0
TECHNICAL PAPERS

# Experimental Study of the Flow in a Simulated Hard Disk Drive

[+] Author and Article Information
Charlotte Barbier, Eric Maslen

Department of Mechanical and Aerospace Engineering, School of Engineering and Applied Science,  University of Virginia, Charlottesville, VA 22904

Joseph A. C. Humphrey1

Department of Mechanical and Aerospace Engineering, School of Engineering and Applied Science,  University of Virginia, Charlottesville, VA 22904jach@virginia.edu

1

Corresponding author.

J. Fluids Eng 128(5), 1090-1100 (Mar 13, 2006) (11 pages) doi:10.1115/1.2236135 History: Received October 13, 2005; Revised March 13, 2006

## Abstract

Instantaneous circumferential and radial velocity components of the air flowing past a symmetrical pair of suspension/slider-units (SSUs) attached to an E-Block/arm were measured in a specially designed corotating disk apparatus simulating a hard disk drive (HDD) using the particle image velocimetry technique. The geometrical dimensions of the components in the apparatus test section were scaled up by a factor of two, approximately, relative to those of a nominal $312$ inch HDD. Most of the measurements were obtained on the interdisk midplane for two angular orientations of the arm/SSUs: (a) One with the tip of the SSUs near the hub supporting the disks; (b) another with the tip of the SSUs near the rims of the disks. Data obtained for disk rotational speeds ranging from $250$ to $3000rpm$ (corresponding to $1250$ to $15,000rpm$, approximately, in a $312$ inch HDD) were post-processed to yield mean and rms values of the two velocity components and of the associated shear stress, the mean axial vorticity, and the turbulence intensity (based on the two velocity components). At the locations investigated near the arm/SSUs, and for disk rotational speeds larger than $1500rpm$, the mean velocity components are found to be asymptotically independent of disk speed of rotation but their rms values appear to still be changing. At two locations $90$ and $29deg$, respectively, upstream of the arm/SSUs, the flow approaching this obstruction displays features that can be attributed to the three-dimensional wake generated by the obstruction. Also, between these two locations and depending on the angular orientation of the arm/SSUs, the effect of the obstruction is to induce a three-dimensional region of flow reversal adjacent to the hub. Notwithstanding, the characteristics of the flow immediately upstream and downstream of the arm/SSUs appear to be determined by local flow-structure interactions. Aside from their intrinsic fundamental value, the data serve to guide and test the development of turbulence models and numerical calculation procedures for predicting this complex class of confined rotating flows, and to inform the improved design of HDDs.

<>

## Figures

Figure 1

Side view of an unobstructed portion of the experimental test section with relevant geometrical variables defined. The top disk is made of glass and the bottom one of anodized aluminum. They are both clamped to a hub of radius Rh and the unit rotates about its central axis. The extra thickness of the bottom disk minimizes any tendency to wobble. The enclosure walls are made of Plexiglass. In this study: Rh=28.6mm, Rd=100mm, a=3mm, b=2mm, H=4.8mm.

Figure 2

Plan view of the experimental test section as seen through the flat top of the Plexiglass enclosure and the top glass disk. The sketch shows to the left the inner (IC) and to the right the outer (OC) E-block/arm/SSUs configurations investigated. The arm/SSUs is located between the two disks. The E-block fits into a cavity extending from the side wall of the enclosure. The cavity is constructed such that its back curved inner surface is an arc of a circle centered at O3 of radius 170mm and its opening spans 39.2deg for the IC and 50.5deg for the OC. For both orientations O2O3¯=128.4mm and O1O2¯=116mm. The hub and disks corotate counter-clockwise in this view.

Figure 3

Sketches and dimensions in mm of the corotating disks flow apparatus (top: Side view) and of the E-block, arm and SSUs (bottom: Top and side views). The E-block/arm/SSUs assembly is also shown to the left of the test section in the top figure. Three weight-saving holes are cut into the arm. The two angular orientations investigated for the E-block/arm/SSUs assembly are shown in Fig. 2.

Figure 4

Sketch defining the circumferential and radial velocity components with respect to cylindrical coordinate systems centered at O1 and O3, respectively. The asymptotic invariance of the flow is checked by reference to mean and rms velocity profiles of these two velocity components obtained at locations 1, 2, 3, 4, and 5 on the interdisk midplane.

Figure 5

(a) Mean and rms values of the radial and circumferential velocity components as a function of radial position and disk rpm at location 1 in Fig. 4 for the outer configuration in Fig. 2. Radius R is given in mm. (b) Mean and rms values of the radial and circumferential velocity components as a function of radial position and disk rpm at location 2 in Fig. 4 for the outer configuration in Fig. 2. Radius R is given in mm.

Figure 6

(a) Mean and rms values of the radial and circumferential velocity components as a function of radial position and disk rpm at location 1 in Fig. 4 for the inner configuration in Fig. 2. Radius R is given in mm. (b) Mean and rms values of the radial and circumferential velocity components as a function of radial position and disk rpm at location 2 in Fig. 4 for the inner configuration in Fig. 2. Radius R is given in mm. (c) Mean and rms values of the radial and circumferential velocity components as a function of radial position and disk rpm at location 3 in Fig. 4 for the inner configuration in Fig. 2. Radius R is given in mm.

Figure 7

Contours of mean velocity magnitude with streamlines superimposed for the inner configuration (left) and outer configuration (right) at 3000rpm

Figure 8

Contours of the mean axial component of vorticity (normal to disk surface) for the inner configuration (left) and outer configuration (right) at 3000rpm

Figure 9

Contours of the turbulence intensity (based on two velocity components) for the inner configuration (left) and outer configuration (right) at 3000rpm

Figure 10

Mean and rms values of the radial and circumferential velocity components as a function of radial position and disk rpm at location 4 in Fig. 4 for the inner configuration in Fig. 2. In this figure R*=(r−Rh)∕(Rd−Rh); see Fig. 1. The dash-dot line in the plot for Uθ* denotes solid body rotation.

Figure 11

Mean and rms values of the radial and circumferential velocity components as a function of radial position and disk rpm at location 5 in Fig. 4 for the inner configuration in Fig. 2. In this figure R*=(r−Rh)∕(Rd−Rh); see Fig. 1. The dash-dot line in the plot for Uθ* denotes solid body rotation.

Figure 12

⟨uruθ⟩* shear stress as a function of radial position and disk rpm at locations 4 (left) and 5 (right) in Fig. 4 for the inner configuration in Fig. 2. In this figure R*=(r−Rh)∕(Rd−Rh); see Fig. 1.

## Discussions

Some tools below are only available to our subscribers or users with an online account.

### Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related Proceedings Articles
Related eBook Content
Topic Collections