Odd, because I would expect the FM to be far closer than the systems.
FM is a function of the aero and the thrust, and that is plain to see for all. I mean, maybe we don’t know the exact speed the control surfaces can deflect or the angle, which would affect how fast it could start to roll, or how far it could pitch up, but it should be close.
The systems, however, could have tons of classified machinations under the surface that would never be allowed in a commercial product. I mean, that woman doing the F-15 cockpit walkthrough a few months back just dismissed half the MFDs as “classified” when we’ve been looking at representations in sims for 30 years now.
Not a wrong concept but thrust is not a constant and the Digital Engine Control Unit of the EFA play a big part of the performance envelope, that have to be put in a function with a the atmosphere and all the rest so…
Quite accurate but probably not more than the systems. Just my 0.02, of course.
I don’t agree here, there are parts in the airflow such as the intake ramps that are still “classified” so modelling to what would be an accurate (although let’s not confuse that with believable) would be impossible to get Correct.
The systems could be modelled much more accurately and effectively based on what we know from similar systems. 5th gen fighters are just so far beyond anything previously flown that we can’t accurately model them. Look at post stall behaviour in DCS aircraft at the moment, its close enough to be believable but still not true to life. 5th gen fighters can still have meaningful control long after the airflow disruption becomes to powerful and random to model. We can’t even get a meaningful spin on a 172 in MS2020.
No one models outside the envelope well. Set that as a bar and you’ll always be disappointed.
However, what you’re talking about are limitations. We know the aero, we know thrust at sea level. Everything else takes away from that. So we know it will never be better than X, we just don’t know how much lesser it will be. So it could have a reduction in thrust of Y or Z at altitude, we don’t know, but we know it won’t be BETTER than that sea level value. Is it reduced by 15% or 20%?
If what you’re saying is “I demand exact representations for pressure, alt, temp, and weight impacts on performance”, yeah, you’re going to be presented with guesswork and limitations. When you say “impossible to get correct”, it’s implying that makes it unworthy. You’re picking an arbitrary standard. Does that mean it’s not worth having? Absolutely not. Look at what is now being admitted to as an inaccurate A-10C FM which results in them flying with far less power and agility than they should have. Didn’t stop people from enjoying it for a decade now.
Yet the entire post-stall regime is outside the capabilities of most million-dollar sims (try flying a 747 at 40 kts nose high at 50k feet in one of those sims and surprise, surprise, it doesn’t match what the real thing will do because it wasn’t thought necessary to model it). All these accidents with things like erroneous pitot readings causing planes to fall out of the sky have proven elusive to simulate. Turbulent flow post-stall is HEAVY number crunching. Chaotic numbers and variables make it questionable if it’s even accurate without flight testing.
Bottom line–the distinctions you have made may be important to you, but they are not important to me. Therefore, it is likely that I am not alone, and what you want is outside the “85% solution” that would require unwise investment of resources to get a relatively small return in the overall product that many customers would not be swayed by when it comes to deciding whether to purchase it.
I beg to differ. If what you want is to fly from A to B, drop an egg and fly back to A your scenario fits perfectly. But when you strt fighting the plane, be it BFM or high altitude BVR jousts or SEAD SAM wrestling, you are on the edge of the envelope, and excursions need to be modelled as those are an essential part of what you’re doing. Simulating a fighter jet you need to have great modelling of post-stall behaviour.
I mean, it’d be nice to have, but “close enough” is as about as good as you’re going to get.
There’s a reason real simulators aren’t used for BFM - it doesn’t get it quite right. It’s good enough to teach the sight pictures, setups, etc, but isn’t good for much else past the initial fight’s on.
Out of control functions tend to be heavily scripted into the sim to get the desired effect - Max A/B to this attitude until this altitude, idle one throttle, max the other, full rudder one direction, wiggle your nose, and bam you get a departure into an inverted spin.
If the guys with all the data on control surface movements and logic can’t get it completely right, you’re not going to get a fully accurate representation in a video game. At the end of the day, the best you can really hope for is a believable representation and I think DCS has done that for most modules.
I absolutely agree. The avionics and EW modeling works very well to simulate systems from this area (well as we know them to be from unclassified data).
I dont think you understood my point. With a 3rd and even 4th gen fighter they still fly and fight in mostly classic regimes. 5th gen are SPECIFICALLY DESIGNED to work and be controllable post stall, edge and past the envelope. My point was that as this is such a large part of the aircrafts flight regime we would find it almost impossible to get correct. Yes, I agree that we could probably take off, pootle around and land fairly realistically but as soon as you start pulling and pushing you will be outside what the sim can model.
I might know the physical form of the plane down to the millimiter but the FCS behaviour (canards and all) is not something you just “get”. No matter what you can deduce from the physical shape or power of the engines.
The FCS is a big big part of the virtual representation of the plane and I happen to know something about the pain of the EFA FCS representation in a virtual environment.
Unless I misunderstood something you said I still agree that the systems can be more acurate than the Flight Model of this particular bird.
I suspect the best we can hope for is relative performance vs absolute. So as long as we’re not comparing to the AI MiG-15.
In other words, sort of like the Gazelle with its FM that’s mostly correct so as long as you follow the real procedures, but goes off the rails as soon as you start doing unconventional stuff.
I totally agree. Its the best we can expect and hope for. My original point that the systems modeling will be easier to replicate than an accurate flight model still stands.
Having done both on a simplified scale… Systems modeling is tedious, with lots of research going into it, but is mostly just a lot of linear work. Flight modeling, by contrast, is usually a trial and error sort of deal and the initial upfront work isn’t high, but the tweaking work can take the entire duration of development.
Pick your poison because either way, it’s hard work!
