Performance prediction and multi-objective optimization of metal seals in roller cone bits

Abstract

Metal seals are the critical component of roller cone bits in oil and gas drilling. The performance of metal seals directly impacts the service life and efficiency of the entire drilling system. How to accurately predict the sealing performance of metal seals with multi-field coupling and optimize the overall sealing performance under the influence of various factors are the main difficulties of this study. A novel thermal-fluid-solid coupling numerical model was established and verified for the new-generation single energizer metal seals (SEMS2). The fluid film characteristics, interface temperatures, and sealing performance of SEMS2 were investigated under different operating conditions, structural parameters, and material parameters by the co-simulation of MATLAB and ANSYS. Based on the coupling analysis results, the significance analysis of the parameters was performed by using the orthogonal test. Furthermore, the multi-objective optimization of SEMS2 was carried out using the Back Propagation Neural Network (BP NN) and three typical multi-objective evolutionary algorithms (MOEAs). By comparison, the elitist non-dominated sorting genetic algorithm (NSGA-II) performed better than the strength Pareto evolutionary algorithm (SPEA2) and region-based Pareto Envelope based Selection Algorithm (PESA-II). The leakage rate and frictional force of SEMS2 were reduced by 52.3 % and 11.0 %, respectively, after multi-objective optimization. The results could provide theoretical support for designing a high-efficiency and long-life sealing system for onshore and offshore drill bits.