How do I avoid mesh collider overlapping? (see image)

Hello, I´ve searched for days to solve this problem, but I can’t find the right solution. This is the situation: I have a dice prefab with a mesh collider attached. The mesh collider component has a different mesh applied with lower poly count and the option convex is checked.

The rigidbody component options are:

  • mass=1
  • drag&angular drag=0
  • use gravity=true
  • is kinematic=false
  • interpolate=none collision
  • detection=discrete

This is the code where I apply force on rolling the dice:

// cycle through dice instances
		foreach(GameObject dice in diceInstanceList){
			// check if current dice is in dice stack, if not roll it
			if(!stackHandler.Contains(dice)){
				// reset position
				dice.transform.position = diceSpawn.position;
				// reset rotation
				dice.transform.Rotate(rndRotation());
				// setup physics
				dice.rigidbody.isKinematic = false;
				dice.rigidbody.useGravity = true;
				dice.rigidbody.mass = GamePlaySettings.diceMass;
				dice.rigidbody.drag = GamePlaySettings.diceDrag;
				dice.rigidbody.angularDrag = GamePlaySettings.diceAngularDrag;
				dice.rigidbody.velocity = Vector3.zero;
				// show dice
				dice.renderer.enabled = true;
				// push dice towards the board
				dice.rigidbody.AddForce(transform.forward * GamePlaySettings.DiceForce);
			} // - is dice in stack?
		} // - foreach dice

Everything works fine but there are some overlapping issues. I roll (rigidbody.AddForce) 6 dice at the same time and the dice are overlapping each other. Look at the picture:

How do I avoid these issues? What am I doing wrong?

Thanks for your help and time!
Best, Arne

The overlap seems severe. Let me suggest two things to try. Go to:

Edit > Project Settings > Physics

Reduce or set to 0.0, Min Penetration For Penalty.

Go to:

Edit > project Settings > Time

Reduce ‘Fixed Timestep’. Try 0.01 to start.

Given how sever the overlap, I’m not convinced that either of these suggestions will address your problem, but it is worth a try.