PDA

View Full Version : BUG: BI-1 compressibility is not correctly related to Mach number



04-12-2004, 07:25 PM
Note for Ivan: this has already been emailed to Oleg. Thanks to KrasniyYastreb for pointing out this bug!

1. Version AEP 2.00 (Ubi version)

2. Computer: irrelevant. Flight model bug.

3. Bug/Defect Description: BI-1 compressibility is not correctly related to Mach number. Critical Mach incorrectly increases with altitude.

4. Describe steps to reproduce the bug (detailed): see test procedure below....

Set up test in FMB on Crimea map with 100% fuel load and perfect weather, at 1200 noon. I used the Crimea map because in this community Crimea has become a sort of the standard map for flight testing. Note however that the Crimea wx is not equivalent to a '76 Standard Atmosphere, so some conversions were necessary to calculate Mach.

Test procedure: Place BI-1 at test altitude using FMB. Accelerate BI-1 gradually in level flight until aircraft begins to shake and note true airspeed. Speed up and slow down across the shake threshold several times to confirm speed.

Note: full "tuck under" does not occur until faster than initial shake speed.

* True airspeed readings were taken from digital TAS speedometer in wonder-woman view.

* Temperature readings taken from He-111 thermometer at same altitude, same time of day, and same wx conditions on Crimea map.


observed data:
altitude, m / observed temperature, C / shake airspeed, kmh (true)

100 / 25 / 775
1000 / 19 / 811
2000 / 12 / 852
3000 / 6 / 895
4000 / -2 / 948
5000 / -9 / 995
6000*

* BI-1 cannot reach "shake speed" in level flight (max. speed at 6000m is approx. 1015 kmh)

Mach can be calculated with Digital Dutch Standard Atmosphere calculator,
http://www.digitaldutch.com/atmoscalc/

Calculated speed of sound:
altitude, m / observed temperature, C / Speed of Sound, kmh

100 / 25 / 1246
1000 / 19 / 1234
2000 / 12 / 1219
3000 / 6 / 1206
4000 / -2 / 1188
5000 / -9 / 1173

************************************************** **
Final Mach calculations:
altitude, m / Mach number at which BI-1 begins to shake

100 / 0.62
1000 / 0.66
2000 / 0.70
3000 / 0.74
4000 / 0.80
5000 / 0.84
************************************************** **

04-12-2004, 07:25 PM
Note for Ivan: this has already been emailed to Oleg. Thanks to KrasniyYastreb for pointing out this bug!

1. Version AEP 2.00 (Ubi version)

2. Computer: irrelevant. Flight model bug.

3. Bug/Defect Description: BI-1 compressibility is not correctly related to Mach number. Critical Mach incorrectly increases with altitude.

4. Describe steps to reproduce the bug (detailed): see test procedure below....

Set up test in FMB on Crimea map with 100% fuel load and perfect weather, at 1200 noon. I used the Crimea map because in this community Crimea has become a sort of the standard map for flight testing. Note however that the Crimea wx is not equivalent to a '76 Standard Atmosphere, so some conversions were necessary to calculate Mach.

Test procedure: Place BI-1 at test altitude using FMB. Accelerate BI-1 gradually in level flight until aircraft begins to shake and note true airspeed. Speed up and slow down across the shake threshold several times to confirm speed.

Note: full "tuck under" does not occur until faster than initial shake speed.

* True airspeed readings were taken from digital TAS speedometer in wonder-woman view.

* Temperature readings taken from He-111 thermometer at same altitude, same time of day, and same wx conditions on Crimea map.


observed data:
altitude, m / observed temperature, C / shake airspeed, kmh (true)

100 / 25 / 775
1000 / 19 / 811
2000 / 12 / 852
3000 / 6 / 895
4000 / -2 / 948
5000 / -9 / 995
6000*

* BI-1 cannot reach "shake speed" in level flight (max. speed at 6000m is approx. 1015 kmh)

Mach can be calculated with Digital Dutch Standard Atmosphere calculator,
http://www.digitaldutch.com/atmoscalc/

Calculated speed of sound:
altitude, m / observed temperature, C / Speed of Sound, kmh

100 / 25 / 1246
1000 / 19 / 1234
2000 / 12 / 1219
3000 / 6 / 1206
4000 / -2 / 1188
5000 / -9 / 1173

************************************************** **
Final Mach calculations:
altitude, m / Mach number at which BI-1 begins to shake

100 / 0.62
1000 / 0.66
2000 / 0.70
3000 / 0.74
4000 / 0.80
5000 / 0.84
************************************************** **

04-13-2004, 04:35 AM
BI-1 shake threshold may be related to dynamic pressure, not Mach. Is it a bug? Yes, but this might get written off as "limitations of the sim". Keep your fingers crossed!

previous test data:

altitude, m / temperature, C / airspeed, (m/s) / density, (kg/m^3)

100 / 25 / 215.28 / 1.1699 / 100129.4386
1000 / 19 / 225.28 / 1.0717 / 89874.5705
2000 / 12 / 236.67 / 0.9712 / 79495.2155
3000 / 6 / 248.61 / 0.8749 / 70108.5447
4000 / -2 / 263.33 / 0.7919 / 61640.2353
5000 / -9 / 276.39 / 0.7124 / 54019.9121

Recall that dynamic pressure is defined as
Q = 0.5 * rho * V^2.

Calculated dynamic pressure:
altitude, m / Q, Pa
100 / 27110
1000 / 27190
2000 / 27200
3000 / 27040
4000 / 27460
5000 / 27210

In the above table the dynamic pressure at which shake begins is nearly constant at all altitudes.

Loco-S
04-13-2004, 10:55 PM
http://mywebpage.netscape.com/kurbalaganda/PIF1944-compressibility.jpg

http://mywebpage.netscape.com/kurbalaganda/Loco-S.gif
Armis Bela, non venenis geri

04-14-2004, 02:35 AM
Wow, thanks for posting that scan Loco-S.

The problem with the flight model here (in case people are confused or simply don't get it) is that critical Mach should not vary with air density.

The way things are now, aircraft like the P-38 and BI-1 have a very cheap "scripted" feeling when they encouter compressibility. You can just tell it's an ugly kludge.

To fix the problem:

1) eliminate the scripted kludge for BI-1 and P-38. instead, apply proper physics to all aircraft.

2) move CP aft in all aircraft as airspeed approaches M_crit.

3) smoothly blend in trim changes (pitching moment) as airspeed approaches M_crit.

4) link buffeting to Mach, not IAS.

5) keep the current relation between control effectiveness and IAS, but also add in a secondary effect proportional to Mach

6) consider modelling high speed stalls caused by normal shocks. When CL increases in accelerated maneuvering, transonic flow often occurs on the upper wing surface. The normal shock causes flow to separate and can lead to stalling below the low-speed CL_max limit.

KrasniyYastreb
04-17-2004, 12:59 AM
I suspect you are right about compressibility effects being linked to a value dynamic pressure, since this is, after all, how the aerodynamis forces are calculated.

A few questions about your points though:

2) CP? What does it stand for? center of pressure?
3) trim changes already as your airspeed increases. The compressibility effects should appear suddenly as the flow goes supersonic.

5) why?

6) i would think that the shocks tend to be oblique/attached since there are no large obstructions on the upper wing surface.

I would be very interested to see Oleg's reply to this, and if it can be fixed without major changes to the game's engine.

Ugly_Kid
04-17-2004, 01:17 AM
I can't say about Bi-1 in particular but the buffeting in the game IMO is related to structural velocity (a sort of beginning flutter). In most of the aircraft pushing a little bit faster will rip of the wings. In this sense it's a gaming effect not related to Mach at all. (Try diving with slower aircraft, stuka, I-153, shaking but quite in a low Mach).

The Mach effects, drag divergence position of aerodynamic center are steady static values and do not cause buffeting. Buffeting is normally related to the shock waves and flow separation caused by them, I don't think the game simulates this one, I think that if Mach effects are present they effect the previously mentioned static values. This is what I understood when I asked Oleg about this drag and aerodynamic center are affected.

04-17-2004, 04:32 AM
Flutter related to IAS is perfectly OK.

The problem here is that we have several aircraft with "scripted" transonic effects in their flight models: the BI-1 and P-38. In these aircraft the compressibility effects are incorrectly linked to IAS, rather than Mach.

If you're going to model "Mach tuck", then at least do it right. We've currently got two aircraft with very poorly modelled transonic effects, and 60+ remaining aircraft with no transonic effects at all. It really makes the P-38 and BI-1 stick out in a bad way.

All the developers need to do is link the "Mach tuck" and control stiffness functions to Mach instead of IAS.

There was a great deal of discussion about the P-38's problems, so hopefully this next patch will fix it (two weeks?).

BTW, I haven't heard from Oleg on this bug yet. Going to email it a second time if there's no response by Monday....

04-17-2004, 08:09 PM
I just heard Oleg is a proud new father. Family comes before aeronautics, so as far as I'm concerned the Mach/IAS bug can wait until later.

Congratulations Maddox family! http://ubbxforums.ubi.com/infopop/emoticons/icon_smile.gif

BBB_Hyperion
04-18-2004, 04:06 AM
Are you sure about Temperature Values cause when flight tests should be done on Crimea Map i would exspect the values are close to ISA Values.

p0 = 1013,25 hPa

Temp T0 = 288,5 Degrees Kelvin = 15,35 Degrees Celsius

Temperature loss with alt
0.65 k / 100 m

Temp vs alt 0 m = 288,5 K = 15,35 C
100 m = 287,85 = 14,7 C
1000 m = 282 = 8,85 C
2000 m = 275,5 = 2,35 C
3000 m = 269 =- 4,15 C
4000 m = 262,5 =-10,65 C
5000 m = 256 =-17,15 C
6000 m = 249,5 =-23,65 C

Speed of sound at 0 Degrees Celsius
c0 = sqrt(101325/1,2935*1,402)
= 331,4 m/s
Speed of sound at t degrees Celsius
ct = 331,4 m/s * sqrt(1+alpha* theta)
alpha = 1/273,15
theta temp in celsius.

Speed of Sound at ALt
m | m/s km/h
0 340 1224
1000 336,73 1212,22
2000 332,82 1198,15
3000 328,87 1183,93
4000 324,88 1169,57
5000 320,83 1154,99
6000 316,73 1140

I agree there is something off special regarding to high speed drag . Not sure maybe its the IAS TAS conversion or the FM itself over 5k.

Regards,
Hyperion

04-19-2004, 02:38 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by BBB_Hyperion:
Are you sure about Temperature Values cause when flight tests should be done on Crimea Map i would exspect the values are close to ISA Values.<HR></BLOCKQUOTE>

Check for yourself. If you have the sim, you can observe temperatures on the Crimea map.

Crimea is warmer than standard atmosphere.

BBB_Hyperion
04-19-2004, 05:03 AM
What sense would it make to compare real flighttest (normalised) to other than standard values none for me at least so why we should do tests on crimea then ?. Well anyway i agree to your argumentation when the temperature gauge is right.

That there is something off with highpspeed ias / tas is obvious even from my calculation. 1104 km/h TAS speed 840 ias shown with 109 F4 from dive about 4800 m using your table its 0.93 Mach or mine 0.95 Mach . The maximum speed optained by a 109 F4(G wing , tailrudder parts) in dive tests was 0.805 at 7 km !

Regards,
Hyperion

Platypus_1.JaVA
04-19-2004, 09:46 AM
If you haven't noticed... FB is not a sim for supersonic fighters.

1 Judge not, that ye be not judged.
2 For with what judgment ye judge,
ye shall be judged: and with what
measure ye mete, it shall be measured
to you again.

http://server5.uploadit.org/files/JaVAPlatypus-1java.JPG (http://www.1java.org)

BBB_Hyperion
04-19-2004, 10:45 AM
If you havent noticed http://ubbxforums.ubi.com/infopop/emoticons/icon_wink.gif we are not talking about supersonic fighters.

Regards,
Hyperion

Tooz_69GIAP
04-20-2004, 04:37 AM
Man oh man, it's threads like these that make me wish I hadn't dropped outta Aero Engineering at university!!

Oh well, maybe I'll make it back one day http://ubbxforums.ubi.com/images/smiley/53.gif

Tooz

whit ye looking at, ya big jessie?!?!

http://www.baseclass.modulweb.dk/69giap/fileadmin/Image_Archive/badges/69giap_badge_tooz.jpg
Za Rodinu!