Probabilistic pipe strength and toughness estimation through information fusion with bayesian updating

The integrity of ageing pipelines infrastructure systems is a critical concern for safety and economy of United States. The present day techniques for accurate pipe strength determination encounter several gaps in terms of single modality measurements and uncertainties. In view of this, the present study focuses on the design of a novel information fusion framework using multimodality diagnosis for pipe materials to predict the accurate probabilistic strength and toughness estimation under uncertainties. The mechanical property variation such as strength/fracture toughness is assessed in terms of a number of material properties such as chemical composition, microstructure, dislocation density etc., with the use of several in-situ and ex-situ experiments. Advanced data analysis using Gaussian Processing model will be performed for surrogate modeling and uncertainty quantification. Simulation and prototype testing will be carried out for model validation and demonstration. Probabilistic pipe strength and toughness estimation is inferred based on the posterior distribution after information fusion. Inhomogeneity in the material properties is studied with several through thickness studies, to account for the mechanical property variation from surface to bulk. Finally, a microstructure-property based 3-D stochastic reconstruction model will be used to serve as an integrated computational framework for prediction of probablisitic strength. Data training of the model will then be performed to obtain an accurate probabilistic pipe strength and toughness.