Numerical Study of Particle Rebound on the Slider Air Bear-ing Surface

Higher areal density of hard disk requires the ultra-low flying height slider, which increases the probability of contact between the slider and hard disk surface. Micro and nanoscale lubricant droplets and metal particles produced in collision can enter the head-disk interface (HDI) and cause serious damage to the slider and hard disk [1, 2]. Our previous work has numerically investigated the air flow patterns, particle trajectories and interaction between particles and slider surface in HDI, which is of critical importance to the reduce of particle induced damage [3–6]. However, all previous work only considers the collision between the particle and flat surface. In this work, the mechanism of particle rebound in HDI is studied. Three types of particles collision are modeled and simulated, including the P-S (particle and surface), P-E (particle and edge) and P-P (particle and particle) collision. It is assumed that all surfaces of slider are smooth, and all particles are spherical. Collision of spherical particles with other objects is considered an elastic point contact. The subsequent action of particles after collision, which includes the rebound on the surface and adherence to the surface, is calculated with considering the collision energy loss. The trajectories and accumulation/distribution of particles on the air bearing surface (ABS) are visually presented. The particle material is alumina, and all Al
2
O
3
particles have a diameter of 150 nm, a density of 3800 kg$3800/\,\mathrm{kg} \mathrm {m}^{3}$, and an initial velocity of $( U_{p}, V_{p}, W_{p})=( 0.8 U_{d},- 0.2 U_{d},0)$, where $U_{d}$is the local hard disk rotating speed. One hundred particles are released one after another at the leading edge of a femto-sized slider $( 850 \times 700 \,\mu \mathrm {m})$. The initial $x$coordinates $( x_{0})$of particles equal zero. The initial $y$coordinates $( y_{0})$of particles are randomly generated in the range of $0 \mu \mathrm {m}$to $700 \mu \mathrm {m}$. The initial distance between particles and hard disk $( z_{0})$is randomly generated in the range of 150 nm to 250 nm. Firstly, it solves the modified Reynolds equations to get the slider attitude as shown in Fig. 1(a) [7]. Since the air pressure gradient is necessary for calculating the air flow velocity and easy to be calculated on the rectangular mesh, an algorithm is applied to achieve the data transformation between the triangular mesh and the rectangular mesh. Then, it solves the Navier-Stokes equations with the second order slip boundary conditions to get the air flow velocity pattern, as shown in Fig 1(b). Finally, particle movement equations are solved by using the fourth order Runge-Kutta method [4]. Some simulation results are shown in Fig. 2. The comparison of Fig. 1(b)and Fig. 2(a)indicates that particle trajectories basically follow air flow streamlines. Fig. 2(a)also shows trajectories and final distribution of randomly released particles. Sixty-five particles adhere to ABS, twenty particles straightly go through HDI without collision, and fifteen particles go through HDI after the rebound on ABS. Especially, particle trajectories show that particle P2 goes through HDI after a P-E collision, compared to the straightly passage of particle P1. Fig. 2(b)shows the P-E collision causes a large $z$-direction displacement of particle P2, which helps the particle P2 to move away from ABS and go through HDI finally. Correspondingly, a large negative $z -$direction velocity of particle P2 is shown in Fig. 2(c). The P-E collision of particle P2 also causes a greatly reduce of $x$-direction velocity as shown in Fig. 2(d). Dynamic performance of particle rebound and the statistical analysis of accumulation/distribution of particles on ABS will be furtherly presented in the full paper.