This is probably a dumb question, but I'm a mathematician who's been trying to understand the equations of motion for Type IIB supergravity, and I'm not quite sure I understand what's going on with the Einstein equations. Specifically, I'm following the appendix A of this paper, and the equation I'm having trouble with is A.4. Here it is for clarity:

$$R_{\mu \nu} + 2 D_{\mu}D_{\nu}\Phi - \frac{1}{4}H_{\mu \nu}^2 = e^{2\Phi}\left[ \frac{1}{2}(F_1^2)_{\mu \nu} + \frac{1}{4}(F_3^2)_{\mu \nu} + \frac{1}{96}(F_5^2)_{\mu \nu} - \frac{1}{4}g_{\mu \nu}\left( F_1^2 + \frac{1}{6}F_3^2 \right )\right]$$

My understanding is that $F_1$ is a differential 1-form, which we would write in components as say $F_1 = f_i dx^i$, and that $F_1^2$ is defined then to be: $$F_1^2 = f_i f_j g^{ij} $$ My problem is therefore that I expect $F_1^2$ to be a scalar. Indeed, I expect all the terms on the RHS of the equation to be scalars, and so I don't know what $(F_1^2)_{\mu \nu}$ means.

Small addendum: I am fairly sure that $D_{\mu }$ is the covariant derivative with respect to the metric connection, something I would normally call $\nabla_{\mu}$, but I just wanted to confirm this.


2 Answers 2


For a target space $p$-form in local coordinates

$$F_p~=~\frac{1}{p!}F_{\mu_1\mu_2\ldots\mu_p}~\mathrm{d}x^{\mu_1}\wedge \mathrm{d}x^{\mu_2}\wedge\ldots\wedge \mathrm{d}x^{\mu_p},\tag{1}$$

the paper defines a scalar

$$F_p^2~:=~F_{\mu_1\mu_2\ldots\mu_p}~g^{\mu_1\nu_1}~g^{\mu_2\nu_2}\ldots g^{\mu_p\nu_p}~F_{\nu_1\nu_2\ldots\nu_p},\tag{2}$$

and a symmetric covariant tensor

$$(F_p^2)_{\mu_1\nu_1}~:=~F_{\mu_1\mu_2\ldots\mu_p}~g^{\mu_2\nu_2}\ldots g^{\mu_p\nu_p}~F_{\nu_1\nu_2\ldots\nu_p}.\tag{3}$$


$F_{n-1}$'s are curvatures of higher gauge fields $A_{n-1}$'s, so they are n-forms. The expression $(F_{n}^2)_{\mu \nu}$ means schematically $F_{\mu \rho_1 \cdots \rho_{n-1}} F_{\nu}^{\phantom{a} \rho_1 \cdots \rho_{n-1}}$, where the indices are risen with the inverse metric $g^{\mu \nu}$.

  • $\begingroup$ Does this mean that $F_1^2$ is just $F_{\mu} F_{\nu}$? $\endgroup$
    – Mark B
    Feb 11, 2017 at 7:16
  • $\begingroup$ @MarkB Sorry, I made a typo in my answer. $F_1$ is the curvature of $A_1$, $F_1=\mathrm{d} A_1$. $\endgroup$ Feb 11, 2017 at 7:23
  • $\begingroup$ Yep, I'm ok with the fact that $F_1$ is the curvature of a scalar field, $F_1 = d C_0$. But a one-form has only one index, so according to your answer the square of the 1-form flux is simply $F_{\mu} F_{\nu}$, which seems a bit strange to me. $\endgroup$
    – Mark B
    Feb 11, 2017 at 7:27
  • 1
    $\begingroup$ @MarkB Sorry, you are right. There are, of course, no odd-degree gauge fields in Type IIB. The correct expressions are given by (2.4) and (2.5) in the paper you cited. Then, $F_1^2$ is just $\partial_{\mu} C_0 \partial_{\nu} C_0$. $\endgroup$ Feb 11, 2017 at 7:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.