Steering Test Issues

Hello, I have attached a .blend of my dilemma, for some reason when the tire rotates past 360 on Y axis it goes nuts,

I think there are gimblelock issues now in 2.6

Attachments


TestDampen.blend (436 KB)

Help? I am not sure why it is doing this I will try in 2.59

… dilemma, gimblelock (sic)… give us a break. And this in a new thread.

What do you want to do?

Maybe you want to spin a wheel. The wheel spins up from a standstill at a rate that we might commonly see in real life. And maybe we can brake the wheel likewise. It might also be a wheel that appears to steer the vehicle. It might have camber and castor. And the wheel might be one of two or four or more that work differently depending upon whether they function to deliver power, steer and are children of a particular type of vehicle.

It’s all easily doable even without using bullet physics, rigid body joints or vehicle wrappers. These three are valuable for particular purposes.

Do youself a favor and get clear on the basics of being able to smoothly and incrementally rotate a cube object with the keyboard, mouse of joystick and then bring it to a smooth and controllable stop.

Dude. I have made vehicles. The problem is that the wheel randomly flips 180 on an axis that is locked… I am not an ubernewb

Eventually the shaking is so bad it lifts off into the air. Something wrong with 6dof download the
file and hold w

I recreated the scenario, and the problem is that it seems like you’re using the wrong type of joint for something like a vehicle.

Change the joint type to hinge, place the pivot inside the main vehicle mesh, and rotate the pivot 90 degrees on the Z-axis so the wheel can rotate forward.

Though it is possible this could also be a bug with the generic 6DOF joint, if the same setup works without problems in Blender 2.49 or the 2.5 betas, then you should submit a bug report.

Rotating the pivot points so rotation is around the x axis fix it in 6DOF.

But I believe it is a bug! Will investigate more.

I have many other vehicles rigged this way. I think it is a bug. Ill remake the scene in 2.59

It seams to be three bugs involved - From my test case:

  1. instability in rotation round Y axis.

The lower left example try to rotate the wheel around the constraints Y axis

  • when rotation reach 90 degree it goes mad. The green and orange stering
    objects are rotated so rotation is aroun Z and X axis - works fine. In the
    top left example I try to rotate the pivot instead…
  1. Rotating the pivot don’t work.

For all I can tell the pivot is not rotated.

  1. Max Torque on motors are wrong.

To isolate error source all the blue wheel is rotated by motor, the grey by
global Torque. The global torque is 1.0, the max in motors are 0.02 - Yet the
motors accelerate faster!?!?

And the test case: TestDampen.blend (764 KB) (2.60a - update: behave the same in 2.59)

Usage: w - spin wheel, a and d to turn stering.

Would be nice if someone check them - You tend to be blind for Your own code and it’s nice if bugreports have good quality - especially case 2 I wonder if I can have got something wrong.

To reduce error source I have also cleaned up the original file a bit, most important:

  1. Deleted some key frames
  2. Applied scale
  3. Made all objects rigid body
  4. Proper collision bounds
  5. Rotated the wheel so they can use cylinder bounds.

I also use python to build the constraint - did not fix the error.

Reported in bug: http://projects.blender.org/tracker/index.php?func=detail&aid=29643&group_id=9&atid=306

Thank you, from you info I fixed it, Do you think the Y axis wont spin right because its the tracking axis?

Here is the working project !!