Skip to content

MethodOfLines issue with coupled non-linear PDE system #398

@RayanJobs

Description

@RayanJobs

Hi. I am trying to solve a highly-coupled system of nonlinear PDEs. I have attached the code, and I am happy to answer any questions. I have a 2D domain (x,z) and time dependence. My two variables are c (x,z,t), a 2D concentration, and h(x,t), a height. Periodicity is assumed in x and the z domain is transformed onto a fixed domain (such that it is fixed between 0 and 1 and the interface does not move). I keep having issues (current issue is ERROR: LoadError: AssertionError: islinear
Stacktrace). I am not sure why this is happening, in particular because my system of equations will include an additional pde later on.

hb = 0.5
s = 3.0


Ma = 0.01
Ma_s = 0.01
eps = 0.1
cppp = 0.01
vv = 0.1
Pe = 10000.0
α = 0.3

using Plots, ModelingToolkit, MethodOfLines, OrdinaryDiffEq, DomainSets


@parameters x z t 
@variables h(..) c(..)
Dt = Differential(t)
Dx = Differential(x)
Dz = Differential(z)
Dxx = Differential(x)^2
Dzz = Differential(z)^2
Dxxx = Differential(x)^3
Dxxxx = Differential(x)^4
Dxz = Differential(x) * Differential(z)  # Mixed derivative, d^2(dx dz)

brusselator_f(x, z) =  (1 + cppp * cos* x) * exp(-(z.-hb)^2/(2*vv^2))) * ((z .<= 0.5 - 1/s) * 1.0 +  (0.5 .- 1/s .< z .< 0.5 + 1/s) * exp(-1/(1-((z .- (hb - 1.0/s))/(.0/s)) ))/(exp(-1/(1-((z .- (hb - 1.0/s))/(.0/s)))) + exp(-1/((z .- (hb - 1.0/s))/(.0/s)))) + (z .>= 0.5 + 1/s) * 0.0)




#plot(z, brusselator_f(0,z,0), label="T(z)", xlabel="z", ylabel="f(z)", title="Plot of initial condition")






#∇²(u) = Dxx(u) + Dyy(u)

x_min = y_min = t_min = 0.0
x_max = y_max = 1.0
t_max = 10.0


eqs = [Dt(c(x,z,t)) ~ 1/Pe*((2*z*(1/h(x,t)*Dx(h(x,t)))^2-z/h(x,t)*Dxx(h(x,t)))*Dz(c(x,z,t))+(z/h(x,t)*Dx(h(x,t)))^2*Dzz(c(x,z,t))-2*z/h(x,t)*Dx(h(x,t))*Dxz(c(x,z,t))+ Dxx(c(x,z,t))) + 1/(eps^2 * Pe * (h(x,t))^2) * Dzz(c(x,z,t)), 
Dt(h(x,t)) ~ -(h(x,t))^2 * Dx(h(x,t))* Dxxx(h(x,t)) - 1/3 * h(x,t)^3 * Dxxxx(h(x,t)) + h(x,t) * Dx(h(x,t)) * Dx(c(x,1.0, t)) + 1/2 * (h(x,t))^2 *  Dxx(c(x,1.0, t))]

domains = [x  Interval(x_min, x_max),
              z  Interval(y_min, y_max),
              t  Interval(t_min, t_max)]

# Periodic BCs
bcs = [h(x,0.0) ~ 1.0,
       h(0.0,t) ~ h(1.0,t),
       
       c(x,z,0.0) ~ brusselator_f(x, z),
       c(0.0,z,t) ~ c(1.0,z,t),
       Dz(c(x,1.0,t)) ~ 0.0,
       Dz(c(x,0.0,t)) ~ 0.0] 

# Initial condition above models a disturbance

@named pdesys = PDESystem(eqs,bcs,domains,[x,z,t],[c(x,z,t),h(x,t)])

N = 100
order = 2 # This may be increased to improve accuracy of some schemes


discretization = MOLFiniteDifference([x=>N, z=>N], t, approx_order=order)





println("Discretization:")
@time prob = discretize(pdesys,discretization)

println("Solving RN:")
@time sol = solve(prob, TRBDF2(), saveat=0.1)

using JLD2
@save "new_solutionn.jld2" sol

Metadata

Metadata

Assignees

No one assigned

    Labels

    questionFurther information is requested

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions