@geloxo So basically what you’re saying is that the airplane model is not correct…
You also said about 100 reply’s earlier that the wind can’t go from 20 kt to 0. It could be even worse. You can have (for example) 25 kt head wind to 10 kt tail wind in a blink of an eye.
Even worse if we would get in or out of a jetstream or a microburst. I wonder how the reaction would be if they simulated microbursts?
Yeah, it’s definitely mechanical turbulence, and it’s way over the top. I’ve flown that approach in real life on similar days in mid-afternoon, in the heat of the day, and it’s nothing like that. To me it’s the mechanical turbulence that’s the issue.
I don´t think the problem is the flight model. Only the light gusts speeds are wrong: they change in the same way as the strong gusts do and that´s not correct.
A cyclone can suddenly go from 20 kts to 0 kts. A breeze not. And even in the case of cyclones they still need several seconds to do it. You could see that on the real winds data, 100 replies earlier. Wind in game is generated artificially, so everything is possible if the formula says that. Real wind is the result of moving air masses and because they have a mass they can´t accelerate at the speed of light nor push nearby masses to accelerate them as well unless there´s a lot of energy on them, which only happens on the fast moving air masses but not on the breezes.
It´s clear that there are exceptions as well and you can also find superfast currents in the middle of calm winds, sure, but this is not what gusts are simulating now in game. They just simulate the changing on wind speeds and wind direction.
Cheers
Yes, i agree. mechanical turbulence has some issues. But maybe it’s because other aspects of simulated air flow is missing? Maybe mechanical turbulence also will get better when 20KM CFD is implemented that they work on? I know you were in the beta as well and we tested it much right? With those graphs i showed of vertical air movement i don’t remember but i think we came to the conclusion those shoud be more smooth right? Well, i know machanical exists IRL and they exists in the sim. Then i think how much it should be depends on the location we are flying in. Also our heads and eyes compensate for those bumps and makes it smoother. In the sim the camera follow every single bump without dampening.
Funny how that works, experiencing the same thing sometimes as you did but I took the opposite conclusion (mechanical turbulence is fine, wind change model isn’t)… i guess it is too complex to deduce what’s wrong by simply flying the sim without having a deep view of what is happening and how it is modelled
Completely forgot about the 20km CFD, in theory that together with thermals/convection should make for a much better Baseline on which gusts etc can be modelled
Watch this one from the 20:00 mark and see what the plane does when it crosses from over water to over land. There’s also a little moat right before the runway threshold, and I get another really good jolt there as well. To me everything is fine save for the mechanical turbulence.
I’ve never seen something like that in the sim… yes that’s not right
The warrior though isn’t using any of the aerodynamics changes so maybe that makes it worse?
Ultimately i think it will take a while but i still think it’s miles ahead off what it was before
Yes i agree. That 20KM CFD should make a local flow of air that is formed realistic. But i’m not sure how they can handle that together with METAR reported winds. Time will tell. I would preffer to have CFD everywhere but at airports many needs the winds to be exactly as METAR reports.
Ok, so I cherry picked an airport with only little wind (3 kt) and did the same run in Sting S4 and Cessna 152 (the only other “light” aircraft with CFD enabled that I have as I hear that CFD makes this phenomenon worse). EDIT: the 152 does NOT have CFD 
I’m happy to report that at least the gust behavior is tuned down in less windy conditions, but the Sting S4 in VR is still somewhat unpleasant (left-right motion). The Cessna flight was much more enjoyable and did not bother me at all. Maybe the Sting flight model needs some tuning or the sim does not handle very little aircraft correctly, or the VR camera makes things feel worse, I don’t know.
I would just like to add that this thing for me is not “real simmers” vs “gamers”. I’m all for realism and I have zero interest in flying in default preset weather, because live weather is what makes things interesting, but I’m also very skeptical of the live weather because I still find snow and ice in July/August in places that have forests and open lakes and +20C in the summer. AND let’s remember that the flight model / weather is constantly changing with every update, so I don’t believe it’s just “spot on” now.
Agree, they can always improve it. It’s not set in stone. But removing it instead of improve i’m 100% against.
I could be wrong but I don’t believe the 152 has CFD. And they did (arbitrarily) reduce turbulence with winds below 3 knots (I believe was the number they gave.
Correct me if i’m wrong but cessna 172 g1000 has CFD?
Also in su10 they toggled the air to be unlimited at ground in the 172 that has been in the sim since FSX. That also makes the plane behave much different on ground than all of the other planes we are using.
I think that thing can be changed in all of the aircraft though. I’m not sure about that. Maybe it’s ony for testing.
172 G1000 does have it.
You should test this with other planes. The justflight arrow fleet has a problem with its flightmodel for months now. Mainly „twitch jerks“. There are threads about it on the justflight forum and the dev confirmed it, but has yet to fix it or implement the improved flight dynamics.
Try the same flights with the 172, as it is a more standardized representation of the interaction between the outside factors and flight model.
I think you are right, thanks! 152 only seems to have the new propeller simulation and 172 g1000 has both.
Yeah I tested it in the 182 to see how the CFD handled it, and at the time I found it just as bad, but in rewatching the three approaches I’ve shared here, it does look like the 172 is better. Watch from 24:00
I can’t comment on the realism or how good the effect has been implemented, but how it’s always the flight model and outside effect that are interacting. Therefore you can’t just say that one is bad and not look at how maybe the flight model needs adjusting. I don’t know if you have addressed this before, just wanted to point it out.
Third party devs have shown that they don’t always fully understand how to program msfs to its fullest potential and while Asobo isn’t known to deliver the greatest airplanes, atleast their airplanes can give us an idea of a properly implemented flight model.
Edit: Upon reading this again, maybe this should be limited to something like the 172. Not every Asobo aircraft has been praised as much and received flight model updates as frequently.
In theory at least, with SU10 you should be able to apply CFD to the steam gauge 172 by comparing flight model files as they are now unencrypted.
Exactly. I can cite two examples where devs used SDK. Default values for the autopilot, as they likely believed those settings only applied to the AI pilot, rather than the AP.
What else don’t they know how to use?