FLCS gain when airspeed over 400kts & (ALT FLAPS in EXTEND / AIR REFUEL in OPEN)
-
seems their FLCS is based on NASA TP1538.
Well, does not feel even close to a Viper, at least currently.
-
Well, does not feel even close to a Viper, at least currently.
To me it seems like they somewhat “designed” the control law especially for CAT III and TO/LND Gains using descriptions and figures from the flight manual, which is not a proper engineering source anyways. That may explain all the issues in those areas where NASA FLCS does not cover. (i.e. pitch rate in TO/LND Gains being 1/2 of what it should.)
Not to mention the fixed gains and gain schedules that are not designed for high speed regions, and can cause control surface fluctuations and oscillations at high speeds, which it does in DCS. There may be other issues like the pitch integrator etc. etc…
-
What I’ve found is that how a sim “feels” wrt to the actual jet is mostly a matter of the controllers used - and I’d been told that some number of years ago by a pro operator I know. I’ve been getting a LOT of time in RL Trainers over the last few years, and also got an chance to fly the actual jet - which has completely changed my approach to how I handle the Trainer; it’s made me far more aware of a lot of things - like take away the forces of Gs on your body and I can say that NO sim feels “like the jet”. And that Trainers also do a bunch of stuff peculiar to Trainers…the jet was much easier to fly.
Most game controllers are really pretty poor models compared to the actual aircraft input system - chiefly in that they are excessively light. This in and of itself screws up any math in the gains by allowing for non-real inputs and input rates. Rates mostly, I’d gather. And no two of our setups are really alike, so comparisons there aren’t really apples-to-apples either.
-
Most game controllers are really pretty poor models compared to the actual aircraft input system - chiefly in that they are excessively light. This in and of itself screws up any math in the gains by allowing for non-real inputs and input rates. Rates mostly, I’d gather. And no two of our setups are really alike, so comparisons there aren’t really apples-to-apples either.
That’s why I started building this:
http://hc625ma.org/wiki/doku.php?id=start#where_do_i_start
For the stick, it uses 4 springs, I think they will be 8kg each, though I’m not sure since I’ll have to make them out of some others (turns out, you can’t just go and buy a spring with arbitrary parameters). Long lever arm will still mean the pull won’t be quite that heavy, but should feel like something closer to reality. I might try to install a stick shaker at a later date, too. Of course, it won’t feel like the Viper, but it might just feel a bit more like the Hornet than my CH Fighterstick.In fact, this thread was motivated by inadequacy of available controllers:
https://www.benchmarksims.org/forum/showthread.php?39274-Idea-Open-source-HOTAS
I was thinking more of price when I made it, but there’s no reason a DIY HOTAS shouldn’t be more accurate to the real thing, as well. In fact, Simchair tries to do just that, though it’s helicopter focused. I think that with a little bit of invention and elbow grease (and a whole lot of 3D printer filament ), it’s perfectly possible to leave commercial hardware in the dust when it comes to accuracy. -
That’s a real nice looking rig there, Dragon1-1! Well done.
Some of the finer aspects of doing something like this for a fighter is that the control forces may vary with G - I know that the spec control force for a P-51 stick was/is 25 lbf/G and that seems to have become a standard…so at 4G your stick force should be 100 lbs. This seems like quite a bit until you actually get to fly a fighter - bearing in mind that under 4G, the amount of force you actually end up applying to the stick is assisted by the G force you are under itself. The Trainers I’ve flown model this, but I have no idea how they actually do this (magnetic damping?..maybe)…but I do know that after two hours of flying one I’m about exhausted…which is how it should be. And if you pay attention, you can note the degree to which getting tired out can affect your flying…which is something I also find fascinating - there are a TON of very minor things that can happen that have nothing to do with the model and more to do with YOU that can trip you up enough to get waxed by impacting or limiting your performance.
A Viper is probably the easiest of sims to get a good implementation of forces in - just start with a force-stick and calibrate it properly, and add the proper cockpit geometry. I now consider my CH stick to be about the worst of the lot, given that I can about move it with a feather…it’s WAY too fast to even be considered for “simulating” anything.
-
That’s a real nice looking rig there, Dragon1-1! Well done.
Well, that’s just an older version of the rig made by someone else. Mine will look different, I’m still waiting for the bearings to arrive (I do have a massive 8-lever TQS just about done, though). Will post pics when they do.
I’ve been thinking about replacing the springs with pneumatic cylinders later on, and hooking them up to an Arduino-controlled air supply. This would allow quite a bit of force feedback to be put in, though pneumatics won’t do high frequency vibrations (stick shaker would still be relevant). This would also allow almost 100% realistic force trim in helos. The problem is figuring out how to actually drive the things from the sim, and dealing with the inevitable noise. That, and I’m pretty sure the plastic parts can’t take a 100lbf load. This is another issue with simulating realistic forces, even with my 8kg or so springs the gimbal is mostly steel and a lot of mighty thick plastic.
The Viper seems easy until you get to the “get a force stick” part. All solutions to that are really expensive, I think the only options here are either the RealSimulator base with a Cougar or Warthog grip or a stick pulled out of a real Viper. The only “cheap” force sensing stick is nothing like the Viper’s.
-
I realized when I first caught the fever about 5 years ago that I was going to put as much as a nice car in $$$ into my project before I even started it…so I’ve been taking the “time is $$$” approach…hence the 5 years. I understand well that “easy” = $$$. And that nothing is easy.
Problem with pneumatics is just as you say - noise and response time. Plus the spring constant that arises due to comprehensibility of the gas (unless you’re also planning to use that to advantage)…which bring you to using hydraulics, and yet another set of issues. I have a feeling that electro-magnetics solves a lot of these the most effectively…probably everything but the $$$ factor.
I had some developmental experience as a virtual “test pilot” for the T-45A Trainer…they have a motion seat that is a combination of mechanical shakers and pneumatic cushions. Which I thought was cool, until they did some developmental updates and the thing started beating the CRAP out of me when I tried to stall the airplane (which is about impossible unless you know just how to get it to enter a stall)…which is what the jet does, apparently. I got used to it.
-
What I’ve found is that how a sim “feels” wrt to the actual jet is mostly a matter of the controllers used - and I’d been told that some number of years ago by a pro operator I know. I’ve been getting a LOT of time in RL Trainers over the last few years, and also got an chance to fly the actual jet - which has completely changed my approach to how I handle the Trainer; it’s made me far more aware of a lot of things - like take away the forces of Gs on your body and I can say that NO sim feels “like the jet”. And that Trainers also do a bunch of stuff peculiar to Trainers…the jet was much easier to fly.
Most game controllers are really pretty poor models compared to the actual aircraft input system - chiefly in that they are excessively light. This in and of itself screws up any math in the gains by allowing for non-real inputs and input rates. Rates mostly, I’d gather. And no two of our setups are really alike, so comparisons there aren’t really apples-to-apples either.
Yes, this is most prominent when I’m used to my 5-year-old X65F which I set 16lbsf for roll axis and 22lbsf for pitch axis just to mimic the stick force in an F-16, as we know pulling 9G in an F-16 requires 25lbsf and a max roll command requires 16.7lbsf. Then I tested out a X52 stick when I visited a friend and suddenly it feels like I’m flying another plane.
There’s actually a guidance in MIL-STD-1797A detailing stick force requirements in an aircraft. Stick force per g can impact flying qualities and it shouldn’t be too light or too heavy. Here’s a pilot rating chart for different stick force in a T-33 aircraft. Level I indicates best flying qualities.
-
While I’m randomly reading the block diagram of F-16 DFLCS, I noticed that the FLCS will actually switch to pitch-rate feedback (blended with AOA > 10°) if dynamic pressure is below 600(?) psf, and either ALT FLAPS switch in EXTEND position or AIR REFUEL switch in OPEN position, and most weirdly, in Standby Gains. Here’s the source:
Closer look:
(F-16 Digital Flight Control System Functional Block Diagrams. Data obtained from Lt. Bruce Peet, F-16 System Program Office, Wright-Patterson OH.) quoted in DTIC ADA202599 and DTIC ADA189675.
Yes the text is very hard to recognise but it can be decoded as:
Whether it’s 600psf I’m not sure, but it’s the closest to 400kts airspeed. This switch genuinely switches between Nz feedback and pitch-rate feedback (blended with AOA > 10°) with a first order filter used as a fade-in function. Notice that Standby Gains requirement, which is not present in the HAF Dash-1 manual.
I wish anyone could check the CM-1 manual, but I’ll just quote the HAF CJ-1:
So my question is whether the above quoted statement stands true for F-16CJ and F-16CM, since I remembered in BMS that there’s no such (400kts) requirements.
first, you missed the BAR (line) above StandbyGain text , that means in those logical diagram **!**StandbyGain
That means that this conditions occurs if NOT in standbygain
The precise value you are looking might be 500 Psi = 370 knots , that is the value where the TEF is automatically fully retracted even with TEF switch EXTEND .
but that’s hard to tell
-
first, you missed the BAR (line) above StandbyGain text , that means in those logical diagram **!**StandbyGain
That means that this conditions occurs if NOT in standbygain
The precise value you are looking might be 500 Psi = 370 knots , that is the value where the TEF is automatically fully retracted even with TEF switch EXTEND .
but that’s hard to tell
Thanks a lot, so the diagram seems to agree with Dash-1.
The precise value I’m looking for is where FLCS switches back to Cruise Gains with Air Refuel Switch at OPEN. Dash-1 says it’s 400 knots so that’s why I suspect it’s 600psf. In BMS it seems that FLCS stays at Takeoff & Landing Gains forever when Air Refuel Switch at OPEN position, regardless of airspeed.
-
Thanks a lot, so the diagram seems to agree with Dash-1.
The precise value I’m looking for is where FLCS switches back to Cruise Gains with Air Refuel Switch at OPEN. Dash-1 says it’s 400 knots so that’s why I suspect it’s 600psf. In BMS it seems that FLCS stays at Takeoff & Landing Gains forever when Air Refuel Switch at OPEN position, regardless of airspeed.
BMS FLCS was developed on the Analog FLCS model , which has 99% of the same features than the Digital, the one you mention is part of the remaining 1%
The inverted anti spin of the digital is included in BMS but this 370 / 400 treshold was not.
about the 370 or 400 , what i know is that the TEF is 0 @ 370 kts with TEF extended…so it wouldnt make any sense to keep the FLCS in LG above…
this would not be the first time there is a mismatch between the real and the -1…
But in reality, keeping LG gain above 400 kts is not really a problem …the Analog FLCS didnt have this feature after all …
-
-
-