Rigid body - pivoting about a fixed point
 

I'm trying to simulate a rigid body pivoting about a fixed point that is not the centre of mass.

http://i.imgur.com/W66m2.png

The rigid body is a hatch, hinged at the origin about the z axis. I’ve created the coord frame ‘hatch pivot’ at this location. I have water flowing up from the bottom of the domain causing the hatch to rotate upwards.

I’ve based my mass and moment of inertia on a density of 8000kg/m3
M = 50mm * 5mm * 10mm * 8000kg/m3 = 0.02 kg
Izz = ML^2/3 = 1.667e-5 kg/m2

When I leave the centre of mass as automatic, the hatch rotates about the origin as expected, but I’m concerned that the gravity is acting through the pivot point rather than the centroid, and providing no resistance to rotation.

When I specify the rigid body centre of mass initial condition as the centroid of the hatch (-25mm, 0, 0), the object seems to rotate about this point, rather than the origin.

I need to account for the weight force acting on the hatch, but the 6-DOF solver doesn't seem to allow for this. As a work-around I've tried to apply an external torque in the rigid body dynamics tab
Torque (Z component) = MgL/2 cos(theta_z)
= .02[kg]*g*0.025[m]*cos(rbstate(Euler Angle Z)@Hatch)

This works, but don't think it will handle the buoyancy correctly. I eventually need to adapt this to irregular shaped hatchs, different densities etc.

Can anyone provide some insight into getting the centre of mass working correctly?

Thanks,
David