisting property.

Alarming things..the Hellcat and why the heck does the P-38J have such a higher terminal speed than the L and L late...methinks it is the same reason the J climbs 1000 ft/min too slow in game!

IL2-chuter
10-12-2005, 08:29 PM
Originally posted by Aaron_GT:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">The center of gravity is always ahead of the center of lift.

Wasn't one of the problems aflicting the P39 was that the CoL was pretty much at the CoG in a stall (or at least much closer than most aircraft of the period), hence the spin problems. </div></BLOCKQUOTE>


Cg/CoL relationship in the P-39 was typical for the aircraft of the period, they had no reason to make it differently. What was drastically different was the extremely low polar moment. The engine wasn't away from the CG where it could mitigate large yaw and pitch swings at low airspeeds where flight controls might be inadequate. In the game you may have experienced different aircraft swinging and wobbling their noses back and forth rapidly during takeoff. Imagine this is because there is no weight away from the CG. Now, imagine hanging 1500lbs or more on the front of the plane and what effect this might have on the nose wobbling about. You might assume the frequency would be dramatically reduced. (Now, when your game plane is taking off, try to imagine what your weighty engine is going through while the aircraft is wobbling and how much energy must be expended to accellerate it back and forth.)

And, though the in game ground handling might give you other ideas, the P-39 (and P-38, etc.) didn't have steerable nosewheels, although this is entirely off even my topic.


PS. I'm off to HyperLobby to convert the heathen and shoot down the rest, cya there. http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif

Jetbuff
10-13-2005, 04:37 PM
One more time: If you're doing power-off dive testing in a CSP aircraft you must reduce prop-pitch to it's lowest setting. (0%)

If you do not do this, you are allowing the prop to windmill and slow you down. If you have any doubts about this take 15 minutes to set up a P-47 at 3000m in a 90 degree dive and test it at both 100% pitch and 0% pitch and compare the top speed at impact. If the difference is less than 50kph IAS, I'll eat my hat.

Grey_Mouser67
10-13-2005, 05:12 PM
I don't disagree, but if the effect is the same for all aircraft...shouldn't the results in the end be the same, relatively speaking?

I'm looking at the results from a relative standpoint as opposed to an absolute standpoint. During my examination, all aircraft props were left in auto...maybe there is a good explanation why a CSP would dive slower than a Luftwaffe one for example, but I can't think of one.

IL2-chuter
10-14-2005, 12:51 AM
CSP props won't adjust off the bottom stop below 1300 or so rpm (varies with specific engine/prop combos, some at 1700 rpm). This is why one must run the engine up to cycle the prop.

The variables 1)total prop area, 2)propeller pitch and 3)propeller rpm will vary the drag and vary between aircraft.

On the other hand . . . I have NO IDEA how the game models this.

Aaron_GT
10-14-2005, 01:31 AM
The engine wasn't away from the CG where it could mitigate large yaw and pitch swings at low airspeeds where flight controls might be inadequate.

ah, that makes more sense. Thanks.

Jetbuff
10-14-2005, 01:54 AM
Originally posted by Grey_Mouser67:
I don't disagree, but if the effect is the same for all aircraft...shouldn't the results in the end be the same, relatively speaking?

I'm looking at the results from a relative standpoint as opposed to an absolute standpoint. During my examination, all aircraft props were left in auto...maybe there is a good explanation why a CSP would dive slower than a Luftwaffe one for example, but I can't think of one.
You are making an incorrect assumption. LW planes do not use a CSP system when set to auto. (the 109's never do actually)

In the 190 on auto, you are using the fabled Kommandogerat system. (sp?) This system essentially links rpm to throttle setting. i.e. when you reduce your throttle it assumes you wish to also decrease rpm and coarsens the pitch. In other words, on auto, the 190's kommandogerat is setting pitch (or RPM actually) much lower than the 100% setting of a CSP - it's like having an assistant setting the prop-pitch for you for optimal performance.

For an apples to apples comparison you should test the 190 on manual at 0% pitch. I got a difference of 10kph IAS in terminal dive speed between the 190A-5 and the P-47D-10 from 3000m that way. However, if you wish to compare to real world tests, the 190 was never flown on manual pitch to my knowledge while the P-47's CSP is the only available system.

For the 109 it's even more different since it has a kommandogerat-like system on auto that defaults to a variable pitch prop on manual. I think the D-9's and 110's also share this feature.


Disclaimer: I am not making any assertions as to the accuracy of the FM. I am simply pointing out that a faulty test methodology may lead to incorrect conclusions. I am 100% positive that the P-47D10 outdives the 190A5 well before terminal speeds in a power-off dive from 3000m so long as you reduce prop-pitch setting to 0% in the P-47 to prevent your prop from slowing your free-fall. However, you do not have to take my word for it as it is easily reproducable.

effte
10-14-2005, 03:21 PM
Originally posted by IL2-chuter:
On the other hand . . . I have NO IDEA how the game models this.

As it should. I've tested it. Possibly with the magnitude of the effect excepted.