Direct algorithm and its application to slider air bearing surface optimization

Summary form only given. Slider air bearing surface (ABS) designs that satisfy very strict multi-objective goals are of great importance for magnetic hard disk drives. Finding such optimal designs is a strongly nonlinear problem. In slider ABS optimization, the evaluation of the objective function for every single sample design takes substantial computation time. It is desirable to reduce the number of sample designs evaluated without losing the global property of the optimization algorithm. In this paper, we make use of the DIRECT (DIviding RECTangle) optimization technique, which is a deterministic global optimization technique developed by Jones et al. (1993) and which is used to find the minimum of a Lipschitz continuous function without knowing the Lipschitz constant.