PDA

View Full Version : Which motion cues to simulate ?



louisgr
10-11-2007, 06:32 PM
Hi guys,

Since it's my first time at this, I'd like your views on which motion cues to simulate. Obviously the more axis the better but I want to know if some cues are either not possible with 2 axis or don't add much to the realism.

At this point I have coded and partially tested the following cues. I have a 2 axis platform and I'm using a washout algorithm. I decide to move the platform on the basis of changes in some parameters. For example, a very small variation of the pitch will not move the platform at all but a moderate one will. The platform returns to center when the variation of a parameter becomes below a certain theshold.

Cues I simulate on the ground.
---------------------------------------------

(After each cues the platform returns back to center (washout))

- When the engine starts so does the vibration (I plan to use a bass shaker).

-Pushing the throttle forward pitches up the platform. If the throttle is above 90% and stays there for 7 seconds (arbitrary number) I assume it is a taking off so I force a washout the platform to prepare for takeoff rotation.

- using the brakes pitches down the platform (slightly).

-Turning left while taxiing rolls (slightly) the platform to the right and vice versa.

-Touching down on the ground (landing) will roll the platform left or right (slightly) depending on the attitude of the plane just before touching the ground.


Cues I simulate in the air.
-------------------------------------

<!--StartFragment -->After each cue, the platform returns to an offset position then to center. This idea comes from Roland at http://www.simprojects.nl/

- Pitching the aircraft makes the platform go up or down then washes out to an offset as mentioned above. Then, when the joystick moves the aircraft to the center the platform goes from the offset point to the center.


- Rolling the aircraft makes the platform go left or right then washes out to an offset as mentioned above.

I am sure that some of you simulate the above cues differently and I'd like to have your opinions on this.

Many thanks in advance.


Louis

IanH1960
10-14-2007, 07:32 AM
Hi Louis,

The question of how to derive motion cues is an interesting one. All I can do is let you know how I went about it when I wrote my cue software.

I went for what is primarily a force cueing approach, but with additional cues to simulate certain movements, ie I try to move the platform into positions which induce forces on the pilot that match (as best they can in a stationary platform) the actual forces acting on the pilot in the real flight. For a 3 DOF platform these are (in the aircraft's local axis system) fore/aft direction acceleration & weight related forces, the same in the lateral (sideways) direction and local vertical heave (Gforce) related acceleration induced forces.

On top of these I also add elements to cue pitch and roll rate - ie so that when the visual cue from the screen shows a rapid bank angle change say, then there is an element of the cue that induces a roll in the correct direction to match.

The acceleration induced force elements are calculated from the full acceleration vector of the aircraft which I split into forward, sideways and vertical components to match the current orientation of the aircraft in 3D space. The calculation also builds-in the effects of gravitational acceleration - ie pilot weight effects which produce fore/aft or sideways force components when the pilot is not sitting perfectly upright.

This might all sound a bit complicated but the great advantage is that all of the flight movements reported by the simulator are fed into the force cues for the platform via the acceleration vector, regardless of what the aircraft is doing or where it is. So the cues are provided by the same algorithm if the aircraft is on the ground taxing or in the air manoeuvring (or when crashing into the ground). And if the simulator's flight model is a good one the cue derivation should be too.

Some of the individual effects are smoothed and some are washed out - as you know it is necessary to adjust the cues to keep within the platform limits but it is also necessary to prevent false cues from coming in - eg a rapid pitch movement of the platform to cue a rapid increase in forward acceleration might introduce a feeling of pitch angle rate which doesn't match the visual pitch angle cue from the screen and so feel strange to the pilot. Here the accel effect is filtered to slow the rate it is applied so that the pilot doesn't notice the platform rotation. Similarly holding a bank angle in the platform introduces a sideways force on the pilot that doesn't match what is really happening in the flight and so might not feel right to the pilot. Here the bank angle is allowed to develop so the the pilot feels the aircraft rolling but it is then washed out to mid so that any false force cue is minimised. There are other examples.

I have found it helps to add-in some "manufactured" effects because the simulator doesn't provide good data on the actual movement. For example FS9 will quite often simulate flat-pan flying conditions - ie there is zero heave movement of the aircraft when in reality there probably will be quite a lot esp with light aircraft - so I have provided the means to add-in recorded flight movements which make flying feel a bit more realistic in these situations. Similarly there are runway effects that would have to be added-in if they are wanted because fs9 doesn't report them very well.

The need for some of this manufactured cue is probably sim-dependent. I'm trying to get my software running with X-Plane at the moment and there seems to be a deal more movement in their flight model.

There's some of this stuff on my site and the software manual says a bit about it too. But there does seem to be a number of approaches in use at the moment regarding cue development.

I should say the force cues work best when the pilot can't see his real movements - ie on enclosed platforms.

Ian

http://buggies.builtforfun.co.uk/Sim/index.html