[Tutorial] How to enable Force Feedback in Beamng Experimental 0.3.6.7

We will need:
- Notepad, Notepad ++ or a similar program;
- Beamng Experimental 0.3.6.7
- The steering wheel with the support of FF;
- The good will.
We have to go to the game folder and then to the: \ Beamng.Drive \ lua \ vehicle. Let's find hydros.lua file. We can use a file or redo my original. Experiment with the settings. Remember my file is set up under the steering wheel Logitech Drivig Force GT. I wish you luck. Sharing your settings.

 

The more important question is, whats your first impression of it?

 

Fulfilled my dream. As a little too nervously steering wheel. Plus it's very easy to setup.

 

I figure this should be a good discussion thread for force feedback in general?

I played a bit today with it, with my Driving Force Pro. The Driving Force Pro is 900 deg and has the same body/motor as the Driving Force GT, though the electronics may be different.

I've noticed great steering behavior and bumps with the Grand Marshall, but it being completely dead with some other cars. I think I have found two reasons why:

1) I don't think the force from the hydro is being scaled depending on the weight of the car/steering ratio, thus lighter cars have lighter ffb. Just seeing a possible pattern here.

2) Steering geometry is not correct on all the cars. The GM has about 6 deg steering axis inclination (steering axis projected to xz plane), and 6 deg caster (steering axis projected to yz plane). I looked for actual Crown Victoria suspension data but was only able to find a circle track forum playing with other similar cars, and the numbers for SAI seemed to be more like 9 to 12 while caster is good. So basically, it's correct to real life. The scrub radius is another variable to look at but I haven't gone that far yet.

Increasing these values has a major effect on the feedback to the road in a real car, since road forces are primarily by wheel scrub radius, mechanical and pneumatic trail (basically, the moment arm created between the contact patch and steering axis at the ground). Well, it turns out that increasing caster and SAI also has a marked effect on ffb forces in the game! It increases the self steering behavior and sensation of the steering getting heavier at high speed, then going light when the tires lock up.

When looking at the Covet, it has only 1.3 deg SAI and 3.8 deg caster, and gives nearly no feedback. The internet tells me (found a pic of an alignment sheet) That SAI for an EF Civic or CRX should be 6.8 to 8.8 degrees. This is the type of suspension the Covet was based on.

Essentially we have a new layer of complexity to the game that is very sensitive to the tire model, suspension and steering geometry. For example, wheel shake gets transmitted through FFB. It was mostly a visual issue before, but now it's having consequences elsewhere. What is cool is that changes of just a few degrees in the game can actually be felt, big time!

Feature Suggestion Time!

Add a scale to represent the steering rack ratio, ie how much mechanical advantage the steering should have.

IRL some cars just have poor steering feel. For example my 60s Corolla has very low caster and SAI, and takes 4 turns lock to lock. But this is a game, where one can't feel g forces or other cues, and FFB is supposed to be helping us with that. So, introduce some factor to account for the variance in SAI and Camber etc, to give nice Porsche like feedback for all the cars

 

I have a question for you: I can not find the file hydros.lua ruler responsible for centering the steering wheel. There is an option?

EDIT: OK. I found.

 

I encourage you to share your configuration files Force Feedback. We are happy to test it.

 

I'm having a hard time with my Driving Force Pro getting rid of jitter. Like the wheel is trying too hard to maintain a position, catching and releasing rather than just exerting a force when, for example, I am countersteering out of a slide. I've ended up going the opposite way with a huge low freq coefficient, and almost nothing for a high freq coefficient, with lots of smoothing. But I'd be interesting to see what numbers you've arrived at given you have the same wheel.