Question Idea for more realistic 6DOF platform?

Most platforms(whatever DOF) I experienced or saw on the internet have a pivot point beneath the seat.

But here come with the problem of this setting. The actual G feeling comes later than it should be.


Let's take an example-
IRL: if you make a left turn, your body sucks into the right part of the seat and feels G instantly.

Sim: if you turn left, seat actually become further from your body, it takes time until your body fully sticks to the seat and lean to right.



So here's an idea: How about making pivot point higher from the platform?

In case pivot point is above your head, it may make the body feel in the same sequence of the real life.

For someone who can't understand what I'm saying, check below.


It's not a big deal to make 6DOF platform working this way, only needs some basic mathematics.



I don't know whether someone already tried or not. How do you guys think about this?

this can already be done with software - so the pivot is around your chest

there are plugins to make this work - see the hexapod software for example

platform pivot point should correspond to the center of gravity of the body, in seated driver it is around the navel .

there are simulators on the market that correctly implement this, for example http://www.evolve-motorsport.com/motion-simulator/

there must be a simple way to adjust the output knowing the dimensions of the sim and the distance to the COG

does anyone have any suggestions for a formula ?

i guess you have to ask the question does it really matter for a fast acting racing car sim compared to a flight sim

I think that best pivot point isn't necessary the center of gravity but the points where humans best perceive motion/acceleration/rotation...and these would be inner ear and gut in my opinion.

But there is also another thing to consider here.
If you move pivot point off the center (above or below upper joint plane), you lose range of motion of the platform for rotations around X and Y axis.

You can get around this to an extent by placing the seat not on the top of platform but rather lower it like in this link:

It is common to move the pivot points to the cog area

But you need a custom seat/frame to make it work

A Stewart platform has a flat base that let's you fit any seat / car / kart / motorbike / plane/ etc that you want

Ok, if we are talking about a flat base then there is no option for lowering the seat...

A suggestion for a formula question you asked in previous post then:
I myself am using quaternions for rotating specific points (platform top joints) around arbitrary pivot point.

A piece of example code (I am using QtCreator):

for (int i=0;i<6;i++)
{
...

QQuaternion point;
point.setScalar(2.0);
point.setX(param1);
point.setY(param2 - rotPointY);
point.setZ(-rotPointZ);

rotationX = QQuaternion::fromAxisAndAngle(1.0, 0.0, 0.0, pitchIn);
rotationY = QQuaternion::fromAxisAndAngle(0.0, 1.0, 0.0, rollIn);
rotationZ = QQuaternion::fromAxisAndAngle(0.0, 0.0, 1.0, yawIn);
rotation = rotationZ*rotationY*rotationX;

QQuaternion resultPoint = rotation*point*rotation.conjugated();

...​

Quick explanation of the code:
- I do a for loop for all six joints
- param1 and param2 are X and Y distances of a point(joint) to original pivot point.
- rotPointY and rotPointZ are offset positions of pivot point in relation to original pivot point.
- pitchIn, rollIn and yawIn are inputs
- after calculation resultPoint holds X,Y and Z values for specific joint

Is it that simple ! Should be easy to integrate into a plug in or the code to index the motors

Well, you will still need to calculate lengths of actuators based on these points (or crank arm rotations if you are not using linear actuators).
But considering the points themselves, it is that simple, yes.

The funny thing about quaternions is that they are really simple to use...
Understanding how they work on the other hand is quite difficult (something about complex numbers in 4 dimensions...)

Just to be clear - I don't understand it at all, but as long as it works it's OK for me...

does anyone have any idea how i can fix the data stream so it will compensate for the distance between the cor and cog (centre of rotation and centre of gravity)

I am building a version of the SMC3 that works with 6 or more actuators and can fix the data live on the fly

simtools outputs numbers for the positions of each actuator

a bit like this

[300],[400],[500],[600],[700],[800]

my sim is a 6dof using gearmotors and levers

some movements will have no effect

Heave - no effect
Surge - no effect
Yaw - no effect
Sway - no effect ?
roll - ??
pitch - ??

am i right with the axis manipulation ?

i may be able to detect if there is roll and pitch by examining if one axis is going up and one is going down


any suggestions greatly appreciated !

Hi Noorbeast - thanks for your help

for RF2 it looks like this

here is the tuning centre - I don't seem to be able to access the Axis Assignments whilst the game is running

Hi Noorbeast - everything is working as expected

what i want to do is compensate for the difference between COG and COR

(i can code this if i can work out how to do it)

@youbuddy - youbuddy do you have any suggestions how to fix/code this

regards