Stress Intensity Factor Influence Coefficients Research Articles

Improvements in Article A-3000 of Appendix A for Calculation of Stress Intensity Factor in Section XI of the 2015 Edition of ASME Boiler and Pressure Vessel Code

Analytical evaluation procedures for determining the acceptability of flaws detected during in-service inspection of nuclear power plant components are provided in Section XI of the ASME Boiler and Pressure Vessel Code. Linear elastic fracture mechanics based evaluation procedures in ASME Section XI require calculation of the stress intensity factor. Article A-3000 of Appendix A in ASME Section XI prescribes a method to calculate the stress intensity factor for a surface or subsurface flaw by making use of the flaw location stress distribution obtained in the absence of the flaw. The 2015 Edition of ASME Section XI implements a number of significant improvements in Article A-3000. Major improvements include the implementation of an alternate method for calculation of the stress intensity factor for a surface flaw that makes explicit use of the Universal Weight Function Method and does not require a polynomial fit to the actual stress distribution and the inclusion of closed-form equations for stress intensity factor influence coefficients for cylinder geometries. With the inclusion of the explicit weight function approach and the closed-form relations for influence coefficients, the procedures of Appendix A for the calculation of stress intensity factors can be used more efficiently. A review of improvements that have been implemented in Article A-3000 of Appendix A in the 2015 Edition of ASME Section XI is provided in this paper. Example calculations are provided for illustration purpose.

Stress intensity factor influence coefficients for internal surface cracks in thick-walled cylinders

Using the boundary integral equation method for three-dimensional elastostatics, stress intensity factor influence coefficients for a wide range of internal semi-elliptical surface cracks in cylinders with radius ratios 2·0 and 2·5 are obtained. Only one crack shape is considered, namely, a semi-ellipse with ellipse aspect ratio 0·8, but the crack depths treated range from 20% to 80% of the wall thickness. The influence coefficients obtained correspond to the stress intensity factor, K I , solutions for the cracks subjected to four different stress distributions—uniform, linear, quadratic and cubic pressure variations—on the crack faces. They can be superposed to obtain K I solutions for other stress distributions, such as those caused by internal pressure, thermal gradients and autofrettage. Numerical results are presented for all these load situations. In the case of autofrettage, the effects on K I for a wide range of overstrain conditions are considered.

Stress-Intensity Factors for Internal and External Surface Cracks in Cylindrical Vessels

The purpose of this paper is to present stress-intensity factor influence coefficients for a wide range of semi-elliptical surface cracks on the inside or outside of a cylinder. The crack surfaces were subjected to four stress distributions: uniform, linear, quadratic, and cubic. These four solutions can be superimposed to obtain stress-intensity factor solutions for other stress distributions, such as those caused by internal pressure and by thermal shock. The results for internal pressure are given herein. The ratio of crack depth to crack length from 0.2 to 1; the ratio of crack depth to wall thickness ranged from 0.2 to 0.8; and the ratio of wall thickness to vessel radius was 0.1 or 0.25. The stress-intensity factors were calculated by a three-dimensional finite-element method. The finite-element models employ singularity elements along the crack front and linear-strain elements elsewhere. The models had about 6500 degrees of freedom. The stress-intensity factors were evaluated from a nodal-force method. The present results were also compared to other analyses of surface cracks in cylinders. The results from a boundary-integral equation method agreed well (±2 percent), and those from other finite-element methods agreed fairly well (±10 percent) with the present results.