Using the Divergence Theorem to Solve Vector Calculus Problems

Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
3 replies · 3K views
SP90
Messages
19
Reaction score
0

Homework Statement



attachment.php?attachmentid=47293&stc=1&d=1337106295.png


Homework Equations



So I have that [itex]v \otimes n = \left( \begin{array}{ccc}<br /> v_{1}n_{1} & v_{1}n_{2} & v_{1}n_{3} \\<br /> v_{2}n_{1} & v_{2}n_{2} & v_{2}n_{3} \\<br /> v_{3}n_{1} & v_{3}n_{2} & v_{3}n_{3} \end{array} \right) [/itex]

The Attempt at a Solution



I've tried applying the Divergence theorem for Tensors:
[itex] \int_{\partial B} ( v \otimes n )n dA = \int_{B} \nabla \cdot ( v \otimes n ) dV[/itex]

But that doesn't lead anywhere particularly useful. I thought it might be worth noting that [itex]\nabla \cdot ( v \otimes n ) = \frac{dv_{1}}{dx_{1}}n_{1}+\frac{dv_{2}}{dx_{2}}n_{2} + \frac{dv_{3}}{dx_{3}}n_{3}[/itex] but I can't seem to get anywhere near [itex]\nabla v[/itex]

And this problem isn't homework, it's just an optional exercise, but it's frustrated me for a while and I figured I should get some pointers.
 

Attachments

  • Screen Shot 2012-05-15 at 19.24.21.png
    Screen Shot 2012-05-15 at 19.24.21.png
    13.8 KB · Views: 605
Last edited:
Physics news on Phys.org
SP90 said:
I thought it might be worth noting that [itex]\nabla \cdot ( v \otimes n ) = \frac{dv_{1}}{dx_{1}}n_{1}+\frac{dv_{2}}{dx_{2}}n_{2} + \frac{dv_{3}}{dx_{3}}n_{3}[/itex] but I can't seem to get anywhere near [itex]\nabla v[/itex]

No it's not! I think you will find (assuming n is constant) that [itex]\nabla \cdot ( v \otimes n ) = n \cdot (\nabla v) + n (\nabla \cdot v)[/itex]

Especially note that it is a vector and not a scalar
 
Isn't [itex]\nabla \cdot v \otimes n = (n_{1}(\frac{dv_{1}}{dx_{1}}+\frac{dv_{2}}{dx_{2}}+\frac{dv_{3}}{dx_{3}}), n_{2}(\frac{dv_{1}}{dx_{1}}+\frac{dv_{2}}{dx_{2}}+\frac{dv_{3}}{dx_{3}}), n_{3}(\frac{dv_{1}}{dx_{1}}+\frac{dv_{2}}{dx_{2}}+\frac{dv_{3}}{dx_{3}}))[/itex]

Which is [itex]n \cdot \nabla v[/itex]?

This would make sense since it gives that

[itex]\int_{\partial B} ( v \otimes n )n dA = \int_{B} (\nabla v) ndV[/itex]

And then since n is just so constant vector, the result follows.

Is that right? Or am I missing something?
 
SP90 said:
Isn't [itex]\nabla \cdot v \otimes n = (n_{1}(\frac{dv_{1}}{dx_{1}}+\frac{dv_{2}}{dx_{2}}+\frac{dv_{3}}{dx_{3}}), n_{2}(\frac{dv_{1}}{dx_{1}}+\frac{dv_{2}}{dx_{2}}+\frac{dv_{3}}{dx_{3}}), n_{3}(\frac{dv_{1}}{dx_{1}}+\frac{dv_{2}}{dx_{2}}+\frac{dv_{3}}{dx_{3}}))[/itex]

Which is [itex]n \cdot \nabla v[/itex]?

This would make sense since it gives that

[itex]\int_{\partial B} ( v \otimes n )n dA = \int_{B} (\nabla v) ndV[/itex]

And then since n is just so constant vector, the result follows.

Is that right? Or am I missing something?

Of course not; you can't just get rid of the terms that give you trouble... This is what you need to show, but the formula I gave above is still correct.