Low-speed clutch modulation at idle feels too abrupt (hard to hold stable walking-speed without throttle)

Simulation & Modelling
1

Body:

Hi BeamNG.tech team,
I’m building a driver-training simulator and I’m testing low-speed maneuvers (reverse parking, tight turns). I’m trying to control vehicle speed at idle using clutch modulation only (no throttle), but the usable “bite/transition zone” feels very small and the vehicle behavior is quite abrupt.

What I expect (real car behavior):

  • In 1st gear / reverse at idle, gradually releasing the clutch should allow a wider, more controllable low-speed range (stable walking-speed).

Question:

  1. Is there a recommended way to achieve smoother idle-speed low-speed control via clutch modulation (for training use)?

  2. Is this mainly controlled by driver assists / transmission logic, or per-vehicle clutch/idle controller parameters?

  3. If tunable, which parameters/workflow should I look at (global vs per-vehicle)?

Thanks!

2

Hi @Jcwzstudio,
Similarly to the real world, the clutch response is dependent on the vehicle’s engine torque, gear ratios and whether it is a commuter car, sportscar, racecar, etc. What vehicle are you using for your testing?

If you go into Settings → Gameplay → Driving assistants, we have a few checkboxes that you can turn off if you haven’t already. That would make sure that nothing interferes with the clutch and throttle control.

Additionally, you can tune the response of the clutch pedal globally. The bite point of our cars often happens to be at a high clutch percentage input. If you go into Settings → Controls → Vehicle → Clutch, then there should be an axis input to click on. You can see a graph that displays the clutch input registered by the game, compared to the actual percentage of your pedal. You can experiment with the linearity and deadzone settings to see what suits your setup best. We do not officially recommend specific settings as they are different for everyone’s hardware setup and preferences, but I usually add some deadzone on each side and leave the linearity at 1.

Let me know how it goes and if you have additional questions.

3

Hi BeamNG.tech team,

I’d like to clarify the exact clutch behavior we’re trying to achieve and what we are currently observing in BeamNG.

Goal (real-world behavior we want):
When starting in 1st gear at idle (no throttle), a real manual car usually has a reasonably wide slip/modulation range (the “bite point” region) where the driver can smoothly control vehicle speed from 0 up to the idle-in-1st speed by holding the clutch in the slip region. This is important for driver-training scenarios (e.g., low-speed maneuvering / parking / exam training).

What we observe in BeamNG (both mod and official vehicles):

  • In 1st gear at idle, once the clutch passes a certain point, the vehicle speed quickly “snaps” to a nearly fixed idle-in-1st speed (e.g., ~1–6 km/h depending on the vehicle).

  • For a large portion of clutch travel, the speed stays near that fixed idle speed, and only in a very narrow portion of clutch travel near the end does the speed become meaningfully adjustable between 0 and the idle-in-1st speed.

  • This makes the “controllable half-clutch band” much narrower than in real vehicles.

What we already tried:

  • Disabled driving assists (clutch/throttle/gearbox assists) to ensure no interference.

  • Adjusted clutch axis settings (linearity, deadzones) in BeamNG. This can shift feel slightly, but does not create a wider modulation band.

  • The behavior appears similar across multiple official BeamNG vehicles, not only this mod vehicle.

Questions:

  1. Is this narrow modulation band expected with the current clutch/Drivetrain model?

  2. Are there any supported JBeam parameters (e.g., frictionClutch engagement characteristics, torque capacity, pressure-plate response, engagement curve, etc.) that can widen the stable slip region and make speed modulation possible over a larger clutch travel range?

  3. If there is no direct engagement curve parameter, do you have recommended approaches for driver-training use cases where we need a broader, more “realistic” half-clutch control range?

Thanks a lot — happy to provide a short recording showing the clutch % vs vehicle speed behavior if needed.

Best regards,
JcwzStudio