||CMES: Computer Modeling in Engineering & Sciences, Vol. 11, No. 1, pp. 9-16, 2006
||Full length paper in PDF format. Size = 227,821 bytes
||Statistical Assessment, POD (probability of detection), Effective POD, Steam Generator Tube, Structural Integrity, Monte Carlo method.
||The growth of stress corrosion cracks in steam generator tubes is predicted using the Monte Carlo simulation and statistical approaches. The statistical parameters that represent the characteristics of crack growth and crack initiation are derived from in-service inspection (ISI) non-destructive evaluation (NDE) data. Based on the statistical approaches, crack growth models are proposed and applied to predict crack distribution at the end of cycle (EOC). Because in-service inspection (ISI) crack data is different from physical crack data, a simple method for predicting the physical number of cracks from periodic in-service inspection data is proposed in this study. Actual number of cracks is easily estimated using the method, and the statistical crack growth is simulated using the Monte Carlo method. Probabilistic distributions of the number of cracks and maximum crack size at EOC are obtained from the simulation. Comparing the predicted EOC crack data with the known EOC data the usefulness of the proposed method is examined and satisfactory results are obtained.