PDA

View Full Version : Fw190 acceleration and zoomclimb test - OUT NOW



PapaFly
09-02-2004, 01:23 PM
Aim of my testing was to compare acceleration and zoomclimb of the late-war Fw190 series.
Tested planes:

Fw190-A8-R2 (with the MK108 wing cannons)
Fw190-A9-R2
Fw190-D9 Late (default weapon configuration)

The reason I tested the R2 variants: almost all online players use this configuration.

Plane configuration, power settings:
50% fuel, 110% power + WEP/MW50, radiators closed, propellor on auto pitch.

ALL SPEEDS ARE TRUE AIR SPEED; ALL ALTITUDES ARE TRUE ALT ABOVE GROUND

Map: Crimea, at 15:00 (I didn't choose noon because its a bit hard to zoom straight into the sun)

T H E Z O O M T E S T:

METHOD: My method for testing the zoomclimb capability is completely focused on the dynamics during the vertical part of the zoom. The altitude gained during
the pull-up maneuver was not taken into consideration. This is mainly because its very hard to get accurate data when including the pullup: you
must start your pullup at the same speed + alt and pull up with the same angle of attack. Then there are factors like differences in energy bleed which might have a big
impact on the outcome of the experiment. None of these factors play a role when only looking at a plane zooming vertically. I look at the altitude gained starting
from 500kph and going down to 100kph.

Testing was performed from the quick mission builder,plane started at 2000m above the sea.
I dove the plane right to sea level, contrails on, did a gentle pull-up and zoomed vertically. I made sure the plane reached a stable vertical attitude before
speed dropped to 500kph (TAS). Plane was kept vertical until it sopped.
Then I flipped it over into a vertical dive, pulling out a few meters above sealevel. This maneuver was performed 4-6 times/plane.
Analysis took place with the recorded tracks: while the speed was dropping I recorded the altitudes at which it reached 500, 400, 300, 200 and 100kph.
This raw data was compiled in an exel sheet, where the altitude differences and the standard deviations of the measurements were calculated.
Here's the sheet:
http://www.black-crusade.de/hellhounds/papafly/ZoomtestFw190.xls

Tracks:
http://www.black-crusade.de/hellhounds/papafly/Zoomtest190A8.TRK
http://www.black-crusade.de/hellhounds/papafly/Zoomtest190A9.TRK
http://www.black-crusade.de/hellhounds/papafly/ZoomtestD9late_01.TRK
http://www.black-crusade.de/hellhounds/papafly/ZoomtestD9late_02.TRK

I must mention the fact that the D9 late started overheating after 2mins16secs, killing the engine after 1min - THIS MUST BE CHANGED!!!
So I had to record 2 tracks, because i can't do 4 consecutive zoomclimbs without killing the engine.

RESULTS:

Following jpeg depicts my exel sheet and includes all data:

http://www.black-crusade.de/hellhounds/papafly/Zoomclimbdata.jpg

A500, A400, A300, A200, A100 = Altitude at 500m, 400m and so on...

Each zoomclimb is in a row, e.g. the first zoomclimb of the 190A8:
500kph reached at 1054m
400kph reached at 1446m
200kph reached at 2034m
The altitude differences are in the orange lanes.
A500-200 = altitude difference between 500 and 200kph, in this case 980m.
The means are in the red rows, standard deviations in the green rows.

ZOOMCLIMB COMPARISON 500kph-300kph

http://www.black-crusade.de/hellhounds/papafly/Zoomclimb500-300.jpg
Error bars are included
A8 gains 724m
A9 gains 737m (13m more, about 1.8% increase)
D9 gains 752.3m (28m over the A8, 3.9% increase; 15m over the A9, 2.1% increase)

ZOOMCLIMB COMPARISON 500kph-200kph

http://www.black-crusade.de/hellhounds/papafly/Zoomclimb500-200.JPG
Error bars included
A8 gains 985.5m
A9 gains 1004.8m (19.3m more, about 2.0% increase)
D9 gains 1029m (43.5m over the A8, 4.4% increase; 24.2m over the A9, 2.4% increase)

OK, the object viewer gives following takeoff weight and power data:

190A8
4250kg/2050hp---Thrust/weight = 2.1kg/hp

190A9
4410kg/2300hp---Thrust/weight = 1.92kg/hp

190D9 late
4350kg/2240hp---Thrust/weight = 1.94kg/hp

I know the object viewer figures aren't accurate, I hope one of you can provide better data, but
I guess you get the point: how come that the plane with the best thrust/weight ratio is not even close to the D9 regarding
its zooming performance? And yes I heard that talk about the A9 using a high alt propellor.


THE ACCELERATION TEST:

METHOD: I created a mission with the plane standing at the start of the runway, crimea map.
I applied 110% power + WEP/MW50 while pressing the brake. All plane parameters equal to those in the zoomtest.
After max rpm and stable prop pitch were reached i released the brakes and kept the plane centered on the runway,
elev. trim neutral, joystick in neutral position. I accelerated up to the opposite end of the runway.
Then tracks were viewed, and i took screenshots from a static camera placed 1200m above the middle of the runway.
(if anyone wants the missions I can mail'em)
Screenshots were taken at 200 and 300kph. As a control I recorded the speed reached at the other end of the runway.
Every plane was tested 3 times, and the control speeds were exactly the same, so we basically have no standard deviation
for this experiment, its extremely reliable and easy to reproduce.
I loaded the screenshots into Adobe PS7, assigning different layers. Then I marked the exact locations were the planes reached
200/300kph.

Tracks:
http://www.black-crusade.de/hellhounds/papafly/Acceleration190A8.TRK
http://www.black-crusade.de/hellhounds/papafly/Acceleration190A9.TRK
http://www.black-crusade.de/hellhounds/papafly/Acceleration190D9late.TRK

http://www.black-crusade.de/hellhounds/papafly/190accelerationtest.jpg

As you can see the A8 accelerates faster than the A9, while the D9 late accelerates much better than both of them.
Actually the D9 late needs about 65% of the way the A8 and A9 need to reach 200kph.
The speeds reached at the end of the runway:

A8 - 329kph
A9 - 324kph
D9 - 352kph


...is it me or is there something wrong with the A9 performance?

AND A BIG BIG THANK YOU TO Jabo_HH_Gotcha for putting my stuff on his webspace :-)

PapaFly
09-02-2004, 01:23 PM
Aim of my testing was to compare acceleration and zoomclimb of the late-war Fw190 series.
Tested planes:

Fw190-A8-R2 (with the MK108 wing cannons)
Fw190-A9-R2
Fw190-D9 Late (default weapon configuration)

The reason I tested the R2 variants: almost all online players use this configuration.

Plane configuration, power settings:
50% fuel, 110% power + WEP/MW50, radiators closed, propellor on auto pitch.

ALL SPEEDS ARE TRUE AIR SPEED; ALL ALTITUDES ARE TRUE ALT ABOVE GROUND

Map: Crimea, at 15:00 (I didn't choose noon because its a bit hard to zoom straight into the sun)

T H E Z O O M T E S T:

METHOD: My method for testing the zoomclimb capability is completely focused on the dynamics during the vertical part of the zoom. The altitude gained during
the pull-up maneuver was not taken into consideration. This is mainly because its very hard to get accurate data when including the pullup: you
must start your pullup at the same speed + alt and pull up with the same angle of attack. Then there are factors like differences in energy bleed which might have a big
impact on the outcome of the experiment. None of these factors play a role when only looking at a plane zooming vertically. I look at the altitude gained starting
from 500kph and going down to 100kph.

Testing was performed from the quick mission builder,plane started at 2000m above the sea.
I dove the plane right to sea level, contrails on, did a gentle pull-up and zoomed vertically. I made sure the plane reached a stable vertical attitude before
speed dropped to 500kph (TAS). Plane was kept vertical until it sopped.
Then I flipped it over into a vertical dive, pulling out a few meters above sealevel. This maneuver was performed 4-6 times/plane.
Analysis took place with the recorded tracks: while the speed was dropping I recorded the altitudes at which it reached 500, 400, 300, 200 and 100kph.
This raw data was compiled in an exel sheet, where the altitude differences and the standard deviations of the measurements were calculated.
Here's the sheet:
http://www.black-crusade.de/hellhounds/papafly/ZoomtestFw190.xls

Tracks:
http://www.black-crusade.de/hellhounds/papafly/Zoomtest190A8.TRK
http://www.black-crusade.de/hellhounds/papafly/Zoomtest190A9.TRK
http://www.black-crusade.de/hellhounds/papafly/ZoomtestD9late_01.TRK
http://www.black-crusade.de/hellhounds/papafly/ZoomtestD9late_02.TRK

I must mention the fact that the D9 late started overheating after 2mins16secs, killing the engine after 1min - THIS MUST BE CHANGED!!!
So I had to record 2 tracks, because i can't do 4 consecutive zoomclimbs without killing the engine.

RESULTS:

Following jpeg depicts my exel sheet and includes all data:

http://www.black-crusade.de/hellhounds/papafly/Zoomclimbdata.jpg

A500, A400, A300, A200, A100 = Altitude at 500m, 400m and so on...

Each zoomclimb is in a row, e.g. the first zoomclimb of the 190A8:
500kph reached at 1054m
400kph reached at 1446m
200kph reached at 2034m
The altitude differences are in the orange lanes.
A500-200 = altitude difference between 500 and 200kph, in this case 980m.
The means are in the red rows, standard deviations in the green rows.

ZOOMCLIMB COMPARISON 500kph-300kph

http://www.black-crusade.de/hellhounds/papafly/Zoomclimb500-300.jpg
Error bars are included
A8 gains 724m
A9 gains 737m (13m more, about 1.8% increase)
D9 gains 752.3m (28m over the A8, 3.9% increase; 15m over the A9, 2.1% increase)

ZOOMCLIMB COMPARISON 500kph-200kph

http://www.black-crusade.de/hellhounds/papafly/Zoomclimb500-200.JPG
Error bars included
A8 gains 985.5m
A9 gains 1004.8m (19.3m more, about 2.0% increase)
D9 gains 1029m (43.5m over the A8, 4.4% increase; 24.2m over the A9, 2.4% increase)

OK, the object viewer gives following takeoff weight and power data:

190A8
4250kg/2050hp---Thrust/weight = 2.1kg/hp

190A9
4410kg/2300hp---Thrust/weight = 1.92kg/hp

190D9 late
4350kg/2240hp---Thrust/weight = 1.94kg/hp

I know the object viewer figures aren't accurate, I hope one of you can provide better data, but
I guess you get the point: how come that the plane with the best thrust/weight ratio is not even close to the D9 regarding
its zooming performance? And yes I heard that talk about the A9 using a high alt propellor.


THE ACCELERATION TEST:

METHOD: I created a mission with the plane standing at the start of the runway, crimea map.
I applied 110% power + WEP/MW50 while pressing the brake. All plane parameters equal to those in the zoomtest.
After max rpm and stable prop pitch were reached i released the brakes and kept the plane centered on the runway,
elev. trim neutral, joystick in neutral position. I accelerated up to the opposite end of the runway.
Then tracks were viewed, and i took screenshots from a static camera placed 1200m above the middle of the runway.
(if anyone wants the missions I can mail'em)
Screenshots were taken at 200 and 300kph. As a control I recorded the speed reached at the other end of the runway.
Every plane was tested 3 times, and the control speeds were exactly the same, so we basically have no standard deviation
for this experiment, its extremely reliable and easy to reproduce.
I loaded the screenshots into Adobe PS7, assigning different layers. Then I marked the exact locations were the planes reached
200/300kph.

Tracks:
http://www.black-crusade.de/hellhounds/papafly/Acceleration190A8.TRK
http://www.black-crusade.de/hellhounds/papafly/Acceleration190A9.TRK
http://www.black-crusade.de/hellhounds/papafly/Acceleration190D9late.TRK

http://www.black-crusade.de/hellhounds/papafly/190accelerationtest.jpg

As you can see the A8 accelerates faster than the A9, while the D9 late accelerates much better than both of them.
Actually the D9 late needs about 65% of the way the A8 and A9 need to reach 200kph.
The speeds reached at the end of the runway:

A8 - 329kph
A9 - 324kph
D9 - 352kph


...is it me or is there something wrong with the A9 performance?

AND A BIG BIG THANK YOU TO Jabo_HH_Gotcha for putting my stuff on his webspace :-)

faustnik
09-02-2004, 01:52 PM
I have been told the answers to the 190A9 performance questions are found in the design of the high altitude prop on the FB model.

I'm sure the standard "A8" prop was used too.

http://pages.sbcglobal.net/mdegnan/_images/FaustSig
www.7Jg77.com (http://www.7jg77.com)
CWoS FB forum. More Cheese, Less Whine. (http://www.acompletewasteofspace.com/forum/viewforum.php?f=31)

XyZspineZyX
09-02-2004, 06:49 PM
Is a test from a standing start a viable acceleration test?

For one, IL-2's takeoff modeling is horrid; almost no torque effect, a less-than-believable 0 to rotation speed sequence, etc.

Wouldn't a full power straight and level extension, following a speed draining turn to a set speed be a better way to do this? I dunno. Just asking.

LEXX_Luthor
09-02-2004, 08:32 PM
Interesting stuff here.

High speed drag may play into level flight acceleration tests that can't be explored
on the runway.

__________________
http://ubbxforums.ubi.com/images/smiley/10.gif Flyable Swedish "Gladiator" listed as J8A ...in Aces Expansion Pack ( AEP )

"You will still have FB , you will lose nothing" ~WUAF_Badsight
"I had actually pre ordered CFS3 and I couldnt wait..." ~Bearcat99
"Gladiator and Falco, elegant weapons of a more civilized age" ~ElAurens
:
"Damn.....Where you did read about Spitfire made from a wood?
Close this book forever and don't open anymore!" ~Oleg_Maddox http://ubbxforums.ubi.com/images/smiley/16x16_smiley-surprised.gif

p1ngu666
09-02-2004, 10:00 PM
hm but in air stuff is very hard, to cruise at say exactly 300.
airracing uses a standing start for this reason, be easy to jump a rolling start.

not that rolling starts or restarts are bad for racing, if u have montoya there anyways http://ubbxforums.ubi.com/images/smiley/59.gif http://ubbxforums.ubi.com/images/smiley/16x16_smiley-very-happy.gif

http://www.pingu666.modded.me.uk/mysig3.jpg
<123_GWood_JG123> NO SPAM!

LEXX_Luthor
09-02-2004, 10:26 PM
testing not racing.

No such thing as "exactly" any speed.

I have no problem maintaining airpeed within +- 10km/hr.

Takes time to get stable, but it can be done.

What you can do is start test recording when you slowly pass the starting speed.

__________________
http://ubbxforums.ubi.com/images/smiley/10.gif Flyable Swedish "Gladiator" listed as J8A ...in Aces Expansion Pack ( AEP )

"You will still have FB , you will lose nothing" ~WUAF_Badsight
"I had actually pre ordered CFS3 and I couldnt wait..." ~Bearcat99
"Gladiator and Falco, elegant weapons of a more civilized age" ~ElAurens
:
"Damn.....Where you did read about Spitfire made from a wood?
Close this book forever and don't open anymore!" ~Oleg_Maddox http://ubbxforums.ubi.com/images/smiley/16x16_smiley-surprised.gif

JaBo_HH--Gotcha
09-03-2004, 12:20 AM
Any time m8 ! Any time !

In regard of the A9 the difference was a more powerful engine and the same prop as of the d9 (wider blades) this actually made the a9 climbing as good as the d9 up to 3000 if I recall right.
From there on the superiority of the Jumo (d9) begun to shine over the BMW (better high alt performance).

Basically both planes should perform quite similar down low. Now I don't expect them to fix anything but just making the d9 slower... http://ubbxforums.ubi.com/infopop/emoticons/icon_biggrin.gif

http://www.g-c-p.de/sigbib/hh/gotcha.jpg

Kwiatos
09-03-2004, 10:35 AM
Maby test A-6/D-9 versus Spitfire MK V and IX?

PapaFly
09-04-2004, 05:26 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Stiglr:
Is a test from a standing start a viable acceleration test?

For one, IL-2's takeoff modeling is horrid; almost no torque effect, a less-than-believable 0 to rotation speed sequence, etc.

Wouldn't a full power straight and level extension, following a speed draining turn to a set speed be a better way to do this? I dunno. Just asking.<HR></BLOCKQUOTE>

A standing start is absolutely viable, every test were you have exactly the same conditions for every plane IS VIABLE. For instance, when flying the Fw in combat, you very often zoom up vertically until you reach some 0-100kph, and then you dive back on your opponents: so what you basically do is accelerate your crate from around 0kph on. And as you can clearly see the A9 can hardly keep up with the A8 in the 0-300kph range.
Now let's follow this thought to further consequences and compare an A8 and an A9 doing repeated zoomclimbs followed by dives: the A9 has a very tiny margin during the zoomclimb, followed by an inferior initial acceleration during the dive. And initial acceleration is exactly what you need in this situation, you don't have 3000m of air between you and the bandit down low, so you need to get as much speed as possible while accelerating some 1000-2000m.
Now you do these consecutive attacks a few times and you see that the A9 isn't much better compared to the A8: it should be better during zoom, and it should absolutely shine during the dive: don't forget that it has better thrust/weight AND more weight. Energy bleed (important while pulling out of the dive into the next zoom)is yet left out of consideration, and I can only hope that the A9 energy bleed isn't any worse than that of the A8.

To the takeoff modelling, yaw effects ans similar ****:
1. yaw effects are ALWAYS weak, they're not weaker during start, and - THINK OF IT, WATCH THE TRACKS - i kept the planes well centered during start, so weak yaw modelling has absolutely no effect on my tests.

2. No matter how ground handling/start is modeled, a plane with better thrust/weight ratio(A9) should still accelerate faster than one with a lower ration. Don't you see these planes are very very near to each other regarding weight, size, drag a.s.o. The only major difference is the engine - thrust to weight. How can you explane the acceleration difference between D9 and A9? Wouldn't you expect the A9 to accelerate like a D9 or better? And no matter what they're tellin us about the prop: the A9 MUST accelerate better than the A8 period.

Some of you suggest flying the planes at the same speed and accelerating then....well this could give you some porked data with no real relevance for the online dogfight: IS ANYBODY OUT THERE SO STUPID TO GET ATTACKED WHILE CRUISING WITH 250KPH, AND THEN TRYING TO EXTEND???
Of course not: in 99% of the cases you already have applied full throttle i.e. your rpm is already at max when a bandit tries to catch you. When you fly a certain speed and then apply full throttle you may get huge acceleration differences which come from different times needed to wind up the engine.
No no no, check my test: at the moment when I release the brakes the engines are already stably running at max rpm. No delays here, just power speaking.

I think a better approach would be to sit on the ramp(directed towards the sea), hit the brake, firewall the engine, wep on, wait for stable rpm then release brake and accelerate up to the end of the runway, then do a very gentle takeoff to 5m above ground, keep this altitude and accelerate up to max speed.
Then you can check how much time you need from 0-100/100-200/200-300/300-400/400-500kph IN MAXIMUM ACCELERATION REGIME. This would be more relevant for the online dogfight.

Also, we need someone to test the maximum sustained turn capabilities of the A9, A8, D9.
A9 and D9 should be better sustained turners compared to A8, cause they should be able to keep up higher speeds (better thrust/weight), and higher speeds in sustained turns give you more lift resulting in increased degrees/second.

S!

LEXX_Luthor
09-04-2004, 07:14 AM
Good Point.

robban's old climb test started from stop at airfield.

Its just that for acceleration tests you must go beyond ground taxi test to reach high speeds
where high speed drag is more important than thrust/mass ratio.

__________________
http://ubbxforums.ubi.com/images/smiley/10.gif Flyable Swedish "Gladiator" listed as J8A ...in Aces Expansion Pack ( AEP )

"You will still have FB , you will lose nothing" ~WUAF_Badsight
"I had actually pre ordered CFS3 and I couldnt wait..." ~Bearcat99
"Gladiator and Falco, elegant weapons of a more civilized age" ~ElAurens
:
"Damn.....Where you did read about Spitfire made from a wood?
Close this book forever and don't open anymore!" ~Oleg_Maddox http://ubbxforums.ubi.com/images/smiley/16x16_smiley-surprised.gif

WWMaxGunz
09-04-2004, 08:02 AM
I've got to agree with Papa there on viability and relevance issues.
Test is easy to make compared to coming up with correct relevances.

I still wonder how many players see 300 and at gut level feeling that is 300mph fast?


Neal

clint-ruin
09-04-2004, 03:37 PM
I think that non-grounded tests might be a better way to put this to Oleg if you're looking for changes to be made.

I'm thinking of JTDs tests where Oleg said that the reason for the apparent dispersion of .50 cals was due to recoil and landing gear modelling.

Quite how 2x .50 cals being fired from a B-239s cowl would produce the same result as a P-39 or P-47 with vastly different gun placement / gear placement and total weight was never actually explained by Oleg. But if you're going to ask for something to get changed it helps not to leave the guy an easy out to dismiss it with. Who knows, you might even find that there really is something going on with the gear modelling as far as acceleration is concerned.

http://users.bigpond.net.au/gwen/fb/leninkoba.jpg

PapaFly
09-05-2004, 09:03 AM
Still, no matter how bad the groundhandling is modeled, these planes have identical gears, the only relevant difference being the thrust/weight ratio. As it is now, I'm sure the A9 has an absolutely pathetic acceleration rate, and I expect this to be changed.

What do you guys think of the zoombclimb performance? Don't you think that the A9 should perform more like the D9 late? And if they modeled the wrong prop, then they should change it, cuz as it is now, the A9 is **** AND NOTHING ELSE. If you have ever tried taking off in an A9 while your field is under attack then you know what I'm talking about...

Sidenote to other guys making tests: you guys must fly the same test at least 3 times, calculate means and standard deviation. you mus t show that your test is reproducible. I work as a biochemist in cancer research, and no researcher will EVER trust data coming from a single test. You try presenting such data to other scientists during a seminar and you get virtually ripped to pieces, lol. Still I love to see more and more people starting testing planes. I know that repeating the same experiment several times is more timeconsuming, but this is - imo - the only way to go.

WWMaxGunz
09-05-2004, 04:44 PM
There was one guy wrote a simple straight & level autopilot with DevideLink but
last I read he was having problems, it is not so simple. With something like
that worked out, repeatability could be more assured tighter toleranced.


Neal

PapaFly
09-20-2004, 04:27 PM
Bump! Developers should see this!

PapaFly
09-20-2004, 04:28 PM
Ummmm, uhhhh huh huh huh, BUMP!

JaBo_HH--Gotcha
09-21-2004, 02:16 AM
BUUUUuuuuuuUUUUUMP !

http://www.g-c-p.de/sigbib/hh/gotcha.jpg

carguy_
09-21-2004, 09:14 AM
Why all the bumps?

Just send it all to Oleg via email so it can actually be noticed by 1C.

http://carguy.w.interia.pl/tracki/sig23d.jpg

Self-proclaimed dedicated Willywhiner since July 2002

moksha
09-21-2004, 09:48 AM
good point

*bump*

PapaFly
09-24-2004, 01:20 PM
bump

PapaFly
10-02-2004, 07:27 AM
bump

lbhskier37
10-02-2004, 08:36 AM
don't bother no in here wants to hear this, I've tried many times before. Only solution is to just fly and get better than the Spitty pilots and shoot them down anyways.