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

Reynolds Transport Theorem derivation question

  1. Aug 9, 2011 #1

    I'm struggling to understand one step in the derivation of the Reynold's Transport Theorem. I get as far as:

    \frac{d}{dt} \int_{V(t)} f(\mathbf x, t) \, dV = \frac {d} {dt} \int_{V_0} f(\mathbf X,t) J(\mathbf X,t) \, dV = \int_{V_0} J \frac {\partial f} {\partial t} + f \frac {\partial J} {\partial t} \, dV

    From here, I want to write the RHS as:

    \int_{V_0} \frac {\partial f} {\partial t} J + f \, ({\nabla \cdot} {\mathbf v}) J \,\, dV = \int_{V(t)} \frac {\partial f} {\partial t} + f \, ({\nabla \cdot} {\mathbf v}) \,\, dV

    However, all of my sources inexplicably replace the \begin{equation} \frac {\partial f} {\partial t} \end{equation} with a \begin{equation} \frac {Df} {Dt} \end{equation} Could anyone explain their reasoning or my error?

  2. jcsd
  3. Aug 9, 2011 #2
    \begin{equation} \frac {Df} {Dt} \end{equation} is just the material derivative (the derivative with respect to time while keeping the initial position constant).
  4. Aug 9, 2011 #3
    But isn't the material derivative given by the total derivative wrt time, \begin{equation} \frac {df} {dt} = \frac {\partial f} {\partial t} + (\mathbf v \cdot \nabla) f \end{equation} whereas the integrand over the control volume just contains a partial derivative?
  5. Aug 9, 2011 #4
    Yes, it is if f is a scalar field. But in the Reynolds transport theorem, f is a vector field. For a vector field V, the material derivative is defined as follows:

    \begin{equation} \frac {Df} {Dt} = \frac {\partial V} {\partial t} + (\mathbf V \cdot \nabla) V \end{equation}
  6. Aug 10, 2011 #5
    Afraid I still don't follow. Why is the material derivative involved at all? In the ordinary course of physics we never swap a partial derivative for a total derivative, and as I understand it, the material derivative gives:
    \frac {df(\mathbf x (\mathbf X,t),t)} {dt}
    which is not the integrand. I'm fairly certain the transport theorem is equally valid for scalars and vectors, so that shouldn't have made a difference either.
  7. Aug 10, 2011 #6
    You are taking the material derivative because the volume is a material fluid (a function of initial position and time where there is no flux of mass) as opposed to a control volume which is just a function of position with a fixed volume. Basically, you can still use the transport theorem on a control fluid, you just need to take the partial derivative instead of the material derivative. The same thing goes for a scalar field; the theorem is still applicable, you just need to consider that when deriving it.
  8. Aug 13, 2011 #7
    Thanks for pointing out that distinction, it clears things up.
  9. Aug 13, 2011 #8
    Glad to be of assistance.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook