Structural-Parametric Synthesis of Controlled Magnetorheological Seals

Abstract

This paper presents a structural-parametric synthesis methodology for magnetorheological fluid seals (MFS) based on multi-objective optimization. The two primary operational characteristics of MFS, namely, maximum retainable pressure and friction torque, are treated as conflicting objective functions, since geometric modifications that enhance sealing capacity simultaneously increase mechanical losses. The design problem is formulated as a mixed-integer multi-objective optimization in which the continuous design variables (tooth width and height, inter-tooth groove width, and radial clearance) are considered jointly with the number of magnetic field concentrators as a discrete integer variable. This formulation distinguishes the proposed approach from existing single-criterion and fixed-topology methods reported in the literature. The MOGA genetic algorithm was employed to generate the Pareto-optimal solution set. Analysis of the resulting Pareto front demonstrates that the radial clearance exerts the dominant influence on both objective functions, while the number of teeth determines the overall performance scale of the seal. Three representative Pareto-optimal configurations are examined, covering boundary and intermediate trade-off regions of the front. The obtained results confirm that the structural-parametric synthesis framework provides a rigorous basis for informed design decisions across a broad spectrum of operational requirements.