I was working on a physics project and have to solve this nonlinear ODE:-

$\dfrac{dy}{dx}=\dfrac{ky^3 +ey}{axy(y+b) + c(y-b)/\sqrt{x}}$,

where $a, b, c, k$ and $e$ are constants.

I haven't had much experience solving nonlinear ODEs, but I did try obtaining a Taylor series solution. It turned out to be far too complicated though. I do know there are simpler numerical methods, but I'm looking for an analytical solution. Can someone help me with this?

  • $\begingroup$ Do you really need to solve it or qualitative behaviour is fine too? $\endgroup$ – Evgeny Feb 14 '14 at 17:51
  • $\begingroup$ Are you sure that an analytical solution exists? Unfortunately many nonlinear ODEs can not be solved analytically. $\endgroup$ – Doubt Feb 14 '14 at 17:52
  • $\begingroup$ @Doubt: I'm not familiar with uniqueness and existence theorems yet, so I can't say for sure. And I'm not aware of any other techniques that would tell me that, either. $\endgroup$ – Train Heartnet Feb 14 '14 at 18:09
  • $\begingroup$ @Evgeny: I do need to solve it, as the function obtained has to be used elsewhere. Behavior of the solution won't suffice. However, if such a solution does not exist, that would be valuable information. $\endgroup$ – Train Heartnet Feb 14 '14 at 18:12

For the simple case where $a, b,c,k ,e = 1$, Mathematica gives the solution implicitly as \begin{align} \int_1^{y(x)}\frac{\xi-2}{\xi\left(1+\xi^2\right)^{7/4}\exp\left(\frac{3}{2}\arctan\xi\right)}\,d\xi = \frac{2x^{3/2}}{3\left(1+y^2(x)\right)^{3/4}\exp\left(\frac{3}{2}\arctan y(x)\right)} + c \end{align} There does not seem to be a closed-form expression for the integral, unfortunately. Probably this form is no more helpful than the original ODE in terms of solving the problem numerically.

  • $\begingroup$ Thank you for trying! I guess this is the closest one could get to a solution. But just out of curiosity (and desperateness), if I had an initial condition like $y(0)=0$, would that help in any way? $\endgroup$ – Train Heartnet Feb 14 '14 at 19:53
  • 1
    $\begingroup$ No, the initial condition is not likely to make things easier. In fact, the initial condition $y(0) = 0$ may not be physically possible since $dy/dx|_{x=0,\,y=0}$ is undefined. Easier would be $y(0) = 1$, which gives $c = 0$. If $y(0) = 0$ should be physically possible, perhaps revisit the derivation of the ODE. $\endgroup$ – Doubt Feb 14 '14 at 20:29
  • $\begingroup$ On further reflection, I realized that neither $y(0)=1$, nor $y(0)=0$ are physically possible. The initial conditions are way more complicated than that. Thank you so much for all the help you've done though. I am accepting your answer, as that's the closest in terms of a solution. :) $\endgroup$ – Train Heartnet Feb 16 '14 at 4:57

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.