Co-Simulation Development for Improved Cavitation Predictions in Oil-Hydraulics Systems : An investigation into the cavitating flow behavior of repetitive water hammers.

Abstract: Numerical modeling of cavitation using computational software is a highly pursued topic of research due to its impact in different industrial sectors. While some industrial applications such as wastewater treatment and mineral processing are known to advantageously use this phenomena, it remains an unwanted process in others where it is known to induce vibration, reduce performance and cause structural damage. The main objective of the current research study is to investigate the accuracy to which cavitating flow behavior inside oil-hydraulic systems can be computationally modeled, what limitations exist and how to improve numerical predictions. An experimental test-rig has been built in the preceding years with plexiglass tube to observed the vapor formations during cavitation and the pressure readings at three points have been recorded. The current study uses a computational model with the same geometry as the experimental test-rig, and uses the experimentally recorded pressure values for validating numerical results. Two main software are used to setup the simulation framework. The first is Hopsan, one open source simulation software for hydraulic systems developed by Link\"oping University and the second is ANSYS Fluent, a commercial software for modeling complex fluid flow applications. Four different orifices are used to create different outlet pressures. For orifices of diameter 2 mm, 3 mm, and 5 mm, good correlation between numerical and experimental results were observed. Further investigations into complex cavitating flow behavior of repetitive water hammers were also carried out. Different valve profile movements were used to investigate what the impact of having and not having vapor bubbles in the plexiglass tube would have on the pressure distribution when oil starts to re-circulate in the system. Furthermore, repetitive water hammer flow behavior for oscillations of 2, 3, and 4 water hammers were investigated. This investigation revealed several important findings.  The first is that if valve opens to the point that the flow starts to re-circulate in the system while vapor bubbles already exist in the plexiglass tube, massive pressure peaks, as high as 350 bar, will be created in the plexiglass pipe. The strength of this pressure surge will be dependent on the amount of vapor in the pipe when flow is re-introduced. The second is that if the valve starts to re-open (move backwards) while no vapor exists in the plexiglass tube, this movement will result in the formation of vapor. However, this vapor only lasts for a small duration and disappears before the valve reaches a point that allows flow move again. The third and final finding for repetitive water hammers was that the strength of the pressure surges will reduce with each sequential water hammer.