The philosophy of FFB design.

This explanation is a more succinct and elegant distillation of things I have come to realize over years of sim driving. The ability to drive so intuitively via FFB signaling is the one aspect I prefer in AC over BeamNG. As is my wont, I pray the devs consider this philosophy when making incremental improvements to Beam in the years to come.

 

Not sure if he concluded anything detailed but my take is its the same reason i made a FFB mod. G-forces communicate more than road forces. Read an old paper on it where it was stated that, through a wheel, the g-forces should count for about 80% of the forces.

With my mod (not trying to promote it, its rather an expression for my thoughts) im trying to increase the side g-forces since ive always felt a lack of those in Beamng. Im also trying to use lateral and up/down g-forces to scale the overall ffb strength, in order to communicate the wheight of the vehicle. Its abit tough though since i dont have the actualy background data (or where to find it) so theres alot of trial and error.
In my head 80% sounds about right, and have that sprinkled with non g-forces. Then scaling the output force using the lateral and up/down g-forces should be a decent way to have a decent ffb. Ive tried scaling it so that less lateral g-force gives less ffb strength and more (say when braking) gives more strength (imagine braking and cornering yould like stronger rather than lighter ffb) but i might end up having it just add/scaling the strength up and not down, since theres applications where less lateral forces isnt optimal to be used for scaling ffb strength down.
Now, i dont know if some or all of the g-forces need a curve to them, can give quite an erratic behaviour otherwise so thats another part of all this.

 

I think the real key to what makes some FFB
designs so communicative MAY be the force being strongly linked to how much or how little weight/grip is on the front wheels, more than the lat G maybe.

What I know for sure is that when I’m bouncing down a rough gravel road the wheel on my truck is mostly still and calm and my body is shaking. Somehow when my body is still in your sim chair and the wheel is shaking the brain interprets as the same signal. I think this interchangeability is what makes good FFB intuitive.
--- Post updated ---
I’m having an epiphany/a suspicion about how games with informative FFB may do it.

Put aside lateral G, weight transfer, tire friction. It may be just this simple: Rotation. When you do something to impart more rotational force to the car steering torque increases and vis versa. That would cover everything.,Turn the wheel and torque increases until you reach understeer territories, where forces stop increasing and decline. Trail braking? Again the rotation imparted to the car increases so does steering resistance. Too much brake, decrees in rotation steering goes light. Oversteer? Strong kickback towards countersteer, your signal he rear end is light. I could be wrong, but this could be the one simple master value you need for informative FFB. Derived from the more complex physics interactions in the underlying game of course.

 

So TIL is that peak lateral force actually occurs AFTER self-aligning torque peaks, meaning that the steering force will have started to plateau/weaken slightly before the car is turning its absolute best. This is not a problem for real drivers because they can feel the rotational force they are creating, but is an example of why good FFB is *not* a 1 to 1 literal copy of what you’d feel through a real car’s wheel.
I’m still thinking over my idea sim wheel steering torque needs to be linked simply to rotational force being inflicted on the car. I haven’t seen any problem with it yet, let me know if you do.

 

If im not misunderstanding your point than the g-forces is actually that rotational force your talking about.
Ideally id see tire grip communicated from all axles/tires but that could draown the ffb in data. I wouldnt say just front tires is a good thing since then yould not know when the rear wheels give way.

 

Sorry for the rambling un-clarity.

Rotational force, or "yaw moment" as I think it is called, and lateral G are not the same. With a lot of grip you can be going around a skidpad at speed with relatively small yaw moment and high lateral G. On the other side you can floor a powerful RWD car from stationary and spin it almost in place, producing a lot of yaw moment and relatively little lateral g.

I'm pretty sure the only clue a steering wheel alone can give that the rear is getting light (unless you have some wheel vibration effect on tire slip, which also would be good to have*) is forceful countersteer.

Now this is all very tentative, but I'm wondering if certain games, mostly AC, make their very intuitive FFB by tying wheel torque strongly to the yawing force being imparted to the car. I think this one physical data point might cover much or all of what one wants to communicate with FFB.

Example: You are in a turn. You lightly trail brake, increasing front grip and thus yaw moment being imparted to the car. Thus steering torque at the wheel increases. But then you push the brake too hard, causing front tire lock/understeer and the yaw moment being imparted to the car decreases drastically, causing wheel forces to lighten drastically.

Or you pull the handbrake in a corner and yaw moment is suddenly very high, meaning torque and countersteering correction forces are suddenly very high.

*Having some consistently detectable wheel vibration on tire slip/lockup/scrub would by itself be a HUGE improvement to BeamNG FFB's usefulness, and I know that one can get that data from the game from looking at SimHub.