1
$\begingroup$

I once tried to read the Millennium Problems statement about the Navier-Stokes equations, decided it was beyond me, and left it at that.

But now I am studying general relativity and through Physics SE learned that the Einstein Field Equations once didn't have exact solutions, but now do ever since the Schwartzchild solution. This sparked my curiosity and wonder about why solutions haven't been found for the Navier-Stokes equations. If it has been done once, perhaps someone will do it for these.

Can someone give me a Navier-Stokes for Dummies version of why exact solutions haven't been found?

$\endgroup$
7
$\begingroup$

There are known solutions to the Navier-Stokes equations. A simple example would be laminar shear-driven flow between two moving plates. Just as in the case of Einstein's equations, the known solutions regard simple situations with particular boundary conditions; a general solution that covers all possible cases is not known in either case. One should not expect such a thing ever to be found, since these systems can exhibit chaos; the difficulty of finding a general solution is the same as for the n-body problem in Newtonian gravity.

However, the millennium problem for the Navier-Stokes equations is something different. It doesn't ask for a single solution, it asks whether a solution always exists, for any initial and boundary conditions. Then, if solutions do always exist, it asks whether they will always have certain technical properties. But it doesn't ask us to find the solution, because we know that in general this will not be possible.

I'm no expert, but I believe the reason this is harder for the Navier-Stokes equations than for Einstein's equations is that the Navier-Stokes equations for an incompressible fluid are inherently non-local; the incompressibility constraint means that something happening in one part of the fluid can instantly affect every other part of the system. (This is, of course, an approximation to the physical reality.) In contrast, general relativity is inherently local.

$\endgroup$
  • 2
    $\begingroup$ Nathaniel beat me to it! All I would add is that Terence Tao has an excellent (as usual :-) article on the problem on his blog. The article should be accessible to most physicists if perhaps not the general public. $\endgroup$ – John Rennie Jan 23 '15 at 6:33

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.