Aurélia Vallier - LTH

Lund university / Energy Sciences / Fluid Mechanics

Computations of cavitating flow on hydrofoils

Aurélia Vallier

Division of Fluid MechanicsDepartment of Energy Sciences

Lund Institute of Technology, Lund Sweden

1


2

OUTLINE

• Introduction

• Modelling

• Problem Setup

• Results

• Conclusions


INTRODUCTION

3

• Cavitation - what is it? - where does it occur? - different types of cavitation - why do we want to understand it?


INTRODUCTION

4

Cavitation - what is it?

Boiling

Cavitation

Phase diagram

Definition : Formation of vapor cavities in a liquid due to pressure drop.


INTRODUCTION

5

Cavitation – where does it occur?


INTRODUCTION

6

Cavitation – different types of cavitation

(a) Travelling bubble Attached cavitation - Sheet cavitation (c) -Cloud cavitation (b)(d) Supercavitation

(e) Vortex cavitation

σ=p∞− pv

12

ρu∞2

σ c

σσ c


INTRODUCTION

7



INTRODUCTION

8




σ=p∞− pv

12

ρu∞2

σ c

σσ c


INTRODUCTION

9



INTRODUCTION

10




σ=p∞− pv

12

ρu∞2

σ c

σσ c


INTRODUCTION

11



INTRODUCTION

12




σ=p∞− pv

12

ρu∞2

σ c

σσ c


INTRODUCTION

13



INTRODUCTION

14




σ=p∞− pv

12

ρu∞2

σ c

σσ c


INTRODUCTION

15

Cavitation – why do we want to understand it?

− Negative effects : - noise,

- vibration, - material damages (erosion due to the collapse) - performance reduction

Understand cavitation inception and evolution avoid (at least control and reduce) these effects

− Difficult to predict because it depends on - flow conditions - water nuclei - surface roughness


16

MODELLING

Inception: - Pressure drop - Nuclei content

Evolution: - Two phase flow - Bubble dynamics - Turbulence -> mixing


17

MODELLING

Inception: - Pressure drop - Nuclei content

Evolution: - Two phase flow - Turbulence -> mixing - Bubble dynamics

σ=p∞− pv

1/2ρu∞2

Model 1 fluidBarotropic equation of state

Model 2 fluidsVOFInterface trackingTransport equationMass transfert

Rayleght-Plesset equation

ρ R R32

R24νRR=p v− p t −

2SR

2γR0

− pv− p0 R0

R3Γ

Nuclei density n0Nuclei radius R

Turbulence model


VOF ---> vapor volume fraction α

MODELLING Sauer Model

18

∂

∂ t∇ .u =0

∂u∂ t

u . ∇u =−∇ p∇ 2u−∇

∣∇∣

∂α∂ t

∇ .α u =ml

ρ=ρv αρl 1−α

μ=μv αμl1−α

Mass transfer rate = creation of vapour + destruction of vapour

m0

m−0


VOF ---> vapor volume fraction α

MODELLING Sauer Model

19

∂

∂ t∇ .u =0

∂u∂ t

u . ∇u =−∇ p∇ 2u−∇

∣∇∣

∂α∂ t

∇ .α u =ml

ρ=ρv αρl 1−α

μ=μv αμl1−α =

n0 .43R3

1n0 . 43R3

d Rd t

= 23

p R −p∞

l

ddt

=1−

n0 .43R2 R

1n0 . 43R3

∇ .u=−1

∂∂t

u ∇ .=−1

∂∂ t

=l−v

∂∂ t

Vapor= small spherical bubbles


PROBLEM SETUP

20

InletUx=8m/s∇p=0

Outlet∇U=0p=0 Pa

WallU=0 m/s∇p=0

Naca0015 profile (chord length=0.15m , angle of attack=8°)2D geometry (1m*0.5m)Re=1.2 *10⁶


RESULTS

21

-Spallart Allmaras model

-Wall function (y+= 30)


RESULTS

22

-Spallart Allmaras model

-Wall function (y+= 30)


RESULTS



PROBLEM SETUP

25

InletUx=8m/s∇p=0

Outlet∇U=0p=0 Pa

WallU=0 m/s∇p=0

Injection50 or 500 particles / time step Dp=0.1,1,10 or 50μmUp=0Solution methodLPT one way coupling


RESULTS

Statistical average of the nuclei concentration over 40 periods

Dense-at the point of impact(collision)-in the low pressure area(close to the leading edge and above the suction side)

Very dilutedin the boundary layer of the suction side

→Nuclei concentration don't explain cavitation inception

→ Importance of the surface rugosity ?


PROBLEM SET UP

Non uniform distribution of the nuclei concentration:-nuclei on the boundary layer (thickness = 6 cells)-nuclei on the first cell close to the surface


CONCLUSION

Cavitation is a complex phenomena which- involve multiphase flow, phase transition, no symmetry (need to simulate in 3D), turbulence, instabilities, chock waves, bubbles dynamic, fluid quality, surface rugosity...- and occurs in complex geometry

No empirism free cavitation modelNo comprehensive capabilities to model development of a type of cavitation to an other.

Rather good results for the prediction oh sheet cavitation.Improvement would be achieved if we take account for the influence of the fluid quality, the surface rugosity and the turbulence.


Thank you

29


short about LPT

30

Movie LPT ST<<1 and St>>1

d x p

d t= up

m p

d up

d t=FD=−m p

up−u

p

R e p=D p∣up−u∣

Stk=p

fluid time scale

Aurélia Vallier - LTH

Documents