I have been working with some mathematical models in biology and fluid mechanics. My problem is about the interpretation of a condition that I found for a vector representing the velocity of a fluid. The exact question is the next:

Let $\mathbf{u=(}u_{1},u_{2},u_{3})$ be a vector field representing the velocity of a fluid. After making some accounts with models using partial differential equations, I found that the matrix \begin{equation} -A:=\frac{1}{2}\left( \frac{\partial u^{j}}{\partial x_{i}}+\frac{\partial u^{i}}{\partial x_{j}}\right) _{i,j=1,2,3}% \ \end{equation} should be positive definite. Does this condition has a physical interpretation in fluid dynamics or tensors?

Any comment or reference will be highly appreciate!

  • 1
    $\begingroup$ In the fluids literature, the symmetric part of the velocity gradient is usually called the "rate of deformation tensor" or the "velocity strain" tensor. Your $A$ is the negative of that. $\endgroup$ – Willie Wong Nov 7 '14 at 8:59
  • 1
    $\begingroup$ I would be highly suspicious of such a model. Biological fluids are usually close to incompressible. In that case the matrix A is traceless and therefore definitely NOT positive definite! $\endgroup$ – Michael Renardy Nov 7 '14 at 11:18

I don't know if this is the answer that you have been looking for, but let me offer a rather trivial observation.

Your tensor is the Lie derivative of the metric tensor with respect to the vector field $\textbf u$. Loosely speaking, the Lie derivative $\mathcal{L}_\textbf{u}$ has an interpretation as a derivative with respect to "dragging" a tensor along the flow defined by $\textbf{u}$. A simple manipulation shows that for any vector field $\textbf{v}$, \begin{equation} \mathcal{L}_\textbf{u} (\textbf{v} \cdot \textbf{v}) = 2 \textbf{v} \cdot \mathcal{L}_\textbf{u} \textbf{v} - 2 \textbf{v} A \textbf{v} \,, \end{equation} where the matrix $A$ is the one you have defined. A flow $\textbf{u}$ with positive definite $(-A)$ has the property that \begin{equation} \mathcal{L}_\textbf{u} (| \textbf{v} |^2 ) \geq 2 \textbf{v} \cdot \mathcal{L}_\textbf{u} \textbf{v} \end{equation} for any vector field $\textbf{v}$.

I hope this observation has some use to you.

| cite | improve this answer | |

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.