CFD
-
-
@wildthing At last report it's an ongoing evaluation as to the benefits of when to apply them, as the sim keeps adding things and breaking things faster than it adds them.
I believe the most recent plan is to start evaluating once SU9 is stable, but there's more coming in SU10 that would be of interest.
-
@jmarkows exactly - SU9 is just testing the waters, we will wait and see how it develops. There is no point in rushing to rebuild the flight model only to have to rebuild it again after the next SU, and the one after that etc.
The documentation in the SDK is still not brilliant for the core flight model, we can only hope that the documentation for CFD is a vast improvement to give us a fighting chance of implimenting it both accurately and in a timely manner.
-
@delta558 Please continue to beat Asobo over the head with that feedback. I know I've said it before on here, but it's frustrating as hell for the end users so I cannot imagine what it must be like for you guys.
I don't remember if this is you on the MSFS forums or not, but since they apparently don't care when the users complain, maybe if bigger devs like Just Flight start raising hell they'll straighten their act.
Now I see the Concorde is down with some fuel system issues that were identified in the beta, but went unfixed, and I agree with Dean's Facebook post entirely; what the hell is the point of releasing a beta if you're not going to fix issues identified during it?
-
@jmarkows yep, that's me! I've not exactly been silent, I tend to speak my mind with the hope that it comes across as intended - constructive criticism. However, I get the feeling that they see it as 'unwarranted criticism', despite the fact that the entirity of my sim work for the last 15 years has been in flight dynamics🤣.
A shame, and I cannot speak for others but from personal discussions it is a widespread feeling amongst some established and some new FD designers and (as reflected in the post I replied to) is probably a large part of the reason for delays.
-
@delta558 I appreciate you weighing in over there on the flight model and the constant changes. It is frustrating for me and you I am sure. It could be of particular interest as to the PA-38 I would think, which may be the first JF aircraft to implement the CFD.
I am of the opinion that MS/Asobo need to take a time out and sit down and document their flight model (and probably other things in the SDK). Although it could be that even they do not fully know what flying characteristics come out of their flight model roulette wheel.
As the other thread said, put your family first and don't let the current moving goalposts as to the flight model get you down. This too shall pass as they say.
-
@delta558 I appreciate you weighing in over there on the flight model and the constant changes. It is frustrating for me and you I am sure. It could be of particular interest as to the PA-38 I would think, which may be the first JF aircraft to implement the CFD.
I am of the opinion that MS/Asobo need to take a time out and sit down and document their flight model (and probably other things in the SDK). Although it could be that even they do not fully know what flying characteristics come out of their flight model roulette wheel.
As the other thread said, put your family first and don't let the current moving goalposts as to the flight model get you down. This too shall pass as they say.
No worries at all - in an attempt to stop hype before it starts, the PA38 will not impliment CFD until it is (a) consistent and (b) stable. Currently it is neither, SU10 has been put back until August and I think the flight model will be too far advanced / aircraft too close to release to suddenly rewrite the entire thing! Having said that, I don't think the current flight model is too shabby given the constraints etc and will shortly be in the hands of the testers (including real world PA38 pilots, who have already given initial feedback to get it to where it currently stands). It spins quite well!
I would love to see Asobo document the CFD properly. The SDK for flight modelling is vaguely improved from what it was, but still makes very little sense from either an aerodynamic or geometrical viewpoint in most areas and anything that gets noted as a bug seems to get a scalar. Not confidence-building, quite frankly, and I think your 'roulette wheel' analogy may be far too close to the truth for their liking. I asked over six months ago about jet rpm behaviour (adjusting the spool rate) and the fundamental error they built in has been ignored as much as my question has. To touch further on the CFD, if it is ever to be taken seriously then we need the ability to account for shape which is currently not there - take the PA38: using the internal forces debugger you can see where it believes the fuselage to be. No taper towards the rear, no bulge for the canopy, the fuselage maintains it's width from nose to tail. The only definition we have for the fuselage is the length, diameter and centre point (in relation to datum). So all you can define is a cylinder. What use is that for this GA aircraft, which does not conform. Or for virtually all military aircraft which will not conform. In fact, probably the only aircraft a cylinder is useful for is an airliner. It's just not accurate enough, and implimenting CFD whilst keeping the developer restricted in what they can define is just another PR stunt, unfortunately.
Their world is brilliant, so much is what we wanted to see during the FSX / P3D years. From someone who spends their entire time on flight dynamics, despite some obvious improvements in certain areas this is a game, and a very 'gamey' game! Despite the drawbacks in scenery and immersion of ALL the others, I still refer to them casually as 'sims'.
-
No worries at all - in an attempt to stop hype before it starts, the PA38 will not impliment CFD until it is (a) consistent and (b) stable. Currently it is neither, SU10 has been put back until August and I think the flight model will be too far advanced / aircraft too close to release to suddenly rewrite the entire thing! Having said that, I don't think the current flight model is too shabby given the constraints etc and will shortly be in the hands of the testers (including real world PA38 pilots, who have already given initial feedback to get it to where it currently stands). It spins quite well!
Very good news on the PA38, and I agree, I wouldn't implement CFD either if it wasn't ready for prime time.
As long it's spinning the way it should, isn't overpowered in climb etc. I'm good with it. I've seen other flight modellers struggle with MSFS in getting climb vs. cruise speed/power to work well together, so good luck.Perhaps Asobo can create the ability to define multiple cylinders. At least then there could be some semblance of proper airflow modelling. Although imagine modelling a slab-sided aircraft like a Sepecat Jaguar.
I really do think Asobo should, after SU10, make SU11 a documentation release, or, round out the areas of the SDK that are currently vague. I realize Asobo have said this is a 10 year project, but, if Asobo are going to add something, they need to fully document it right then in the SDK, not later. I remember a bgl scenery compiler in the late 90's called FSASM by Adam Szofran. Mr. Szofran's documentation was PERFECT. Extremely precise, all commands and effects extensively documented. This allowed us to really do it right, and properly use every function he allowed the compiler to handle.
And finally, I'm of the mind that Asobo needs to not just add something to put a notch in their belt (CFD) which it seems like that's what they've done with SU9's version of CFD.
-
@mace_rb There was ( is? ) a very neat mod for Kerbal Space Program of all things, which analysed the shape of your vehicle using voxels in order to calculate wave drag, along with prroviding some half decent aerodynamics ( and some analysis tools that probably did more for my understanding of fluid dynamics than a year at uni ). I think that was done in something like 2015, too. I was rather expecting something similar here, I can't believe they're really basically giving you a tube - what is there for wings? anything at all?