Masafumi Kawai

Numerical Analysis of the Unsteady Cavitating Flows in a 2D-Cascade and a 3D-Inducer

The unsteady cavitating flow in a turbopump inducer has been analyzed using CFD simulation results. An accurate three-dimensional detached eddy simulation (DES) code which took into account the cavitating phenomena was used. The three-dimensional code is based on the numerical method for compressible flow using low Mach number flow approximation. The cavitation is modeled by the source/sink of vapor phase in the liquid flow. The code has been applied for cavitating flows in both a 2D cascade and a 3D inducer. The computational results showed similar unsteady behaviors to experimental measurements in terms of the direction in cavity propagations and tendency in cavity developments. Mechanism for the unsteady cavitating flow such as rotating cavitation is investigated by analyzing the numerical results. One model is proposed to describe the reason why the cavity appears to propagate in the forward direction.

NUMERICAL PREDICTION OF TURBULENT FLOW OVER A SURFACE-MOUNTED CUBE

Theoretical and experimental studies are reported on the three-dimensional turbulent flow over a surface-mounted cube, oriented with one face perpendicular to the uniform incident flow. A triaxial hot wire probe was used to measure the mean and fluctuating velocities. Flow visualization technique was used to observe the extent of the recirculating flow. The numerical method consists of a finite-different procedure for solving the momentum and continuity equations with two additional conservation equations for the turbulent kinetic energy and its dissipation rate. The modified version of k-e model employs a new dissipation rate equation whose production term is made more sensitive to streamline curvature effects. The new implicit solution procedure was used to achieve a saving in c.p.u time for calculation. The prediction obtained is compared with present and other available experimental data. The results predicted are in general accord with experimental data, but there are certain discrepancies.

Numerical Simulation of Sand Erosion in Steam/Water Separator

ABSTRACT This paper presents an investigation into a phenomenon Several wall and particle materials and also the combinthat happened on the wall surface of a Steam / Water Separator (SWS). It was reported that erosion caused from unknown solid particle took place on the SWS wall. In order to capture this sand erosion phenomenon numerically, the SWS flow field was solved, and then particle trajectory and wear quantity were calculated, based on the CFD results. Several wall and particle materials and also the combinations of them are assumed. Furthermore, the particle diameter was varied from 10 -6 to 10 -2 m. The numerical results insist that the particle, which could be the factor of the phenomenon, is limited in its diameter range and its material. The present study will be an aid to clarify the cause of sand erosion in a SWS. INTRODUCTION A Steam / Water Separator (below SWS) is today widely used in a power plant, for the purpose of separating steam and water of the two phase flow. In a SWS, the liquidized and gaseous components in the steam are separated, flow field can be reproducedmaking use of the difference of centrifugal force caused by the swirling motion, and finally the liquidized component is collected and removed. It was reported that a solid particle incidentally mixed into the steam caused the sand erosion to a dSWS. In the present study, sand erosion phenomenon in a SWS is investigated by simulating the turbulent flow field and the trajectory of a solid particle. To reproduce the strongly swirling turbulent flow in the SWS, Reynolds stress turbulence model (RSM) is employed. Using the Neilson-Gilchrist model, the wear quantity on the inner surface of the SWS is estimated.ations of them are assumed. Furthermore, the particle diameter was varied from 10

Development of a New HVAC Flow System for the Next Generation Automated People Mover

IHI has succeeded in development of a new generation Automated People Mover (APM). One of the features of this APM is the composite car body which enables stylish exterior, spacious cabin and light weight. This feature demanded the development of a new Heating, Ventilating, and Air-Conditioning (HVAC) system, which does not have any devices on the roof of the car, to avoid heavy weight on the composite body and to keep the cabin ceiling high for passenger comfort. Consequently we adopted a new HVAC flow system in which the air is supplied upwards from the bottom of the side windows while all the mechanical devices are placed under the floor. However, when the prototype was first designed, it caused congested flow channel due to its location and we had to have ducts that were too small. It was unexpectedly difficult to keep the space for HVAC system like in an ordinary layout. So at first, the prototype vehicle did not have enough HVAC performance due to lack of air flow. Therefore we performed full review of the duct system. We made desk analysis, duct model test, experiment on the prototype, and CFD simulation, for the improvement of the new HVAC flow system. Finally, we have established an excellent system with the new layout which has been validated on the prototype vehicle. In this paper, we describe the process of developing a new HVAC system, based on the results of actual measurement as well as CFD simulation.Copyright

J0501-3-5 航空機用大型防音壁の空力設計と実機検証([J0501-3]流体機械の研究開発におけるEFD/CFD(3))

Aircraft are required to be tested prior to flight when its engines are maintained in an airport. The test is usually done in an exhaust gas diffusing device, called a blast fence. Since the blast fence has poor noise suppression performance, when noise problems take place in an airport, a noise suppression facility, or run-up facility (RUF) is used. RUFs are roughly categorized into two kinds: closed and open [1] (intermediate types [2][3] also exist). The former has better noise suppression ability, but it is difficult to establish good aerodynamic characteristics. The latter, with a typical one shown in Fig. 1, surrounds the aircraft by walls of suitable height, area and material. With this type it is rather easy to directly take the atmosphere into the engine. However, since engines are designed for open-air conditions, aerodynamic problems may still arise in the open-type RUF. Among the existing RUFs, there are several implementations of closed type (hush house). On the other hand, there are only a few examples of large scale open-type RUFs which surround the entire commercial airliner.[1] There were some cases for closed-type RUFs in which it had been difficult to establish good enough aerodynamic performance. Hence the development of the procedure for aerodynamic design is needed Is this paper, the aerodynamic characteristics of an open-type RUF, as shown in Fig. 1, have been studied by wind tunnel experiment and computational fluid dynamics (CFD) to provide a new design procedure. The target of the aerodynamic performance was set as “operability nearly equal to open air.” Then the validation tests were performed for the actual RUF constructed based on the new design procedure.

Prediction of three-dimensional turbulent flows in a dump diffuser

A finite volume method for the solution of three-dimensional incompressible steady Navier-Stokes equations based on a general curvilinear coordinate system was employed to Study the characteristics of turbulent flow in dump diffuser of gas-turbine combustor. The standard k-E turbulence model is used to characterize the effect of turbulence. in order to achieve a saving in CPU time for calculation, present calculation was performed by lending itself to vector computer architecture of the Supercomputer FACOM VP-50. This method is applied to prediction of turbulent flow in a three-dimesional dump diffuser with and without the fuel nozzle. The calculated results are compared with the corresponding experimental data obtained in this work. General features of the flow pattern are adequately predicted although discrepancies in detail seem to indicate deficiencies in the turbulence model used in present Study.