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

Vector identity question

  1. Apr 15, 2015 #1
    Does ∇⋅A = A ⋅∇? If not then, what does the latter actually equal?
  2. jcsd
  3. Apr 15, 2015 #2
    if [itex]\vec{A}=(A_x,A_y,A_z)[/itex] then [itex]\vec{A}\cdot\nabla=A_x\frac{\partial}{\partial x}+A_y\frac{\partial}{\partial y}+A_z\frac{\partial}{\partial z}[/itex] that is A⋅∇ is still an operator and can be applied to a scalar or a vector. For example applied to scalar φ, [itex](\vec{A}\cdot\nabla)\phi=A_x\frac{\partial\phi}{\partial x}+A_y\frac{\partial\phi}{\partial y}+A_z\frac{\partial\phi}{\partial z}[/itex]
  4. Apr 16, 2015 #3
    So they are equal? Since it is multiplication, they are the same then?
  5. Apr 16, 2015 #4
    No they are not equal, [itex]\nabla\cdot\vec{A}[/itex] is not an operator itself , its the result of the operator [itex]\nabla[/itex]applied to vector [itex]\vec{A}[/itex] via the dot product and it is a scalar quantity refered as the divergence of vector A. It is [itex]\nabla\cdot\vec{A}=\frac{\partial{A_x}}{\partial x}+\frac{\partial{A_y}}{\partial y}+\frac{\partial{A_z}}{\partial z}[/itex].
  6. Apr 16, 2015 #5
    Nice thank you! Are you familiar with vector potentials in physics? I am working on proving the following but I have been running into trouble. Trying to prove that ∇xA=B!

    Attached Files:

  7. Apr 16, 2015 #6


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I assume that what you want to prove is the second equality in the second line. The proof is fairly long and tedious. You have to use a definition of the cross product, and the product rule. What definition of the cross product are you using? Are you familiar with this one (using the Levi-Civita symbol and Einstein's summation convention)?
    $$(\vec A\times\vec B)_i=\varepsilon_{ijk} A_j B_k.$$
  8. Apr 16, 2015 #7
    For that you can also use the definition of the operator ∇ in spherical coordinates and the identity [itex]\nabla\times(A\times B)=A(\nabla\cdot B)-B(\nabla\cdot A)+(B\cdot\nabla)A-(A\cdot\nabla)B[/itex] which can be simplified in this case where [itex]A=m[/itex] is a constant vector, by setting the 2 medium terms to zero.
    Last edited: Apr 16, 2015
  9. Apr 17, 2015 #8
    I found out this is not so long afterall if one knows how to do it in spherical coordinate system. However extra caution is needed when working with operators like [itex]\vec{m}\cdot\nabla[/itex] in spherical coordinates. The derivatives of the spherical unit vectors are not always zero.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Discussions: Vector identity question
  1. Vector identity (Replies: 1)