Validation and Rules-of-Thumb for Computational Predictions of Liquid Slosh Dynamics

摘要

During flight, the sloshing of liquid propellant onboard the vehicle can have significant effects on vehicle stability due to the large amount of propellant mass. Because of the stability effect, propellant sloshing is included in the guidance, navigation and control (GNC) analysis to account for the dynamics of the propellant sloshing motion. Typically, the sloshing propellant is simplified either as a mechanical system in the form of a swinging pendulum or as a spring-mass system. In either case, the important parameters governing the dynamics of the system are the mass, natural frequency, and damping ratio. The sloshing parameters of frequency and damping of the slosh system are dependent on the type of propellant, tank geometry, and the liquid level in the tank. The parameters are typically derived from the large amount of experimental data currently available for conventional tank shapes. Unconventional tank shapes, however, require either experimental or computational work to develop new sloshing parameters. Because of the cost of experimental slosh testing, it may be preferred to use computational fluid dynamics (CFD) solvers to predict sloshing dynamics. While previous studies have shown the applicability of CFD to predict sloshing parameters, a consistent methodology to conduct sloshing simulations has not yet been reported. Specifically, there is currently no guidance in the literature for selecting the necessary physics in the simulation, nor is there a suggestion for a starting point of the computational grid resolution. In this paper, the CFD solver FLOW-3D is used to simulate propellant sloshing in oblate spheroidal and cylindrical tanks. The slosh frequency and damping factor predicted with the simulations are validated with experimental data for a range of propellant fill levels and with and without tank baffles. The required physics for sloshing simulations are discussed and rules-of-thumb are presented to provide a starting point for defining the computational grid resolution.