Determination of the Fatigue State of the Material under Dynamic Loading by Thermal Video Image Processing

In this study we propose an infrared thermography and artificial neural networks (ANNs) based method for obtaining the fatigue state of the steel specimen under bending fatigue. We use the thermal images (TIs) of the specimen under bending fatigue, taken by FLIR E45 infrared camera at 1 Hz. By processing TIs using ANNs we obtain temperatures of the characteristic spots of the specimen surface. Based on these temperatures we fit curve on which we fix the temperatures of all spots of the specimen surface. The region of hottest spots of the TI obtained by this method allows us to kept under control the probably crack region of the tested specimen until it is fractured.