A Simplified Dynamic Model for the Analysis of the Slider Off-Track Motion Due to Head-Disk Interactions

Abstract

A five-degree-of-freedom model was developed for the analysis of the off-track motion of the magnetic head slider in a hard disk drive. The air bearing was integrated into the dynamic system by combining its stiffness and damping matrices with those of the suspension. Simulation was conducted for the slider in intermittent contact with circumferentially located bumps on a rotating magnetic disk. For a single bump, the excitation to the transverse displacement of the slider is close to that of an impulse. However, for multiple bumps in a sequence, the excitation gives an effect similar to that of a step force function. The maximum transverse displacement increases almost linearly with both the coefficient of friction and the skew angle. The average contact force is determined by the maximum contact force, the contact time ratio, and the shape factor of the contact force, which change with the bump spacing and the rotational speed of the disk. The steady-state transverse displacement is related to the average contact force. As the bump spacing decreases, the average contact force increases, resulting in a greater transverse displacement. Based on the system dynamic characteristics alone, changing the rotational speed of the disk has only a small impact on the average contact force and, thus, on the transverse displacement. At zero skew angle with the bump path close to a rear pad edge, significant transverse motion occurs because of the excited roll mode and the coupling between the roll angle and the transverse displacement. The off-track motion of the slider is dominated by the rotational mode of the actuator arm and the sway mode of the suspension, as verified by comparing the results of the transverse displacement from the 5DOF model to that from a 2DOF model of the transverse motions of the actuator arm and suspension. The effects of the roll angle on the transverse displacement through coupling were found to be responsible for the difference in the transverse displacements obtained from the two models.