On one of the string theory notes I read that the conformal symmetry of the bosonic string theory is implied by virasoro constraints. Could someone explain what is the relationship between conformal symmetry and the virasoro constraints?
-
1What exactly do you mean by "Virasoro constraints"? The Virasoro algebra is just a central extension of the conformal algebra, so I'm uncertain what there is to explain (unless the issue here is that you don't know what the Virasoro algebra is). – ACuriousMind Apr 24 '23 at 20:52
-
Related: https://physics.stackexchange.com/q/425371/2451 – Qmechanic Apr 24 '23 at 20:53
-
@ACuriousMind Virasoro constraints are constraint equations that we get from stress-energy tensor in Bosonic string theory. It's common terminology used in string theory books ex. Polchinki String theory. – htr Apr 24 '23 at 23:02
-
2Please explain what you understand in the context of this question what these constraints are. The "Virasoro constraints" I know are just Gupta-Bleuler like equations such as $(L_n-c)\lvert \psi\rangle = 0$ where $L_n$ is a Virasoro generator. Again, since the Virasoro algebra is just a centrally extended conformal algebra, I don't really know what you want an answer here to explain. – ACuriousMind Apr 24 '23 at 23:42
1 Answers
In bosonic string theory, we apply reparametrisation invariance and Weyl invariance to put the worlsheet metric in conformal gauge, i.e. $h_{\mu\nu}=\eta_{\mu\nu}$. However, there exists a residual symmetry, which in light cone-coordinates is $\sigma^+\to f(\sigma^+)$ and $\sigma^-\to f(\sigma^-)$. Moreover, in light cone coordinates, the conservation laws of the energy-momentum tensors read $\partial_-T_{++}=0$ and $\partial_+T_{--}=0$. These facts imply an infinite number of conserved charges:
$$ L_f=2T\int_0^l d\sigma \,f(\sigma^+)T_{++}(\sigma^+)$$
For a closed string, taking into account the periodic boundary conditions, one chooses $f$ in the form $f(\sigma^{\pm})=\exp(\frac{2\pi i}{l} m\sigma^{\pm})$. The Virasoro generators are defined to be the corresponding charges at $\tau=0$:
$$ L_n=-\frac{l}{4\pi^2}\int_0^ld\sigma\, e^{-\frac{2\pi i}{l}n\sigma}T_{--}$$
$$ \bar{L}_n=-\frac{l}{4\pi^2}\int_0^ld\sigma\, e^{\frac{2\pi i}{l}n\sigma}T_{++} $$
The Poisson brackets are
$$ \{ L_m,L_n\}=-i(m-n)L_{m+n}, \hspace{3mm} \{ \bar{L}_m,\bar{L}_n\}=-i(m-n)\bar{L}_{m+n}, \hspace{3mm}\{ \bar{L}_m,L_n\}=0$$
This algebra is calles Virasoro algebra. If we replace Poisson brackets by Lie brackets, a realisation of the Virasoro algebra is furnished by the vector fields
$$\bar{L}_n=e^{\frac{2\pi i}{l} n\sigma^+}\partial_+, \hspace{3 mm} L_n=e^{\frac{2\pi i}{l} n\sigma^-}\partial_- $$
If we define the variable $z=e^{\frac{2\pi i}{l} n\sigma^-}$, then $L_n=iz^{n+1}\partial_z$, which are the generators of the conformal group in two dimensions.
-
Thank you for the answer, I already understood what you wrote except the very last line. Could you extend your answer for defining a conformal group in 2-D and how it is generated by $L_n$? – htr Apr 30 '23 at 14:37
-
Well, this is just basic knowledge about the conformal group in 2D, you can read about it in any textbook. – Ruben Campos Delgado Apr 30 '23 at 14:45