Ultrasonic sensing classification of foundry pieces applying neural networks

This work describes an inspection and classification system based on ultrasonic sensing that is developed for application to foundry pieces which will later be machined within the automobile industry. In the manufacturing of foundry pieces, small defects can sometimes be found after casting, which although not critical for the final use of the piece, affect the manufacturing processes detrimentally, especially in machining tasks. The inspection system tries to solve this problem in an automated way applying ultrasonic identification techniques. In this field, ultrasonic tools appear to be a powerful technique for the quality control of processes because of their capacity to recognise and classify pieces. The ultrasonic signal reflected by a piece is mathematically treated and two neural network designs are developed to perform the discrimination between machinable and nonmachinable piece. This work forms part of the investigative tasks of a collaboration project between the Engineering Control Group of the University of Cantabria and FUNDIMOTOR S.A. (Nissan Group).