Executive Summary : | Objective: The present work is proposed to explore data assimilation methods for developing models for prediction of creep degradation of steel by taking into account the aforementioned parameters. The data assimilation involves collection and processing of material property database available in public domain as well as generation of results from both experiments and simulations Summary: Big data assimilation is a multi-principal analysis cum simulation methodology. It is widely used to develop a statistical correlations between the data obtained from various sources. Statistical correlations and analysis can be used to predict the desirable properties of the considered system. Specifically, in the proposed investigation, we will develop a model for the data assimilation algorithms to predict creep degradation of steel using data from simulations (DFT, MD, PF and FEM) and real laboratory experiments. Using these simulation methodologies, we will extract huge data related to stress-strain-time responses, crystal structures, grain growth, grain orientations, Young’s modulus, von-Misses stress, mean stresses etc. Experimental investigations will provide data for the validation of simulation and for the data assimilation analysis. Bayesian and three/four dimensional variational data assimilation (3D-Var, 4D-Var) analysis will be carried out to predict accurately creep degradation in steel. The proposed application of data assimilation algorithm in steel will facilitate and provide a robust database for the development of high performance steel with desirable creep properties. |
