Design and development of a high-speed motor for a vacuum pump

Abstract

Turbomolecular pumps require a motor to drive them at speeds of up to 90,000 rpm. These high-speed motors are typically brushless permanent magnet motors that are specifically designed for the application. Intellectual property rights over the design lock the company into one single supplier. Also low manufacturing volumes make the motor expensive to produce. To stay competitive continuous cost reductions or product improvements are necessary. This can only be achieved by looking at new materials, alternative manufacturing methods and simplified assembly processes. The aim of this project is to replace an existing laminated, high-speed (60,000rpm) motor with a new low-cost design. Special considerations need to be given to the motor design with regards to minimising losses due to the high operating speed and the fact that the motor operates in a vacuum. A machine design employing a simple, 3-tooth segmented stator made from soft magnetic composites (SMC) and using non-overlapping coils, and a `deep' plastic bonded magnet on the rotor is proposed to deliver low rotor losses and low manufacturing costs. Four SMC prototype motors have been built, which have led to the following discoveries: (1) Bulk eddy currents in the SMC material cannot be neglected and need to be taken into account separately as a function of the actual component size and shape. (2) A process is suggested to improve the iron loss calculation in SMC, which is evaluated against the prototypes built. (3) SMC material properties are adversely affected by prototype machining, leading to higher iron losses than initially expected. (4) The segmented SMC design has proven to be commercially very attractive. During testing a large sensitivity of the magnetically supported pump shaft to the inherent unbalanced magnetic pull (UMP) force of the 3-tooth, 2-pole design was discovered. This led to a practical and theoretical study into the effects of UMP in this application. An alternative design that avoided the inherent UMP was required, and a 6- tooth, 4-pole motor has been designed and built. For reasons of minimising risks laminations were chosen as the stator material rather than SMC. Test results of this motor in the pump have been successful and the motor has been selected to go into the next generation of turbomolecular pumps.