16 year old solves 300 year old problem set by Isaac Newton

  1. mobile.news.com.au/breaking-news/world/german-teen-shouryya-ray-solves-300-year-old-mathematical-riddle-posed-by-sir-isaac-newton/story-e6frfkui-1226368490521


    I am still trying to figure out what the original problem was. Any thoughts on this?
  2. jcsd
  3. I find it hard to believe a problem like that "stumped" mathematicians (and physicists too, I guess) for this long, only to be solved by a 16 year old kid.
  4. AlephZero

    AlephZero 7,248
    Science Advisor
    Homework Helper

    Whatever it was, it won him 2nd prize in a competition and it has been published - but I don't read German well enough to spend time finding any more. http://en.wikipedia.org/wiki/Shouryya_Ray
  5. Hurkyl

    Hurkyl 15,987
    Staff Emeritus
    Science Advisor
    Gold Member

    Read the talk page. At this moment:

    The article's only citation is to an Indian news website which repeats the claims of the British tabloid The Daily Mail. This is not a reliable source. —Psychonaut (talk) 11:23, 27 May 2012 (UTC)
    yeah, I'm seeing this story everywhere but Can't find any details on the actual math involved. (talk) 11:38, 27 May 2012 (UTC)
    Where is the maths problem and what was his solution? (talk) 11:44, 27 May 2012 (UTC)
    ha I guess I am not the only one looking for the problem. It's just annoying when you hear something like an unsolved problem in physics and they don't tell you the actual problem. — Preceding unsigned comment added by (talk) 12:17, 27 May 2012 (UTC)
    Page should be deleted and recreated some time in the future if the story turns out to be true. It's too soon and Wikipedia is not a news source. There should also be a verifiable citation of the nature of the two problems in question and that they actually were regarded as unsolven previously. (talk) 14:02, 27 May 2012 (UTC)​
  6. The problem is to calculate the trajectory of a body thrown at an angle in the Earth's gravitational field and in a Newtonian fluid. His paper claims to be the first analytical solution to the problem.
  7. In "Mathematical Aspects of Classical and Celestial Mechanics", Arnold & co. claim that this problem was solved by Legendre for a wide class of power law resistance terms of the form [tex]c v^\gamma[/tex]. The extract is attached.

    Arnold claims that the 1st order equation which the system reduces to is soluble by the method of variation of parameters, but when he says something like this you always get the impression he's ducking out. But what do you know, Wolfram alpha solves it so I assume the method must work eventually.

    Maybe the solution here is for more complicated force laws, or for a particle which perhaps has angular momentum or something? Of course, it's also possible that everyone (outside of Russia) simply forgot that the solution had ever been found.

    Attached Files:

  8. How do they know he figured out the actual solution if it has stumped mathematicians for so many years? That said, things like this have happened. There was a woman who, purely by random chance, figured out how to solve some kind of mathematical color theorems that had stumped mathematicians for many years.
  9. AlephZero

    AlephZero 7,248
    Science Advisor
    Homework Helper

    This seems to be the primary source for the story. http://www.jufo-dresden.de/projekt/teilnehmer/matheinfo/m1 Computer translation:

    Without seeing his actual competition entry, comparing it with any previous work is just speculation IMO. Perhaps the press is ignoring the second problem because it doesn't have an nice headline like "Indian kid is smarter than Newton".
  10. K^2

    K^2 2,470
    Science Advisor

    Alpha solves it because it interprets your equation for u[k], rather than u[a]. Partial u with respect to alpha is zero, in this case, so naturally, solution is just the remainder of your equation, which is a non-differential equation. You really should never rely on Alpha to interpret your equations correctly. Always double-check. Better yet, skip Alpha and use Mathematica.

    Mathematica does solve this equation down to an integral which probably cannot be evaluated analytically.
  11. mfb

    Staff: Mentor

    Solving problems can be tricky - checking the solution is usually much easier.

    While the computer translation in AlephZero's post is a bit funny, it contains everything relevant. The actual problem and the solution are not given.
  12. I've also been searching high and low for his paper, to no avail, though I did run across one photo of him holding his equation, which looked quite simple for such a vicious problem. (The drag on a projectile is a function of the velocity squared (with caveats), and the velocity decreases based on the drag. The current method of solving the problem is iterative interpolation using data from standard reference projectiles.)

    From some other references, I gather he approached it as a damping problem, mentioning attempts at it by Hertz and Stokes.
  13. russ_watters

    Staff: Mentor

    I'm having deja vu - this just happened a few months ago.
  14. AlephZero

    AlephZero 7,248
    Science Advisor
    Homework Helper

    I took the computer translation as meaning they were essentially two separate problems, the second one being Hertzian contact with the wall (including some model of energy loss during the impact).

    If he has achieved anything significant on the contact/impact problem, I would be professionally interested in seeing it. Modelling this numerically as part of a larger mechanical system is usually a PITA.
  15. D H

    Staff: Mentor

    It's always a good idea to check whether Mathematica made some amazingly dumb mistake. And it did.

    Here is the correct solution: http://www.wolframalpha.com/input/?i=u'(a)+k*u(a)*tan(a)++k*c*arccos(a)=0. Note that the solution contains a definite integral.

    What about that definite integral? As an indefinite integral, Wolfram alpha just gives up . http://www.wolframalpha.com/input/?i=integrate+(arccos(x)/cos(x)^k)*dx As a definite integral it times out.
    Last edited: May 28, 2012
  16. Did a 16 year old solve a centuries-old problem by Newton?

    It seems like either something important (I'm not sure what) has just happened, or this is just another baseless tempest in a teapot manufactured by the media. But see here. They're claming that a 16-year old kid named Shouryya Ray just solved a problem posed Newton centuries ago, concerning the trajectory of a particle in the Earth's gravitational field subject to air resistance. They're also claiming that in the course of his work, he solved a problem of linear damping in a Newtonian fluid posed by Stokes in 1850 and another linear damping problem concerning collision of a ball and a wall posed by Hertz in 1858. Apparently for this work he won 2nd place in the national high school science competition in Germany.

    Here's the abstract or description of his work (via Google Translate):
    What's going on here? Can anyone find out any details about this if it's significant, or this is just a much ado about nothing?
  17. K^2

    K^2 2,470
    Science Advisor

  18. jtbell

    Staff: Mentor

    Threads merged.
  19. D H

    Staff: Mentor


    My main point still stands: Wolfram alpha (Mathematica) occasionally makes some absolute howlers. It assumed [itex]\frac{du}{da}[/itex] meant [itex]\frac{\partial u(k)}{\partial a}[/itex]. It then assumed that since u is a function of k that this means [itex]\frac{\partial u(k)}{\partial a}=0[/itex]. That is a howler.
  20. K^2

    K^2 2,470
    Science Advisor

    That's purely Alpha. It does best it can to interpret the input. How the heck is it supposed to know what u is a function of? In Mathematica, you would have to enter it explicitly.

    Code (Text):
    DSolve[u'[a]+k*u[a]*Tan[a] +k*c*ArcCos[a]==0, u, a]
    This way, there can be no ambiguity. But there is no room for user error, either. If you put = instead of ==, Mathematica isn't going to try and guess what you meant. It will actually treat that whole expression as 0 from there on, because that's what your code requested. Alpha tries to be peasant-friendly, so it will obviously resolve ambiguities in favor of simplicity.
Know someone interested in this topic? Share this thead via email, Google+, Twitter, or Facebook

Have something to add?