## Solve a non linear differential equation systems with only boundary conditions

### violette (view profile)

on 6 May 2014
Latest activity Answered by Bill Greene

### Bill Greene (view profile)

on 6 May 2014
Hi everybody,
I am currently trying to solve a differential equation of the form :
function dy= myfun(r,y)
N=200e3;
g=1.6861;
dy=zeros(2,1);
dy(2)=y(1).*r.*(N*pi/60)^2;
dy(1)=(y(1)./(g.*y(2))).*y(1).*r.*(N*pi/60)^2;
end
Only , I know from my physical problem only what happened on the edge (at r2=10mm). I tried then to use bvp4c, but I have the feeling they would need anyway at least one initial condition.
guess=[val1;val2];
solinit=bvpinit(linspace(0,r2),40),guess)
sol=bvp5c(@interdisc_space_compressible,@bc,solinit)
where val1=y1(0) et val2=y2(r2) However, I would like to set val1=y1(r2) and val2=y2(r2) instead!
Any hint??
Thanks!!!

### Bill Greene (view profile)

Answer by Bill Greene

### Bill Greene (view profile)

on 6 May 2014

You have a system of two first order differential equations so you need just two boundary conditions. I would guess that specifying y1(r2) and y2(r2) is fine.
What kind of problem are you running into? If you post your complete example, we might be able to help.
Bill