How does variable density and viscosity affect Reyleigh-Taylor instability?

Click For Summary

Homework Help Overview

The discussion revolves around modifying a computational program to study the effects of variable density and viscosity on Rayleigh-Taylor instability. The original poster is working on a thesis related to combustion and seeks assistance in adapting existing code to include a scalar variable representing liquid fraction.

Discussion Character

  • Mixed

Approaches and Questions Raised

  • Participants have raised concerns about the clarity and readability of the provided code, suggesting that it lacks sufficient comments and structure. There are questions about the specific modifications needed to incorporate the scalar variable phi and how it relates to existing parameters in the code.

Discussion Status

Some participants have offered guidance on improving the clarity of the original poster's code, emphasizing the importance of proper formatting and detailed explanations. The conversation is ongoing, with multiple interpretations of the problem being explored.

Contextual Notes

There is mention of a diffuse interface at phi = 0.5, and the original poster has indicated that surface tension will not be included in the modifications. The participants are also questioning the adequacy of the problem statement and the definitions of variables used in the code.

MrOne
Messages
1
Reaction score
0

Homework Statement


import time
import numpy as np
from Output import *
from PoissonSolver import *
tmps1=time.clock()

# Physical parameters
Re = 10.

# Numerical parameters
CFL = 0.9
dt = 0.001/4
t = 0.0
tmax = 1.5

# Make mesh
from MeshGeneration import *
mesh = Mesh(2.0,1.0,50,25)

# Output parameters
dtAff = 0.01 # s

# Boundary condition
# Cavity
#BC = {"type":{"Left": "Wall", "Right": "Wall","Bottom": "Wall", "Top": "MovingWall"},
# "u":{"Top": 1.0},
# "v":{}}
# Forcing Vortex
# BC = {"type":{"Left": "MovingWall", "Right": "MovingWall","Bottom": "MovingWall", "Top": "MovingWall"},
# "u":{"Top": 1.0, "Bottom": -1.0},
# "v":{"Left": 1.0, "Right": -1.0}}
# Channel
BC = {"type":{"Left": "Inflow", "Right": "Outflow","Bottom": "Wall", "Top": "Wall"},
"u":{"Left": 1.0},
"v":{"Left": 0.0},
"p":{"Right": 0.0}}
#BC = {"type":{"Left": "Wall", "Right": "Wall","Bottom": "Outflow", "Top": "Inflow"},
# "u":{"Top": 0.0},
# "v":{"Top": -1.0}}

# Initial condition
from InitNS import *
u, v, p = InitRest(mesh)

if BC["type"]["Left"] == "Inflow":
u[0,:] = BC["u"]["Left"]
if BC["type"]["Top"] == "Inflow":
v[:,mesh.Ny] = BC["v"]["Top"]

# Define fluxes
from NS_fluxes import GRAD, STRESS, TRANS, DIV

# Create Matrix for Poisson Solvre
A, bp = CreatePoissonMatrix(mesh, BC)

savefigNS(u,v,p,t,mesh)

while t<tmax:
ustar, vstar = np.copy(u), np.copy(v)
utrans, vtrans = TRANS(u,v,mesh,BC)
ustress, vstress = STRESS(u,v,mesh,BC)

ustar[1:-1,:] = u[1:-1,:] + dt*(utrans+1.0/Re*ustress)
vstar[:,1:-1] = v[:,1:-1] + dt*(vtrans+1.0/Re*vstress)
b = DIV(ustar,vstar,mesh)
p = PoissonSolver(A,b,bp,mesh)
#p = PoissonSolver(b,BC,mesh)
uprime, vprime = GRAD(p,mesh)
u, v = ustar-uprime, vstar-vprime

t = t + dt
# Plot
if (abs(t % dtAff) < dt):
savefigNS(u,v,p,t,mesh)
tmps2=time.clock()
print "Temps d'execution = %d\n" %(tmps2-tmps1)

Homework Equations


Hello everyone, i need your help to modify this program to include an scalar phi corresponding to a liquid fraction. i want to include the modification of the equations in order to take into account a variable density and viscosity. Here, no surface tension will be included. The diffuse interface will be at phi = 0.5.
I want to test the case of Reyleigh-Taylor instability

The Attempt at a Solution



Sorry everyone, I am not in this domain, I am doing a thesis in combustion but i need this program in my works.
Thank you in advance for your help, i will appreciate that.
 
Physics news on Phys.org
Did you really expect anyone to be able to figure out what your coding is all about without any real detailed explanation?
 
Some points to help you get help:
Please use code tags, and comment your code. Why? Well, for example python is sensitive to leading spaces. Which is a good thing.
Your question is currently a wall of hard-to-read and syntactically incorrect code.

https://www.quora.com/Why-is-Python-space-sensitive
 
Hello mr 1, :welcome:

As you can read in the PF guidelines, 'Hello everyone' does not constitute a relevant equation, and 'Sorry everyone' does not count as an attempt at solution. So please do better than that.

And under 1. I expect at least a problem statement and I'm not referring to
MrOne said:
modify this program to include an scalar phi corresponding to a liquid fraction
Does that mean there is already a tensor ##\phi## in there and you want it simpler ?

And how about a desscription and a list of variables and their meaning ?

I can put ##[##code=python##]## and ##[##\code##]## around your post, but that - of course - does not restore the missing spaces:

Python: 
import time
import numpy as np
from Output import *
from PoissonSolver import *
tmps1=time.clock()

# Physical parameters
Re = 10.

# Numerical parameters
CFL = 0.9
dt = 0.001/4
t = 0.0
tmax = 1.5

# Make mesh
from MeshGeneration import *
mesh = Mesh(2.0,1.0,50,25)

# Output parameters
dtAff = 0.01 # s

# Boundary condition
# Cavity
#BC = {"type":{"Left": "Wall", "Right": "Wall","Bottom": "Wall", "Top": "MovingWall"},
# "u":{"Top": 1.0},
# "v":{}}
# Forcing Vortex
# BC = {"type":{"Left": "MovingWall", "Right": "MovingWall","Bottom": "MovingWall", "Top": "MovingWall"},
# "u":{"Top": 1.0, "Bottom": -1.0},
# "v":{"Left": 1.0, "Right": -1.0}}
# Channel
BC = {"type":{"Left": "Inflow", "Right": "Outflow","Bottom": "Wall", "Top": "Wall"},
"u":{"Left": 1.0},
"v":{"Left": 0.0},
"p":{"Right": 0.0}}
#BC = {"type":{"Left": "Wall", "Right": "Wall","Bottom": "Outflow", "Top": "Inflow"},
# "u":{"Top": 0.0},
# "v":{"Top": -1.0}}

# Initial condition
from InitNS import *
u, v, p = InitRest(mesh)

if BC["type"]["Left"] == "Inflow":
u[0,:] = BC["u"]["Left"]
if BC["type"]["Top"] == "Inflow":
v[:,mesh.Ny] = BC["v"]["Top"]

# Define fluxes
from NS_fluxes import GRAD, STRESS, TRANS, DIV

# Create Matrix for Poisson Solvre
A, bp = CreatePoissonMatrix(mesh, BC)

savefigNS(u,v,p,t,mesh)

while t<tmax:
ustar, vstar = np.copy(u), np.copy(v)
utrans, vtrans = TRANS(u,v,mesh,BC)
ustress, vstress = STRESS(u,v,mesh,BC)

ustar[1:-1,:] = u[1:-1,:] + dt*(utrans+1.0/Re*ustress)
vstar[:,1:-1] = v[:,1:-1] + dt*(vtrans+1.0/Re*vstress)
b = DIV(ustar,vstar,mesh)
p = PoissonSolver(A,b,bp,mesh)
#p = PoissonSolver(b,BC,mesh)
uprime, vprime = GRAD(p,mesh)
u, v = ustar-uprime, vstar-vprime

t = t + dt
# Plot
if (abs(t % dtAff) < dt):
savefigNS(u,v,p,t,mesh)
tmps2=time.clock()
print "Temps d'execution = %d\n" %(tmps2-tmps1)