Results 1 to 7 of 7

Math Help - Second order PDE

  1. #1
    Member
    Joined
    Jan 2009
    Posts
    145

    Second order PDE

    So I'm working on a solution to this PDE,
     y * u_{xx} + 3y * u_{xy} = -3u_x;
     y != 0

    I started by classifying the PDE and found it to be hyperbolic. So, I attempted to reduce it to canonical form, making the substitutions

     \eta = y - 3x and  \xi = y

    Then, I obtained that
     u_x = -3 *u_{\eta}
     u_{xx} = 9*u_{\eta \eta}
     u_{xy} = -3( u_{\xi \eta} + u_{\eta \eta})

    Substituting this back into the PDE, I end up with the equation

     u_{\eta \xi} = -u_{\eta} * \frac{1}{\xi}

    At about this point, I'm getting stuck on finding a general solution to the PDE by simplifying it further. I suspect the problem is with my substitution, specifically with  \xi , but I'm not really sure what's wrong with it, other than it looking a bit suspect. Can anyone point me in the right direction?
    Last edited by Math Major; April 4th 2010 at 12:18 PM.
    Follow Math Help Forum on Facebook and Google+

  2. #2
    MHF Contributor
    Jester's Avatar
    Joined
    Dec 2008
    From
    Conway AR
    Posts
    2,392
    Thanks
    56
    Quote Originally Posted by Math Major View Post
    So I'm working on a solution to this PDE,
     y * u_{xx} + 3y * u_{xy} = -3u_x;
     y != 0

    I started by classifying the PDE and found it to be hyperbolic. So, I attempted to reduce it to canonical form, making the substitutions

     \eta = y - 3x and  \xi = y

    Then, I obtained that
     u_x = -3 *u_{\eta}
     u_{xx} = 9*u_{\eta \eta}
     u_{xy} = -3( u_{\xi \eta} + u_{\eta \eta})

    Substituting this back into the PDE, I end up with the equation

     u_{\eta \xi} = -u_{\eta} * \frac{1}{\xi}

    At about this point, I'm getting stuck on finding a general solution to the PDE by simplifying it further. I suspect the problem is with my substitution, specifically with  \xi , but I'm not really sure what's wrong with it, other than it looking a bit suspect. Can anyone point me in the right direction?
    Actually, you're almost there. Separate and integrate

     <br />
\frac{u_{\eta \xi}}{u_{\eta}} = - \frac{1}{\xi}<br />

    so

     <br />
\ln u_{\eta} = - \ln \xi + \ln f'(\eta), f \text{\;is\; arbitrary}<br />

    so

     <br />
u_{\eta} = \frac{f'(\eta)}{\xi}<br />

    Then integrate again.
    Follow Math Help Forum on Facebook and Google+

  3. #3
    Member
    Joined
    Jan 2009
    Posts
    145
    Thank you so much for that. It helped a lot to see that step worked out. Sorry for asking again, but could you possibly take a look at what I've done here and tell me where I'm going wrong? I'm sure it's something stupid.

    I'm working on solving
     u_{xx} + u_{xy} - 2u_{yy} = 3( u_x + 2u_y + 6x - 3y)

    Once again, the equation is hyperbolic so I found appopriate substitutions to be

     \xi = y - 2x and  \eta = y +x

    So, taking the derivatives I find
     u_x = -2u_{\xi} + u_{\eta}
     u_xx = 4(u_{\xi \xi} + u_{\xi \eta}) + u_{\eta \eta}
     u_xy = u_{\eta \eta} -u_{\xi \eta} - 2u_{\xi \xi}
     u_y = u_{\xi} + u_{\eta}
     u_yy = u_{\xi \xi} + 2u_{\xi \eta} + u_{\eta \eta}

    Plugging these back into the pde, I end up with the equation (in canonical form),

     u_{\xi \eta} = \xi - u_{\eta}
    I tried integrating with respect to eta to get

     u_{\xi} = \xi * \eta - u + B'(\xi)
    For some arbitrary B. However, I can't solve this ODE. I tried using an integrating factor, but that got messy quickly. I might have made an algebraic mistake in the reduction, but I've gotten the same canonical form twice.

    If someone could take a look and show me where I went awry, I'd be most apprecitive.
    Follow Math Help Forum on Facebook and Google+

  4. #4
    MHF Contributor
    Jester's Avatar
    Joined
    Dec 2008
    From
    Conway AR
    Posts
    2,392
    Thanks
    56
    Quote Originally Posted by Math Major View Post
    Thank you so much for that. It helped a lot to see that step worked out. Sorry for asking again, but could you possibly take a look at what I've done here and tell me where I'm going wrong? I'm sure it's something stupid.

    I'm working on solving
     u_{xx} + u_{xy} - 2u_{yy} = 3( u_x + 2u_y + 6x - 3y)

    Once again, the equation is hyperbolic so I found appopriate substitutions to be

     \xi = y - 2x and  \eta = y +x

    So, taking the derivatives I find
     u_x = -2u_{\xi} + u_{\eta}
     u_xx = 4(u_{\xi \xi} {\color{red}{+}\,} u_{\xi \eta}) + u_{\eta \eta}
     u_xy = u_{\eta \eta} -u_{\xi \eta} - 2u_{\xi \xi}
     u_y = u_{\xi} + u_{\eta}
     u_yy = u_{\xi \xi} + 2u_{\xi \eta} + u_{\eta \eta}

    Plugging these back into the pde, I end up with the equation (in canonical form),

     u_{\xi \eta} = \xi - u_{\eta}
    I tried integrating with respect to eta to get

     u_{\xi} = \xi * \eta - u + B'(\xi)
    For some arbitrary B. However, I can't solve this ODE. I tried using an integrating factor, but that got messy quickly. I might have made an algebraic mistake in the reduction, but I've gotten the same canonical form twice.

    If someone could take a look and show me where I went awry, I'd be most apprecitive.
    I got a negative sign where there's a plus (in red above) but I agree with your final transformation. From what you have

     <br />
u_{\xi} + u = \xi \eta + F(\xi)\; (F\, \text{arbitrary})<br />

    Integrating factor e^{\xi} so

     <br />
\frac{\partial}{\partial \xi} \left( e^{\xi} u\right) = \xi e^{\xi} \eta + F(\xi)e^{\xi}<br />

    Since F is arbitrary then let F(\xi) e^{\xi} = G'(\xi) where G is arbitrary. Then integrate both sides.
    Follow Math Help Forum on Facebook and Google+

  5. #5
    Member
    Joined
    Jan 2009
    Posts
    145
    Oh, duh, I was getting hung up on trying to find an anti-derivative for  F(\xi) e^{\xi} when it's just as much the derivative of some arbitrary function too.

    Thank you so much for your help.
    Follow Math Help Forum on Facebook and Google+

  6. #6
    Member
    Joined
    Jan 2009
    Posts
    145
    Are you positive on that negative sign?

     u_x = \xi_x u_{\xi} + \eta_x u_{\eta} = -2u_{\xi} + u_{\eta}

    From here,

     u_{xx} = -2 * ( \xi_x u_{\xi \xi} + \eta_x u_{\xi \eta} ) + \xi_x u_{\eta \xi} + \eta_x u_{\eta \eta}

     = -2(-2u_{\xi \xi} + u_{\xi \eta}) - 2u_{\xi \eta} + u_{\eta \eta}

     = 4u_{\xi \xi} - 4u_{\xi \eta} + u_{\eta \eta}

    Or did I make a mistake in that?
    Follow Math Help Forum on Facebook and Google+

  7. #7
    MHF Contributor
    Jester's Avatar
    Joined
    Dec 2008
    From
    Conway AR
    Posts
    2,392
    Thanks
    56
    Quote Originally Posted by Math Major View Post
    Are you positive on that negative sign?

     u_x = \xi_x u_{\xi} + \eta_x u_{\eta} = -2u_{\xi} + u_{\eta}

    From here,

     u_{xx} = -2 * ( \xi_x u_{\xi \xi} + \eta_x u_{\xi \eta} ) + \xi_x u_{\eta \xi} + \eta_x u_{\eta \eta}

     = -2(-2u_{\xi \xi} + u_{\xi \eta}) - 2u_{\xi \eta} + u_{\eta \eta}

     = 4u_{\xi \xi} - 4u_{\xi \eta} + u_{\eta \eta}

    Or did I make a mistake in that?
    I agree in what you have in this post.
    Follow Math Help Forum on Facebook and Google+

Similar Math Help Forum Discussions

  1. [SOLVED] Re-writing higher order spatial derivatives as lower order system
    Posted in the Differential Equations Forum
    Replies: 11
    Last Post: July 27th 2010, 09:56 AM
  2. Replies: 1
    Last Post: October 27th 2009, 05:03 AM
  3. Replies: 2
    Last Post: February 23rd 2009, 06:54 AM
  4. Replies: 2
    Last Post: November 25th 2008, 10:29 PM
  5. Replies: 1
    Last Post: May 11th 2007, 04:01 AM

Search Tags


/mathhelpforum @mathhelpforum