Stability of the oil-air boundary in fluid dynamic bearings of hard disk drives

Ferdinand Hendriks, Hitachi Global Storage Technologies

Modern hard disk drives rely on fluid dynamic bearings (FDB) in which the rotating disk pack "floats" on a thin film of oil. In MPI 2004 the motion of the oil-air interface (the boundary) was computed for the case of a herringbone groove bearing. We assumed that the boundary is compact and that its motion is described by the average velocity in the lubrication layer at the oil-air interface, without (much) regard for capillary effects and fluid inertia. This assumption works well in many Hele-Shaw flows and this approach also captures many of the attributes of actual FDBs. For example the MPI team successfully produced an expression for the mean boundary deflection of sinusoidal grooves and demonstrated the tendency of the mean boundary to form "fingers." What makes FDBs different from most Hele-Shaw flows is the high shear of the oil film, which causes the location of the oil-air interface (the contact line) at the rotor to be quite different from that at the stator.

1. Problem I: The problem I wish to pose is to investigate the actual shape of the boundary surface without using the averaging assumption. For convenience the actual shape of the bearing grooves is left free. Any shape will do as long as the bearing produces pressure.
   
   
2. Problem II: I am also interested in the stability of an annular capillary interface between a stationary inner cylinder and a rotating outer cylinder. I suspect that under certain conditions this flow can be unstable but would love to be proven wrong.