ABSTRACT
Prestressed concrete bridges experience complex load scenarios throughout their lifespan which impacts their load capacity and overall performance. To ensure optimal functionality and safety it is crucial to understand the diverse responses from the structure under varying load. By deploying sensors in strategically placed locations, a numerical representation of varying responses can be attained during operation. Variation in temperature can affect the readings of a sensor. This can lead to false positive structural damage responses. Thus, it is important to differentiate between variations in structural properties induced by temperature fluctuations, and variations from structural damage. This paper presents a novel approach for mapping the temperature effect in structural responses. By employing an artificial neural network (ANN) and measurement data from a prestressed concrete bridge located in northern Sweden, a correlation between the temperature and the sensor values is achieved.
