Hydrodynamic effects on modeling and control of a high temperature active magnetic bearing pump with a canned rotor

Embedding instrumentation and control (I&C) at the component level in nuclear power plants can improve component performance, lifetime, and resilience by optimizing operation, reducing the constraints on physical design, and providing on-board prognostics and diagnostics. However, there are difficulties associated with using embedded I&C in nuclear power plants (NPP). First is that regulatory constraints make deployment of new technologies difficult. This difficulty is partially alleviated by developing and refining these technologies for passively safe reactor designs. The second challenge is the extreme environments that some components that would benefit from embedded I&C operate in. In this paper, we will use a high temperature (700 °C) canned rotor pump as an ideal testbed for the development of embedded I&C that can operate in extreme environments. This pump utilizes active magnetic bearings (AMBs) to eliminate the need for mechanical bearings and shaft seals. Successfully utilizing embedded I&C in extreme environments requires developing a deep understanding of the system's physics and dynamics and using that knowledge to overcome material and performance limitations on actuators and sensors imposed by the environment. In this paper, we will develop a coupled dynamic model of a canned rotor pump that incorporates hydrodynamics, rotordynamics, and active magnetic bearing dynamics. Then we will compare two control design methods, one that uses a simplified decoupled model of the system and another that utilizes the full coupled system model. It will be seen that utilizing all the available knowledge of the system dynamics in the controller design yield an order of magnitude improvement in the magnitude of the magnetic bearing response to disturbances at the same level of control effort, a large reduction in the settling time of the system, and a smoother control action.