New stability patch = UFO

[rossmum]

AI flight models have nothing to do with the player module flight models, they're separate SFMs.

[Barra1]

Is the AI SFM created by the developer or ED?

[Tarres]

Base is the old LOMAC MiG-21 mod, by the developer.

[Sideburns]

On 7/16/2020 at 11:45 PM, Hiromachi said:

Alright Shmal, I looked through your report and to be completely fair the speed at which you try to compare things is way below normal flying curve. At speeds you push that aircraft I land

But in all seriousness, I took a look through my documentation and most of graphs for sustained turn, turn radius and turn time end at Ma 0.5. So in order to get something for a speed around 0.31 - 0.32 such data would need to be extrapolated (and even that would not be precise enough, we're basically moving from what is known and described to what is unknown and undescribed in documents). Since that is beyond my skills, Im going to forward your report to Dolphin over the weekend for him to review it.

 

Thank you for the input !

 

1 year anniversary  Any update on this? The MiG21's low speed handling still seems generous.

[Hummingbird]

The MiG-21Bis does appear to have an advantage over the F-5E at low speeds in DCS atm, that much I can confirm. This is in rather stark contrast to the available performance charts, which indicate that the F-5E should enjoy a noticable advantage in sustainable load factor up until around M 0.8.

 

For comparison:

 

 

 

 

 

 

[Shmal]

Yes, I completely agree with this another argument from Hummingbird!

Real F-5 at 0.3M has to pull 2.2g in sustained turn. I think at 0.33M it has to be 2.4g.

Real MiG-21 has to pull evidently not more but even lesser, because with 57° swept wing she has to have higher AOA and lower G in the same turn. Ok

On the chart she has the same 2.2g. At 0.33M probably the same 2.4g.

But in DCS MiG-21 able to pull 2.8-3g at 0.33M (400km/h approx). 

[Shmal]

I am tired of looking for the sources to prove it!

Here is another one proof from Soviet Manual of MiG-21bis (Izdelie 75A)

Edited October 25, 2021 by Shmal

[m4ti140]

So far you've drawn a bunch of squiggles, not shown a proof. Particularly the one with sustained turn time is pretty absurd to call an extrapolation.

[Noctrach]

Generally idk why were nitpicking percentages on arbitrary feelings of handling.

We've got a user posting in-game results from automated flight tests right here every couple patches https://dcs.silver.ru/74,9,IAS_kts,turnrate You can directly recognize all the important key points of both charts in there.

I'd say these two jets are pretty much where they should be within the sim. I'm not surprised users doing their own tests find results on either side of the truth. After all, the F-5 was considered to be simulating the MiG-21 nearly perfectly...

Seems to me like we're talking about regions that are extrapolated even in the real charts because they are pretty much nonsensical to discuss for combat operations. Below mach 0.3 is where you start preparing for landing...

Things will never be simulated perfectly, they're good in all reasonable areas of the flight model.

Edited October 19, 2021 by Noctrach

[Conker4]

5 hours ago, m4ti140 said:

So far you've drawn a bunch of squiggles, not shown a proof. Particularly the one with sustained turn time is pretty absurd to call an extrapolation.

idk how much you know about aerodynamics but a wing does not allow you to turn faster especially with drag, airflow seperation and a wing that is not designed for low speed... This aircraft has a primitive delta wing from the 50's. The graphs are there to show you the trend in load on the aircraft. If continued, the load would lower and infact would lower at a faster rate as the airflow starts to separate from the wing more and more. At the end of the day the FM is broken there is no two ways about it. IRL in US testing the F-5 was superior in a rate fight, in soviet testing the F-5 was again superior, the performance graphs that we have say the F-5 is superior in the rate, soviet doctrine and training to foreign users of the aircraft directly state that they should not engage in sustained turning fights with the F5. In game this is not the case. The devs, on the bug list has recognised that the FM below 500kph is broken and they are yet to assign someone to the task.

[Conker4]

2 hours ago, Noctrach said:

Generally idk why were nitpicking percentages on arbitrary feelings of handling.

We've got a user posting in-game results from automated flight tests right here every couple patches https://dcs.silver.ru/74,9,IAS_kts,turnrate You can directly recognize all the important key points of both charts in there.

I'd say these two jets are pretty much where they should be within the sim. I'm not surprised users doing their own tests find results on either side of the truth. After all, the F-5 was considered to be simulating the MiG-21 nearly perfectly...

Seems to me like we're talking about regions that are extrapolated even in the real charts because they are pretty much nonsensical to discuss for combat operations. Below mach 0.3 is where you start preparing for landing...

Things will never be simulated perfectly, they're good in all reasonable areas of the flight model.

 

tldr, you don't actually care about the aircraft at all.

I think its rather important to stack aircraft where they should be relative to each other. Even if all aircraft stats are wrong I think its important to have a mig-21 that can not rate better than an F-5 because... it could not out rate the F-5.

If the plane is magically contradicting real life at a speed where we do not have the information to state it with proof, then trends should be followed and aerodynamic simulations are to be conducted if the latter is not acceptable.

FYI its the multiplayer crowd that is pointing out the issues because any issues are exploited and reasonable areas of the flight model are not what people care about or use when vs real people.

 

[Noctrach]

1 hour ago, Conker4 said:

tldr, you don't actually care about the aircraft at all.

I think its rather important to stack aircraft where they should be relative to each other. Even if all aircraft stats are wrong I think its important to have a mig-21 that can not rate better than an F-5 because... it could not out rate the F-5.

If the plane is magically contradicting real life at a speed where we do not have the information to state it with proof, then trends should be followed and aerodynamic simulations are to be conducted if the latter is not acceptable.

FYI its the multiplayer crowd that is pointing out the issues because any issues are exploited and reasonable areas of the flight model are not what people care about or use when vs real people.

 

Look at the graph and tell me how it rates better.

You people pick weird edge case hills to die on.

[m4ti140]

3 hours ago, Conker4 said:

idk how much you know about aerodynamics

More than you

Quote

a wing does not allow you to turn faster especially with drag, airflow seperation and a wing that is not designed for low speed

Do you understand how a delta stalls?
 

Quote

The graphs are there to show you the trend in load on the aircraft. If continued, the load would lower and infact would lower at a faster rate as the airflow starts to separate from the wing more and more

Sea above, ever heard of leading edge vortices? Do you even understand that these are sustained turn rate diagrams and engine thrust vs drag plays an important role, ESPECIALLY IN A DELTA? I'm not saying it can't nosedive there, I'm saying drawing squiggles is not a good enough extrapolation. Notice how the characteristic changes in that region in the simulation. The FM doesn't have the turn rate diagram coded in, it handles much lower level data that need to be adjusted until simulation results check out.
 

Quote

At the end of the day the FM is broken there is no two ways about it.

You're yet to delver reliable data to support this claim

Edited October 19, 2021 by m4ti140

[Conker4]

17 minutes ago, m4ti140 said:

More than you

Do you understand how a delta stalls?
 

Sea above, ever heard of leading edge vortices? Do you even understand that these are sustained turn rate diagrams and engine thrust vs drag plays an important role, ESPECIALLY IN A DELTA? I'm not saying it can't nosedive there, I'm saying drawing squiggles is not a good enough extrapolation. Notice how the characteristic changes in that region in the simulation. The FM doesn't have the turn rate diagram coded in, it handles much lower level data that need to be adjusted until simulation results check out.
 

You're yet to delver reliable data to support this claim

 

COPE

[m4ti140]

Quality answer

Anyway, it's hard to draw any conclusions about what's wrong from sustained turn diagrams, especially if they conflict between sources. Devs most likely have additional data, cause last public document I've seen where they attempted to identify the FM from data that is easily accessible concluded in massive errors coming up in verification chapter. Any other attempt will have similar results. Get a single source with enough data to identify at least drag, lift and thrust curves in this region, otherwise it is hard to even conclude what is wrong, yet alone give devs a valid reason to revisit this when half of the systems are broken in one way or another and sometimes need redoing from scratch. Previous FM had the lift suddenly completely zero out as if the wings disappeared as soon as you hit critical AoA, if you think that was more realistic be my guest. The FM definitely breaks down past 35-40 degrees, cause vortex breakdown doesn't seem to happen and you can just stay there forever if you have lift, hovering on engine. But it has even less data on when and how it should depart, cause this is past the point where you would usually already lose control.

It's evident from the existing portion of the data that it's overperforming. It might seem slight there, but if FM is adjusted to match performance in normal flight regimes, the performance at 0.3 will very likely fix itself as the underlying issue is removed. One possible reason is that the lift coefficient in transition region is too high, possibly the potential lift is just modelled as linear and abruptly cuts off just like in old FM, and they just added vortex lift on top of that. The thing with delta is that as you get into vortex lift regime the lift curve will go up even steeper before it flattens out. Another is that the engine overperforms at low speeds. If you can find data on the latter in particular, that would either confirm the reason and rule it out. The former was most likely calculated or identified from other data, I don't know what data they had to get it from. See if you can extract the lift curve from the game, that will give some idea what might be off. Either way, anecdotal evidence and guesswork will go nowhere, this isn't an arcade game, nobody will "nerf" FM to make it "feel right".

[BIGNEWY]

Please treat each other with respect when posting 

[Sideburns]

14 hours ago, Noctrach said:

Look at the graph and tell me how it rates better.

You people pick weird edge case hills to die on.

Check out your own link to dcs.silver.ru, particularly the low speed section. Doesn't it seem odd to you, and contrary to the evidence in this thread and reviews by test pilots, that the MiG21 has an increase in sustained turn rate whereas other jets lose it in this low speed range? This is also in fairly accessible speed range and not some strange, rarely encountered edge case either. If a dogfight ends up in a grovelling 2c or 1c you could likely end up in these speed ranges. You people seem to pick weird flight models to die defending / get attached too.

 

21 hours ago, m4ti140 said:

So far you've drawn a bunch of squiggles, not shown a proof. Particularly the one with sustained turn time is pretty absurd to call an extrapolation.

If you consider these extrapolations a bunch of squiggles, and don't consider flight manuals as reliable data then it is just as well you are not the one reviewing the data and making decisions on the flight model. There's been some good evidence presented here which afaik has been under review by the M3 team. Feelings don't come into it, at least not for putting the argument together for the MiG21's low speed FM to be reviewed.

If you've got data to support the MiG21 having this exceptional low speed handling feel free to present it, or perhaps provide your own extrapolation of how you think the line should go? At the moment the evidence and reasonable extrapolations suggests the MiG21 is significantly overperforming at low speeds.

 

@Hiromachiany update on the FM review, last official response on this issue was some time ago.