Loading [MathJax]/extensions/TeX/boldsymbol.js

Isotropization of the magnetic universe in Horndeski theory with G3(X,ϕ) and G5(X)

Figures(1)

Get Citation
Ruslan K. Muharlyamov, Tatiana N. Pankratyeva and Shehabaldeen O.A. Bashir. Isotropization of the magnetic universe in the Horndeski theory with G_3(X,ϕ) and G_5(X)[J]. Chinese Physics C. doi: 10.1088/1674-1137/ad65de
Ruslan K. Muharlyamov, Tatiana N. Pankratyeva and Shehabaldeen O.A. Bashir. Isotropization of the magnetic universe in the Horndeski theory with G_3(X,ϕ) and G_5(X)[J]. Chinese Physics C.  doi: 10.1088/1674-1137/ad65de shu
Milestone
Received: 2024-05-12
Article Metric

Article Views(1083)
PDF Downloads(17)
Cited by(0)
Policy on re-use
To reuse of subscription content published by CPC, the users need to request permission from CPC, unless the content was published under an Open Access license which automatically permits that type of reuse.
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Email This Article

Title:
Email:

Isotropization of the magnetic universe in Horndeski theory with G3(X,ϕ) and G5(X)

  • 1. Department of General Relativity and Gravitation, Institute of Physics, Kazan Federal University, Kremlevskaya str. 18, Kazan 420008, Russia
  • 2. Department of Higher Mathematics, Kazan State Power Engineering University, Krasnoselskaya str. 51, Kazan 420066, Russia
  • 3. Department of General Relativity and Gravitation, Institute of Physics, Kazan Federal University, Kremlevskaya str. 18, Kazan 420008, Russia; Department of physics, Faculty of Science, University of Khartoum, Khartoum 11115, Sudan

Abstract: We study the isotropization process of Bianchi-I space-times in Horndeski theory with G3(X,ϕ)0 and G5=const/X. A global unidirectional electromagnetic field interacts with a scalar field according to the law f2(ϕ)FμνFμν. In Horndeski theory, the anisotropy can develop in different ways. The proposed reconstruction method allows us to build models with acceptable anisotropy behavior. To analyze space-time anisotropy, we use the relations ai/a (i=1,2,3), where ai are metric functions, and a(a1a2a3)1/3.

    HTML

    I.   INTRODUCTION
    • Large-scale magnetic fields in the Universe are important and enigmatic phenomena. The influence of these fields on the evolution of the Universe has been the subject of many studies over the years [111]. The origin of the magnetic component of the Universe is unknown, but there are various hypotheses. Here, we adhere to the hypothesis regarding the primordial origin of the magnetic field, that is, the field may arise before or during primary inflation. There is academic interest in the interaction of the magnetic field with the scalar field that causes inflation. In addition, the combined influence of these fields on the development of the Universe expands the possibilities of observing its dark sector.

      To explain the accelerated expansion of the Universe and other observational facts, various modifications of gravity theory are used. A prominent representative is Horndeski gravity (HG) [12]. HG is constructed in such a way that the motion equations are of the order of a derivative no higher than the second. In this sense, HG is the most general variant of the scalar-tensor theory of gravitation. The action density for HG can be represented as follows [13]:

      LH=g(L2+L3+L4+L5),

      (1)

      L2=G2(ϕ,X),L3=G3(ϕ,X)ϕ,

      L4=G4(ϕ,X)R+G4X(ϕ,X)[(ϕ)2(μνϕ)2],

      L5=G5(ϕ,X)Gμνμνϕ16G5X[(ϕ)33ϕ(μνϕ)2+2(μνϕ)3],

      (2)

      where g is the determinant of the metric tensor gμν, R is the Ricci scalar, and Gμν is the Einstein tensor. The factors Gi (i=2,3,4,5) are arbitrary functions of the scalar field ϕ and canonical kinetic term X=12μϕμϕ. We consider the definitions GiXGi/X, (μνϕ)2μνϕνμϕ, and (μνϕ)3μνϕνρϕρϕμϕ.

      The electromagnetic component is represented in the form

      LFϕ=14[f(ϕ)]2FμνFμν,

      (3)

      where Fμν is the electromagnetic field. Selection f=const corresponds to minimal interaction between the electromagnetic and scalar fields.

      In recent decades, the technical basis of observational science has undergone significant development; for example, the Wilkinson Microwave Anisotropy Probe (WMAP) [14], Planck satellites [15], and the Dark Energy Spectroscopic Instrument (DESI) [16, 17]. Observational data indicate that the modern Universe is almost isotropic. In Ref. [18], constraints were obtained on the isotropy of the Universe in a general test using Planck’s data on the temperature and polarization of cosmic microwave background (CMB) radiation. Аnomalies have been observed on a large scale of CMB radiation; therefore, the early Universe may have been anisotropic. The Bianchi Universe can explain these CMB anomalies [1921].

      A cosmological model containing the global magnetic field is necessarily anisotropic because the magnetic field vector specifies a preferred spatial direction. Here, we consider Bianchi type-I space-time (BI). BI models have been studied from different perspectives [2233]. In general relativity (GR), a scalar field is not anisotropic matter, whereas in HG, the scalar field can increase the anisotropy level (the "anisotropization" process) [34, 35]. An important criterion of the viability of any anisotropic model is a sufficiently low level of anisotropy at certain stages of Universe evolution. It was argued in [36] that from the perspective of particle production, a significant decrease in anisotropy should occur early on and no later than the beginning of primary nucleosynthesis. Ref. [37] analyzed the effects of cosmic anisotropy on the primordial production of 4He. The modern Universe is close to an isotropic state [38], and the issue of isotropization in the modified theories of gravity has been studied in many studies [3944]. In this study, we search for models with acceptable anisotropy behaviors using a reconstruction method that is commonly used by researchers [4551]. There are five arbitrary functions Gi(X,ϕ), f(ϕ) in action densities LH, LFϕ. This number of degrees of freedom ensures an effective reconstruction method.

      Here, we develop a reconstruction algorithm for case G5=const/X, which distinguishes this study from previous ones [4648]. The function G5(X,ϕ) provides a non-minimal kinetic coupling to the spacetime curvature [52, 53], which may appear in some Kaluza-Klein theories [54, 55]. We also select the following HG functions:

      G2=εXV(ϕ),G30,G4=116π,ε=±1.

      (4)

      We previously considered the theory with G5=const/X in Ref. [56], in which G3=0 was assumed. The anisotropic properties of the models were analyzed based on the ratio σ/H, where σ is the shear scalar. Cosmological models with a constant small anisotropy value and those with a decreasing anisotropy to a constant small value were obtained. Here, we determine the anisotropy of the Universe based on the ratios ai/a, where ai are metric functions in (5). As a result, we obtain models with anisotropy that can have a value arbitrarily close to zero during Universe expansion. A comparative analysis of different isotropization criteria can be found in Ref. [6].

    II.   FIELD EQUATIONS
    • We consider the homogeneous and anisotropic Bianchi I metric

      ds2=dt2+a21(t)dx21+a22(t)dx22+a23(t)dx23.

      (5)

      The field equations of HG have the form

      18πG00=G2G2X˙ϕ23G3XH˙ϕ3+G3ϕ˙ϕ25G5XH1H2H3˙ϕ3G5XXH1H2H3˙ϕ5+T(em)00,

      (6)

      18πGii=G2˙ϕdG3dtddt(G5X˙ϕ3HjHk)G5X˙ϕ3HjHk(Hj+Hk)+T(em)ii,

      (7)

      1a3ddt(a3˙ϕ[G2X2G3ϕ+3H˙ϕG3X+H1H2H3(3G5X˙ϕ+G5XX˙ϕ3)])==G2ϕ˙ϕ2(G3ϕϕ+G3Xϕ¨ϕ)f(ϕ)fϕ(ϕ)2FγδFγδ,

      (8)

      μ[a3f2(ϕ)Fμν]=0.

      (9)

      The dot denotes the t-derivative, Hi=˙ai/ai, and the average Hubble parameter is H=133i=1Hi˙a/a with a=(a1a2a3)1/3. In equation (7), there is no summation over the indices i; the triples of indices {i,j,k} take values of {1,2,3}, {2,3,1}, or {3,1,2}. The Einstein tensor components are

      G00=(H1H2+H2H3+H3H1),

      (10)

      Gii=(˙Hj+˙Hk+H2j+H2k+HjHk).

      (11)

      The stress–energy tensor of the electromagnetic field is expressed as

      T(em)μν=f2(ϕ)(14δμνFγδFγδ+FνβFμβ).

      (12)

      We assume that there are homogeneous electric and magnetic fields with the same direction x3. The electromagnetic field tensor Fγδ has non-vanishing components,

      F03=F30=qea3f2(ϕ),F21=F12=qm,

      (13)

      where qe and qm are constants. The electric and magnetic field strengths are determined by the equalities

      E2=F03F30=q2ea21a22f4(ϕ),B2=F21F21=q2ma21a22.

      (14)

      We assume that there is no electric field, that is, qe=0.

      The tensor T(em)μν has non-zero components:

      T(em)00=T(em)33=T(em)11=T(em)22==f2(ϕ)B22=Ψ(ϕ)2a21a22,

      (15)

      where

      Ψ(ϕ)q2mf2(ϕ)>0.

      (16)

      We parameterize the three scale factors as follows:

      a1=aeβ++3β,a2=aeβ+3β,a3=ae2β+.

      (17)

      Then, the Hubble parameters in the directions x1, x2 , and x3 are given by

      H1=H+˙β++3˙β,H2=H+˙β+3˙β,H3=H2˙β+.

      (18)

      In view of (17) and (18), Eqs. (6), (7), and (8) have the consequences

      38π(H2σ2)=Ψ(ϕ)2a4e4β+G2+˙ϕ2G2X+3G3XH˙ϕ3G3ϕ˙ϕ2++˙ϕ3(5G5X+G5XX˙ϕ2)×(H2˙β+)[(H+˙β+)23˙β2],σ2˙β2++˙β2,

      (19)

      18π(2˙H+3H2+3σ2)=Ψ(ϕ)6a4e4β+G2+G3ϕ˙ϕ2+G3X˙ϕ2¨ϕ++ddt[G5X˙ϕ3(H2σ2)]+2G5X˙ϕ3(H3+˙β3+3˙β+˙β2),

      (20)

      ˙β8π+G5X˙ϕ3(2˙β+˙βH˙β)=Ca3,

      (21)

      ˙β+8π+G5X˙ϕ3(˙β2˙β2+H˙β+)=13a3Ψ(ϕ)dtae4β++C+a3,

      (22)

      ˙ϕ{G2X2G3ϕ+3H˙ϕG3X+(3G5X˙ϕ+G5XX˙ϕ3)(H2˙β+)×[(H+˙β+)23˙β2]}=1a3(G2ϕ˙ϕ2(G3ϕϕ+G3Xϕ¨ϕ)Ψϕ2a4e4β+)a3dt+Cϕa3,

      (23)

      where Cϕ, C+ , and C are integration constants.

      The system in (21) and (22) is nonlinear for ˙β±. Therefore, the "anisotropization" process [34, 35] by the scalar field is possible even in the absence of an anisotropic source Ψ(ϕ)ae4β+.

      We use the following criterion for isotropization of the cosmological model:

      aiaconstasv+,

      (24)

      where v=a3 is the volume factor.

      Next, we apply the reconstruction method. Using (24), we set a3/a=1, or

      β+=0.

      (25)

      Therefore,

      H3=H.

      (26)

      We will study the behavior of relations a1/a and a2/a.

      From equalities (14), (17), and (25), it follows that the magnetic field strength decreases monotonically with Universe expansion.

      B=consta2.

      (27)

      Next, let us set

      G5=132πγX,

      (28)

      C=0.

      (29)

      The choice in (28) allows us, without unnecessary technical difficulties, to obtain interesting cosmological models based on the nonlinearity of Eqs. (21) and (22) relative to ˙β±. As shown below, function (28) allows for acceptable anisotropy behavior in cosmological models.

      Then, from (21) and (22), it follows that

      H=γ˙ϕ,

      (30)

      ˙β=8πγ˙ϕ3a3[dtΨ(ϕ)a+3C+].

      (31)

      The shear scalar σ2 and Hubble parameters take the form

      σ2=8πγ˙ϕ3a3[dtΨ(ϕ)a+3C+],

      (32)

      H1,2=H±8πγ˙ϕa3[dtΨ(ϕ)a+3C+].

      (33)

      From (30), the connection between the scalar field ϕ and scale factor a is

      ϕ=1γln(ac1),ora=c1eγϕ.

      (34)

      The system in (19) and (23) takes the form

      3γ˙ϕ2{G3X˙ϕ2γ8π}=G2˙ϕ2G2X+G3ϕ˙ϕ2γ2˙ϕ28πΨ(ϕ)2a4,

      (35)

      3γ˙ϕ2{G3X˙ϕ2γ8π}=˙ϕ2G2X+2G3ϕ˙ϕ22γ2˙ϕ28π+(Cϕ3γC+)˙ϕa3+˙ϕa3dta3×[G2ϕ˙ϕ2(G3ϕϕ+G3Xϕ¨ϕ)12a4(Ψϕ+2γΨ)].

      (36)

      Eq. (20) can be ignored because it is automatically fulfilled via the Bianchi identities. We have four unknown functions (a(t),G2,G3,Ψ) and two independent equations, (35) and (36). Function ϕ(t) is determined from (34). Therefore, we have two degrees of freedom.

      We make the first assumption for the integral in (36):

      G2ϕ+˙ϕ2(G3ϕϕ+G3Xϕ¨ϕ)+c0e2γϕ2a6(Ψϕ+2γΨ)=HUa(a)a2.

      (37)

      By specifying function U(a), we exhaust one degree of freedom. Equation (36) becomes

      3γ˙ϕ2{G3X˙ϕ2γ8π}=˙ϕ2G2X+2G3ϕ˙ϕ22γ2˙ϕ28π+˙ϕa3(Cϕ3γC+U(a)).

      (38)

      Thus, we obtain the system in (35), (37), and (38).

      Let us consider the following model:

      G3=η2lnXC+χ2X+G(ϕ),G2=εXV(ϕ).

      (39)

      The form G3lnX is associated with the existence of the black hole solution with a scalar hairy black hole [57, 58]. The logarithmic form G3 was also studied in Refs. [59, 60]. From Eqs. (35), (37), and (38), it follows that

      ˙ϕ[γ28πε3γη3γχ|˙ϕ|+2Gϕ]=U(a)Cϕ+3γC+a3,

      (40)

      V+Ψ(ϕ)2a4=˙ϕ2[2γ28πε23γη3γχ|˙ϕ|+Gϕ],

      (41)

      Vϕ+12a4(Ψϕ+2γΨ)=γ˙ϕUaa2˙ϕ2Gϕϕ.

      (42)

      If we choose G(ϕ) and U(a) and consider (34), from (40) and (34), we obtain the function ˙ϕ(a). Therefore, the right-hand sides of Eqs. (41) and (42) are expressed through the scale factor a:

      V(ϕ(a))+Ψ(ϕ(a))2a4=S(a)>0,

      (43)

      Va+12a4(Ψa+2Ψa)=N(a),

      (44)

      where

      S(a)˙ϕ2[2γ28πε23γη3γχ|˙ϕ|+Gϕ],

      (45)

      N(a)1γa[γ˙ϕUaa2˙ϕ2Gϕϕ].

      (46)

      The system in (43) and (44) has the solution

      Ψ=a53(NSa),

      (47)

      V=S+a6(SaN).

      (48)

      Thus, two degrees of freedom are reduced to the choice of two functions, U(a) and G(ϕ). Below, we provide examples of these functions for which isotropization condition (24) is satisfied. We restore functions V(ϕ)and Ψ(ϕ) and analyze the resulting models.

      Functions G3lnX and G51/X have a exotic form. However, these functions and the kinetic part X of the well-known function G2=εXV make a homogeneous and standard contribution to the field equations, namely, ˙ϕ2. In Eqs. (40) and (41), combinations of the parameters γ28π, ε, and γη arise. Thus, in addition to functional freedom, U(a) and Gϕ), we obtain parametric freedom. By adjusting these parameters, we obtain cosmological solutions with certain properties, including anisotropic ones.

    III.   MODEL \boldsymbol\chi\boldsymbol\neq \bf 0 , \boldsymbol U(a)=\bf 0 , \boldsymbol G=\boldsymbol\xi {\rm e}^{-k\phi}
    • Here, we assume

      \begin{aligned} U(a)=0,\,\, G=\xi {\rm e}^{-k\phi}, \,\, \xi>0,\,\,k>0, \,\, C_\phi=C_+=0. \end{aligned}

      (49)

      In this case, from Eqs. (40), (45), and (46),

      \begin{aligned}[b] |\dot{\phi}|&=\dfrac{1}{3\gamma\chi}\left[\dfrac{\gamma^2}{8\pi}-\varepsilon-3\gamma\eta-2\xi k {\rm e}^{-k\phi}\right]\\& =\dfrac{1}{3\gamma\chi}\left[\dfrac{\gamma^2}{8\pi}-\varepsilon-3\gamma\eta-2\xi k \left(\dfrac{c_1}{a}\right)^{k/\gamma}\right], \end{aligned}

      (50)

      \begin{aligned}[b] S(a)=&\dfrac{1}{9\gamma^2\chi^2}\left[\dfrac{\gamma^2}{8\pi}-\varepsilon-3\gamma\eta-2\xi k \left(\dfrac{c_1}{a}\right)^{k/\gamma}\right]^2\\& \times\left[\dfrac{\gamma^2}{8\pi}+\dfrac{\varepsilon}{2}+\xi k \left(\dfrac{c_1}{a}\right)^{k/\gamma}\right], \end{aligned}

      (51)

      \begin{aligned}[b] N(a)=&-\dfrac{\xi k^2}{9\gamma^3\chi^2 a}\left(\dfrac{c_1}{a}\right)^{k/\gamma} \\& \times\left[\dfrac{\gamma^2}{8\pi}-\varepsilon-3\gamma\eta-2\xi k \left(\dfrac{c_1}{a}\right)^{k/\gamma}\right]^2. \end{aligned}

      (52)

      Equations (47) and (48) provide the functions \Psi(\phi) and V(\phi) :

      \begin{aligned}[b] \Psi=&\dfrac{2c_1^4 k^2\xi {\rm e}^{(4\gamma-k)\phi}}{27\chi^2\gamma^3}\cdot\left[\varepsilon+\dfrac{2\gamma^2}{8\pi}+2\xi k {\rm e}^{-k\phi}\right]\\& {\times \left[\varepsilon-\dfrac{\gamma^2}{8\pi}+3\gamma\eta+2\xi k {\rm e}^{-k\phi}\right],} \end{aligned}

      (53)

      \begin{aligned}[b] V=&\dfrac{1}{18\gamma^2\chi^2}\left[\varepsilon+\dfrac{2\gamma^2}{8\pi}+2\xi k {\rm e}^{-k\phi}\right]\\& {\times \left[\varepsilon-\dfrac{\gamma^2}{8\pi}+3\gamma\eta+2\xi k {\rm e}^{-k\phi}\right]}\\& {\times\left[\varepsilon-\dfrac{\gamma^2}{8\pi}+3\gamma\eta+2\xi k {\rm e}^{-k\phi}\left(1-\dfrac{k}{3\gamma}\right)\right].} \end{aligned}

      (54)

      The inequalities \xi>0 , k>0 , and

      \begin{aligned} \varepsilon+\frac{2\gamma^2}{8\pi}\geqslant0,\,\,\varepsilon-\frac{\gamma^2}{8\pi}+3\gamma\eta\geqslant0,\,\,k<3\gamma,\,\, \gamma>0,\,\,\chi<0 \end{aligned}

      (55)

      lead to the constraint \Psi>0 and V>0 . Hereafter, we adhere to these conditions.

      The Hubble parameters take the form

      \begin{aligned} H=H_3=\frac{1}{3|\chi|}\left[\varepsilon-\frac{\gamma^2}{8\pi}+3\gamma\eta\right.\left.+2\xi k\left(\frac{c_1}{a}\right)^{k/\gamma}\right], \end{aligned}

      (56)

      \begin{aligned} H_{1,2}=H\left[1\pm\sqrt{\dfrac{16\pi \xi k^2}{3\gamma^3}\left(\dfrac{c_1}{a}\right)^{k/\gamma}\cdot\dfrac{\dfrac{\varepsilon+\dfrac{2\gamma^2}{8\pi}}{3-k/\gamma}+\dfrac{2\xi k}{3-2k/\gamma}\left(\dfrac{c_1}{a}\right)^{k/\gamma}}{\varepsilon-\dfrac{\gamma^2}{8\pi}+3\gamma\eta+2\xi k\left(\dfrac{c_1}{a}\right)^{k/\gamma}}}\,\right], ~~ k<3\gamma/2. \end{aligned}

      (57)

      To analyze the anisotropy level, we write the relation

      \begin{aligned}[b] \dfrac{\sigma^2}{H^2}=&\dfrac{8\pi}{3a^{3}H} \displaystyle{\int} {\rm d}t\, \dfrac{\Psi(\phi)}{a} =\dfrac{16\pi \xi k^2}{9\gamma^3}\left(\dfrac{c_1}{a}\right)^{k/\gamma}\\&\cdot\dfrac{\dfrac{\varepsilon+\dfrac{2\gamma^2}{8\pi}}{3-k/\gamma}+\dfrac{2\xi k}{3-2k/\gamma}\left(\dfrac{c_1}{a}\right)^{k/\gamma}}{\varepsilon-\dfrac{\gamma^2}{8\pi}+3\gamma\eta+2\xi k\left(\dfrac{c_1}{a}\right)^{k/\gamma}}. \end{aligned}

      (58)

      The scale factor has the form

      \begin{aligned} a=a_3=c_1\left\{\frac{2\xi k}{3|\chi|H_\infty}\left[\exp\left(\frac{kH_\infty t}{\gamma}\right)-1\right]\right\}^{\gamma/k},\,\, t\geq 0, \end{aligned}

      (59)

      H_\infty\equiv\frac{1}{3|\chi|}\left(\varepsilon-\frac{\gamma^2}{8\pi}+3\gamma\eta\right).

      We use the following approximations (H_\infty\neq0) :

      \begin{aligned} a=a_3\propto t^{\gamma/k } \,\, \text{as}\,\, \, H_\infty t\ll 1; \,\,\, a=a_3\propto {\rm e}^{H_\infty t} \,\, \text{as}\,\, \, H_\infty t\gg 1. \end{aligned}

      (60)

      In the case \gamma<k<3\gamma/2 , the phase of Universe expansion without acceleration is replaced by a phase of accelerated expansion. When k<\gamma , power-law inflation is replaced by the exponential inflation over time. The scalar field ϕ has the time dependence

      \begin{aligned} \phi=\frac{1}{k}\ln\left\{\frac{2\xi k}{3|\chi|H_\infty}\left[\exp\left(\frac{kH_\infty t}{\gamma}\right)-1\right]\right\}. \end{aligned}

      (61)
    • A.   General case

    • Now, we consider the qualitative anisotropic behavior of the model for the following case:

      \begin{aligned} \varepsilon+\frac{2\gamma^2}{8\pi}>0,\,\,\,\varepsilon-\frac{\gamma^2}{8\pi}+3\gamma\eta>0. \end{aligned}

      (62)

      For H_\infty t\gg 1 , the following approximation is valid:

      \begin{aligned} H_{1,2}\approx H_\infty \left[1\pm \sqrt{\frac{8\pi k\left(\varepsilon+\dfrac{2\gamma^2}{8\pi}\right)}{3\gamma^3\left(3-k/\gamma\right)}}\cdot {\rm e}^{-\tfrac{kH_\infty t}{2\gamma}}\right] \end{aligned}

      (63)

      \begin{aligned} a_{1,2}\approx s_{1,2}\cdot {\rm e}^{H_\infty t}\cdot\exp\left\{\mp 4\sqrt{\dfrac{2\pi \left(\varepsilon+\dfrac{2\gamma^2}{8\pi}\right)}{3k\gamma\left(3-k/\gamma\right)}}\cdot {\rm e}^{-\tfrac{kH_\infty t}{2\gamma}}\right\}. \end{aligned}

      (64)

      Consequently, the model isotropization condition is satisfied,

      \begin{aligned}[b] \frac{a_{1,2}}{a}\sim &\exp\left\{\mp 4\sqrt{\frac{2\pi \left(\varepsilon+\frac{2\gamma^2}{8\pi}\right)}{3k\gamma\left(3-k/\gamma\right)}}\cdot {\rm e}^{-\tfrac{kH_\infty t}{2\gamma}}\right\}\\&\rightarrow \text{const}\,\, \,\text{as}\,\, \, H_\infty t\rightarrow +\infty. \end{aligned}

      (65)

      At early times ( H_\infty t\ll 1 ), we obtain

      \begin{aligned} H_{1,2}\approx\pm\sqrt{\frac{8\pi |\chi|}{ k^2(3-2k/\gamma)}}\cdot \frac{1}{t^{3/2}}+\frac{\gamma}{k}\cdot\frac{1}{t} \end{aligned}

      (66)

      \begin{aligned} a_{1,2}\approx s_{1,2}\cdot t^{\gamma/k}\cdot\exp\left\{\mp4\sqrt{\frac{2\pi|\chi|}{k^2(3-2k/\gamma)}}\cdot\frac{1}{t^{1/2}}\right\}. \end{aligned}

      (67)

      The Universe is expanding along the x_1 and x_3 axes: H_{1,3}>0 , and the Hubble parameter H_2 takes on negative and positive values. The Universe contracts in early times and expands in late times along the x_2 axis. A change in the sign of H_2 indicates a possible bounce of the scale factor a_2 . At the beginning t=0 , the Universe is an infinite straight line along the x_2 axis: a_1(0)=a_3(0)=0 , a_2(0) =+\infty . If there is a bounce, the model can only be applied to the early Universe before the end of primary inflation.

      In the case

      \begin{aligned} \frac{1}{3-2k/\gamma}\left(\varepsilon-\frac{\gamma^2}{8\pi}+3\gamma\eta\right)= \left(\varepsilon+\frac{2\gamma^2}{8\pi}\right)\frac{1}{3-k/\gamma}>0, \end{aligned}

      (68)

      using (62), we obtain an example of an exact solution in elementary functions:

      \begin{aligned} a_{1,2}=s_{1,2}\cdot a\cdot\exp \left\{\mp8\cdot\sqrt{\frac{\pi \xi}{3\gamma(3-2k/\gamma)}}\cdot\left(\frac{c_1}{a}\right)^{k/(2\gamma)}\,\right\}. \end{aligned}

      (69)

      The scale factors a_i(a) are shown in Fig. 1, revealing a bouncing scale factor a_2 . The function a_2 falls from a greater value at the beginning, bounces to the minimum value

      Figure 1.  Profile of a_{i} ; k=7\gamma/5 , \xi=3\gamma/(320\pi) , c_1=1 .

      \begin{aligned} a_{2\min}=c_1\cdot\left(\frac{16\pi \xi k^2}{3\gamma^3(3-2k/\gamma)}\right)^{\gamma/k}, \end{aligned}

      (70)

      and then rises again at the end.

    • B.   Special cases

    • Let us consider the following borderline cases:

      1. If we set \varepsilon+\dfrac{2\gamma^2}{8\pi}=0 and \varepsilon-\dfrac{\gamma^2}{8\pi}+3\gamma\eta>0 , then

      \begin{aligned} \varepsilon=-1, \,\, \gamma=2\sqrt{\pi}, \,\, \eta>\frac{1}{4\sqrt{\pi}}, \,\, k<3\sqrt{\pi}, \end{aligned}

      (71)

      and scale factors a_{1,2} are expressed through the scale factor a:

      \begin{aligned}[b] a_{1,2}=&s_{1,2}\cdot a\times \exp\left\{\mp4\sqrt{\dfrac{\pi}{3k(3\sqrt{\pi}-k)}}\right.\\&\left.\cdot\sqrt{-3/2+6\eta\sqrt{\pi}+2\xi k\left(\dfrac{c_1}{a}\right)^{k/(2\sqrt{\pi})}}\right\} \end{aligned}

      (72)

      From (59) and (71),

      \begin{aligned} a=c_1\left\{\frac{2\xi k}{3|\chi|H_\infty}\left[\exp\left(\frac{kH_\infty t}{2\sqrt{\pi}}\right)-1\right]\right\}^{2\sqrt{\pi}/k },\,\, t\geq 0, \end{aligned}

      (73)

      H_\infty=\frac{4\eta\sqrt{\pi}-1}{2|\chi|}.

      In the case 2\sqrt{\pi}<k<3\sqrt{\pi} , the phase of Universe expansion without acceleration is replaced by a phase of accelerated expansion. When k<2\sqrt{\pi} , power-law inflation is replaced by exponential inflation over time.

      The anisotropic behavior of the model does not differ qualitatively from the previous one. The model becomes isotropic over time, and the scale factor a_2 has a bounce. At the beginning t=0 , the Universe is an infinite straight line along the x_2 axis.

      2. If we set \varepsilon-\dfrac{\gamma^2}{8\pi}+3\gamma\eta=0 ( H_\infty= 0 ) and \varepsilon+\dfrac{2\gamma^2}{8\pi} > 0, the scale factors a_{1,2} are expressed through the scale factor a:

      \begin{aligned}[b] a_{1,2}=&s_{1,2}\cdot a\cdot\left[ \sqrt{1 +\dfrac{\nu^2 }{A^2_0}\left(\dfrac{a}{c_1}\right)^{k/\gamma}}+ \sqrt{\dfrac{\nu^2}{A^2_0}\left(\dfrac{a}{c_1}\right)^{k/\gamma}}\,\right]^{\pm\tfrac{2|\nu|\gamma}{k}}\\& \times\exp \left\{\mp\dfrac{2\gamma}{k}\cdot\sqrt{\nu^2+A^2_0\left(\dfrac{c_1}{a}\right)^{k/\gamma}}\,\right\}, \end{aligned}

      (74)

      \begin{aligned} \nu^2\equiv\dfrac{8\pi k}{3\gamma^3}\cdot \dfrac{\varepsilon+\dfrac{2\gamma^2}{8\pi}}{3-k/\gamma}\,,\,\, A_0^2\equiv\dfrac{16\pi \xi k^2}{3\gamma^3(3-2k/\gamma)}. \end{aligned}

      (75)

      From (58), it follows that

      \begin{aligned} \frac{\sigma^2}{H^2}=\frac{\nu^2}{3}+\frac{A^2_0}{3}\left(\frac{c_1}{a}\right)^{k/\gamma}. \end{aligned}

      (76)

      The isotropization condition (24) is not satisfied,

      \begin{aligned} \frac{a_{1,2}}{a}\sim a^{\pm |\nu|}\,\,\, \text{as} \,\,\, a \rightarrow +\infty. \end{aligned}

      (77)

      However, when |\nu|\ll 1 , functions a_{1,2}/a change within a narrow range, a^{\pm |\nu|}\approx \text{const} ; therefore, the anisotropy level tends to a constant low value,

      \begin{aligned} \left|\frac{\sigma}{H}\right|\rightarrow \frac{|\nu|}{\sqrt{3}}\ll 1\,\,\, \text{as} \,\,\, a \rightarrow +\infty. \end{aligned}

      (78)

      A small anisotropy is allowed by observational data.

      The Hubble parameter and scale factor are, respectively, given by

      \begin{aligned} H=\frac{2\xi k}{3|\chi|}\left(\frac{c_1}{a}\right)^{k/\gamma}, \end{aligned}

      (79)

      \begin{aligned} a=c_1\left[\frac{2\xi k^2}{3|\chi|\gamma}\cdot t\right]^{\gamma/k},\,\, t\geqslant 0. \end{aligned}

      (80)

      In the case k/\gamma<1 , we have power-law inflation. The scalar field ϕ has the time dependence

      \begin{aligned} \phi=\frac{1}{k}\ln\left(\frac{2\xi k^2 t}{3|\chi|\gamma}\right). \end{aligned}

      (81)

      At early times, the model behaves in the same manner as the previous ones.

    IV.   MODEL \boldsymbol\chi= \bf 0 , \boldsymbol G=\boldsymbol\xi\cdot\boldsymbol\phi
    • Let us consider the following model:

      \begin{aligned} \chi=0,\,\, G=\xi\cdot\phi. \end{aligned}

      (82)

      In this case, from Eqs. (40), (45), and (46) it follows that

      \begin{aligned} \dot{\phi}=\frac{\widetilde{U}}{\alpha a^3},\,\, \widetilde{U}\equiv U(a)-C_\phi+3\gamma C_+, \end{aligned}

      (83)

      \begin{aligned} S(a)= \frac{\beta \widetilde{U}^2}{\alpha^2a^6},\,\, N(a)=\frac{\widetilde{U}\widetilde{U}'_a}{\alpha a^6}\,, \end{aligned}

      (84)

      where

      \begin{aligned} \alpha\equiv\frac{\gamma^2}{8\pi}-\varepsilon-3\gamma\eta+2\xi,\,\,\beta\equiv\frac{2\gamma^2}{8\pi}-\frac{\varepsilon}{2}-3\gamma\eta +\xi>0. \end{aligned}

      (85)

      Equations (47) and (48) provide the functions \Psi(\phi) and V(\phi) :

      \begin{aligned}[b] \Psi&=\dfrac{a^{-1}}{3\alpha^2}\left[\left(\dfrac{\alpha}{2}-\beta\right)(\widetilde{U}^2)'_a+\dfrac{6\beta \widetilde{U}^2}{a}\right]\\&\; {=\dfrac{{\rm e}^{-2\gamma\phi}}{3c_1^2\alpha^2}\left[\left(\dfrac{\alpha}{2}-\beta\right)\gamma^{-1} [\widetilde{U}^2(a(\phi))]'_\phi+6\beta \widetilde{U}^2(a(\phi))\right],} \end{aligned}

      (86)

      \begin{aligned}[b] V&=\dfrac{1}{6\alpha^2}\left(\beta-\dfrac{\alpha}{2}\right)\dfrac{(\widetilde{U}^2)'_a}{a^5}\\&\; {=\dfrac{\gamma^{-1}{\rm e}^{-6\gamma\phi}}{6\alpha^2 c_1^6}\left(\beta-\dfrac{\alpha}{2}\right)[\widetilde{U}^2(a(\phi))]'_\phi\,.} \end{aligned}

      (87)

      The Hubble parameters \sigma^2/H^2 take the form

      \begin{aligned} H=H_3=\frac{\gamma \widetilde{U}(a)}{\alpha a^3}, \end{aligned}

      (88)

      \begin{aligned} H_{1,2}=H\left[ 1 \pm \sqrt{\frac{8\pi}{3\gamma^2}\left( \alpha -2\beta + \frac{6\beta}{\widetilde{U}}\int\frac{{\rm d}a \cdot \widetilde{U}}{a} + \frac{9\alpha \gamma C_+}{\widetilde{U}} \right)}\,\right], \end{aligned}

      (89)

      \begin{aligned} \frac{\sigma^2}{H^2}=\frac{8\pi}{9\gamma^2}\left(\alpha-2\beta+ \frac{6\beta}{\widetilde{U}}\int\frac{{\rm d}a \cdot \widetilde{U}}{a}+\frac{9\alpha \gamma C_+}{\widetilde{U}}\right). \end{aligned}

      (90)

      Choosing a simple function

      \begin{aligned} U(a)=Aa^n, \,\, n>0 \end{aligned}

      (91)

      and C_+=C_\phi=0 , we obtain the Hubble parameter

      \begin{aligned} H=\frac{\gamma A}{\alpha a^{3-n}}. \end{aligned}

      (92)

      Moreover, the condition

      \begin{aligned} \frac{\gamma A}{\alpha}>0 \end{aligned}

      (93)

      ensures Universe expansion. Therefore,

      \begin{aligned} a=\left[\frac{\gamma A(3-n)}{\alpha}\cdot t\right]^{1/(3-n)}. \end{aligned}

      (94)

      If n<3 , then t\geq0 . In the case

      \begin{aligned} 2<n<3, \end{aligned}

      (95)

      the Universe is expanding under acceleration. If

      \begin{aligned} n>3, \end{aligned}

      (96)

      then t<0 and we obtain the Big Rip in time at t=0 :

      \begin{aligned} a=\left[\frac{\alpha}{\gamma A(n-3)}\right]^{1/(n-3)}\cdot\left(\frac{1}{-t}\right)^{1/(n-3)},\,\, a(0)=+\infty. \end{aligned}

      (97)

      The scalar field ϕ has the time dependence

      \begin{aligned} \phi=\frac{1}{\gamma}\ln\left\{\frac{1}{c_1} \left[\frac{\gamma A(3-n)}{\alpha}\cdot t\right]^{1/(3-n)}\right\} . \end{aligned}

      (98)

      Equations (86) and (87) provide the functions \Psi(\phi) and V(\phi) :

      \begin{aligned} \Psi={\rm e}^{2(n-1)\gamma\phi}\cdot\frac{c_1^{2(n-1)}A^2}{\alpha^2} \left[2\beta+\frac{n}{3}(\alpha-2\beta)\right], \end{aligned}

      (99)

      \begin{aligned} V={\rm e}^{(2n-6)\gamma\phi}\cdot \frac{nc_1^{2n-6}A^2\left(2\beta-\alpha\right)}{6\alpha^2}. \end{aligned}

      (100)

      The inequalities

      \begin{aligned} 2\beta+\frac{n}{3}(\alpha-2\beta)>0,\,\,2\beta-\alpha>0 \end{aligned}

      (101)

      lead to the constraints \Psi>0 and V>0 .

      The magnitude \sigma^2/H^2 takes the form

      \begin{aligned} \frac{\sigma^2}{H^2}= \frac{8\pi}{3n\gamma^2}\left[2\beta+\frac{n}{3}(\alpha-2\beta)\right]=\text{const}, \end{aligned}

      (102)

      where

      \begin{aligned} \frac{8\pi}{3n\gamma^2}\left[2\beta+\frac{n}{3}(\alpha-2\beta)\right]>0. \end{aligned}

      (103)

      The anisotropy is small, |\sigma/H|\ll 1 , if

      \begin{aligned} \left|\frac{8\pi}{3n\gamma^2}\left[2\beta+\frac{n}{3}(\alpha-2\beta)\right]\right|\ll 1. \end{aligned}

      (104)

      The Hubble parameters take the form

      \begin{aligned} H_{1,2}=H\left(1\pm \sqrt{\frac{8\pi}{n\gamma^2}\left[2\beta+\frac{n}{3}(\alpha-2\beta)\right]}\,\right)\,. \end{aligned}

      (105)

      The scale factors a_{i} are expressed through the scale factor a:

      \begin{aligned} a_{1,2}=s^{\pm}_{0}\cdot a^{1\pm \sqrt{\tfrac{8\pi}{n\gamma^2}\left[2\beta+\tfrac{n}{3}(\alpha-2\beta)\right]}\,},\,\, a_3=a. \end{aligned}

      (106)

      If condition (104) is satisfied, the Universe is close to the isotropic state:

      \begin{aligned} \frac{a_{1,2}}{a}=s^{\pm}_{0}\cdot a^{\pm \sqrt{\tfrac{8\pi}{n\gamma^2}\left[2\beta+\tfrac{n}{3}(\alpha-2\beta)\right]}\,}\approx \text{const}. \end{aligned}

      (107)

      If there is no Big Rip, the model is applicable to the era of primary inflation or late acceleration. If there is a Big Rip, the model is limited to late acceleration.

      Inequalities (85), (93), (101), (103), and (104) lead to the following system:

      \begin{aligned} \frac{2\gamma^2}{8\pi}-3\eta\gamma-\frac{\varepsilon}{2}+\xi>0, \end{aligned}

      (108)

      \begin{aligned} \frac{\gamma A}{{\gamma^2}/({8\pi})-3\gamma\eta-\varepsilon+2\xi}>0, \end{aligned}

      (109)

      \begin{aligned} \frac{\gamma^2}{8\pi}-\gamma\eta>0, \end{aligned}

      (110)

      \begin{aligned} \frac{\gamma^2}{8\pi}(4-n)+\eta\gamma(n-6)-\varepsilon+2\xi>0, \end{aligned}

      (111)

      \begin{aligned} \frac{8\pi}{n\gamma^2}\left[\frac{\gamma^2}{8\pi}(4-n)+\eta\gamma(n-6)-\varepsilon+2\xi\right]\ll 1. \end{aligned}

      (112)

      Next, we provide an example of the parameters for which this system of inequalities holds. We select the parameters

      \begin{aligned} \eta=0,~~ \varepsilon=1,~~ \xi=0,~~ n>2. \end{aligned}

      (113)

      Considering (111), (112), and (113), we set

      \begin{aligned} 4-n-\frac{8\pi}{\gamma^2}=\mu^2>0, \end{aligned}

      (114)

      where |\mu|\ll 1 . Other inequalities take the forms

      \begin{aligned} \frac{2\gamma^2}{8\pi}-\frac{1}{2}>0,~~\frac{\gamma A}{\dfrac{\gamma^2}{8\pi}-1}>0. \end{aligned}

      (115)

      Two cases are allowed:

      1. \gamma A<0 . Then, 1<\dfrac{8\pi}{\gamma^2}<4-n ; therefore, 2<n< 3-\mu^2 . The Universe is expanding with acceleration. The Big Rip model is excluded, n<3- .

      2. \gamma A>0 . Then, \dfrac{8\pi}{\gamma^2}<1 , \dfrac{8\pi}{\gamma^2}<4-n ; therefore, 3-\mu^2< n<4 . The Universe is expanding with acceleration. The Big Rip model is not excluded.

    V.   MODEL \boldsymbol\chi\neq \bf 0 , \boldsymbol G=\boldsymbol\xi\cdot\boldsymbol\phi
    • Here, we assume

      \begin{aligned} \chi\neq0,\,\, G=\xi\cdot\phi, \,\, C_\phi=C_+=0. \end{aligned}

      (116)

      Assuming

      \frac{\gamma^2}{8\pi}-\varepsilon-3\gamma\eta+2\xi=0,

      \gamma>0,\,\, \dot{\phi}>0,\,\,\chi<0,\,\,U>0,

      using (30) and (40), we obtain

      \begin{aligned} \dot{\phi}=\left(\frac{1}{3\gamma|\chi|}\cdot\frac{U}{a^3}\right)^{1/2}, \end{aligned}

      (117)

      \begin{aligned} H=\left(\frac{\gamma}{3|\chi|}\cdot\frac{U}{a^3}\right)^{1/2}>0, \end{aligned}

      (118)

      that is, we obtain a model of the expanding Universe. In the case

      \begin{aligned} \frac{2\gamma^2}{8\pi}-\frac{\varepsilon}{2}-3\gamma\eta+\xi=0, \end{aligned}

      (119)

      from Eqs. (45) and (46), it follows that

      \begin{aligned} S(a)=\frac{1}{(3\gamma|\chi|)^{1/2}}\cdot\left(\frac{U}{a^3}\right)^{3/2}, \end{aligned}

      (120)

      \begin{aligned} N(a)=\left(\frac{1}{3\gamma|\chi|}\cdot\frac{U}{a^3}\right)^{1/2}\frac{U'_a}{a^3}, \end{aligned}

      (121)

      then

      \begin{aligned} \Psi=\frac{a^5}{3}\cdot\left(\frac{1}{3\gamma|\chi|}\cdot\frac{U}{a^3}\right)^{1/2}\cdot \left[\frac{U'_a}{a^3}-\frac{3}{2}\left(\frac{U}{a^3}\right)'_a\right], \end{aligned}

      (122)

      \begin{aligned} V=\left(\frac{1}{3\gamma|\chi|}\cdot\frac{U}{a^3}\right)^{1/2} \cdot\left[\frac{U}{a^3}-\frac{a}{6}\left(\frac{U'_a}{a^3}-\frac{3}{2}\left(\frac{U}{a^3}\right)'_a\right)\right]. \end{aligned}

      (123)

      Choosing the function

      \begin{aligned} U=A^2a^{-2n+3},\,\, A>0, \end{aligned}

      (124)

      we obtain the Hubble parameter

      \begin{aligned} H=\left(\frac{\gamma }{3|\chi|}\right)^{1/2}\cdot \frac{A}{a^n}. \end{aligned}

      (125)

      Therefore,

      \begin{aligned} a=\left(\sqrt{\frac{\gamma}{3|\chi|}}\cdot nA\cdot t\right)^{1/n}. \end{aligned}

      (126)

      If n>0 , then t\geq0 . In the case 0<n<1 , the Universe is expanding under acceleration. If n<0 , then t<0 and we achieve the Big Rip in time at t=0 :

      \begin{aligned} a=\left[\sqrt{\frac{\gamma}{3|\chi|}}\cdot |n|A\right]^{-1/|n|}\cdot\left(\frac{1}{-t}\right)^{1/|n|},\,\, a(0)=+\infty. \end{aligned}

      (127)

      The scalar field ϕ has the time dependence

      \begin{aligned} \phi=\frac{1}{\gamma}\ln\left\{\frac{1}{c_1} \left(\sqrt{\frac{\gamma}{3|\chi|}}\cdot nA\cdot t\right)^{1/n}\right\} . \end{aligned}

      (128)

      Equations (34), (122), and (123) provide the functions \Psi(\phi) and V(\phi) :

      \begin{aligned} V=\frac{(3-n)A^{3}}{6(3\gamma|\chi|)^{1/2}a^{3n}}=\frac{(3-n)A^{3}}{6(3\gamma|\chi|)^{1/2}c_1^{3n}}\cdot {\rm e}^{-3n\gamma\phi}, \end{aligned}

      (129)

      \begin{aligned} \Psi=\frac{\left(3+n\right)A^{3}a^{4-3n}}{3(3\gamma|\chi|)^{1/2}}= \frac{\left(3+n\right)A^{3}c_1^{4-3n}}{3(3\gamma|\chi|)^{1/2}}\cdot {\rm e}^{(4-3n)\gamma\phi}. \end{aligned}

      (130)

      The inequalities

      \begin{aligned} -3<n<3 \end{aligned}

      (131)

      lead to the constraints \Psi>0 and V>0 .

      The Hubble parameters take the form

      \begin{aligned} H_{1,2}=H\left[1\pm\sqrt{\frac{8\pi |\chi|^{1/2}A(3+n)}{3^{1/2}\gamma^{3/2}(3-2n)}}\cdot\frac{1}{a^{n/2}}\,\right], \end{aligned}

      (132)

      where a new requirement arises, n<3/2 . The scale factors a_{i} are expressed through the scale factor a:

      \begin{aligned} a_{1,2}=s_{1,2}\cdot a\cdot\exp\left[\mp\frac{2}{na^{n/2}} \cdot\sqrt{\frac{8\pi |\chi|^{1/2}A(3+n)}{3^{1/2}\gamma^{3/2}(3-2n)}}\,\right]. \end{aligned}

      (133)

      The model isotropization condition is satisfied for n>0 :

      \begin{aligned}[b] \frac{a_{1,2}}{a}=&s_{1,2}\cdot\exp\left[\mp\frac{2}{na^{n/2}} \cdot\sqrt{\frac{8\pi |\chi|^{1/2}A(3+n)}{3^{1/2}\gamma^{3/2}(3-2n)}}\,\right]\\&\rightarrow \text{const} \,\,\, \text{as} \,\,\, a \rightarrow +\infty. \end{aligned}

      (134)

      Because n>0 , the Big Rip is excluded. The anisotropic properties of model (133) are similar to those of model (69). The function a_2 falls from a greater value at the beginning, bounces to the minimum value

      \begin{aligned} a_{2{\rm min}}=\left(\frac{8\pi |\chi|^{1/2}A(3+n)}{3^{1/2}\gamma^{3/2}(3-2n)}\right)^{1/n} \end{aligned}

      (135)

      and then rises again at the end. At the beginning, t=0 , the Universe is an infinite straight line along the x_2 axis: a_1(0)=a_3(0)=0 , a_2(0) =+\infty . The model can be applied to the early Universe before the end of primary inflation.

    VI.   CONCLUSION
    • We construct anisotropic models in BI for a subclass of HG:

      \begin{aligned}[b] G_4&=1/(16\pi),\,\,G_2=\varepsilon X-V(\phi),\\ G_5&=-\dfrac{1}{32\pi \gamma X},\,\, G_3=\dfrac{\eta}{2}\ln\dfrac{X}{C}+\chi\sqrt{2X}+G(\phi). \end{aligned}

      (136)

      with the non-minimal interaction by the law f^2(\phi)F_{\mu\nu}F^{\mu\nu} .

      Using the reconstruction method, we present functions \Psi(\phi)=q_m^2f^2(\phi) and V(\phi) , for which the isotropization criterion \lim\limits_{a\rightarrow \infty} a_i/a=\text{const} is satisfied.

      The function G_5\sim 1/X results in the possibility of an anisotropic bounce. At the beginning, t=0 , the Universe is an infinite straight line along the x_2 axis: a_1(0)= a_3(0)=0, a_2(0) =+\infty .

      Combinations of the parameters \varepsilon , \gamma , \eta , \chi , and G(\phi) allow for different possibilities for Universe development: power-law inflation, exponential inflation, the Big Rip, and two-phase expansion. In all cases, the anisotropy behavior is acceptable.

      In conclusion, we find a method of obtaining exact solutions for a large subclass of HG with an electromagnetic field.

Reference (60)

目录

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return