Abstract
The flow systems in nuclear power plants and aircraft engine fuel pipelines are often subjected to extreme high-pressure conditions, which can induce cavitation and severely affect essential system components. Orifice plates are the most typical infrastructures representing the flow principle of the aboved flow systems. In this paper, a modified cavitation model combining local flow characteristics, supervised learning, and genetic algorithms is proposed to investigate the cavitation flow characteristics of orifice plates under high-pressure conditions. The modified cavitation model eliminates the influence of the bubble diameter and nucleation site volume fraction on mass flow rate and achieves dimensionality reduction at the physical level. The relationship between evaporation/condensation coefficients and mass flow rate is constructed by supervised learning, and the two coefficients are determined using the genetic algorithm. The mass flow rates are calculated by the modified cavitation model within a 5% experiment error, proving the accuracy of the modified cavitation model. The effect of the diameter ratio (the diameter of the pipe to the orifice plate) and pressure drop on the mass flow rate are obtained based on the validated cavitation model. Finally, an empirical formula for calculating the mass flow rate based on the diameter ratio and pressure drop is derived. The modified cavitation model shows great potential for cavitation prediction applications for throttling devices such as nuclear power safety valves and aircraft engine nozzles.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call