Data Storage Systems Center

Due to their small size (from 10s of millimeters to 100s of nanometers), testing the dynamics and vibrations behavior of miniature structures poses different challenge. High accuracy measurements for the entire structure must be made within a wide frequency bandwidth. This requires specific test apparatus to be designed, and cutting edge instrumentation to be utilized. Similarly, in terms of theoretical modeling, high fidelity reduced order (analytical) models are needed for rapid yet accurate modeling of dynamic behavior.

This presentation outlines experimental instrumentation used in Multiscale Manufacturing and Dynamics Laboratory for measurement of dynamic behavior of miniature structures with picometer resolution and within 20 MHz frequency bandwidth. Various sample vibration data for structures ranging from polymer nano-fibers, MEMS membranes, and meso-scale manufacturing tool are provided. A short description of high-fidelity reduced order dynamic modeling for beams in bending and torsion is then described. The experimental capabilities also include measurement of error motions (radial and axial) of rotating structures.

Actuator bias in disk drives is the steady-state force or torque needed to hold the head at a particular track. It is important to model for increasing seek performance. This project examines two bias phenomena. The first phenomenon is the change in the magnitude of the bias hysteresis loop, which is the separation between forward and reverse segments of the loop composed of sequential seeks of a single length (Fig. 1). The magnitude decreases with the length of the individual seeks. We hypothesize that the decrease is due to the change in lubrication regime from boundary lubrication to elastohydrodynamic lubrication, because longer seeks have higher peak velocities.
The second phenomenon is the transition from the steady state of one seek length to another. Fig. 2 shows repeated long seeks reaching a constant hysteresis value. The transition to short seeks occurs at the red marker, and the bias measured for short seeks drops initially. We hypothesize that high pressure causes lubricant glassing during the long seeks. Subsequent short seeks disrupt the glassy state, and result in lower bias measurements while lube is plentiful. The bias then approaches the separation level expected from the Fig. 1. When many short seeks precede a transition to long seeks in Fig. 3, we believe that the reason the bias initially increases, is the depleted lubricant supply during short seeks and the bearing must move far enough to entrain sufficient lubricant to reach the expected long seek bias level.

Problem: The head of hard disk makes an angle with the track dependent on the position on the platter. This head skew dictates the track width and so the capacity of the hard disk.

Approach: The skew problem can be solved by a in plane rotary MEMS actuator between the slider and the suspension. The actuator consist of multimorph beams which move the rotor due to temperature difference.

Design constrains: Rotation of at least 10 degrees Rotate fast enough Stiff enough to not cause extra movement of the head Shock resistance