Complete Newbie Here, I have a design question

I will start with I have 0 experience in this field. Just tried a $80k rig at my local simcade and fell in love with the idea of a motion sim. I have been reading the forums on the 6 DoF rigs with the homemade actuators and it seems very doable. My only issue is reddit has been saying for racing sims (mostly what I play) these rigs are too much and have too late of latency to have a genuine experience as it doesn't replicate G forces very well. My question is, has anybody built one like a gyroscope? Like a seat in a ring, in a ring, in a ring. Each ring their own axis. If you need G's the ring would spin. Am I too stoned or am I on to something? Would the ability to spin allow G force simulation or am I missing something big in the physics of centripetal forces?

We had that discussion about cardanic suspension (seat in a ring in a ring...) many times before. Newbies who don't know how a motion simulator actually works think this would be better than a commonly used stewart platform. But for almost all cases (cars, bikes, general aviation...) it's not. And here's why:

The friction coefficient for rubber on asphalt on a dry road is somewhere around 1. This means if the centrifugal force in a turn gets higher than the gravity then traction is lost. So the maximum inclination of the combined force vector can't go higher than 45° to the side. You can simulate this by tilting the seat 45° in the roll direction. A F1 car with very sticky and hot tires can go a bit higher but not much.

You can even simulate loopings or barrel rolls with a plane without full 360° rotations necessary. This is because the force pressing the pilot into the seat still points downward in the cockpit reference frame during those maneuvers. Only in aerobatic flight or in fighter jets you have enough thrust to go straight up or do a continous knife-edge or inverted flight. But if you want to simulate a fighter jet in a realistic way you also have to generate g-forces substantially over 1g. So you'd need a large centrifuge. And nobody except the NASA or air force hat the budget for that.

To create the illusion of high forces during turns or braking (car racing) it's much easier to use belt tensioners and helmet-pullers. The cases where a cardanic support gives any advantages are quite limited and not worth the trouble and costs, IMHO.

Thank you for your information and guidance. I have been going through the forum a lot (did not sleep at all last night in my excitement). You seem to be the legend on here. I do have another question; I had my mind set on the linear actuator design originally but then I went through Smitty's Rotary 6Dof rig, and I like the simplicity of the motors rotating rather than building a 6 actuators. I was curious as to if there was an advantage to one over the other rather than ease of link arm assembly and also cleanliness as the linear rigs seem a little cleaner with the floor space open on the bottom. Are there feedback advantages? Do the motors have to be more powerful?

Well, at the end it's all a question of money. Both systems, linear and rotary, can be built either cheap or with high quality components, which of course has a big influence on the quality of movement, precision, noise etc. you get.

Cheap gearboxed are not designed for high precision positioning but for driving loads mainly in one direction (wheel chairs, for example). For simulators you need motion with quickly changing direction and speeds so backlash and cogging can become a problem. Helical low-backlash planetary gears for servo drives are specially designed to overcome this but are a lot more expensive. Same for linear actuators. Precision ground ball screws run almost totally silent but nobody can afford them.

Positional accuracy is not the problem as the circular curve can be completely compensated in the software. But a changing "transmission ratio" (angular vs. linear movement) leads to a changing inertial load and affects tuning of the control loop which can cause inaccurate velocity and acceleration (vibrations, lag). This is not a problem with strong motors as you can make the lever arm long and limit the angle to a range where the non-linearity is managable (say +/- 40°).

I would also say that the complexity of both linear and rotary actuators are compareable in terms of number of moving parts. But with a rotary design you can delegate a big part of the work to the company that makes the gear. You pay a bit more but you need a lot less special parts. The crank arm and joints can be assembled in a DIY workshop without the need of a CNC machine.

What is your budget ?