Nested Loops Not Changing with Time

I am working on a problem that simulates flow break up. My code is supposed to show how the fluid breaks up over time. I have run into the issue that the flow doesn't change with time. I was wondering if a new set of eyes could help me figure out why this is happening.

for n=1:10000000
   
 for j=2:N-2
    un(j)=uo(j)+dt*S(j)...
         +3*dt*mu/(rho*favg(j))*(f(j+1)-f(j))/dN*(uo(j+1)-uo(j-1))/(2*dN)...
         +3*dt*mu/rho*(uo(j+1)-2*uo(j)+uo(j-1))/dN^2 ...
         -dt*uo(j)*(uo(j+1)-uo(j-1))/(2*dN);
    un(1)=-un(2); %symmetric BC
    un(N-1)=-un(N-2); %symmetric BC
    uo(j)=un(j);
    fn(j)=f(j)-dt*f(j)*((uo(j)-uo(j-1))/dN)-dt*uo(j)*((f(j+1)-f(j-1))/(2*dN));
    hn(j)=(fn(j)^0.5);
    f(j)=fn(j);
    h(j)=hn(j);
 end     
hmin=min(h);
if hmin<bkthd
    break 
end 
Time=n*dt;
figure(1)
plot(h)
hold on 
plot(-h)
title(['Time=', num2str(Time), ''])
end

Thanks in advance!

Answer 1

[the cyclist]

Nothing in your inner for loop depends on n. Things depend on N. So it looks like that inner loop just does the same calculation over and over again.

[Monique Embury]

So I modified my code to include n, but it seems to still not be working with replacing values... any suggestions?

for n=1:nt
  
 for j=2:N
    un(n+1,j)=uo(n,j)+dt.*S(n,j)...
         +3*dt*mu/(rho*favg(n,j)).*(f(n,j+1)-f(n,j))/dN.*(uo(n,j+1)-uo(n,j-1))/(2*dN)...
         +3*dt*mu/rho*(uo(n,j+1)-2*uo(n,j)+uo(n,j-1))/dN^2 ...
         -dt*uo(n,j).*(uo(n,j+1)-uo(n,j-1))/(2*dN);
    un(n+1,1)=-un(n+1,2); %symmetric BC
    un(n+1,N+1)=-un(n+1,N); %symmetric BC
    uo(n,j)=un(n+1,j);
    uo(n,1)=un(n+1,1);
    uo(n,N+1)=un(n+1,N+1);
    fn(n+1,j)=f(n,j)-dt*f(n,j).*((uo(n,j)-uo(n,j-1))/dN)-dt*uo(n,j).*((f(n,j+1)-f(n,j-1))/(2*dN));
    fn(n+1,1)=fn(n+1,3);
    fn(n+1,N+1)=fn(n+1,N-1);
    f(n,j)=fn(n+1,j);
    favg(n,j-1)=0.5*(f(n,j-1)+f(n,j));
    fz(n,j)=0.5*(f(n,j+1)-f(n,j-1))/dN;
    fzz(n,j)=(f(n,j+1)-2*f(n,j)+f(n,j-1))/(dN)^2;
    fz(n,1)=0;
    fzz(n,1)=(2*f(n,2)-2*f(n,1))/(dN)^2;
    fz(n,N+1)=0;
    fzz(n,N+1)=(2*f(n,N)-2*f(n,N+1))/(dN)^2;
    H(n,j)=0.25*((f(n,j).*(2-fzz(n,j))+(fz(n,j)).^2)./(f(n,j)+0.25*(fz(n,j)).^2).^(3/2));
    H(n,1)=0.25*((f(n,1).*(2-fzz(n,1))+(fz(n,1)).^2)./(f(n,1)+0.25*(fz(n,1)).^2).^(3/2));
    H(n,N+1)=0.25*((f(n,N+1).*(2-fzz(n,N+1))+(fz(n,N+1)).^2)./(f(n,N+1)+0.25*(fz(n,N+1)).^2).^(3/2));
    S(n,j)=-2*sigma/rho*(H(n,j+1)-H(n,j))/dN;
    S(n,1)=-2*sigma/rho*(H(n,2)-H(n,1))/dN;
    h(n,j)=(f(n,j).^0.5);
    
 end   
 
end


 

[the cyclist]

Your code is too complex for me to just look at it and figure out what you intend, or what might be going wrong. If you want to upload your full code, such that we could actually run it, that would help.

I can tell you some things I would do myself, to try to debug.

Separate the different segments visually, and write comments for each section (or possibly each line) so that it is very clear what it is supposed to be doing, e.g. ...

% Describe the H calculation
H(n,j)=0.25*((f(n,j).*(2-fzz(n,j))+(fz(n,j)).^2)./(f(n,j)+0.25*(fz(n,j)).^2).^(3/2));
H(n,1)=0.25*((f(n,1).*(2-fzz(n,1))+(fz(n,1)).^2)./(f(n,1)+0.25*(fz(n,1)).^2).^(3/2));
H(n,N+1)=0.25*((f(n,N+1).*(2-fzz(n,N+1))+(fz(n,N+1)).^2)./(f(n,N+1)+0.25*(fz(n,N+1)).^2).^(3/2));

% Describe the S calculation
S(n,j)=-2*sigma/rho*(H(n,j+1)-H(n,j))/dN;
S(n,1)=-2*sigma/rho*(H(n,2)-H(n,1))/dN;

% Describe the h calculation
h(n,j)=(f(n,j).^0.5);

Perhaps make more descriptive variable names. (Although maybe these are obvious to you.)

Pull lines like these:

uo(n,1)=un(n+1,1);
uo(n,N+1)=un(n+1,N+1);

outside of the loop over j, since they have no dependence on j. (I wonder if this is part of the problem.)

Use the debugger to stop the code at the very first line of the for loop, to see if what you get what you expect, rather than waiting until the entire loop is done to check the final results.

[Monique Embury]

Thank you for your suggestions! I have attached a copy of my code, I would greatly appreciate it if you could take a look into it.