A Detailed Observation of Hydrofoil Cavitation and a Proposal for Improving Cavitation Model

2010 ◽  
Author(s):  
Motohiko Nohmi ◽  
Naoya Ochiai ◽  
Yuka Iga ◽  
Toshiaki Ikohagi
Keyword(s):  
Free Surface ◽  
High Speed ◽  
Bubble Dynamics ◽  
Leading Edge ◽  
Free Surface Flow ◽  
Surface Flow ◽  
Experimental Results ◽  
Cavitation Model ◽  
Suction Surface ◽  
Cloud Cavitation

Cavitation of a hydrofoil is observed in detail by using a high speed video camera. A paint removal test is also carried out in order to evaluate cavitation aggressiveness for erosion. 2D hydrofoil profile is Clark Y 11.7% and its angle of attack is seven degrees. Cavitation number is σ = 1.08. The experimental results are compared with cavitation CFD. Numerous features of unsteady cavitation are observed such as cyclic fluctuation of the sheet cavity, existence of the glassy cones on a sheet cavity, generation of the cloud cavitation from the sheet cavity and the isolated bubbles traveling over the suction surface of the blade. The isolated traveling bubbles and their collapses are thought to be one of the main causes of the severe paint removals. The isolated traveling bubbles are derived from the flowing cavitation nucleus or from abrupt onset at the leading edge of the blade. For computing these complicated phenomena, combination of grid scale bubbles (GSB) and sub grid scale bubble model (SGSB) are proposed. GSB shall be computed by using the computational scheme for the free surface with phase change model. SGSB can be computed with conventional cavitation model. The breakup of GSB generates SGSB, and the coalescence of SGSB makes GSB. Upper limit of void fraction of SGSB is estimated in the range of five or ten percent from the simple speculation of the structure of packed spheres. The two types of cavitation bubble inception model are also discussed based on the generation of the isolated bubbles observed in the experiments. To verify the proposed concepts of cavitation model, a traveling air bubble over a hydrofoil is computed by using the free surface flow scheme of Volume of Fluid (VOF) approach. Cavitation on the hydrofoil is also computed by VOF approach with boiling model concerning the heat transfer. Both the computed results show qualitatively similar characteristics of the bubble dynamics to those in experimental results.

A model of flow separation at a free surface

1973 ◽  
Vol 57 (1) ◽  
pp. 129-148 ◽  
Author(s):  
M. S. Longuet-Higgins
Keyword(s):  
Free Surface ◽  
Flow Separation ◽  
Leading Edge ◽  
Free Surface Flow ◽  
Static Solution ◽  
Surface Flow ◽  
Effective Density ◽  
State Of Rest ◽  
Two Sides ◽  
Spilling Breaker

Flow separation can be observed (1) at the leading edge of a spilling breaker or ‘white-cap’, (2) at the lower edge of a tidal bore or hydraulic jump and (3) upstream of an obstacle abutting a steady free-surface flow. At the point of flow separation there is a discontinuity in the slope of the free surface. The flow upstream of this point is relatively smooth; the flow downstream of the discontinuity is turbulent.In this note, a local solution for the flow in the neighbourhood of the discontinuity is derived. The turbulence is represented by a constant eddy viscosity N, and the tangential stress across the interface between the laminar and turbulent zones is expressed in terms of a drag coefficient C. It is shown that the inclinations of the free surface of the two sides of the discontinuity depend on C only, and are independent of N and g. As C increases from zero to large values, so the inclination of the free surface in the turbulent zone increases from 10° 54′ to 30°. In the laminar zone the inclination of the free surface simultaneously decreases from 10° 54′ to 0°, the densities in the two zones being assumed equal.Owing to the possible entrainment of air at the separation point, the effective density ρ′ in the turbulent zone may be less than the density ρ in the laminar zone. When these densities are allowed to be different it is found that the possible flows are of two distinct types. Flows of the first type, called ‘quasi-static’, are contiguous to a state of rest. Flows of the second type, called ‘dynamic’, are contiguous with the frictional flows described above, for which ρ′ = ρ At a given positive value of C there exists generally only one quasi-static solution. There is also just one dynamic solution provided ρ′/ρ > 0·50012. On the other hand, if ρ′/ρ < 0·5 there may be either two or no dynamic flows, depending on the value of C; and when 0·5 < ρ′/ρ > 0·50012 there may be three such flows.The inclination of the free surface is studied as a function of C and ρ′/ρ.

Numerical analysis of high-speed liquid lithium free-surface flow

2012 ◽  
Vol 87 (5-6) ◽  
pp. 569-574 ◽  
Author(s):  
Sergej Gordeev ◽  
Volker Heinzel ◽  
Robert Stieglitz
Keyword(s):  
Free Surface ◽  
Numerical Analysis ◽  
High Speed ◽  
Free Surface Flow ◽  
Surface Flow ◽  
Liquid Lithium

Scale Effects on the Cavitation Development in Fluid Machinery

2011 ◽  
Author(s):  
Weiping Yu ◽  
Xianwu Luo ◽  
Yao Zhang ◽  
Bin Ji ◽  
Hongyuan Xu
Keyword(s):  
Turbulence Model ◽  
High Speed ◽  
Scale Effects ◽  
Design Procedure ◽  
Leading Edge ◽  
Characteristic Size ◽  
Cavitation Model ◽  
Fluid Machinery ◽  
Cloud Cavitation ◽  
Sheet Cavitation

The prediction of cavitation in a design procedure is very important for fluid machinery. However, the behaviors of cavitation development in the flow passage are believed to be much different due to scale effects, when the characteristic size varies greatly for fluid machines such as pumps, turbines and propellers. In order to understand the differences in cavitation development, the evolution of cavity pattern in two hydro foils were recorded by high-speed video apparatus. Both foils have the same section profile, and their chord lengths are 70mm and 14mm respectively. For comparison, the cavitating flows around two foils were numerically simulated using a cavitation model based on Rayleigh-Plesset equation and SST k-ω turbulence model. The experiments depicted that for both hydro foils, there was attached sheet cavitation near the leading edge, which separated from the rear part of the cavity and collapsed near the foil trailing edge. There was clear cloud cavitation in the case of the mini foil. The results also indicated that the numerical simulation captured the cavitation evolution for the ordinary foil quite well compared with the experiments, but could hardly predict the cloud cavitation for the mini foil. Thus, it is believed that both the cavitation model and the turbulence model should be carefully treated for the scale effect on cavitation development in fluid machinery.

The Application of Non-Axisymmetric Profiled End-Walls for Axial Flow Turbines in the Embedded Stage Environment

2013 ◽  
Author(s):  
Andreas Lintz ◽  
Liping Xu ◽  
Marios Karakasis
Keyword(s):  
Axial Flow ◽  
Leading Edge ◽  
Surface Flow ◽  
Secondary Flows ◽  
Experimental Results ◽  
Front Part ◽  
Flow Conditions ◽  
Suction Surface ◽  
Inlet Conditions ◽  
End Wall

In this paper, an assessment of the effectiveness of non-axisymmetric profiled end-walls in the embedded stage environment at varying inlet conditions is presented. Both numerical and experimental results were obtained in a three-stage model turbine which offers flow conditions representative of embedded blade rows in a typical high pressure steam turbine. The end-wall profile design was carried out using automatic optimization in conjunction with 3D RANS CFD. The design target is to reduce the end-wall losses by reducing the loading in the front part of the passage, which resulted in a single trough close to the blade suction surface in the leading edge region. 5-hole probe traverses and surface flow visualization show that the intensity of the secondary flows is reduced by about 10%, but overall loss is only reduced slightly. Experimental results have been obtained for the cylindrical end-wall and three different trough depths. With increasing depth, transitional effects at the end-walls might come into play, increasing the total pressure loss in the boundary layer region. The effects of the end-wall design is similar at positive and negative incidence, despite the reduced loading in the front part of the passage at negative incidence. At very high negative incidence angles, such as those occurring at the stator tip with rotor shroud leakage flows, the mechanism of secondary flow generation changes, so that a design under nominal inlet flow conditions shows no effect on the exit flow field.

Free Surface Flow in High Speed Fiber Drawing With Large-Diameter Glass Preforms

2004 ◽  
Vol 126 (5) ◽  
pp. 713-722 ◽  
Author(s):  
Zhiyong Wei ◽  
Kok-Meng Lee ◽  
Serge W. Tchikanda ◽  
Zhi Zhou ◽  
Siu-Ping Hong
Keyword(s):  
Boundary Condition ◽  
Free Surface ◽  
High Speed ◽  
Stokes Equations ◽  
Large Diameter ◽  
Free Surface Flow ◽  
Surface Flow ◽  
Staggered Grid ◽  
Fiber Drawing ◽  
Measured Profile

This paper presents a complete two-dimensional (2D) thermofluid model for predicting the neck-down shape in the fiber drawing process. This model uses the controlled draw tension to calculate the Neumann boundary condition at the furnace exit; thus, it does not require specifying the speed (or diameter) of the fiber as most previous studies did. The model presented here can be applied to optimization of the high-speed draw process with large-diameter preforms. In this study, the radiative transfer equation is directly solved for the radiation fluxes using the discrete ordinate method coupled with the solution of the free surface flow, which does not assume that the glass is optically thick and does not neglect the glass absorption at the short-wavelength band. The artificial compressibility method is used to solve the Navier-Stokes equations. A staggered-grid computation scheme that is shown to be efficient and robust was used to reduce the computation load in solving the complete 2D model. The neck-down profile of a large preform (9 cm dia) drawn at a relatively high speed of 25 m/s was experimentally measured. The measured profile well matches that derived numerically. Results also show that the free surface calculated using the Dirichlet boundary condition deviates considerably from the measured profile, particularly near the furnace exit where the actual diameter (and, hence, the speed of the glass) is essentially unknown. Although the difference between the numerical results obtained from the full and semi-2D models was small, this difference could be significant if the location at which the glass converges to 125 μm dia is of interest, especially when the preform has a large diameter drawn at a high speed.

scholarly journals Laminar free-surface flow into a vertical cylinder

1975 ◽  
Vol 3 (1) ◽  
pp. 51-68 ◽  
Author(s):  
Thomas G. Smith ◽  
J.O. Wilkes
Keyword(s):  
Free Surface ◽  
Vertical Cylinder ◽  
Free Surface Flow ◽  
Surface Flow
Sign in / Sign up

Export Citation Format

Share Document