1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Grad x (grad x B)

  1. Mar 11, 2014 #1
    1. The problem statement, all variables and given/known data

    Show that:

    ∇x(∇xB) = (B∇)B - ∇ (1/2B2)

    2. Relevant equations

    r = (x,y,z) = xiei

    ∂xi/∂xj = δij

    r2 = xkxk

    δij = 1 if i=j, 0 otherwise (kronecker delta)
    εijk is the alternating stress tensor and summn convn is assumed.

    3. The attempt at a solution

    On the LHS I simplified to get:


    but was unsure what to do next because the RHS contains only first order derivatives

    On the RHS I was able to get to:

    (B∇)B - ∇ (1/2B2) = B(∂Bi/∂i)-B
    = B(∂Bi/∂i-1)

    I feel like I'm just not seeing some simple trick, or there is a rule that I don't remember/haven't learned. This is for my Classical Mechanics class BTW.
  2. jcsd
  3. Mar 12, 2014 #2


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    There must be something wrong in your problem statement or how can you get an expression which is quadratic in [itex]\vec{B}[/itex] taking derivatives of an expression that contains only one [itex]\vec{B}[/itex]? The correct equation to prove is
    [tex]\vec{\nabla} \times (\vec{\nabla} \times \vec{B})=\vec{\nabla} (\vec{\nabla} \cdot \vec{B}) - \vec{\nabla}^2 \vec{B},[/tex]
    which holds, however, only in Cartesian coordinates!
  4. Mar 12, 2014 #3
    that's what I was thinking, but the assignment is what I wrote above
  5. Mar 12, 2014 #4


    User Avatar
    Gold Member

    (∇xB)xB = (B∇)B - ∇ (B²/2)

    is famous in MHD
  6. Mar 12, 2014 #5
    that's still not what i'm asking. but maybe showing a proof might help me out a bit
  7. Mar 13, 2014 #6


    User Avatar
    Gold Member

    Showing a proof of what?
    What are you asking, actually?
    The original identity is obviously wrong (∇x(∇xB) = (B∇)B - ∇ (1/2B2) is wrong).
    Shall we advise you to complain to your teacher?

    The proof of the second identity, (∇xB)xB = (B∇)B - ∇ (B²/2), is straightforward by using components representation.

    Using the notation "eik" for the Levi-Civita tensor,
    using 'F,l" to denote the derivative of F with respect to xl,
    (∇xB)xB can be developed as follows:

    = eijk (ejlm Bm,l) Bk
    = - eikj elmj Bm,l Bk
    = -(eil ekm - eim ekl) Bm,l Bk
    = - Bk,i Bk + Bi,k Bk
    = - (Bk²/2),i + Bi,k Bk

    which ends the proof.

    Reading you initial post:

    "εijk is the alternating stress tensor ..."
    "On the LHS I simplified to get: εijk∂2/∂xj∂xk"

    I have the feeling you lack some basic understanding, since it makes almost no meaning.
    I don't know if your question is part of a math course or a physics course (electrodynamics).
    In any case, you need to go back to the basics.
    The strange thing is that the identity "(∇xB)xB = (B∇)B - ∇ (B²/2)" is indeed related to the Maxwell stress tensor in electrodynamics (if B is the magnetic field). The second term is then called the magnetic pressure.
    As you posted in the "Calculus & Beyond Homework" section I wonder how you could have mixed that "math exercise" with electrodynamics. Is Google the reason?
    Last edited: Mar 13, 2014
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted