Improvements to B31.8 Dent Strain Estimation and Assessment of Dent Formation Induced Cracking

Abstract Local plastic deformation associated with the dent formation can potentially cause ductile damage leading to cracks. Plastic strain is considered as a key indicator to dent formation-induced cracking. The estimation of local plastic strain in a dent is mainly based on the calculation of dent profile curvatures. The ASME B31.8 effective strain equations [1] provide a basis to convert the geometric curvatures into strains. In this paper, a variety of methods used to estimate dent profile curvatures were evaluated for calculating the effective strain. Hypothetical dent profiles generated from FEA simulations were used for the evaluation. The effective strains obtained by using the B31.8 equations were compared with the FEA results on equivalent plastic strains for all the simulated dent profiles. It was shown that the estimation of plastic strain was sensitive to the methods used for extracting the dent profile curvatures. A method based on the 4th order polynomial curve fitting on a total of 9 points selected near the dent apex was shown to provide more consistent results compared with other methods investigated in the present paper. Unconstrained dents pose a particular challenge when using the B31.8 effective strain equations to estimate the plastic strain. The measured dent profiles from pipeline inline inspection (ILI) for an unconstrained dent can be significantly different from the indentation dent profiles due to plastic re-rounding. The estimated plastic strains based on the current dent profile after re-rounding with the B31.8 equation display a significant scatter compared with the FEA simulation results. A regression model was developed to improve the strain estimates based on the B31.8 equations using the current dent profiles after re-rounding. The regression model was intended to compensate for the re-rounding and correct the strain estimates from the B31.8 equations by including the current dent geometries and pipe internal pressure. Trained with FEA simulation results for unconstrained dents, the regression equation was shown to significantly improve the strain estimates compared with the FEA results.