Question 12.70: (a) Construct a tensor D^μν (analogous to F^μν) out of D and...

(a) Construct a tensor D^{\mu \nu} (analogous to F^{\mu \nu}) out of D and H. Use it to express Maxwell’s equations inside matter in terms of the free current density J_{f}^{\mu}.

\left[\text { Answer: } D^{01} \equiv c D_{x}, D^{12} \equiv H_{z}, \text { etc. } ; \partial D^{\mu \nu} / \partial x^{\nu}=J_{f}^{\mu} .\right]

(b) Construct the dual tensor \left.H^{\mu \nu} \text { (analogous to } G^{\mu v}\right). \left[\text { Answer: } H^{01} \equiv H_{x}, H^{12} \equiv\right.\left.-c D_{z}, \text { etc. }\right]

(c) Minkowski proposed the relativistic constitutive relations for linear media:

D^{\mu \nu} \eta_{\nu}=c^{2} \epsilon F^{\mu \nu} \eta_{\nu} \quad \text { and } \quad H^{\mu \nu} \eta_{\nu}=\frac{1}{\mu} G^{\mu \nu} \eta_{\nu}

where  is the proper^{30} permittivity, μ is the proper permeability, and \eta^{\mu} is the 4-velocity of the material. Show that Minkowski’s formulas reproduce Eqs. 4.32 and 6.31, when the material is at rest.

D =\epsilon E                         (4.32)

B =\mu H                                 (6.31)

(d) Work out the formulas relating D and H to E and B for a medium moving with (ordinary) velocity u.

The Blue Check Mark means that this solution has been answered and checked by an expert. This guarantees that the final answer is accurate.
Learn more on how we answer questions.

(a)  \left.\begin{array}{l} D =\epsilon_{0} E + P \text { suggests } E \rightarrow \frac{1}{\epsilon_{0}} D \\H =\frac{1}{u_{0}} B – M \text { suggests } B \rightarrow \mu_{0} H\end{array}\right\} but it’s a little cleaner if we divide by \mu_{0} while we’re at it, so that

E \rightarrow \frac{1}{\mu_{0} \epsilon_{0}} D =c^{2} D , B \rightarrow H . Then: D^{\mu \nu}=\left\{\begin{array}{cccc}0 & c D_{x} & c D_{y} & c D_{z} \\-c D_{x} & 0 & H_{z} & -H_{y} \\-c D_{y} & -H_{z} & 0 & H_{x} \\-c D_{z} & H_{y} & -H_{x} & 0\end{array}\right\}

Then (following the derivation in Sect. 12.3.4):

\frac{\partial}{\partial x^{\nu}} D^{0 \nu}=c \nabla \cdot D =c \rho_{f}=J_{f}^{0} ; \frac{\partial}{\partial x^{\nu}} D^{1 \nu}=\frac{1}{c} \frac{\partial}{\partial t}\left(-c D_{x}\right)+( \nabla \times H )_{x}=\left(J_{f}\right)_{x} ; so \frac{\partial D_{\mu \nu}}{\partial x^{\nu}}=J_{f}^{\mu},

where J_{f}^{\mu}=\left(c \rho_{f}, J _{f}\right) . Meanwhile, the homogeneous Maxwell equations \left( \nabla \cdot B =0, E =-\frac{\partial B }{\partial t}\right) are unchanged, and hence \frac{\partial G^{\mu \nu}}{\partial x^{\nu}}=0.

(b)    H^{\mu \nu}=\left\{\begin{array}{cccc}0 & H_{x} & H_{y} & H_{z} \\-H_{x} & 0 & -c D_{z} & c D_{y} \\-H_{y} & c D_{z} & 0 & -c D_{x} \\-H_{z} & -c D_{y} & c D_{x} & 0\end{array}\right\}

(c) If the material is at rest, \eta_{\nu}=(-c, 0,0,0), \text { and the sum over } \nu collapses to a single term:

D^{\mu 0} \eta_{0}=c^{2} \epsilon F^{\mu 0} \eta_{0} \Rightarrow D^{\mu 0}=c^{2} \epsilon F^{\mu 0} \Rightarrow-c D =-c^{2} \epsilon \frac{ E }{c} \Rightarrow D =\epsilon E (\text { Eq. } 4.32),

H^{\mu 0} \eta_{0}=\frac{1}{\mu} G^{\mu 0} \eta_{0} \Rightarrow H^{\mu 0}=\frac{1}{\mu} G^{\mu 0} \Rightarrow- H =-\frac{1}{\mu} B \Rightarrow H =\frac{1}{\mu} B (\text { Eq. } 6.31).

(d) In general, \eta_{\nu}=\gamma(-c, u ), \text { so, for } \mu=0:

D^{0 \nu} \eta_{\nu}=D^{01} \eta_{1}+D^{02} \eta_{2}+D^{03} \eta_{3}=c D_{x}\left(\gamma u_{x}\right)+c D_{y}\left(\gamma u_{y}\right)+c D_{z}\left(\gamma u_{z}\right)=\gamma c( D \cdot u ),

F^{0 \nu} \eta_{\nu}=F^{01} \eta_{1}+F^{02} \eta_{2}+F^{03} \eta_{3}=\frac{E_{x}}{c}\left(\gamma u_{x}\right)+\frac{E_{y}}{c}\left(\gamma u_{y}\right)+\frac{E_{z}}{c}\left(\gamma u_{z}\right)=\frac{\gamma}{c}( E \cdot u ) , so

D^{0 \nu} \eta_{\nu}=c^{2} \epsilon F^{0 \nu} \eta_{\nu} \Rightarrow \gamma c( D \cdot u )=c^{2} \epsilon\left(\frac{\gamma}{c}\right)( E \cdot u ) \Rightarrow D \cdot u =\epsilon( E \cdot u ).                         [1]

H^{0 \nu} \eta_{\nu}=H^{01} \eta_{1}+H^{02} \eta_{2}+H^{03} \eta_{3}=H_{x}\left(\gamma u_{x}\right)+H_{y}\left(\gamma u_{y}\right)+H_{z}\left(\gamma u_{z}\right)=\gamma( H \cdot u ),

G^{0 \nu} \eta_{\nu}=G^{01} \eta_{1}+G^{02} \eta_{2}+G^{03} \eta_{3}=B_{x}\left(\gamma u_{x}\right)+B_{y}\left(\gamma u_{y}\right)+B_{z}\left(\gamma u_{z}\right)=\gamma( B \cdot u ) , so

H^{0 \nu} \eta_{\nu}=\frac{1}{\mu} G^{0 \nu} \eta_{\nu} \Rightarrow \gamma( H \cdot u )=\frac{1}{\mu} \gamma( B \cdot u ) \Rightarrow H \cdot u =\frac{1}{\mu}( B \cdot u ).                         [2]

Similarly, for µ = 1:

D^{1 \nu} \eta_{\nu}=D^{10} \eta_{0}+D^{12} \eta_{2}+D^{13} \eta_{3}=\left(-c D_{x}\right)(-\gamma c)+H_{z}\left(\gamma u_{y}\right)+\left(-H_{y}\right)\left(\gamma u_{z}\right)=\gamma\left(c^{2} D_{x}+u_{y} H_{z}-u_{z} H_{y}\right)

 

=\gamma\left[c^{2} D +( u \times H )\right]_{x},

F^{1 \nu} \eta_{\nu}=F^{10} \eta_{0}+F^{12} \eta_{2}+F^{13} \eta_{3}=\frac{-E_{x}}{c}(-\gamma c)+B_{z}\left(\gamma u_{y}\right)+\left(-B_{y}\right)\left(\gamma u_{z}\right)=\gamma\left(E_{x}+u_{y} B_{z}-u_{z} B_{y}\right)

 

=\gamma[ E +( u \times B )]_{x}, \quad \text { so } \quad D^{1 \nu} \eta_{\nu}=c^{2} \epsilon F^{1 \nu} \eta_{\nu} \Rightarrow

 

\gamma\left[c^{2} D +( u \times H )\right]_{x}=c^{2} \epsilon \gamma[ E +( u \times B )]_{x} \Rightarrow D +\frac{1}{c^{2}}( u \times H )=\epsilon[ E +( u \times B )].                                        [3]

\begin{aligned}H^{1 \nu} \eta_{\nu} &=H^{10} \eta_{0}+H^{12} \eta_{2}+H^{13} \eta_{3}=\left(-H_{x}\right)(-\gamma c)+\left(-c D_{z}\right)\left(\gamma u_{y}\right)+\left(c D_{y}\right)\left(\gamma u_{z}\right) \\&=\gamma c\left(H_{x}-u_{y} D_{z}+u_{z} D_{y}\right)=\gamma c[ H -( u \times D )]_{x}\end{aligned},

\begin{aligned}G^{1 \nu} \eta_{\nu} &=G^{10} \eta_{0}+G^{12} \eta_{2}+G^{13} \eta_{3}=\left(-B_{x}\right)(-\gamma c)+\left(-\frac{E_{z}}{c}\right)\left(\gamma u_{y}\right)+\left(\frac{E_{y}}{c}\right)\left(\gamma u_{z}\right) \\&=\frac{\gamma}{c}\left(c^{2} B_{x}-u_{y} E_{z}+u_{z} E_{y}\right)=\frac{\gamma}{c}\left[c^{2} B -( u \times E )\right]_{x}, \quad \text { so } \quad H^{1 \nu} \eta_{\nu}=\frac{1}{\mu} G^{1 \nu} \eta_{\nu} \Rightarrow\end{aligned}

 

\gamma c[ H -( u \times D )]_{x}=\frac{1}{\mu} \frac{\gamma}{c}\left[c^{2} B -( u \times E )\right]_{x} \Rightarrow H -( u \times D )=\frac{1}{\mu}\left[ B -\frac{1}{c^{2}}( u \times E )\right].                                  [4]

Use Eq. [4] as an expression for H, plug this into Eq. [3], and solve for D:

D +\frac{1}{c^{2}} u \times\left\{( u \times D )+\frac{1}{\mu}\left[ B -\frac{1}{c^{2}}( u \times E )\right]\right\}=\epsilon[ E +( u \times B )];

D +\frac{1}{c^{2}}\left[( u \cdot D ) u -u^{2} D \right]=\epsilon[ E +( u \times B )]-\frac{1}{\mu c^{2}}( u \times B )+\frac{1}{\mu c^{4}}[ u \times( u \times E )].

Using Eq. [1] to rewrite (u · D):

\begin{aligned}D \left(1-\frac{u^{2}}{c^{2}}\right) &=-\frac{\epsilon}{c^{2}}( E \cdot u ) u +\epsilon[ E +( u \times B )]-\frac{1}{\mu c^{2}}( u \times B )+\frac{1}{\mu c^{4}}\left[( E \cdot u ) u -u^{2} E \right] \\&=\epsilon\left\{\left[1-\frac{u^{2}}{\epsilon \mu c^{4}}\right] E -\frac{1}{c^{2}}\left[1-\frac{1}{\epsilon \mu c^{2}}\right]( E \cdot u ) u +( u \times B )\left[1-\frac{1}{\epsilon \mu c^{2}}\right]\right\}\end{aligned}

Let \gamma \equiv \frac{1}{\sqrt{1-u^{2} / c^{2}}}, \quad v \equiv \frac{1}{\sqrt{\epsilon \mu}} . Then

D =\gamma^{2} \epsilon\left\{\left(1-\frac{u^{2} v^{2}}{c^{4}}\right) E +\left(1-\frac{v^{2}}{c^{2}}\right)\left[( u \times B )-\frac{1}{c^{2}}( E \cdot u ) u \right]\right\}.

Now use Eq. [3] as an expression for D, plug this into Eq. [4], and solve for H:

H – u \times\left\{-\frac{1}{c^{2}}( u \times H )+\epsilon[ E +( u \times B )]\right\}=\frac{1}{\mu}\left[ B -\frac{1}{c^{2}}( u \times E )\right];

H +\frac{1}{c^{2}}\left[( u \cdot H ) u -u^{2} H \right]=\frac{1}{\mu}\left[ B -\frac{1}{c^{2}}( u \times E )\right]+\epsilon( u \times E )+\epsilon[ u \times( u \times B )].

Using Eq. [2] to rewrite (u · H):

\begin{aligned}H \left(1-\frac{u^{2}}{c^{2}}\right) &=-\frac{1}{\mu c^{2}}( B \cdot u ) u +\frac{1}{\mu}\left[ B -\frac{1}{c^{2}}( u \times E )\right]+\epsilon( u \times E )+\epsilon\left[( B \cdot u ) u -u^{2} B \right] \\&=\frac{1}{\mu}\left\{\left[1-\mu \epsilon u^{2}\right] B +\left[\epsilon \mu-\frac{1}{c^{2}}\right][( u \times E )+( B \cdot u ) u ]\right\}\end{aligned}

H =\frac{\gamma^{2}}{\mu}\left\{\left(1-\frac{u^{2}}{v^{2}}\right) B +\left(\frac{1}{v^{2}}-\frac{1}{c^{2}}\right)[( u \times E )+( B \cdot u ) u ]\right\}.

Related Answered Questions