Disk logo

Talke Lab Head/Disk Interface Research


The future of computer technology lies not only in increasing the computing power but also in providing the maximum amount of information at the lowest cost. For these reasons, there is a large demand for high capacity storage devices with low access times. As of today, hard disk drives are the only high capacity storage devices that can meet these requirements at the lowest cost.

Figures 1, 2 and 3 show schematic views of the slider-disk interface. The binary information is written and read from the disk using a magnetic head. This head is situated at the trailing edge of the slider (see Figure 2) which is in turn supported above the disk by the suspension arm which moves the slider around the disk (see Figure 1). The most commonly used heads are the inductive type wherein the binary information (1s and 0s) on the disk is distinguished by measuring the induced voltage at the transition between two adjeccent bits in a given track (see Figure 3). In the past few years there has been a great interest in using magetoresistive (MR) heads. The MR heads detect the 1s and 0s by measuring the change in the resistance of the magnetic element. Thus with MR heads the readback voltage is independent of the rotational speed of the disk.
Head/Disk Interface
Head/disk schematic Head/disk schematic Head/disk schematic
Figure 1 Figure 2 Figure 3
Courtesy Ed Grochowski, IBM
The separation between the magnetic read/write element and the hard disk during operation, known as the flying height, is one of the important parameters that controls the performance and durability of a hard drive. In order to increase the recording density it is necessary to decrease the flying height so that the signal to noise ratio obtained from the read element is within an acceptable range. Ideally, zero spacing is preferred. However, zero spacing or contact recording would lead to higher friction and wear at the head-disk interface, hence degrading the performance of the hard drive.

Decreasing the flying height is not the only way to improve the performance of a hard disk drive. Better magnetic materials and signal processing allow more bits per inch (linear density) while better servo techniques and more sensitive heads allow more tracks per inch (track density). Figure 4 shows the increase in linear, track, and areal density over the last few years. Figure 5 shows how the overall size of the drives have decreased as the areal density has increased.
Head/Disk Historical Overview
Performance graph Form factor graph
Figure 4:
Performance
Figure 5:
Form Factors
Large version Large version
Courtesy Ed Grochowski, IBM
In present-day hard drives the flying height is between 20 and 60 nanometers (nm), which is of the same magnitude as the roughness of the disk. Hence there is partial contact between the head and the disk during the operation of the hard drive. In the partial or semi- contact regime an ideal head-disk interface would consist of a slider where 99.999% of the weight is supported by the air bearing force that is developed beneath the slider as the disk spins. In other words, the normal load on the disk due to the slider is only 0.001% of the total weight of the slider, resulting in lower friction and wear at the head-disk interface. However, control over the interactions between the head and the disk during flying is limited by tolerances in the slider and suspension geometry and the materials and processing parameters used in the industry.

Figures 6-9 show various types of sliders that are used in hard drives. The first two types of sliders (Figs. 6 and 7) have flying heights ranging from 60 to 120 nanometers. This flying height is much greater than the peak to valley surface roughness of typical disks. Hence these sliders operate in the flying or non- contact regime. Sliders shown in figure 8 and 9 operate in semi-contact regime. The major difference between the sets of sliders is the presence of a third pad at the trailing edge and shortening of the two side rails. Shortening of the side rails helps in reducing the flying height and the addition of third pad decreases the contact area between the slider and the disk. The sliders shown in figures 7 and 9 have a cavity at the leading edge which creates negative (below atmospheric) pressure in the cavity and which reduces the pitch of the slider. Another advantage of using negative pressure sliders is the uniformity in the flying height from the inner to the outer diameter of the disk.
Modern Slider Designs
Slider picture Slider picture Slider picture Slider picture
Figure 6:
Two-rail
taper-flat
Figure 7:
Shaped-rail
negative pressure
Figure 8:
Tripad
Figure 9:
Negative pressure 
tripad
Prof. Talke's group is involved in studying the various ways to improve the tribological and mechanical performance of the head-disk interface. Equal emphasis is given to conduct both numerical and experimental work in the field of mechanics and tribology of the head-disk interface.

Below are descriptions of some of the projects we are currently engaged in.


PROXIMITY RECORDING INVESTIGATIONS USING PHASE DEMODULATED INTERFEROMETRY AND ACOUSTIC EMISSION ANALYSIS

Researcher: Thomas McMillan, Graduate Student, AMES Dept.
Advisor: Frank E. Talke, Professor, CMRR & AMES Dept.

Hard drive manufactures are starting to incorporate proximity recording type sliders in their drives in order to achieve higher storage densities with inductive heads. The proximity recording slider is designed to maintain a small area near the read-write element in constant contact with the disk, and thus enabling smaller bit size and ultimately larger storage densities. This approach to increasing storage density puts considerable amount of strain on controlling wear at the slider-disk interface, because a slight variation in contact load and contact area could greatly affect the drive survivability.

Acoustic Emission has become a popular method to monitor slider-disk interactions, because the AE sensor detects the energy associated with slider-disk contact via stress waves originating from the slider. These stress waves are associated with the slider resonant frequencies, and it is apparent that the acoustic emission signal contains detailed information about the type of slider-disk interactions, and hence information about the impact behavior.

The impact behavior can also be studied with a Phase Demodulated Interferometer (PDI). The current PDI has been adapted to measure either pitch or spacing variation on a 50% size slider during "take-off" or during other transient events.

It is the intent of this research to investigate the proximity recording slider behavior with Acoustic Emission and Phase Demodulated Interferometry during "take-off" and during normal operating conditions. 


TRIBOLOGY OF CARBON COATED SLIDERS

Researcher: Sai S. Varanasi, Graduate Student, AMES Dept.
Collaborator: James L. Lauer, Professor
Advisor: Frank E. Talke, Professor, CMRR & AMES Dept.

Sliders coated with diamond-like carbon (DLC) are becoming a standard in the magnetic recording industry because of the improved tribological performance of the DLC coated slider-disk interface. DLC overcoats are chemically inert, hard and wear resistant and have low friction when sliding against themselves.

In this study the friction and wear behavior of DLC overcoated sliders is investigated as a function of DLC overcoat thickness using contact-start-stop and constant speed drag tests. A substantial improvement in the frictional behavior of the DLC coated slider-disk interface in comparison with uncoated slider-disk interface is observed. The dependence of the Raman signal on the DLC film thickness is studied. Raman spectra obtained from different locations on the air bearing surface are analyzed before and after wear testing to study wear of the DLC overcoat. Wear rate of the DLC overcoat has been found to be nonuniform along the air bearing surface. It is also found that Raman spectroscopy can be used to study changes in DLC film thickness as a function of wear if the overcoat thickness is below a critical thickness of 10 nm. Above the critical thickness Raman spectroscopy cannot detect small changes in the film thickness during wear. The intensity of the fluorescence spectra of the sliders is also a function of the DLC overcoat thickness. 


TRIBOLOGY ISSUES WITH PERFLUORO- POLYETHER/PHOSPHAZENE MIXED LUBRICANTS

Researcher: Vinod Sharma, Postdoctoral Researcher, CMRR
Collaborator: Dr. Quock Ng, Maxtor Corporation
Advisor: Frank E. Talke, Professor, CMRR and AMES Dept.

Higher storage densities in computer hard drives can be achieved by lowering the flying height between the recording head and the magnetic disk. Currently, a majority of disk drives use so-called "proximity recording tri-pad sliders" that make extended contact with the thin film disk. The increased tribological interaction between proximity recording slider and thin film disk has reestablished interest in alternative lubricants. A new class of lubricant, phosphazene, had shown to reduce the stiction of the head/disk interface during start-up of the disk drive and their lubricating property seemed to improve as the temperature and humidity are increased. Phosphazenes are nearly spherical molecules of about 1.2 nm diameter and they seem to protect the carbon overcoat by forming a single monolayer. Since the conventional lubricants are long chain polymers with open space between the chains, it seems advantageous to mix phosphazenes with conventional lubricants for better coverage of the lubricant on thin film disks.

This study investigates the tribological properties of the interface between tri-pad slider and nitrogenated carbon overcoated disk lubricated with a mixture of AM2001 and phosphazene (mixed lubricated disks). The tribological behavior of tri-pad sliders was explored by monitoring the time dependent frictional force and acoustic emission (AE) signal during ramp-up of the disk from rest to the operating speed. The frictional behavior of the tri-pad slider/mixed lubricated disk interface was studied by performing stiction tests with different dwell times and contact start stop (CSS) tests. Similar tests were also performed on disks lubricated with AM2001 and disks lubricated with phosphazene, to establish a reference for the comparison of tribological properties of mixed lubricated disks. 


CONTACT FORCE CALCULATIONS FOR TRIPAD SLIDERS

Researcher: Haesung Kwon, Postdoctoral Researcher, CMRR
Associate Researcher: Michael H. Wahl, Assistant Project Scientist
Advisor: Frank E. Talke, Professor, CMRR & AMES Dept.

Recently, a new class of sliders has gained popularity that operates at very low flying heights with part of the airbearing surface in continuous contact with the disk. One well-known representative of this class of so-called proximity sliders is the tri-pad design that is based on a modified two-rail taper-flat slider with shortened rails and a small third pad at the trailing edge center carrying the magnetic transducer. The typical tri-pad slider is designed to fly with high pitch in order to decrease the head/disk separation at the rear pad. However, lowering the rear pad increases the likelihood of asperity contacts. The existence of acoustic emission signals even at high linear velocities confirms that continuous asperity contacts occur between the third pad and the disk [1].

Previous numerical investigations of the tri-pad design [2] have revealed a strong dependence of minimum flying height on the edge blend of the third pad suggesting that even small wear rates could cause significant variations in the minimum flying height. Since the wear rate is directly proportional to the contact force, it is important to know the contact force during all flying conditions.

This study investigates the quasi-steady contact force under various operating conditions as a function of asperity height. The contact model assumes small, elastic deformations of asperities that are characterized by an empirical stiffness/spacing function [3]. The calculation of the contact force is based on the principle that a uniform distribution of asperity heights causes a quasi-steady contact force as a result of the temporal average of the individual contacts between asperities. It is further assumed that even at contact the airbearing is fully established since the spatial distribution of the asperities is usually quite sparse. The maximum effective asperity height hm is defined as the flying height at which the temporal average of asperity contacts is zero.

References
[1] Machcha, A., McMillan, T., Tang, W.T. and Talke, F.E., "The Tribology of Tri-Pad Sliders with Hydrogenated and Nitrogenated disks", submitted for pres. and publ. at the IEEE Intermag Conf., 1996.
[2] Stenberg, H., Wahl, M.H., and Talke, F.E., "Comparison of Numerical and Experimental Flying Characteristics of Tri-Pad Sliders", IEEE Transactions, Vol. 31, No. 6, pp. 2970- 2972, 1995.
[3] Oden, J.T. and Martins, J.A.C., "Models and Computational Methods for Dynamic Friction Phenomena", Computer Methods in Appl. Mech. and Eng., Vol. 52, pp. 527-634, 1985. 


SUSPENSION MODELLING OF STANDARD AND INTEGRATED SUSPENSIONS FOR PICO SLIDERS

Researcher: Joe Jen, Postdoctoral Researcher, CMRR
Advisor: Frank E. Talke, Professor, CMRR & AMES Dept.

As the track density of hard disk drives increase, more and more attention needs to be paid to the design of the suspension spring, since suspension resonances and in-plane suspension motions limit the track density that can be achieved. In recent years, the trend in suspension design has been toward smaller suspensions, and much research and development work is going on in the areas of suspension design, integration of electrical wires from the head on the suspension, and optimization of suspension design to reduce sway modes and other undesirable suspension resonances. Recently, a new integrated suspension has been proposed by Hutchinson featuring integrated leads, and a different design approach toward using a different gimbal structure in a rectangular form factor has been implemented by Fujitsu. Although the standard suspension, such as a 850 Hutchinson suspension, is commonly used in present day drives, it is of importance to know the advantages and disadvantages inherent in various designs to optimize the suspension design for future high density small form factor drives using pico sliders. Joe's work is directed towards the understanding and optimization of various trade-off in the design of suspension springs currently under design and evaluation in the industry.

The modeling of the standard 50 percent suspension, the new integrated Hutchinson suspension, as well as the new suspension design from Fujitsu is studied using finite element analysis, and a comparison of the dynamic performance of these suspensions will also be carried out on the basis of the models. In addition, frequency response measurements of these suspensions will also be done using a suspension resonance tester based on Laser Doppler Interferometry for the purpose of experimental verification. After comparing and evaluating the numerical and experimental results, we plan to investigate scaling of suspensions to accommodate 25 percent sliders in conjunction with proximity recording and negative pressure sliders.