Experimental and numerical investigation of cavitating two-phase flows and cavitation-induced erosion
Motivation and Objectives
The formation of vapor bubbles in a liquid due to pressure reduction is called “cavitation”. Flows involving cavitation feature a series of unique physical properties such as discontinuous jumps in the speed of sound from O(1000) m/s to O(1) m/s, a jump in density of up to 4 orders in magnitude, and intense compressibility effects, such as the formation of intense shock-waves with post-shock pressures of more than 1GPa. Flows involving cavitation occur in a wide range of technical systems. In particular, injection systems for combustion engines, high pressure hydraulics, naval propellers and biomedical applications are prone to cavitation and cavitation-induced material erosion. Our objective is to develop efficient and accurate simulation approaches for predicting all dominating phenomena in cavitating flows including shock-wave formation and propagation, with the goal to provide the groundwork for the design optimization of future technical devices.
Approach to Solution
We perform fundamental experiments using a shock-tube and state-of-the art high speed cameras/sensors to investigate collapse processes of gas und vapor bubbles embedded in a liquid-like gel. These experiments are used to enhance physical understanding of involved fluid dynamics and serve as reference data to our numerical investigations. Since about one decade, mathematical models and numerical approaches for efficient and accurate predictions of cavitating flow phenomena are developed at the institute. A series of numerical approaches, including state-of-the art Large-Eddy Simulation (LES) schemes enable high performance computing with linear scaling on HPC systems, such as SuperMUC. Our approaches are “monolithic” in a sense that all involved fluid components (liquid, vapor, inert gases) are handled in a consistent way. Shock wave formation due to collapsing vapor patterns is resolved by application of time-steps smaller than one nanosecond. The resulting loads on material surfaces – and thus the potential of material erosion -are obtained without the need for additional models. Fundamental research is funded by the European Union (Project “CaFE”), while applied research is performed in collaboration with several automotive suppliers, the U.S. Office of Naval Research and with the European Space Agency.