Jump to content


Photo
- - - - -

Flight Model Verification and Data Analysis 101


  • Please log in to reply
445 replies to this topic

#81 Vati

Vati
  • Posts: 820

Posted 14 May 2010 - 17:03

Memorial flight planes (in france) for example the SPAD XIII is far less tolerant to spins that replicas that fly out there.
Makes one wondering why ROF is so spin proof, as we know they had access to Memorial flight :)
  • 0

#82 Ogami_musashi

Ogami_musashi
  • Posts: 859

Posted 14 May 2010 - 17:14

Because the memorial flight SPAD recovers less easily; The departure is another topic and even if the recover of ROF spad is too gentle the reason is simple (and i can tell you that because i've seen the discussion between noeqb and memorial flight): You can't ask people to try out spins in planes restored with countless volunteers work hours and brave pilots that fly planes that not only had viscious aero characteristics but in addition could break up under area loads.

Instead, aerobatics (pitts et al) stall behavior was studied in addition to engineering literature.


As i said countless times, spin behavior in sims is too complex to be rendered 100% realistically.
If you don't have load of aero coefficients from test flights simulating it as real as it gets is impossible.

There's nothing shocking in spins in ROF to me, the recovery procedures are not fantasist as some say, see the camel, the camel spin that requires aft spin is a flat spin, and that type of spin in real life requires the same.

The question is does the camel departed in flat spin? maybe yes, maybe not, here we hit the limit of FMs.

But as i said also, FM WILL be improved in the following months.
  • 0

#83 Eldur

Eldur
  • Posts: 220

Posted 14 May 2010 - 17:19

That's interesting Eldur, I wonder if they had more instruments on board to check the real planes' performance? I work in instrumentation (electronics) so this is interesting to me - sorry if my question sounds sarcastic mate in the er, rather delicate climate, I don't mean to be awkward at all. But I have been thinking lately about how they tested planes back then, without the meters and indicators you mention

It doesn't matter too much really because we have climb rate as height/time, hmm no inclinometer - that's an important instrument because they used to paint a white line on an airfield building and instruct rookies to line the wings up to the line for correct lift angle at takeoff (WW2)

Ming

It would be interesting to know how things were measured back then. Data like "x min y sec to z m" shows that obviously they did not have gauges that show climb/descent rate. But how could you ensure you're holding your altitude while testing the speed at 4000m, with a gauge where 1000m is something like 5cm? You can't tell if you're steady at 4000 or going up and down +- 20m with these instruments. Real pilots can at least feel changes in attitude as well as G forces. We can't. I can't exactly hold my altitude without checking my flight path with F11 repeatedly. At sea level or close to, yes, but not at 500m ore even more. And this makes accurate speed measurements impossible. I'd just aks for some gauges to do so because the simple gauges also are not found in the planes. Just something like a "Test mode" that shows flight parameters we have no instruments for. At least they have to be able to re-check their own programming which makes me suspect that there's at least soem kind of debug mode which shows these data.
  • 0

#84 Hellbender

Hellbender
  • Posts: 3321
  • LocationMadrid, Spain (originally Brussels, Belgium)

Posted 14 May 2010 - 17:22

Col. Hartney says the N.28 could out-turn a Dr.1. Well that means nothing to me. But if Rickenbacker also says that the N.28 can out-turn a Dr.1 then I'll begin to believe it. And then if a German ace from one of the circuses further states that he was out-turned by a N.28 while flying his Dr.1 then I will take it as fact.


You've obviously never heard of a German Dr.I flyer speaking of being outturned by a Nieuport 28.

…exactly because they were outturned!

It's the only reasonable explanation.

:lol:
  • 0

J5_Hellbender


#85 ImPeRaToR

ImPeRaToR
  • Posts: 7902

Posted 14 May 2010 - 17:23

you can engage auto level :)
  • 0

#86 Hellbender

Hellbender
  • Posts: 3321
  • LocationMadrid, Spain (originally Brussels, Belgium)

Posted 14 May 2010 - 17:37

you can engage auto level :)

Away with you and your logical proposal towards the betterment of empirical research!

Only witch burners are allowed in this thread.


What we really need is… a duck!

http://www.youtube.c...h?v=zrzMhU_4m-g">
  • 0

J5_Hellbender


#87 BroadSide

BroadSide
  • Posts: 2057

Posted 14 May 2010 - 17:56

I wonder if Emperator will float….
  • 0

#88 WW1EAF_Ming

WW1EAF_Ming
  • Posts: 2565

Posted 14 May 2010 - 18:04

It would be interesting to know how things were measured back then

Yes it would Eldur, and out of interest I'll be examining a WW1 test-flight reports archive next week with a view to finding out stuff like that. There's a ton of documents to look at apparently but I'll see if I can dig out anything interesting (and relevant). No promises you understand because things can look good on paper but when you turn up it's only 3000 invoices for horse-feed. Oats :)

Ming
  • 0

#89 WW1EAF_Ming

WW1EAF_Ming
  • Posts: 2565

Posted 14 May 2010 - 18:15

Hopefully Neoqb will learn from Oleg's experience and NEVER make a change based on the wishes of the mob

That's not going to happen Eric, we all of us learned from that experience and neoqb's customers will be able to see for themselves how accurately RoF simulates the flight and characteristics of WW1 aircraft

It is possible sometimes for flight-sim players to throw out the baby rather than the bathwater in moments of madness but neoqb insist that their customers are brought onboard for the ride and all will be well :)

Ming
  • 0

#90 ImPeRaToR

ImPeRaToR
  • Posts: 7902

Posted 14 May 2010 - 18:24

I wonder if Emperator will float….

Who is that then? :roll:
  • 0

#91 WW1EAF_Ming

WW1EAF_Ming
  • Posts: 2565

Posted 14 May 2010 - 18:54

Is it The Evil Twin :)

Ming
  • 0

#92 Marco_._

Marco_._
  • Posts: 2594

Posted 14 May 2010 - 19:01

ming you are purple now….but thankfully, you "didnt" change your attitude :lol:
  • 0

#93 WW1EAF_Ming

WW1EAF_Ming
  • Posts: 2565

Posted 14 May 2010 - 19:09

Hi Tvrdi yes it's me here mate, I know you to be a very decent chap. You can be headstrong but your heart is definitely in the right place :)

I've set you a challenge nearby so you can become a virtual aeronautical engineer. And there are plenty more planes when that one is tested :)

Ming
  • 0

#94 Marco_._

Marco_._
  • Posts: 2594

Posted 14 May 2010 - 19:16

Hi Tvrdi yes it's me here mate, I know you to be a very decent chap. You can be headstrong but your heart is definitely in the right place :)

I've set you a challenge nearby so you can become a virtual aeronautical engineer. And there are plenty more planes when that one is tested :)

Ming

Your right, Im headstrong and later I bear the guilt. After all Tvrdi means Hardy on croatian :lol:

sry Im offtopic….roger out
  • 0

#95 WW1EAF_Ming

WW1EAF_Ming
  • Posts: 2565

Posted 14 May 2010 - 19:36

Image

Don't shoot the messenger :)

Ming
  • 0

#96 WW1EAF_Ming

WW1EAF_Ming
  • Posts: 2565

Posted 14 May 2010 - 19:38

Off-topic is the old regime mate, everything but rudeness is compulsory :)

And anyway, all good threads end in food

Ming
  • 0

#97 J2_squid

J2_squid
  • Posts: 3815

Posted 14 May 2010 - 19:43

Attached File  laurel-stan-oliver-hardy.jpg   40.45KB   68 downloads

Sorry, couldnt resist :lol:
  • 0

#98 mike_espo

mike_espo
  • Posts: 253

Posted 14 May 2010 - 21:21

Wow! :lol: Brings back memories of IL2 where the FMs are under dispute!
  • 0

#99 IRFC_SmokinHole

IRFC_SmokinHole
  • Posts: 1861

Posted 15 May 2010 - 02:08

Replicas usually show significant differences with the actual planes.

Memorial flight planes (in france) for example the SPAD XIII is far less tolerant to spins that replicas that fly out there.

The reason lies in quality of materials, fabrication and joints.

No question. And since some planes, famously Fokker, were planned from a disassembled hand-built original, each unit produced at different locations was different. "Replica" is the appropriate term for planes from this period because there are no surviving flying models with a majority of original major components still intact.
  • 0

#100 IRFC_SmokinHole

IRFC_SmokinHole
  • Posts: 1861

Posted 15 May 2010 - 02:17

And Ming, re. Oleg's experience and Neoqb's committment not to repeat history, I hope you are right. Having witnessed a bit of the Neoqb obstinance in the face of a rock-throwing mob (I include myself) during DF beta, I am confident too. It's a trait I am really learning to like.
  • 0

#101 BroadSide

BroadSide
  • Posts: 2057

Posted 15 May 2010 - 02:45

Sorry Imperator

Image
  • 0

#102 Hellbender

Hellbender
  • Posts: 3321
  • LocationMadrid, Spain (originally Brussels, Belgium)

Posted 15 May 2010 - 10:39

More witches!
  • 0

J5_Hellbender


#103 J2_squid

J2_squid
  • Posts: 3815

Posted 15 May 2010 - 11:15

Its a fair cop
  • 0

#104 ImPeRaToR

ImPeRaToR
  • Posts: 7902

Posted 15 May 2010 - 12:33

I know this is a joke but I did actually find test data for both the Pfalz D.IIIa and the D.VII :roll:

Obviously there are no numbers on the stall characteristics of the Pfalz D.IIIa but if it had been the great aircraft it is in ROF surely more pilots would have liked it. ;)

It was at least en par with the D.Va, but that was mediocre at best. Rambling again I guess…
  • 0

#105 J2_squid

J2_squid
  • Posts: 3815

Posted 15 May 2010 - 12:36

Can you show us that data Imp?
  • 0

#106 ImPeRaToR

ImPeRaToR
  • Posts: 7902

Posted 15 May 2010 - 13:00

Did actually :D Can't remember where, with all these threads about FMs it is hard to keep track.

I found a french test of an OAW D.VII with D.IIIaü, sent the link to An. Petrovich actually since it was a really interesting result and included details for the throttle operation in case he can be convinced to model it for the D.VII and hopefully the D.XII. And after that, say in a year or so, they could release an Albatros D.Va with D.IIIaü for completeness sake, once more pressing planes have been finished.

Ah, found the D.IIIa thread:
http://www.theaerodr...8421-post4.html">http://www.theaerodr...8421-post4.html

a few things to keep in mind that I realised so far:

The germans ignored sea level for HP and top speed. it was irrelevant to them for testing (and it makes sense). This is shown here:


German figures (not sourced)

Max. speed 165 km. hr. (103.12 mph). Climb to 1,000 m. (3,280 ft.) in 3.25 min., to 2,000 m (6,560 ft) in 7.25 min., to 3,000 m (9,840 ft.) in 11.75 min.

As you can see there is no altitude given. THis is NOT sea level however:


British figures from a captured D.III (160 hp Mercedes)

Max. speed at 10,000 ft. 102.5 mph, at 15,000 ft. 91.5 mph. Climb to 5000 ft in 6 min., to 15,000 ft in 41 min. 20 sec.


As you can see, 102mph are 165kh/h, but at 3km! And these are Pfalz D.III, they don't even have the D.IIIa which was rated 170/175PS (two models I believe) by the germans while the engine in this one was rated 160.

Like I said my theory is that the german PS rating were referring to a specific altitude (maybe 3km again?), not sea level. This also explains why the BMW was rated 185PS and not 230+, the sea level power of the overcompressed engines were theoretical anyway since full power at low altitudes was going to damage the engines.


165km/h at 3000m (presumably IAS) is a lot faster than what our current D.IIIa can do, we barely have 165km/h at sea level. So this does not help to model it correctly, it's only a clue that the plane is too slow.

The climb rate is higher as well for the german test: We have:

3000 m — 12 min. 51 sec

german test says:

3,000 m (9,840 ft.) in 11.75 min.

MiG-77 tested in 1.010:

3000m in 12 mins 30 secs.



If there really is no complete data for the Pfalz D.IIIa, the performance of the Albatros D.V and D.Va could be used for reference. They had about the same speed and climb rate given the same engine, just like now. Pfalz climbed slightly better than D.Va, D.Va was slightly faster, iirc.

However they should not have the same engine ;) If they keep the D.Va a mid 1917 plane like now at least the Pfalz needs a new FM since it currently has the performance of an engine it never used.




If you look at this list the Pfalz D.IIIa speed can't be sea level, all the other planes with the same engine are alot faster (180+km/h). At sea level, the D.IIIaü had similar power so they can be included in this comparision, they also all have 180+ speeds:
This data is from "The Benighted Rolands", Air Enthusiast Quarterly, Three. by Dan-San Abbott and Peter M. Grosz.
Max airspeed:
Fok.D.VII 170Ps Merc. D.IIIa, 116 mph (186km/hr).
Fok.D.VII 180Ps Merc. D.IIIaü, 118 mph (189km/hr).
Fok.D.VIIF 185 BMWIIIa, 124mph (198km/hr).
Rol.D.VIa 170Ps Merc.D.IIIa, 118 mph(189Km/hr).
Rol.D.VIb 195Ps Benz Bz.IIIaü, 124mph (198km.hr).
Pfalz D.IIIa 170Ps Merc.D.IIIa, 103mph (165km/hr).
Pfalz D.VIII 160Ps ShIIIa, 114mph(190km/hr).
Pfalz D.XII 180 Ps Merc.D.IIIaü, 119mph (190km/hr).
Pfalz D.XII 185 BMWIIIa, 115mph (184km/hr).
Max weight and time to climb to 5000m:
Fok.D.VII 170D.III, 2115lb (960kg) ,38'.
Fok.D.VII180 D.IIIaü, 1940 lb (880kg, 31'.
Fok.D.VIIF 185 BMW, 1995lb (906kg), 14'.
Rol.D.VIa 170 D.IIIa, 1860lb (845kg0, 25'.
Rol.D.VIb 195 BzIIIaü, 1875lb (850kg, 19'.
Pfal.D.IIIa 170Ps D.IIIa, 2015lb (915kg), 33'.
Pfal.D.VIII 160 ShIII, 1625lb (738kg), 15'.
Pfal.D.XII 180 D.IIIaü, 1975lb (896kg),31'.
Pfal.D.XII 185 BMWIIIa, 1985lb (902kg),17.6'

  • 0

#107 Kwiatek

Kwiatek
  • Posts: 680

Posted 15 May 2010 - 13:20

Here i started some time ago topic about Pfalz DIII flight characteristic:

http://www.theaerodr...bad-really.html">http://www.theaerodr...com/forum/aircr … eally.html


Some notes:

"German pilots variously criticized the Pfalz’s heavy controls, low speed, lack of power, or low rate of climb compared to the Albatros. The D.III slipped in turns, leading to crashes when unwary pilots turned at very low altitudes. Moreover, the Pfalz stalled sharply and spun readily. Recovery from the resulting flat spin was difficult, though some pilots took advantage of this trait to descend quickly or evade enemy aircraft"


Here are CLmax for these planes :

1.462 - Pfalz D.IIIa
1.370 - Fokker D.VII
1.313 - Albatros D.V

Analyzing Pflaz wing airfoil i see that even if it get high critical AoA ~ 16 deg. it lose very quickly lift without so much warning and has sharp polar Clmax/AoA. From charts it look that e.x Fokker airfoil get critical AoA ~ 17 deg but also has much more gentle polar. Albatros airfoil get lower critical AoA ~ 14,5 deg but also polar is more gentle then Pflaz. Also longer wingspan of Plfaz DIII could result worse roll rate and the same worse manouverbility in change directions of flight."
  • 0

#108 ImPeRaToR

ImPeRaToR
  • Posts: 7902

Posted 15 May 2010 - 13:25

That's about the D.III with the D.III engine, the extra power did make it less awkward to fly. But it is hard to believe that it suddenly became the aircraft we have in ROF. Especially since the performance of the one we have actually matches the D.III :D


Btw, I am guessing that the data in my post is victim of a mix up. the plane the british tested is probably a D.IIIa with D.IIIa engine. That would explain why it has such a good performance compared to the one in our game.

So if I understand this correctly the Pfalz could pull high AoAs but would depart very suddenly if it taken past that, kinda like the current Dolphin is modelled maybe?
  • 0

#109 Eldur

Eldur
  • Posts: 220

Posted 15 May 2010 - 17:57

So basically the Pfalz D.IIIa should be faster and a lot more "b1tchy" in handling. Stall like a FW-190 and the need of lots of rudder in turns :D
Even if it will be faster then, it will be worse. Not fast enough for good BnZ vs SPAD, Dolphin and SE5a, and not that good turn fighter (Albies better).
  • 0

#110 Marco_._

Marco_._
  • Posts: 2594

Posted 15 May 2010 - 18:03

Stall like a FW-190 and the need of lots of rudder in turns :D

like all other planes except N17 :mrgreen:
  • 0

#111 ImPeRaToR

ImPeRaToR
  • Posts: 7902

Posted 15 May 2010 - 18:25

So basically the Pfalz D.IIIa should be faster and a lot more "b1tchy" in handling. Stall like a FW-190 and the need of lots of rudder in turns :D
Even if it will be faster then, it will be worse. Not fast enough for good BnZ vs SPAD, Dolphin and SE5a, and not that good turn fighter (Albies better).

Sounds like the perfect plane for somebody that likes a challenge :) With the better speed and its current dive handling it might still be a nightmare for S.E.5a and Spad actually but it will have to be scared of camels, while atm, imo, after the Dr.I, it is the second best plane to fight camels in.
  • 0

#112 Hellbender

Hellbender
  • Posts: 3321
  • LocationMadrid, Spain (originally Brussels, Belgium)

Posted 15 May 2010 - 18:40

Image


(I keeed, I keeed)
  • 0

J5_Hellbender


#113 J2_squid

J2_squid
  • Posts: 3815

Posted 15 May 2010 - 21:05

Lol thats hilarious Hellbender :lol:
  • 0

#114 Chill31

Chill31
  • Posts: 1891

Posted 15 May 2010 - 21:11

Can I say politely that this is out of control will likely not lead to any changes…I think that it is justified that no changes would be made at this point…

If you want to improve rise of flight, I would recommend you stick to numerical data. Leave the handling characteristics based on anecdotal evidence for a later effort…

I would look at things like airspeeds, time to climb, etc for the moment. Like the fact that the Fokker DR1 is faster than the Fokker D7 at 4km…
  • 0

#115 WW1EAF_Ming

WW1EAF_Ming
  • Posts: 2565

Posted 15 May 2010 - 23:25

I would recommend you stick to numerical data

Quite right Chill, although in this thread it is not a recommendation, it is a requirement

Reminder:
1) Provide numerical data, citing its source so it can be independently checked.
Ming
  • 0

#116 J2_squid

J2_squid
  • Posts: 3815

Posted 15 May 2010 - 23:47

Like the fact that the Fokker DR1 is faster than the Fokker D7 at 4km…

Chill mate, show me the data! lol.
  • 0

#117 Chill31

Chill31
  • Posts: 1891

Posted 16 May 2010 - 02:20

Squid, that is how it is currently in game…you can run your own test to verify…
  • 0

#118 Hellbender

Hellbender
  • Posts: 3321
  • LocationMadrid, Spain (originally Brussels, Belgium)

Posted 16 May 2010 - 11:05

If you want to improve rise of flight, I would recommend you stick to numerical data. Leave the handling characteristics based on anecdotal evidence for a later effort…

I agree with you, Chill.


More so, accurate numerical data goes beyond single plane performance. It's in plane numbers.

If you're in a single Camel fighting a single Dr.I, it's your own bloody fault that the performance of your plane isn't like in the books. You should make a point of it not to engage an enemy without an energy and a numerical advantage.


In the end, if performance is all about having the best low altitude turner that can win any 1 on 1 fight, then you don't want a sim, you want AirQuake.


It's a completely different discussion altogether, but I think that's where neoqb misunderstood the community. Yes, we all wanted Dogfight mode to create better historical "war type" missions where you can take off at any time during the game. The fact that we also got CTF mode is surprising to say the least and, I believe, a major aggravator towards the issue highlighted in bold above.

And here's another problem of Rise of Flight. The community frowns upon everything that isn't a completely fair 1 on 1 fight. I remember =IRFC= being accused of "aces stacking" on ToW. I assure you that while we do have a few very talented pilots (Darling says hi), not all of us are so-called aces (including yours truly), but we do fly as a team as often as possible, even on public servers. And I do realise TeamSpeak is also to be blamed, but what can you expect?
  • 0

J5_Hellbender


#119 Mogster

Mogster
  • Posts: 3919

Posted 16 May 2010 - 11:45

The FM's I'd like to see looked at are the early ones, Spad13, Pfalz D3a and the Albi's.

There's a long thread over at SimHQwith people commenting how easy those 3 are to fly to the max compared to the other aircraft in the sim. As others have said, there's something odd when people are choosing the Pfalz D3 and Albi over the Fokker D7 online.

The later FM's seem to be much more detailed, maybe Neoqb realised that the DR1 and Camel would come under intense scrutiny when they were released and pulled the FM stops out. Both seem excellent and compare well with historical accounts of either aircraft.

The Pfalz and Albatros just seem to be yank and bank machines at the moment requiring very little rudder input to achive max turn rate. Its just a matter of initiating a turn and hauling the stick back as far as you can, you just go round and round in tight circles. The Spad has issues as well, its another basic trainer in disguise.
  • 0

#120 J2_squid

J2_squid
  • Posts: 3815

Posted 16 May 2010 - 12:12

Crikey Chill your right…

Attached File  DR1 @ 3000m.jpg   582.75KB   275 downloads

Attached File  DVII @ 4000m.jpg   1.06MB   274 downloads

Now I dont know if its the DR1 that is too fast, or the DVII that is too slow. Probably both.

Unfortunatley I cant find any data on speeds for either of them at 4000m. Can anyone else?

Also here are the references Neoqb use for the DVII. I have no 3, but can anyone lay their hands on any of the others?

Attached File  DVII REFERENCES.jpg   88.72KB   267 downloads

These are the references for the DR1. Again I have a couple of the books but can anyone find the others?

Attached File  DR1 References.jpg   133.12KB   266 downloads
  • 0


0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users