A simple model of magnetic universe without singularity associated
with a quadratic equation of state

The geometry of spacetime in general relativity is specified by the metric tensor $g_{\mu\nu}$ which gives the spacetime interval $ds$ between two infinitesimally separated events, that is⁵⁵5The $g_{\mu\nu}$ may also be viewed as gravitational potentials since the effect of gravity is to modify the curvature of spacetime.

$$ds^{2}=g_{\mu\nu}dx^{\mu}dx^{\nu}.$$

(5)

Here and below the Greek indices $\mu,\nu$ etc run over the spacetime coordinates (ranging from $0$ to $3$) while the Latin indices $i,j$ etc run only over the space coordinates (ranging from $1$ to $3$). We assume summation over repeated indices.

We first discuss the Maxwell equations in a general curved spacetime and then focus on Friedmann-Lemaître-Roberston-Walker (FLRW) models of cosmology. Electrodynamics in curved spacetime is most conveniently formulated by giving the action for the electromagnetic fields and their interaction with charged particles. For the sake of generality, we also include the contribution of a perfect fluid (or matter field). The total action of the system, which is the sum of the Einstein-Hilbert action of general relativity $+$ the action of the perfect fluid $+$ the action of the electromagnetic field $+$ the action describing the interaction between the charges and the electromagnetic field can be written as

$$S=\frac{c^{4}}{16\pi G}\int R\sqrt{-g}\,d^{4}x+\int\mathcal{L}_{m}\sqrt{-g}\,d^{4}x+\int\mathcal{L}({\cal F})\sqrt{-g}\,d^{4}x-\int A_{\mu}J^{\mu}\sqrt{-g}\,d^{4}x,$$

(6)

where $R$ is the Ricci curvature scalar, $g={\rm det}(g_{\mu\nu})$ is the determinant of the metric tensor, $A^{\mu}=(U/c,{\bf A})$ is the electromagnetic quadripotential, and $J^{\mu}$ is the quadricurrent density. We have assumed that the Lagrangian density $\mathcal{L}({\cal F})$ of the electromagnetic field is an arbitrary function of the electromagnetic invariant

$${\cal F}=\frac{1}{4\mu_{0}}F_{\mu\nu}F^{\mu\nu},$$

(7)

where $F_{\mu\nu}=\partial_{\mu}A_{\nu}-\partial_{\nu}A_{\mu}$ is the Faraday or electromagnetic field strength tensor (one can replace the covariant derivative $D_{\mu}$ by the partial derivative $\partial_{\mu}$ in this equation). In the above equation, $\mu_{0}$ denotes the magnetic permeability in vacuum. The electric permittivity in vacuum is denoted by $\epsilon_{0}$. They are related by $\mu_{0}\epsilon_{0}=1/c^{2}$. Ordinary (linear) Maxwell electrodynamics corresponds to the Lagrangian

$${\cal L}_{\rm Maxwell}=-{\cal F}.$$

(8)

The Einstein equations of the gravitational field can be derived from the principle of least action $\delta S=\delta(S_{g}+S_{M})=0$, where $S_{g}$ and $S_{M}$ are the actions of the gravitational field and all sources of mass-energy (including the electromagnetic field), by performing variations with respect to the metric $g_{\mu\nu}$. This yields

$$R_{\mu\nu}-\frac{1}{2}g_{\mu\nu}R=\frac{8\pi G}{c^{4}}T_{\mu\nu},$$

(9)

where $R_{\mu\nu}$ is the Ricci tensor measuring the curvature of spacetime, $R=R_{\mu}^{\mu}=g^{\mu\nu}R_{\mu\nu}$ is the scalar curvature of spacetime, and $T_{\mu\nu}$ is the energy-momentum tensor of the matter (or electromagnetic field) given by

$$T_{\mu\nu}=\frac{2}{\sqrt{-g}}\frac{\partial(\sqrt{-g}{\cal L})}{\partial g^{\mu\nu}}=2\frac{\partial{\cal L}}{\partial g^{\mu\nu}}-g_{\mu\nu}{\cal L}.$$

(10)

From this formulation, the energy-momentum tensor (10) is automatically symmetric. For a macroscopic body, the energy-momentum tensor can be written as

$$T_{\mu\nu}=(\rho c^{2}+P)u_{\mu}u_{\nu}-Pg_{\mu\nu},$$

(11)

where $\rho c^{2}$ is the energy density, $P$ is the pressure and $u^{\mu}$ is the quadrivelocity such that $u_{\mu}u^{\mu}=1$ (for a fluid at rest $T_{00}=\rho c^{2}$ and $T_{xx}=T_{yy}=T_{zz}=P$).

The conservation of energy and momentum is expressed by

$$D_{\nu}T^{\mu\nu}=0,$$

(12)

where $D_{\nu}=\frac{1}{\sqrt{-g}}\partial_{\nu}\sqrt{-g}$ is the covariant derivative in a curved spacetime. According to this equation, the divergence of the left hand side of Eq. (9) must be zero. This is actually the case because of the contracted Bianchi identities. As a result, the conservation of energy and momentum is contained in the Einstein equations (9). The equations of the gravitational field contain the equations for the matter which produces this field. However, for a complete determination of the distribution and motion of the matter (or electromagnetic field) one must still supplement to the Einstein equations the equation of state of the matter (or electromagnetic field), i.e., an equation relating the pressure $P$ to the energy density $\rho c^{2}$. This equation must be given along with the gravitational field equations.

The fundamental equations of electromagnetism are the Maxwell equations. In a curved spacetime the first pair of Maxwell equations reads (one can replace the covariant derivative $D_{\mu}$ by the partial derivative $\partial_{\mu}$ in this equation)

$$\partial_{\lambda}F_{\mu\nu}+\partial_{\nu}F_{\lambda\mu}+\partial_{\mu}F_{\nu\lambda}=0.$$

(13)

The field equations determining the second pair of Maxwell equations can be obtained from the principe of least action $\delta S=0$ by varying the electromagnetic potential $A_{\mu}$. This yields

$$D_{\nu}\left[{\cal L}^{\prime}({\cal F})F^{\mu\nu}\right]=\mu_{0}J^{\mu}.$$

(14)

Equations (13) and (14) form the (nonlinear) Maxwell equations. We see that only the second pair of Maxwell equations is affected by a possible nonlinearity of the Lagrangian.

The conservation of charge is expressed by the equation of continuity

$$D_{\mu}J^{\mu}=0.$$

(15)

According to this equation, the divergence of the left hand side of Eq. (14) must be zero. This is actually the case because of the antisymmetry of the Faraday tensor. The conservation of charge is therefore included in the Maxwell equations.

The electromagnetic energy-momentum tensor associated with the Lagrangian ${\cal L}({\cal F})$ is

$$T_{\mu\nu}=-\frac{1}{\mu_{0}}{\cal L}^{\prime}({\cal F})F_{\mu\alpha}F^{\alpha}_{\,\,\,\nu}-{\cal L}({\cal F})g_{\mu\nu}.$$

(16)

We note that

$$T_{\mu}^{\mu}=4{\cal F}{\cal L}^{\prime}({\cal F})-4{\cal L}({\cal F}).$$

(17)

In linear electrodynamics (${\cal L}_{\rm Maxwell}=-{\cal F}$) the energy-momentum tensor of the electromagnetic field has the property that $T=T_{\mu}^{\mu}=0$. With the identity $R=-(8\pi G/c^{4})T$ obtained from the Einstein equations (9) it follows that in the presence of an electromagnetic field without any masses the scalar curvature of spacetime is zero: $R=0$. On the other hand, the identity $T=\rho c^{2}-3P$ obtained from Eq. (11) implies that $P=\rho c^{2}/3$. These relations are no more true for a nonlinear electrodynamics.

If we consider an expanding homogeneous and isotropic cosmological spacetime (background) with a uniform curvature, the line element is given by the Friedmann-Lemaître-Roberston-Walker (FLRW) metric

$\displaystyle ds^{2}=c^{2}dt^{2}-a(t)^{2}\left[\frac{dr^{2}}{1-kr^{2}}+r^{2}\,(d\theta^{2}+\sin^{2}\theta\,d\phi^{2})\right],$

(18)

where $a(t)$ represents the radius of curvature of the $3$-dimensional space, or the scale factor. By an abuse of language, we shall sometimes call it the “radius of the universe”. On the other hand, $k$ determines the curvature of space. The universe is closed if $k>0$, flat if $k=0$, or open if $k<0$.

If the universe is isotropic and homogeneous at all points in conformity with the line element (18), and contains a uniform perfect fluid of energy density $\rho(t)c^{2}$ and isotropic pressure $P(t)$, the energy-momentum tensor $T_{\mu\nu}$ is given by Eq. (11) and we have

$\displaystyle T^{0}_{0}=\rho c^{2},\qquad T^{1}_{1}=T^{2}_{2}=T^{3}_{3}=-P.$

(19)

The Einstein equations

$\displaystyle R^{\mu}_{\nu}-\frac{1}{2}g^{\mu}_{\nu}R-\Lambda g^{\mu}_{\nu}=-\frac{8\pi G}{c^{2}}T^{\mu}_{\nu}$

(20)

relate the geometrical structure of spacetime ($g_{\mu\nu}$) to the material content of the universe ($T_{\mu\nu}$) including the electromagnetic field. For the sake of generality, we have included the cosmological constant $\Lambda$ in the Einstein equations (20). Given Eqs. (18) and (19), the Einstein equations reduce to

$\displaystyle 8\pi G\rho+\Lambda=3\frac{{\dot{a}}^{2}+kc^{2}}{a^{2}},\qquad\frac{8\pi G}{c^{2}}P-\Lambda=-\frac{2a\ddot{a}+{\dot{a}}^{2}+kc^{2}}{a^{2}},$

(21)

where dots denote differentiation with respect to time. These are the well-known Friedmann friedmann1 ; friedmann2 cosmological equations. The Friedmann equations are usually written under the form

$$H^{2}=\frac{8\pi G}{3}\rho-\frac{kc^{2}}{a^{2}}+\frac{\Lambda}{3},$$

(22)

$$2\dot{H}+3H^{2}=-\frac{8\pi G}{c^{2}}P-\frac{kc^{2}}{a^{2}}+\Lambda,$$

(23)

where $H=\dot{a}/a$ is the Hubble parameter. From these equations, one can derive the acceleration equation

$$\frac{\ddot{a}}{a}=-\frac{4\pi G}{3}\left(\rho+\frac{3P}{c^{2}}\right)+\frac{\Lambda}{3}.$$

(24)

The deceleration parameter is defined by

$$q=-\frac{{\ddot{a}}a}{{\dot{a}}^{2}}.$$

(25)

The universe is decelerating when $q>0$ and accelerating when $q<0$.

Combining Eqs. (22) and (23), we obtain the energy conservation equation

$$\frac{d\rho}{dt}+3H\left(\rho+\frac{P}{c^{2}}\right)=0.$$

(26)

This equation can be directly derived from the relation (12) which is included in the Einstein equations through the contracted Bianchi identities. It can be written under the form

$$d\rho+3\frac{da}{a}(\rho+P/c^{2})=0.$$

(27)

For a given barotropic equation of state $P=P(\rho)$, we can solve Eq. (27) to obtain

$$\ln a=-\frac{1}{3}\int\frac{d\rho}{\rho+P(\rho)/c^{2}}.$$

(28)

This equation determines the relation $\rho(a)$ between the energy density and the scale factor. We can then solve the Friedmann equation (22) with $\rho=\rho(a)$ to obtain the temporal evolution of the scale factor $a(t)$.

The energy conservation equation (27) can be rewritten as

$${d}(\rho c^{2}a^{3})=-P{d(a^{3})}.$$

(29)

Introducing the volume $V\propto a^{3}$ and the energy $E=\rho c^{2}V$, Eq. (29) becomes $dE=-PdV$. It can be interpreted as the first principle of thermodynamics for an adiabatic evolution of the universe $dS=0$ lemaitre1927 ; lemaitre1931 ; cosmopoly1 .

The equation of state parameter is defined by

$$w=\frac{P}{\rho c^{2}}.$$

(30)

According to Eq. (26) the energy density decreases with the scale factor when $w>-1$ (null dominant energy condition) and increases with the scale factor when $w<-1$. When $w=-1$ the energy density is constant. The case where the energy density increases with the scale factor corresponds to a phantom universe caldwell ; cosmopoly3 .

Remark: It is possible to develop a useful mechanical analogy to study the Friedmann equation (22). Indeed, it can be cast in the suggestive form

$$E=\frac{1}{2}{\dot{a}}^{2}+V(a),$$

(31)

where

$$E=-\frac{1}{2}kc^{2},\qquad V(a)=-\frac{4\pi G}{3}\rho(a)a^{2}-\frac{\Lambda}{6}a^{2}.$$

(32)

Eq. (31) has the structure of the first integral of motion for a particle of unit mass in a potential $V(a)$. In that case, $E$ represents its conserved energy. The Friedmann equation (22) then has the solution

$$t=\int\frac{da}{\sqrt{2[E-V(a)]}}$$

(33)

determining $a(t)$ in reversed form.

In this paper, we consider a flat universe ($k=0$) in agreement with the inflation paradigm guthinflation and the observations of the cosmic microwave background (CMB) planck2014 ; planck2016 . On the other hand, we set the cosmological constant equal to zero ($\Lambda=0$) because dark energy will be taken into account in the nonlinear electrodynamics. The Friedmann equations (22) and (23) then reduce to the form

$$H^{2}=\left(\frac{\dot{a}}{a}\right)^{2}=\frac{8\pi G}{3}\rho,$$

(34)

$$2\dot{H}+3H^{2}=-\frac{8\pi G}{c^{2}}P.$$

(35)

The acceleration equation (24) becomes

$$\frac{\ddot{a}}{a}=-\frac{4\pi G}{3}\left(\rho+\frac{3P}{c^{2}}\right).$$

(36)

Using Eqs. (34) and (36) we see that the deceleration parameter (25) is related to the equation of state parameter (30) by

$$q=\frac{1+3w}{2}.$$

(37)

The universe is decelerating when $w>-1/3$ (strong energy condition) and accelerating when $w<-1/3$; when $w=-1/3$ the scale factor increases linearly with time cosmopoly1 .

Let us study some general properties of nonlinear electrodynamics in cosmology novello1 ; novello2 ; cgcl ; novello3 ; vollick ; novello4 ; garcia ; kruglov2015a ; kruglov2015 ; kruglov2020 . We assume that the universe is filled with electromagnetic radiation. The electromagnetic field that is of cosmological interest is the cosmic microwave background (CMB). It can be considered as a random field of short coherent radiation wavelength as compared to the cosmological horizon scales. Due to the isotropy of the spatial sections of the FLRW model, an average procedure is needed for compatibility reason if the electromagnetic field is to act as a source of gravity. Using the usual Tolman-Ehrenfest te procedure, we assume that the averaged electromagnetic field obeys the equations

$$\overline{E}_{i}=0,\quad\overline{B}_{i}=0,\quad\overline{E_{i}B_{j}}=0,$$

(38)

$$\overline{E_{i}E_{j}}=\frac{1}{3}E^{2}\delta_{ij},\quad\overline{B_{i}B_{j}}=\frac{1}{3}B^{2}\delta_{ij},$$

(39)

where $\overline{X}$ denotes an average over a volume that is large compared to the radiation wavelength but small compared to the curvature of spacetime. In the following, we omit the averaging bars for notational simplicity.

With these conditions, the average value of the energy-momentum tensor of the electromagnetic field associated with the Lagrangian density ${\cal L}({\cal F})$ [see Eq. (16)] can be written in the form of the energy-momentum tensor for a perfect fluid [see Eq. (11)]. The energy density and the pressure of the radiation are given by novello2 ; novello3

$$\rho c^{2}=-{\cal L}({\cal F})-E^{2}{\cal L}^{\prime}({\cal F}),$$

(40)

$$P={\cal L}({\cal F})+\frac{E^{2}-2B^{2}}{3}{\cal L}^{\prime}({\cal F}),$$

(41)

where $E^{2}$ and $B^{2}$ are the averaged electric and magnetic fields squared, respectively. For convenience we have rescaled ${\bf E}$ by $1/(\epsilon_{0})^{1/2}$ and ${\bf B}$ by $(\mu_{0})^{1/2}$. The electromagnetic invariant takes the form

$${\cal F}=\frac{1}{4}F_{\mu\nu}F^{\mu\nu}=\frac{B^{2}-E^{2}}{2}.$$

(42)

For Maxwell’s linear electrodynamics described by the Lagrangian (8) the foregoing equations reduce to

$$\rho c^{2}=\frac{1}{2}(E^{2}+B^{2}),\qquad P=\frac{1}{6}(E^{2}+B^{2}),$$

(43)

returning the usual equation of state of radiation

$$P=\frac{1}{3}\rho c^{2}.$$

(44)