Detection of uniform structural temperature distributions for a temperature based method of structural health monitoringh

Structural Health Monitoring aims to characterize the state of a structure based on some combinations of sensing technologies and analysis of acquired data. This process normally seeks to find changes in the mechanical response of a structure, primarily bridges in this work, based on strain or displacement data. Temperature acts as a distraction in this type of application of SHM, clouding the mechanical input-output behavior of the structure. Thermal effects on most structures, however, play too large of a role to be ignored or taken out of consideration. Combining the output strain and displacement data with input temperature data has the potential to robustly characterize a structure. To begin implementing this Temperature Driven approach, time periods with a uniform temperature distribution on the structure need to be identified. These time periods will reduce any effects of thermal gradients on the output strain and displacement measurement, leaving a much simpler input-output relationship. An initial algorithm was developed to recognize these uniform temperature distribution from years of data collection. This algorithm can modify the strictness of the required level of temperature uniformity, as perfect uniformity can be impossible to find in some cases. This algorithm has been validated on a real structure, the Streicker Bridge on Princeton University's campus, and the initial findings will be used to upgrade and improve the algorithm.