Neural network applied to prediction of the failure stress for a pressurized cylinder containing defects

An application of the neural network to predict the failure stress of a cyl inder with penetrating flaws under internal pressure is studied in this art icle, The results from the new method are compared with the experimental re sults. In addition, the neural network method is used to analyze the sensit ivity of parameters of a pressurized cylinder with defects. Satisfactory re sults are obtained. It is shown that the neural network could be a potentia l tool in engineering practice. (C) 1999 Elsevier Science Ltd. All rights r eserved.