This study focuses on the development of a framework to predict liquid jet atomization process using Gaussian Process (GP) based machine learning techniques. The machine learning algorithm is trained using results obtained from experimentally validated high-fidelity numerical simulations of diesel jet injected into a quiescent air environment in an Eulerian-Eulerian (E-E) framework. A volume-of-fluid interface capturing algorithm is used in the E-E formulation to treat the liquid-gas interface and adaptive mesh refinement technique is adopted to ensure computational efficiency. The overall algorithm proceeds in four steps- 1) high-fidelity simulations that are validated using experimental measurements; 2) extraction of dominant energetic modes, basis functions and corresponding time coefficients using proper orthogonal decomposition (POD) such that 99% of the energy content is captured; 3) training the Gaussian stochastic process using the modal extracted data from the POD analysis for the range of operating conditions under consideration; and 4) prediction of energetic modes (basis functions and time coefficients) using the learned algorithm, within the bounds of the training range, and comparison with measurements. It should be noted that our machine learning framework can predict both the spatial basis functions and the time coefficients, thus, predicting the entire flowfield in time and space. As a first step, the above methodology is used to predict the dynamic behavior of flow over a circular cylinder for a range of Reynolds numbers from 10 to 200. Galerkin reconstruction of the predicted flowfield shows excellent agreement with the results obtained from high-fidelity data. Next, the framework is applied to diesel injection in quiescent nitrogen environment. The operating conditions consists of a pressure of 30 atm, a temperature of 300K, and injection velocities between 10 to 55 m/s, corresponding to a range of Weber numbers from 11.5 to 348. The GP shows an excellent agreement between simulations and predictions for time marching of cylinder flow and the basis function modes of flow for the liquid jet. The resulting GP emulator is expected to significantly improve the design process of devices based on multiphase flow processes, such as fuel injectors, by enabling otherwise unaffordable studies, including sensitivity analysis and uncertainty quantification, that are critical for the development of such devices.
Ganti, H., & Khare, P. (2018). Prediction of Liquid Jet Atomization Using Gaussian Process Based Machine Learning Techniques. In ICLASS. 14th Triennial International COnference on Liquid Atomization and Spray Systems (pp. 1–8).