Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Derivatives of a radial function

  1. Nov 10, 2003 #1
    I have a scalar function of position only, V(r) where r is the position vector (x, y,z). Since V depends only on position, is it true to say that

    dV/dx = dV/dy = dV/dz ?
    (these should be partial derivatives)

    I am trying to show L = r x [nab] commutes with any radial function V(r) meaning that for any function f

    r x [nab] ( V(r) * f) - V(r) * r x [nab] f = 0

    most of the terms cancel out but I am left with

    (y dV/dz - x dV/dz -zdV/dy + xdV/dy + z dVdx - y dV/dx) * f

    or (y (dV/dz - dV/dx) + x( dV/dy - dVdz ) +z( dV/dx - dV/dy ) ) * f

    If the partial derivatives cancel then I am done. If not I have no clue how to continue but I do know the final answer must be zero.
  2. jcsd
  3. Nov 10, 2003 #2
    What, all the Cartesian components of the gradient of V should equal each other, for an arbitrary function V? If that's what you mean, then no.
  4. Nov 10, 2003 #3
    By the way, L does not commute with V unless V is spherically symmetric, i.e. V ≡ V(r) where r is the distance from the origin (not the position vector).
  5. Nov 11, 2003 #4
    Re: Re: derivatives of a radial function

    Yes, thank you. That is correct and in fact I think the book means that V(r) is a central potential so the spherically symmetric is true in this case. Does that mean the answer is yes - the spherical symmetry means dV/dx = dV/dy = dV/dz ?
  6. Nov 11, 2003 #5
    Re: Re: Re: derivatives of a radial function

    No. (Consider a central potential at a point on the x axis. At that point, dV/dx will generally be nonzero in the x direction since it is the radial direction, but dV/dy = dV/dz = 0 since they are perpendicular to the radial direction.)

    It's hard to incorporate the central constraint in Cartesian coordinates, which is why everyone does this calculation in spherical coordinates.
  7. Nov 12, 2003 #6
    Thanks for keeping me from seizing the easy incorrect answer. When I took the physics out of the problem to simplify it I took out an essential element.

    What I really had was
    x dV/dy - y dV/dx and similar terms so you can have cancelation without having dV/dy = dVdx = 0. Saddly I didn't see the trick but using dV/dy = dV/dr * dr/dy and similar for x gets the terms to cancel. I hope I don't forget this trick.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook