Using Laser-Doppler Vibrometry for the Optimization of Air Flow in the Read/Write Unit of Hard Disk Drives

Laser vibrometers are well-established and dependable tools for dynamic tests on data storage media such as hard disk drives (HDD) or DVD drives. Inside HDDs, the read/write heads are gliding on a type of air bearing that is susceptible to introducing unwanted vibrations to the system under particular conditions. Both Single-point and Scanning Vibrometers from Polytec were employed to chart the stable and unstable flying behavior of the heads inside a disk drive through a glass window that had been fixed onto the housing of the HDD.

An air bearing design was examined, focusing on air flow-induced vibration (FIV) inside hard disk drives. A series of HDDs were built with several varying airflow designs that resulted in stable or unstable air bearing flying behavior. This study was used to identify arrangements that reduced airflow disturbances around the suspension and integrated lead system (ILS).

The points of maximum amplitude resonance were examined at leading and trailing edges (roll and pitch sensitivity) of the head and suspension. Numerous HDD components (called air blocks) were altered, which directly affected the airflow disturbance around the head/suspension and ILS. Single-point and scanning (2-D) Polytec LDVs were put in place to track the stable and unstable flying behavior of the heads inside an HDD through a glass window.

Several techniques designed to dampen the strongest vibration levels. Additionally, Computational fluid dynamics (CFD) modeling was utilized to identify the FIV across different features of the suspension system.

Figure 1. Window in HDD cover.

Measurements

The front cover of the drive was adapted with a clear glass plate detailed in Figure 1 to allow a laser beam to be projected onto the top head in the drive.

The drive was positioned in an LDV laser microscope system with a laser beam projected through the microscope’s optics and then through the drive’s front cover via the glass window. The reflected beam then sent the signal back again. A strong resonance signal band of low frequencies was revealed on the trailing edge of the slider with the laser beam reflecting through the suspension and focused onto one of the two trailing edge corners of the head shown in Figure 2.

Figure 2. LDV on trailing edge corner (outside diameter).

These modulation signals were recognized as some of the vibration modes of the ILS and the suspension. It was deduced that the resonance signal spreads from the ILS, through the head’s bonding pads, and directly to the trailing edge of the head. Work carried out to compare these results has also progressed with a control file with a different airborne suspension (ABS) and airflow features to determine whether the same resonance mode signal is present in this file.

This article will detail a range of techniques to dampen the main resonance modes by altering or eliminating airflow components without changing the ABS design.

Results and Discussion

Before the LDV measurements were carried out, the file was put through magnetic testing in which the file was tracked by a magnetic signal Wallace equation analysis. The analysis gathered the signal’s non-repeatable runout (NRRO) or spectral coherence. The results are detailed in Figure 3, indicating there were strong signals from the head resonance. It was determined that the LDV measurement correlated with the NRRO data.

Figure 3. Band of head resonance.

It was supposed that new airflow components were the source of an unwanted airflow that forced the suspension and ILS to be modulated. The control HDD didn’t possess these components or the problems with the strong head or ILS resonance. The chief airflow component was a long airflow block found downstream of the head/suspension/actuator assembly.

Determining and eradicating the block produced changes to the head resonance band. Figure 4 details an HDD possessing two types of airflow blocks of both short and long lengths. These two airflow blocks were compared to no airflow block in the HDD.

Figure 4. Short and long air blocks.

Figure 5 illustrates the LDV results of this study with different blocks. In this plot, a comparison between the head resonance for short and long airflow blocks and no block is shown. The short block added to the maximum head resonance of 1,000 pm, and the long block weakened the resonance to 300 pm.

Figure 5. Air-flow block‘s effect on resonance.

Regarding no block, the head resonance was found to be 650 pm. It is theorized that the short block was the source of turbulent airflow affecting the suspension system, resulting in oscillation. The long block formed a large, upstream wake showing low turbulence. This data makes it clear that airflow blocks add, to a certain extent, to variations in the conditions of head resonance. The size of the blocks can play a part in other airflow effects, like lateral vibrations of the head that lead to track misregistration (TMR).

Conclusion

Unstable FIV resonance modes were measured with an LDV instrument that revealed the head was vulnerable to a band of vibration when the ILS and suspension were disturbed by airflow turbulence. It was concluded that HDD airflow components, in particular airflow blocks, were influencing the airflow around the ILS/suspension system, resulting in vibrations to be conveyed into the head. The CFD analysis included a finite volume model of the ILS polyimide structure that illustrated vertical excitations by the skewed airflow.

Figure 6. Symmetric ILS Polyimide

This information has been sourced, reviewed and adapted from materials provided by Polytec.

For more information on this source, please visit Polytec.