Issues with the JAS 39 Gripen flight model

[Necronomica]

@Metrallaroja

In game wing area is 30m^2 and canards are treated like an elevator with 4.60m^2

oh wow. yeah that explains a lot actually. I get that there are limitations in the coding but man that’s not going to behave as i should at all.

are there any sort of plans for bigger overhauls to the FM codes? it’s getting to a point where NONE of the future added Jets (newer than what we have at the moment) will be able to be modeled even close to correctly (we even have some that cant be modeled correctly already as it is).

(just a spontaneous thought; can’t the canards be modeled as a wing? Giving them the ability to give variable lift as well as give the plane negative stability where negative pitch of canards cancels out the stability? this isn’t how it completely works IRL but i mean it will at least be closer.)

[Metrallaroja]

I mean area wise, canards area is treated apart from wing area as every other elevator in the game.
And yes it produces lift, every elevator does in game but gripen canards does produce more lift than regular elevators.

[Flame2512]

@SlowHandClap

Granted I’m fairly certain the physics model in WT is as simple as “this top speed and this turn rate”, so it probably won’t make any difference when they do “model” it.

How about you take a look at the actual flight model code. It’s a lot more complicated than that (look at the full file, not just the embedded preview):

gszabi99/War-Thunder-Datamine/blob/master/aces.vromfs.bin_u/gamedata/flightmodels/fm/saab_jas39c.blkx

[Necronomica]

@SlowHandClap

The TWR difference at subsonic and transonic speeds is significantly less than the comparison you see the usual suspects giving, using the F-16s thrust @ 1,400 kph and the Gripen max thrust @ 1,200 kph.

Well, a truth with modification. sure thrust will be closer, but i’m using numbers to give the F-16 the best possible chance just to show that even in those circumstances the gripen performs better.

@SlowHandClap

Whereas the wingloading is a fixed 30% improvement that affects the fm at all speeds.

That is why i don’t like using wing loading as a comparison. wing loading numbers assumes the same lift generated per area of wing. the in game used numbers result in a wing loading that is 48% better on Gripen. but in reality the canards cause dynamic lift so it will be a completely different amount of lift on the same wing att different AoA and speed. having a static wing will cause a (close to) linear increase in lift with speed. the JAS39 is designed to create less lift from canards at higher speeds making it basically supersonically/transonically stable but subsonically unstable. the elevons then have to help compensate somewhat for these changes in collaboration with moving the canards.

@SlowHandClap

Those shouldn’t be percentages, just ratios.

i felt it easier and more understandable to demonstrate a percentage of the ratio instead of writing 1:0,0026 or 10000:26.

@SlowHandClap

GJN includes the canards as part of the 30 m^2.

as written above:

@Metrallaroja

In game wing area is 30m^2 and canards are treated like an elevator with 4.60m^2

which is very weird because then the canards are treated as both pure lift (and not including the positive vortex interactions and extra lift they naturally give) as well as elevators thus producing lift only in certain situations where they also cause drag where they IRL wouldn’t.

(i’m hoping they are modeled as elevators positioned in front of center of lift and not just visually placed there but modeled as elevators behind the center of lift. @Metrallaroja ?)

@SlowHandClap

Worth noting that, as far as I know, relaxed static stability isn’t modelled for the F-16 or Gripen. Granted I’m fairly certain the physics model in WT is as simple as “this top speed and this turn rate”, so it probably won’t make any difference when they do “model” it.

i’m personally both hoping for and advocating for an overhaul of their simulations, as it stands basically none of the future modern jets they add will be able to be modeled correctly or Bug-reported with any sort of reasonability.

[Metrallaroja]

@Necronomica

(i’m hoping they are modeled as elevators positioned in front of center of lift and not just visually placed there but modeled as elevators behind the center of lift. @Metrallaroja ?)

Yes, they are modeled where they are visually.

[Metrallaroja]

Old wiki post but you can get an idea with this:
https://wiki.warthunder.com/Flight_model_creation

[Necronomica]

THANKS! :) Will read when i have time :)

[BBCRF]

They are already modeled incorrectly.

[SlowHandClap]

@Flame2512

How about you take a look at the actual flight model code. It’s a lot more complicated than that (look at the full file, not just the embedded preview):

… Obviously I’ve seen what the fm files look like. It was hyperbole… 🙄

@Necronomica

That is why i don’t like using wing loading as a comparison. wing loading numbers assumes the same lift generated per area of wing.

which is very weird because then the canards are treated as both pure lift (and not including the positive vortex interactions and extra lift they naturally give) as well as elevators thus producing lift only in certain situations where they also cause drag where they IRL wouldn’t.

(i’m hoping they are modeled as elevators positioned in front of center of lift and not just visually placed there but modeled as elevators behind the center of lift. @Metrallaroja ?)

Points raised by @Necronomica are valid concerns, hence my tongue in cheek comment.

[Necronomica]

that’s what i said?

[FeetPics]

[quote=“Necronomica, post:286, topic:86895”]
gross weight by doing (T/W)/Wing Loading we get:
F-16: 0.26%
JAS39: 0.

Why are you using what appears to be max gross weight when all the comparisons of SEP are not based on that? Why are you leaving the Mirage 2000 out of the comparison?

I used the localhost fm command tool to get the weight numbers.

Mirage 2000 has around 229kg/m2 for wing loading based on 50% fuel and 2 missiles.

Gripen has around 275kg/m2 for 50% fuel and 2 missiles.

F-16A has around 347kg/m2 for 50% fuel and 2 missiles.

I say around because in order to get the weight for the fuel in the localhost thing you have to load in with 50% fuel and then adjust it with the localhost tool to get the weight numbers because it defaults to like 8020kg and the fuel percentage on the localhost command is based on with drop tanks…so you have to adjust it.

Thrust Loading
Mirage 2000 1.08 @ 800kph
Gripen = 1.18 @ 800kph
F-16 = 1.38 @ 800kph

These numbers are for 50% fuel and no missiles. This is limitation of calculator program.

Mirage 2000
1.08/229 = .0047 = .47%

Gripen = 1.18 / 275 = .0043 = .43%

F-16A = 1.38 / 348 = .0040 = .40%

I am rounding the numbers but using the full value in calculator.

F-16 / Gripen = .93 ( so F-16 has 93% ratio of TW/WL)
Mirage 2000 / Gripen = 1.093 (so Mirage 2000 has 109% ration of TW/WL)

If we run thrust loading for 500kph…we get the following.

F-16 = 1.18
Mirage = 1.03
Gripen = 1.03

F-16 = 1.18 / 347 = .0034
Mirage 2000 = 1.03 / 229 = .0047
Gripen = 1.03 / 275 = .0037

So even at low speeds we get the following
F-16 / Gripen = 92%
Mirage 2000 / Gripen = 109%

All of these numbers are based on in-game data from the localhost program. The numbers might be slightly off because of rounding errors.

If we just go off of the numbers here…I don’t see any reason why the Gripen would be radically better than either the F-16A or the Mirage 2000 when it comes to specific excess power. Or to be more specific…why it’s bleed rate should be significantly lower.

“Bleed rate” is a term used to express basically the loss of speed while performing a max performance turn. Itis a slightly different term than SEP because SEP curves also cover positive value…i.e being able to accelerate while in more gentle turns.

[FeetPics]

@SlowHandClap

The TWR difference at subsonic and transonic speeds is significantly less than the comparison you see the usual suspects giving, using the F-16s thrust @ 1,400 kph and the Gripen max thrust @ 1,200 kph. Whereas the wingloading is a fixed 30% improvement that affects the fm at all speeds.

Thrust to weight ratio for the F-16 is significantly better than the Gripen.

[Necronomica]

@FeetPics

Why are you using what appears to be max gross weight when all the comparisons of SEP are not based on that? Why are you leaving the Mirage 2000 out of the comparison?

because most T/W i have seen have been based on gross weight.
also, “max gross” what is that?

@FeetPics

Why are you leaving the Mirage 2000 out of the comparison?

because this specific interaction started with only F-16. i have showen both before further up in the thread.

@FeetPics

I used the localhost fm command tool to get the weight numbers.

i was calculating the weight based on IRL publicly available numbers. i might have been slightly off.
i have also shown the Wing loading and T/W ratios of different amounts of weights further up in thread.

@Necronomica

Half fuel (weight of half internal max fuel) and 2 missiles (180kg):
F-16: T/W 1.287 Wing load: 371
Ratio: 0.35%
M2K: T/W 1.051 Wing load: 225
Ratio: 0.47%
JAS39c: T/W 1.007 Wing load: 272
Ratio: 0.37%

@FeetPics

I am rounding the numbers but using the full value in calculator.

thanks for showing this :) i am also doing the same when calculating anything i post.

@FeetPics

If we just go off of the numbers here…I don’t see any reason why the Gripen would be radically better than either the F-16A or the Mirage 2000 when it comes to specific excess power.

that’s the main issue i have, you can’t use JUST those number and take nothing else into account when evaluating SEP and bleed rate.
Firstly it assumes the same amount of lift is generated per wing area of all three planes not only at level flight but also at all AoA (obviously not true, and we can’t know real values but we can estimate).

the reason they loose speed at all is because of drag. which (according to what i can find and have seen) Gripen will have less of for any specific radius turn at a specific speed. as it needs less AoA to achieve that turn.

i’ll just reiterate my main points:
Issues with the Jas39 gripen flight model - #271 by Necronomica

[SlowHandClap]

1. Is the F-16 not a massive outlier in terms of how the engine produces 170% thrust at peak compared to static thrust? Interesting how you’re not upset about that being an outlier 🤔

2. Are you able to plot drag with these graphs?

@FeetPics

I say around because in order to get the weight for the fuel in the localhost thing you have to load in with 50% fuel and then adjust it with the localhost tool to get the weight numbers because it defaults to like 8020kg and the fuel percentage on the localhost command is based on with drop tanks…so you have to adjust it.

Can you clarify this a bit please? You’re selecting 50% on the fuel slider I assume? Why are you using the 50% values from local host if that’s the 50% value with fuel tanks? That will skew things heavily in favour of the F-16.

[BBCRF]

You said for modeling future aircraft. Even now the projects made by the 4th generation are modeled incorrectly, every single one

[Necronomica]

i also said in the same post:

@Necronomica

we even have some that cant be modeled correctly already as it is

[Spychicken]

MiG-29 gains 180% thrust, F-5C gains 170% thrust. The F-16A is not an outlier.

[MiG_23M]

@SlowHandClap

Is the F-16 not a massive outlier in terms of how the engine produces 170% thrust at peak compared to static thrust

No, real world installed thrust charts are available for the engine. Other engines have similar curves. This is not remotely an outlier.

[SlowHandClap]

@Spychicken

MiG-29 gains 180% thrust, F-5C gains 170% thrust. The F-16A is not an outlier.

Well those two examples achieve those numbers at their rip speed, whereas the F-16A achieves 170% @ 150 kph below its rip speed. For two more numbers - the F-5E is only +36%. The 23MLD +52%.

@MiG_23M

No, real world installed thrust charts are available for the engine.

I’ve not seen these for the F-16. Can you share them in the F-16 thread if you haven’t already please.

[FeetPics]

@SlowHandClap

• Is the F-16 not a massive outlier in terms of how the engine produces 170% thrust at peak compared to static thrust? Interesting how you’re not upset about that being an outlier 🤔

It is not a massive outlier.

@SlowHandClap

Are you able to plot drag with these graphs?

I have already posted drag curves. Graph can do SEP for a specific AoA but that takes into account engine power to weight ratio and different planes have different AoA limits for max performance turns.

@SlowHandClap

Can you clarify this a bit please? You’re selecting 50% on the fuel slider I assume? Why are you using the 50% values from local host if that’s the 50% value with fuel tanks? That will skew things heavily in favour of the F-16.

If you select 50% of the slider it will fill the tank to 50% of internal capacity. If you select 50% in localhost command it will fill to 50% of total possible fuel including drop tanks.