Help with variation of the 3-dimensional ##\sigma##-model action

user1139
Messages
71
Reaction score
8
Homework Statement
See below.
Relevant Equations
See below.
Consider the following action

$$S=\int\mathrm{d}^3x\sqrt{h}\left[R^{(3)}-\frac{1}{4}\mathrm{Tr}\left(\chi^{-1}\chi_{,i}\chi^{-1}\chi^{,i}\right)\right]$$

where ##h## is the determinant of the 3-dimensional metric tensor ##h_{ij}## and ##R## is the Ricci scalar.

I want to get the equations of motion

\begin{align*}
\left(\chi^{-1}\chi^{,i}\right)_{;i}&=0,\\
R_{ij}&=\frac{1}{4}\mathrm{Tr}\left(\chi^{-1}\chi_{,i}\chi^{-1}\chi_{,j}\right).
\end{align*}

However, how do I perform the variation on the trace?
 
Physics news on Phys.org
Thomas1 said:
Homework Statement:: See below.
Relevant Equations:: See below.

Consider the following action

$$S=\int\mathrm{d}^3x\sqrt{h}\left[R^{(3)}-\frac{1}{4}\mathrm{Tr}\left(\chi^{-1}\chi_{,i}\chi^{-1}\chi^{,i}\right)\right]$$

where ##h## is the determinant of the 3-dimensional metric tensor ##h_{ij}## and ##R## is the Ricci scalar.

I want to get the equations of motion

\begin{align*}
\left(\chi^{-1}\chi^{,i}\right)_{;i}&=0,\\
R_{ij}&=\frac{1}{4}\mathrm{Tr}\left(\chi^{-1}\chi_{,i}\chi^{-1}\chi_{,j}\right).
\end{align*}

However, how do I perform the variation on the trace?
If your concern is only about the variation of the trace, you better ignore the coupling to gravity. Also, the manipulations hold in any number of dimensions. The relevant identities are \delta \mbox{Tr}(\cdot) = \mbox{Tr}(\delta \cdot), \delta \partial = \partial \delta, \mbox{Tr}(AB) = \mbox{Tr}(BA), and the following important two \delta u^{-1} = - u^{-1} (\delta u) u^{-1}, \ \ \ \ (1)\partial u^{-1} = - u^{-1}(\partial u) u^{-1}. \ \ \ \ (2) To make your life easy, define the vector A^{\mu} = u^{-1} \partial^{\mu} u. So, the NLSM action becomes S = - \frac{1}{4} \int d^{n}x \ \mbox{Tr}(A^{\mu}A_{\mu}). Thus \delta S = - \frac{1}{2} \int d^{n}x \ \mbox{Tr}(A^{\mu} \ \delta A_{\mu}) . Now, use (1) to obtain \delta A_{\mu} = u^{-1} \ \partial_{\mu}\delta u - u^{-1} \ \delta u \ A_{\mu} . Substitute in \delta S, you get \delta S = \frac{1}{2} \int \ \mbox{Tr}\left( u^{-1}\delta u \ A_{\mu}A^{\mu} - A^{\mu} u^{-1} \ \partial_{\mu}\delta u \right). Integrate the 2nd term by part and ignore the surface term to obtain \delta S = \frac{1}{2} \int \ \mbox{Tr}\left( u^{-1}\delta u \ A^{\mu}A_{\mu} + \partial_{\mu}(A^{\mu}u^{-1}) \ \delta u \right). Finally, expand the differentiation in the second term and use (2) to get \delta S = \frac{1}{2} \int \ \mbox{Tr} \left( u^{-1}\delta u \ A^{\mu}A_{\mu} + \partial_{\mu}A^{\mu} \ u^{-1}\delta u - A^{\mu}A_{\mu} \ u^{-1}\delta u \right). The first and the third terms add up to zero (because \mbox{Tr}(ab)= \mbox{Tr} (ba)), and you end up with \delta S = \frac{1}{2} \int \ \mbox{Tr}\left( (\partial_{\mu}A^{\mu}) u^{-1} \ \delta u \right). This gives you the equation of motion \partial_{\mu}A^{\mu} \equiv \partial_{\mu} \left(u^{-1} \ \partial^{\mu}u \right) = 0.
 
  • Like
Likes ergospherical
Thread 'Help with Time-Independent Perturbation Theory "Good" States Proof'
(Disclaimer: this is not a HW question. I am self-studying, and this felt like the type of question I've seen in this forum. If there is somewhere better for me to share this doubt, please let me know and I'll transfer it right away.) I am currently reviewing Chapter 7 of Introduction to QM by Griffiths. I have been stuck for an hour or so trying to understand the last paragraph of this proof (pls check the attached file). It claims that we can express Ψ_{γ}(0) as a linear combination of...
Back
Top