ARTIFICIAL NEURAL-NETWORK TECHNOLOGY AS A METHOD TO EVALUATE THE FAILURE BENDING MOMENT OF A PIPE WITH A CIRCUMFERENTIAL CRACK

The artificial neural network (ANN) method is introduced to evaluate t he failure bending moment of pipes with circumferential cracks. In par ticular examples. it is used to evaluate the failure bending moment of full-scale pipes under internal pressure and external bending moment. The examples show the feasibility that the neural network method can be used to predict the failure bending moment of pipes with circumfere ntial cracks. The neural network method is also used to assess the acc eptability of flawed pipes in plant. A comparison of the assessment re sults of the neural network method with those of the ASME Boiler and P ressure Vessel Code Section XI IWB-3650 is also presented. Copyright ( C) 1996 Elsevier Ltd.