2
$\begingroup$

I wanted to solve the following PDE with initial condition $$ u_t+tu_x=0, $$ $$ u(x,1)=f(x),$$ where $f(x)$ is a given function, using the method of characteristics.

I explain what I have done. First of all the characteristic system is $$ x'(\tau)=t $$ $$ t'(\tau)=1 $$ $$ u'(\tau)=u $$ with initial conditions $$ x(s,1)=s $$ $$ t(s,1)=1 $$ $$ u(s,1)=f(s) $$

It follows that $$x(s,\tau)=\tau^2+s-1,$$ $$t(s,\tau)=\tau,$$ $$u(s,\tau)=e^{\tau-1}f(s).$$ Therefore, $s=x-t^2+1$ and $\tau=t$. Hence $$ u(x,t)=u(x-t^2+1,t)=e^{t-1}f(x-t^2+1).$$

But this is wrong because it does not satisfy the system. What would be right is $$ u(x,t)=e^{t-1}f(x-t^2/2+1/2).$$ But I do not know where this $1/2$ can come from...

Thanks in advance!

$\endgroup$
2
$\begingroup$

The integral of $\tau$ is $\tau^2/2$, not $\tau$. It should be $$ x=\frac{\tau^2}{2}+s-\frac12. $$

$\endgroup$
  • $\begingroup$ Thank you very much! This means that in the characteristic system must be $x'(\tau)=\tau$, not $x'(\tau)=t$, true? In spite of we know $t=\tau$ after solving $t'(\tau)=1$ with initial condition $t(s,1)=1$. $\endgroup$ – S. Proa Feb 1 '16 at 15:54
  • $\begingroup$ Yes, that's it. $\endgroup$ – Julián Aguirre Feb 1 '16 at 17:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.