Skip to Main content Skip to Navigation

Modélisation multiphysique, reconception et optimisation d’une motopompe à rotor noyé

Abstract : Canned motor pumps are one of the safest ways to pump dangerous (radioactive, toxic), expensive and volatile fluids in nuclear and chemical fields. The canned motor pump is a compact unit integrating a hydraulic part and an induction motor with a common shaft. The special feature of the motor is that two non-magnetic cylindrical tube called “the can” are inserted into the air gap to seal the rotor from the stator windings. The pumped fluid passes through the air gap to cool the motor and lubricates the bearings. The new European standards on the motor efficiency (IEC standards), the increase in the service life of nuclear power plants in France, as well as the increased safety of the nuclear power plant installations after the Fukushima catastrophe are the reasons why canned motor pumps have to be designed with better efficiency and service life while fulfilling safety functions. The purpose of this PhD thesis is to develop a multiphysical model combining the electrical and thermal parts of canned motor to determine the efficiency and service life of canned motor. The reliability of the established multi-physical model has been validated by a comparison with the experimental test results carried out on different instrumented prototype motors. An analysis of the influence of parameters that determine the electrical and thermal performances of the canned motor was carried out and indications on the redesign of canned motor pumps were formulated to have motors with better efficiency and longer service life.
Complete list of metadata
Contributor : Thèses UL Connect in order to contact the contributor
Submitted on : Friday, October 23, 2020 - 3:24:56 PM
Last modification on : Friday, August 5, 2022 - 2:54:00 PM


  • HAL Id : tel-02976714, version 1


Amadou Tinni. Modélisation multiphysique, reconception et optimisation d’une motopompe à rotor noyé. Mécanique des matériaux [physics.class-ph]. Université de Lorraine, 2020. Français. ⟨NNT : 2020LORR0103⟩. ⟨tel-02976714⟩



Record views