Structural Health Monitoring of Nuclear Site Facilities Using Optimal Sensor Placement for Damage Detection and Prediction of Failure

Abstract

The structural components of nuclear power plants (NPPs) and their support systems are built to endure a variety of thermal, vibrational, and fatigue loads. These structures are vulnerable to damage from unanticipated events, even with rigorous design considerations. There is a real risk of nuclear radiation leaks, infrastructure destruction, and human casualties. Continuous structural health monitoring of such structures guarantees safety and effective maintenance procedures, preventing disastrous consequences. The use of finite element analysis (FEA) offers a useful perspective on the behavior of structures when they are in use. With fewer sensors, optimal sensor placement (OSP) offers crucial and economical structural insight. Using a reduced order FE model for complete field estimation of the structural behavior is possible when using the OSP approach via FEA. A System Equivalent Reduction Expansion Process (SEREP) and a D-Optimal approach have been employed in tandem for sensor placement and complete field estimation of the measured, respectively. Three test scenarios, where simple structural elements like plates and beams with different boundary conditions are subjected to impact and single point harmonic loads are employed. It has been demonstrated that when strain data is gathered at the optimal sensor locations, it can be used to recreate fairly accurately structural dynamic response when seen at various time instances. The complete field strain/displacement fields for the structures made it possible to pinpoint the location and magnitude of the load. Force prediction error was recorded as low as 11% for a cantilever beam and 13% in case of all sides clamped plate. The problem formulation through this work gives an advantage to utilize the FEA environment provided by COMSOL and using the solution file to extract data in the MATLAB LiveLink environment. Also, execution of D-Optimal function in MATLAB and custom sensor placement studies along with the detection of force through SEREP expansion was the ultimate goal. In order to provide a thorough and in-depth monitoring of NPP structures and to prevent damages and failures due to unidentified cause(s), it is suggested that the above technique be applied to structures like reactors, boilers, and pipes in nuclear facilities.