Search

A hardware-in-the-loop model was developed to represent digital sensing and control of steam generator water-level. The model was created with an intention to serve as a component within a larger, distributed digital systems conceptual testing facility. In the present application, a software model simulates a nuclear pressurized water reactor core, and the core is cooled by a hardware and software model of a Westinghouse U-tube steam generator. The present application is configured using plant specifications consistent with “Plant X”.Software was written in C++¿¿, and hardware components include Phidgets and Measurement Computing digital input-output modules, proportional solenoid valves, a PC cooling pump, and gear pumps. This model assumes perfect implementation of sliding average-core-temperature control. During plant startup or normal operation, plant power precisely determines all steam flow characteristics. Liquid water simulates secondary coolant flow. Water pumped from a tank simulates steam, and recuperating feed water responds to subsequent level readings. Level readings as a function of plant power are measured and lie within the alarm-free region of the narrow range (+/-5% of the level set point). This design has been developed for incorporation within a distributed hardware/software component within a digital systems conceptual testing facility for digital systems testing by network-distributed control.

PROCTOR, WILLIAM CYRUS. Reactor Loose Part Damage Assessments on Steam Generator Tube Sheets. (Under the direction of Joseph Michael Doster). Damage from loose parts inside reactor systems can potentially cause integrity issues that jeopardize the operations of these facilities. Parts such as nuts, bolts, pins, sections of tubing and even hand tools are found inside the primary circuits of PWRs [Michel]. These parts carried by the coolant flow impact structures including the steam generator tube sheets and can cause significant damage leading to the unscheduled shut down of a facility. In this work we assess the behaviors of typical loose parts that may reside in the primary coolant system. Validations of scaled simulations are linked to previous experiments conducted by Shi [Shi]. Monte Carlo simulations of typical impact and energy distributions on a representative steam generator are analyzed and discussed. To obtain a more complete understanding of loose part damage caused to the tube sheet of PWR steam generators, CFD using the ANSYS CFX software package is used to compute detailed three dimensional flow fields within the steam generator inlet plenum. The flow field information is then input into a Monte Carlo program developed as part of this work to predict the trajectory of the loose part. Existing software packages lack the ability to track finite volume, finite mass particles. Additionally, there were no packages available that allowed for detailed manipulation of the collision physics necessary to accurately model impacts. The particle tracking program developed here then allows for the calculation of loose part impact locations and the energy imparted from loose part impacts with the tube sheet surface. Ultimately given this information along with the previous models developed by Shi, damage rates can be estimated aiding in the development of guidelines to improve the decision making process when loose parts are detected in the primary coolant system. As part of previous research, a 1:8 scaled model of the McGuire steam generator inlet plenum and tube sheet was constructed by Shi. This scaled steam generator tube sheet impact pattern experiment was run with two different types of hexagonal nuts and varied fluid inlet velocities. These experiments serve as a benchmark reference for development of the computational models in this work. Simulations of a full scale system similar to that of a Westinghouse Model D steam generator have also been performed. Detailed impact analysis is conducted as a function of coolant temperature, coolant inlet velocity, loose part type, shape, mass, density, initial starting location and initial kinetic energy. No a priori knowledge is assumed for the initial starting location and initial kinetic energy of the parts. Full scale results are compared to the scaled experiment to assess the validity of making predictions using only a scaled simulation.