Actually if you check back to some of the Q&As there are photos of them in C172s doing just what you described.  For that particular plane they did indeed compare numbers to the model. I believe several of the developers got their PPL’s to facilitate this data calibration step.  I’m sure some folks on this thread could confirm this.

I’m sure they wouldn’t have access to every plane in the default hangar, but in terms of validating the model against a specific airfoil, I do believe that was done. I leave it to the aeronautical experts to say how much that vetted the “core model” to allow it to be extrapolated to other planes

Edited August 31, 20214 yr by VFXSimmer

58 minutes ago, VFXSimmer said:

Actually if you check back to some of the Q&As there are photos of them in C172s doing just what you described.  For that particular plane they did indeed compare numbers to the model. I believe several of the developers got their PPL’s to facilitate this data calibration step.  I’m sure some folks on this thread could confirm this.

I’m sure they wouldn’t have access to every plane in the default hangar, but in terms of validating the model against a specific airfoil, I do believe that was done. I leave it to the aeronautical experts to say how much that vetted the “core model” to allow it to be extrapolated to other planes

I can't express an opinion because I have no direct knowledge of their development methods, although I seem to remember that they actually flew some 172 to get a feeling of the real thing. That doesn't hurt.

What is needed for a validation phase is to have very accurate telemetry for a number of different measurements in a systematic testing process and then compare real world data with sim data. That can be a huge, huge investment.

For the extrapolation question, again without direct knowledge, I think that there is a problem with modular sims like this one: you need a programming interface of some sort to allow third parties to develop new aircraft, which means that either you capture the essence of a Universal Flight Model (highly unlikely) or you will have airplanes that fly pretty well and others that are not realistic at all. For this reason, 3rd parties will ask for assorted scalars to account for corrections here and there, and you end up with a hybrid.

Can't really blame them though: as I said, a Universal Flight Model simulation is not available, to my knowledge, anywhere on this planet. At least not on desktop computers. Maybe as research software on highly parallelized supercomputing beasts for airflow simulation, but that's not my field.

Come to think of it, I absolutely know who to ask this last question, I will meet this guy in a few weeks. Will let you know.

A.

2 minutes ago, ADamiani said:

I can't express an opinion because I have no direct knowledge of their development methods, although I seem to remember that they actually flew some 172 to get a feeling of the real thing. That doesn't hurt.

What is needed for a validation phase is to have very accurate telemetry for a number of different measurements in a systematic testing process and then compare real world data with sim data. That can be a huge, huge investment.

 

Thanks ADamiani, I will have to search for the particular video, but I remember them taking actual measurements of "something" and comparing to the sims numbers.  They had a video camera watching the instruments and had a simultaneous view outside, and I'd swear someone may have had a clipboard in their hand.. and we all know that where there's a clipboard, there's science happening 😉

Again, have no info on the specifics at the moment, but I remember it impressed me about their level of interest at the time.

1 hour ago, Bert Pieke said:

And yet, the current C172 flies quite well... so they must have figured out how to correct the earlier mistakes made..

I agree that the early flight model was pretty bad, but for the C172 at least, that is not the case any more. 🙂

The C172 has always been one of my favorites in MSFS2020. I agree it flies pretty nice now, the best between most I would say. But if you look under the hood, you get what I mean. Basically nothing was done in the Source code for the FlightModel. All changes were accomplished using the Scalars. That suited well the C172, but really doesn't work as well for other popular planes, without getting some weird collateral effects

4 minutes ago, Alec said:

The C172 has always been one of my favorites in MSFS2020. I agree it flies pretty nice now, the best between most I would say. But if you look under the hood, you get what I mean. Basically nothing was done in the Source code for the FlightModel. All changes were accomplished using the Scalars. That suited well the C172, but really doesn't work as well for other popular planes, without getting some weird collateral effects

I do not know how to look under the hood, so I can only go by what I feel when I fly the plane..  like not being able to go into the kitchen, but only go by what is being served on the plate..

The C172 flies OK, the Rob Young Turbo Bonanza flies nicely, the P149 flies really well, the Grravel Cub flies great, the CJ4 flies OK.. those are the planes I fly.. so not sure what I am missing.. 😉