Simulation and optimization of a centrifugal compressor of hydrogen in supercritical conditions

• Grant number: 21/04538-1
• Support Opportunities: Scholarships in Brazil - Doctorate (Direct)
• Effective date (Start): December 01, 2021
• Effective date (End): April 30, 2025
• Field of knowledge: Engineering - Mechanical Engineering - Thermal Engineering
• Principal Investigator: Jurandir Itizo Yanagihara
• Grantee: Murillo Souza dos Santos Pereira Neto
• Host Institution: Escola Politécnica (EP). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Host Company: Universidade de São Paulo (USP). Escola Politécnica (EP)
• Associated research grant: 20/15230-5 - Research Centre for Greenhouse Gas Innovation - RCG2I, AP.PCPE

Abstract

As hydrogen contains a high quantity of energy per unit of mass and it does not produce any direct emissions of greenhouse gases, there is a considerable potential towards strengthening the participation of this gas in the global energetic matrix. The transport of this gas, from production to consumption sites, may be performed by pipelines or by shipping and, in either case, compressors capable of providing high flowrates at high pressures are required. A potentially interesting fluidmachinery for accomplishing this task is the centrifugal compressor, which is the object of study at the present research project. In this sense, this project objectives are: to model and simulate the flow inside a compressor in supercritical conditions; to perform a sensitivity analysis of objective functions with respect to geometrical parameters; to optimize the geometry in order to maximize isentropic efficiency and minimize the power required. The flow will be modeled as steady compressible and turbulent, beyond considering the k-É SST turbulence model and an Equation of State, to be posteriorly selected. Geometry and mesh will be generated by using ANSYS Vista CCD and ANSYS TurboGrid software, and a grid independence study will be performed. Computational Fluid Dynamics (CFD) simulations will be performed by using ANSYS CFX and flow solutions will be analyzed numerically and phenomenologically. Regarding the Sensitivity Analysis (SA) and optimization: instead of assessing the CFD model, this works aims to use Response Surfaces (RS) surrogate models for assessment of the objective functions, which makes these analyses much faster. The SA methods to be utilized are Morris Screening and SS-ANOVA and the optimization method selected is the Non-Sorting Genetic Algorithm (NSGA-II). Finally, the optimal geometries will be CFD simulated and a phenomenological analysis will be performed. (AU)