With the progress of super computers in recent years, a number of studies on “Computational Fluid Dynamics” (CFD) have been carried out, and various schemes for Navier-Stokes equations have been presented. Similar methods have also been applied to automotive engineering - aerodynamics, for exampre - in order to determine flow phenomena.In this paper, the application of numerical simulations to the flow cavitation that occurs in some part of orifices in the vehicle hydraulic system, will be discussed. Authors have developed a CFD program for the clarification of flow phenomena in such orifices. Using the relationship between calculated results and measured results of noise levels in such orifices, a new method for estimation of the occurrence of flow cavitation has also been developed. As a result, a new orifice configuration capable of preventing the cavitation has been designed.VARIOUS ATTEMPTS have been made to determine flow phenomena by means of numerical simulations, which are difficult to determine by measurements. Many papers related with this issue, however, have dealt only with the verification of accuracy of simulations in comparison to measured results, without practical applications to design needs.Under this study, CFD is applied to the problem of noise in orifice that presumably occurs due to the cavitation in the vehicle hydraulic system. A stable orifice mounted in the steering gear box of a power steering system is shown in Fig.1. Noise is found in the orifice, but it is very difficult to determine the flow phenomena in the orifice by measurement because of the high pressure and extremely small size. Thus, the mechanism of the noise reduction by the modification of the orifice geometry from one-phase to three-phase has not been clarified. Experimental studies , ,and  have been reported,but discussions focused only on the variation of cavitation phenomena, due to the difference in the geometry of orifice, without determination of flow phenomena.This study thus aims to clarify the mechanism and to find an effective countermeasure for the noise problem with determination of flow phenomena.