Originally Posted by safetypee
Simulators have tilt, but not acceleration...[e]xcept the really clever and expensive ones.........
The false climb illusion is predominantly triggered by acceleration.
To amplify somewhat, Newtonian mechanics is important here to understand what is going on physically.
I realise that this egg-sucking missive will be read by certain grannies who know how to already, but I crave indulgence, for there are obviously those who don't and this note is addressed to them.
From a given starting point and velocity, the integration (integral calculus) of acceleration over time gives a unique path through space. Starting from, say, (velocity vector) 140 kts Vapp and 3° downward FPA wrt Earth coordinates, and point-in-space X some 200 ft above a runway, there is a unique path described by the accelerations experienced by the aircraft.
To reproduce those accelerations, the sim starts from 0 kts (and therefore no need for a reference direction). And it ends up ...... in the same place (somewhere within the small spatial box which bounds its movement limits).
To get the exact path followed by the simulator if it is tracking the aircraft exactly, take the unique path followed by the aircraft, and subtract the starting velocity velocity. from the velocity vector at each point. It does not require much imagination (I hope!) to see that such a path will very quickly pass outside the bounding box of the sim.
Conclusion: the sim cannot reproduce the accelerations experienced by a real aircraft performing a real accelerated manoeuvre except over a short period of time sufficient to retain that manouevre within the bounding box. That is a short period of time!
That all follows from basic first-semester calculus.
Now for the accelerations. Suppose you are sitting more or less upright in an aircraft with a shallow nose-up pitch angle and constant velocity. Let's use pilot-coordinates as our coordinate system (same as aircraft coordinates, but for pilot). Your state, with your weight vector tilted 3° in the negative-x direction, can be veridically reproduced in the sim by tilting you to the same nose-up pitch angle. Suppose you are in a lightly-loaded aircraft and put in TOGA thrust, and pitch up to, let's say 14° (which takes a second or two, we hope). First, your mass vector is being tilted relative to you (as origin of coordinates) from 3° back to 14° along negative-x. Second, you are being accelerated in the postive-x direction by many kNewtons generated by those large motors. The sim can give you a bit of a shove to start with, but it can't move you out of the bounding box, so after giving you the shove, it has got to let up. Whereas those motors go on shoving.
Newtonian mechanics and unique path. The sim can trick your body into thinking it is being momentarily accelerated by doing so, by reproducing the acceleration starting from a different velocity vector. But then your mind must do the rest, keeping you thinking your body is being accelerated when in fact it is not.
If the sim designers have done their job, then your mind is going to be thinking, yes we are still accelerating heavily after 20 seconds. But you ain't. And the real thing is going to feel different. There will be considerably more force on the body in the real thing. After that initial shove, the sim is working entirely on your body weight, whereas the real scenario is body weight plus all those kN from the motors.
Similarly, everybody has "simulated" what it is like to be in a Formula 1 car at the start of a race, by driving away from traffic lights in the Mini. But it ain't the same thing. And you can't reproduce the feeling by replacing the Mini windshields with plasma screens, and tilting the driver's seat back as the Mini accelerates off and the screen views with them. Body weight plus a bit of PS does not equal body weight plus an order of magnitude more PS.
What you can try to do is catalog a list of motion illusions. That is what some aviation psychologists do. And then investigate motions in the sim that appear able to trigger those illusions in most pilots. And then program your sim to generate those motions. You may be successful. You may find ways to make your Mini give the impression to its driver that heshe is really in an F1 car. But it can't be the same. More applied force is *always* felt, unless you're dead and gone.
[BTW, this point is a different one from the discussion about whether sims can mimic aerodynamic behavior, say up to and beyond a fully-stalled wing, which aren't known in detail. Of course you can't program a machine to mimic a behavior whose progress you don't actually know. This seems to be a perennial sticking point also.]
PBL