Hi! I’m trying to learn FEniCS and I thought I’d try applying it to the initial conditions from this arxiv article I found on simulating the 2D Kelvin-Helmholtz instability.
I’ve managed to collect a bunch of ad hoc solutions to various problems I’ve run into while setting this up, and I think I’ve got something now. I’m mostly adapting the example of the FEniCS tutorial’s Navier-Stokes solver. Here’s my code:
from sympy import *
x, y = symbols('x[0] x[1]')
U, C, delta = symbols('U_oo c_n delta')
phi = U * exp(-(y-0.5)**2/delta**2) * (cos(8*pi*x) + cos(20 * pi * x))
ux = U * tanh((2 * y - 1)/delta) + C * phi.diff(y)
uy = - C * phi.diff(x)
ux_expression = ccode(ux).replace("M_PI", "pi")
uy_expression = ccode(uy).replace("M_PI", "pi")
from fenics import *
from mshr import *
import numpy as np
from matplotlib import pyplot as plt
T = 0.05 # final time
num_steps = 50 # number of time steps
dt = T / num_steps # time step size
mu = 0.0 # dynamic viscosity
rho = 1 # density
U_oo = 1
c_n = 1e-3
delta = 1/28
# Create mesh
channel = Rectangle(Point(0, 0), Point(1, 1))
domain = channel
mesh = generate_mesh(domain, 64)
class PeriodicBoundary(SubDomain):
# Left boundary is "target domain" G
def inside(self, x, on_boundary):
return bool(x[0] < DOLFIN_EPS and x[0] > -DOLFIN_EPS and on_boundary)
# Map right boundary (H) to left boundary (G)
def map(self, x, y):
y[0] = x[0] - 1.0
y[1] = x[1]
pbc = PeriodicBoundary()
# Define function spaces
V = VectorFunctionSpace(mesh, 'P', 2, constrained_domain=pbc)
Q = FunctionSpace(mesh, 'P', 1, constrained_domain=pbc)
# Define boundaries
inflow = 'near(x[0], 0)'
outflow = 'near(x[0], 1)'
walls = 'near(x[1], 0) || near(x[1], 1)'
# Define inflow profile
inflow_profile = ('4.0*1.5*x[1]*(1 - x[1]) / pow(1, 2)', '0')
initial_condition_profile = (
ux_expression, uy_expression,
)
initial_condition = Expression(initial_condition_profile,
degree=2,
U_oo = U_oo,
c_n = c_n,
delta = delta,
)
# Define boundary conditions
bcu_free_slip = DirichletBC(V.sub(1), 0.0, walls)
bcp_outflow = DirichletBC(Q, Constant(0), outflow)
bcu = [
bcu_free_slip,
]
bcp = [bcp_outflow]
# Define trial and test functions
u = TrialFunction(V)
v = TestFunction(V)
p = TrialFunction(Q)
q = TestFunction(Q)
# Define functions for solutions at previous and current time steps
u_n = Function(V)
u_ = Function(V)
u_n.interpolate(initial_condition)
u_.interpolate(initial_condition)
p_n = Function(Q)
p_ = Function(Q)
u_n
# Define expressions used in variational forms
U = 0.5*(u_n + u)
n = FacetNormal(mesh)
f = Constant((0, 0)) # is this gravity?
k = Constant(dt)
mu = Constant(mu)
rho = Constant(rho)
# Define symmetric gradient
def epsilon(u):
return sym(nabla_grad(u))
# Define stress tensor
def sigma(u, p):
return 2*mu*epsilon(u) - p*Identity(len(u))
# Define variational problem for step 1
# time stepping
# convective term
# stress tensor - diffusion and pressure
# pressure on boundary
# diffusion
# gravity????
F1 = rho*dot((u - u_n) / k, v)*dx \
+ rho*dot(dot(u_n, nabla_grad(u_n)), v)*dx \
+ inner(sigma(U, p_n), epsilon(v))*dx \
+ dot(p_n*n, v)*ds \
- dot(mu*nabla_grad(U)*n, v)*ds \
- dot(f, v)*dx
a1 = lhs(F1)
L1 = rhs(F1)
# Define variational problem for step 2
a2 = dot(nabla_grad(p), nabla_grad(q))*dx
L2 = dot(nabla_grad(p_n), nabla_grad(q))*dx - (1/k)*div(u_)*q*dx
# Define variational problem for step 3
a3 = dot(u, v)*dx
L3 = dot(u_, v)*dx - k*dot(nabla_grad(p_ - p_n), v)*dx
# Assemble matrices
A1 = assemble(a1)
A2 = assemble(a2)
A3 = assemble(a3)
# Apply boundary conditions to matrices
[bc.apply(A1) for bc in bcu]
[bc.apply(A2) for bc in bcp]
# Create XDMF files for visualization output
# xdmffile_u = XDMFFile('navier_stokes_kh/velocity.xdmf')
# xdmffile_p = XDMFFile('navier_stokes_kh/pressure.xdmf')
# Time-stepping
t = 0
from tqdm import auto
with auto.tqdm(total=num_steps) as pbar:
for n in range(num_steps):
# Update current time
t += dt
# Step 1: Tentative velocity step
b1 = assemble(L1)
[bc.apply(b1) for bc in bcu]
solve(A1, u_.vector(), b1, 'bicgstab', 'hypre_amg')
# Step 2: Pressure correction step
b2 = assemble(L2)
[bc.apply(b2) for bc in bcp]
solve(A2, p_.vector(), b2, 'bicgstab', 'hypre_amg')
# Step 3: Velocity correction step
b3 = assemble(L3)
solve(A3, u_.vector(), b3, 'cg', 'sor')
# Plot solution
plt.figure()
mappable = plot(u_**2, title='Velocity squared')
plt.colorbar(mappable)
# plt.figure()
# plot(p_, title='Pressure')
# # Save solution to file (XDMF/HDF5)
# xdmffile_u.write(u_, t)
# xdmffile_p.write(p_, t)
# Update previous solution
u_n.assign(u_)
p_n.assign(p_)
pbar.set_postfix({'u max': u_.vector().max(), "t": t})
pbar.update(1)
# xdmffile_u.close()
# xdmffile_p.close()
However, on running this within a jupyter notebook, my velocity blows up to infinity somewhere on the mesh - the plot doesn’t display it. I have tried bumping up the number of steps to 500, but that doesn’t seem to have helped.
Is there anything obvious that I’m missing here?