Abstract
We study the spherical collapse of an over-density of a barotropic fluid with linear equation of state in a cosmological background. Fully relativistic simulations are performed by using the Baumgarte–Shibata–Shapiro–Nakamura formalism jointly with the Valencia formulation of the hydrodynamics. This permits us to test the universality of the critical collapse with respect to the matter type by considering the constant equation of state parameter \(\omega \) as a control parameter. We exhibit, for a fixed radial profile of the energy-density contrast, the existence of a critical value \(\omega ^*\) for the equation of state parameter under which the fluctuation collapses to a black hole and above which it is diluting. It is shown numerically that the mass of the formed black hole, for subcritical solutions, obeys a scaling law \(M\propto |\omega - \omega ^*|^\gamma \) with a critical exponent \(\gamma \) independent on the matter type, revealing the universality. This universal scaling law is shown to be verified in the empty Minkoswki and de Sitter space-times. For the full matter Einstein-de Sitter universe, the universality is not observed if conformally flat sub-horizon initial conditions are used. But when super-horizon initial conditions computed from the long-wavelength approximation are used, the universality appears to be true.
Similar content being viewed by others
Notes
We point out here an apparent little typo in [31] where the terms with indices \(\theta \) are missing.
Note that this relation is correct whatever the equation of state and does not requires spherical symmetry.
The terms of \(R^r_r\) and R which are proportional to \(\hat{\Delta }^r\) and \(\partial _r \hat{\Delta }^r\) are in fact included in the \(L_{3(A_a)}\) operator instead of \(L_{2(A_a)}\).
References
Darmois, G.: Les équations de la Gravitation Einsteinienne. Gauthier-Villars (1927)
Lichnerowicz, A.: Problèmes Globaux en Mécanique Relativiste. Hermann et Cie (1939)
Lichnerowicz, A.: J. Math. Pures Appl. 23, 37 (1944)
Lichnerowicz, A.: Bulletin de la Société Mathématique de France 80, 237 (1952)
Fourès-Bruhat, Y.: J. Ration. Mech. Anal. 5, 951 (1956)
Dirac, P.A.M.: Proc. R. Soc. Lond. Ser. A. Math. Phys. Sci. 246, 333 (1958). https://doi.org/10.1098/rspa.1958.0142
Dirac, P.A.M.: Phys. Rev. 114, 924 (1959)
Arnowitt, R., Deser, S.: C. W. Misner. In: Witten, L. (ed.) Gravitation: An Introduction to Current Research (Chap. 7), pp. 227–265. John WIler and Sons Inc, New York, London (1962)
Nakamura, T., Oohara, K., Kojima, Y.: Prog. Theor. Phys. Suppl. 90, 1 (1987)
Nakamura, T.: In: Sasaki, T. M. (Ed.) 8th Nishinomiya–Yukawa Memorial Symposium: Relativistic Cosmology. Universal Academy Press (1994)
Shibata, M., Nakamura, T.: Phys. Rev. D 52, 5428 (1995)
Baumgarte, T.W., Shapiro, S.L.: Phys. Rev. D 59, 1 (1998). https://doi.org/10.1103/PhysRevD.59.024007
Shibata, M., Sasaki, M.: Phys. Rev. D 60, 1 (1999). https://doi.org/10.1103/PhysRevD.60.084002
Shibata, M., Uryu, K.: Progr. Theor. Phys. 107, 265 (2002). https://doi.org/10.1143/PTP.107.265
Pretorius, F.: Phys. Rev. Lett. 95, 1 (2005)
Abbott, B.P., et al.: Phys. Rev. Lett. 116, 1 (2016)
Choptuik, M.W.: Phys. Rev. Lett. 70, 9 (1993)
Gundlach, C., Martín-García, J.M.: Living Rev. Rel. 10, 1 (2007). https://doi.org/10.12942/lrr-2007-5
Evans, C.R., Coleman, J.S.: Phys. Rev. Lett. 72, 1782 (1994)
Maison, D.: Phys. Lett. B 366, 82 (1996). http://www.sciencedirect.com/science/article/pii/0370269395013814
Neilsen, D.W., Choptuik, M.W.: Class. Quantum Grav. 17, 761 (2000). https://doi.org/10.1088/0264-9381/17/4/303
Niemeyer, J.C., Jedamzik, K.: Phys. Rev. D 59, 1 (1999)
Hawke, I., Stewart, J.M.: Class. Quantum Grav. 19, 3687 (2002). https://doi.org/10.1088/0264-9381/19/14/310
Rekier, J., Cordero-Carrion, I., Fuzfa, A.: Phys. Rev. D 91, 024025 (2015). https://arxiv.org/pdf/1409.3476.pdf
Rekier, J., Fuzfa, A., Cordero-Carrion, I.: Phys. Rev. D 93, 043533 (2016). https://arxiv.org/pdf/1509.08354.pdf
Banyuls, F., Font, J.A., Ibanez, J.M., Marti, J.M., Miralles, J.A.: Astrophys. J. 476, 221 (1997). https://doi.org/10.1086/303604
Baumgarte, T.W., Shapiro, S.L.: Numerical Relativity: Solving Einstein’s Equations on the Computer. Cambridge Univ. Press, Cambridge (2010)
Shibata, M.: Numerical Relativity. World Scientific, Singapore (2016)
Noble, S.C., Choptuik, M.W.: Phys. Rev. D 93, 1 (2016)
Cordero-Carrion, I., Cerda-Duran, P.: (2016). https://arxiv.org/abs/1211.5930
Montero, P.J., Cordero-Carrion, I.: Phys. Rev. D 85, 1 (2012)
Alcubierre, M., Mendez, M.D.: Gen. Relat. Gravit. 43, 2769 (2011). http://doi.org/10.1007/s10714-011-1202-x
Alcubierre, M.: Introduction to 3+1 Numerical Relativity. Oxford University Press (OUP), New York (2008)
Arbona, A., Bona, C.: Comput. Phys. Commun. 118, 229 (1999)
Alcubierre, M., González, J.A.: Comput. Phys. Commun. 167, 76 (2005)
Ruiz, M., Alcubierre, M., Núñez, D.: Gen. Rel. Grav. 40, 159 (2008). https://doi.org/10.1007/s10714-007-0522-3
Rezzolla, L., Zanotti, O.: Relativistic Hydrodynamics. Oxford University Press (OUP), Oxford (2013)
Montero, P.J., Baumgarte, T.W., Müller, E.: Phys. Rev. D 89, 1 (2014)
Baumgarte, T.W., Montero, P.J., Cordero-Carrion, I., Müller, E.: Phys. Rev. D 87, 1 (2013)
Leer, B.V.: J. Comput. Phys. 23, 276 (1977)
Harten, A., Lax, P.D., Leer, B.V.: SIAM Rev. 25, 35 (1983)
Einfeldt, B.: SIAM J. Numer. Anal. 25, 294 (1988). https://doi.org/10.1137/0725021
Alcubierre, M., Brügmann, B., Diener, P., Koppitz, M., Pollney, D., Seidel, E., Takahashi, R.: Phys. Rev. D 67, 1 (2003). https://doi.org/10.1103/PhysRevD.67.084023
Gourgoulhon, E.: 3+1 Formalism in General Relativity?: Bases of Numerical Relativity. Springer, Heidelberg (2012)
Bona, C., Massó, J., Seidel, E., Stela, J.: Phys. Rev. Lett. 75, 600 (1995)
Torres, J.M., Alcubierre, M., Diez-Tejedor, A., Nuñez, D.: Phys. Rev. D 90, 1 (2014)
Harada, T., Yoo, C.-M., Nakama, T., Koga, Y.: Phys. Rev. D 91, 1 (2015)
Kodama, H.: Prog. Theor. Phys. 63, 1217 (1980)
Musco, I.: Phys. Rev. D 100, 1 (2019)
D. H. Lyth, K. A. Malik, and M. Sasaki (2005) J. Cosmol. Astropart. Phys. 2005, 4. https://doi.org/10.1088/1475-7516/2005/05/004
Padmanabhan, T.: Structure Formation In The Universe. Cambridge Univ. Press, Cambridge (1993)
Musco, I., Miller, J.C., Polnarev, A.G.: Classical and Quantum Gravity 26, 235001 (2009). https://doi.org/10.1088/0264-9381/26/23/235001
Musco, I., Miller, J.C.: Classical and Quantum Gravity 30, 145009 (2013). https://doi.org/10.1088/0264-9381/30/14/145009
Acknowledgements
The authors warmly thanks Dr. I. Cordero-Carrión for helpful advices and discussions concerning the Valencia formulation and the numerical methods used in this work. F. S. is supported by a FRS-FNRS (Belgian Funds for Scientific Research) Research Fellowship. J.R. is supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Advanced Grant agreement No. 670874). This research used resources of the “Plateforme Technologique de Calcul Intensif (PTCI)” (http://www.ptci.unamur.be) located at the University of Namur, Belgium, which is supported by the F.R.S.-FNRS under the convention No. 2.5020.11. The PTCI is member of the “Consortium des Équipements de Calcul Intensif (CÉCI)” (http://www.ceci-hpc.be).
Funding
Funding was provided by Fonds De La Recherche Scientifique - FNRS (Grant No. FC 17573) and European Research Council (Grant No. 670874)
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A: Recovering the primitive variables from the conserved ones
The conserved variables are defined in such a way:
where the rest-mass density \(\rho \), the energy density e, the pressure p, the velocity \(v_i\) and the Lorentz factor \(W = \frac{1}{\sqrt{1-v^2}}\) are the primitive variables. In general, the inversion of this relation is not analytical and requires a numerical root-finding procedure (see [37] for more details).
In the case of the barotropic equation of state \(p = \omega e\), this inversion can be made analytically. To obtain it, we start by squaring the second equation of (97):
Recalling that \(v^2 = 1-\frac{1}{W^2}\), (98) becomes
The third equation of (97) gives
Reinserting in (99) gives the relationFootnote 2
Using now the equation of state \(p=\omega e\), we obtain a quadratic equation in e:
Its solutions are
and
and
The sign we have to consider depends on the value of \(\omega \).
If \(0 < \omega \le 1\), we have \((\omega - 1)(\tau +D) \le 0\) and thus we have to take the plus to keep a non negative energy density.
If \(\omega > 1\), the positivity of the interior of the root and the fact that
imply that
where the equality holds if and only if \(e = 0\). The latter case is trivial,because it requires \(D = 0\), \(S_i = 0\) and \(\tau = 0\), and in the other cases, this means that the solutions \(e_+\) and \(e_-\) have opposite signs and that only one is positive.
If \(\omega < 0\), the situation is less obvious because both solutions can be positive. For example, if \(\omega \in ]-1;0[\), we have
because all the terms of the numerator are positive. This means that both solutions have the same sign and thus are positive. The choice must be done thanks to the continuity of the solution with time, which can be difficult numerically.
Ones the energy density e is computed, the other variables follow easily:
Appendix B: Splitting for the source terms in the PIRK operators
The evolution equations are written in the form (40) because we are using the PIRK algorithm. The splitting has been chosen to ensure the scheme to be as stable as possible (see [24]). In the first step, the hydrodynamical conserved variables, the cosmological scale factor a, the lapse \(\alpha \), the elements of the conformal 3-metric \(\hat{a}\) and \(\hat{b}\) and \(\psi \) are evolved explicitly. These are thus included in the \(L_1\) operator. In the second step, the extrinsic curvature is evolved. This means that K and \(A_a\) are split into the following \(L_2\) and \(L_3\) operatorsFootnote 3:
Finally, the auxiliary variable \(\hat{\Delta }^r\) is evolved partially implicitly:
Note that general expressions can be found in [31].
Appendix C: Kodama mass and mean energy-density contrast in BSSN variables
The Kodama mass was first defined in [48] but we take [13] and [47] as references.
Recall that, in spherical symmetry, the areal radius is the positive quantity R(t, r) defined by the area A(t, r) of the surface defined by constant t and r coordinates in such a way:
In our BSSN metric (2), the areal radius is simply the square root of its \(\theta \theta \) component:
Consider now the 2-metric \(G_{AB} = \begin{pmatrix} g_{tt} &{} g_{tr}\\ g_{rt} &{} g_{rr} \end{pmatrix}\) with \(A,B\in {t,r}\). We define the Kodama vector by
where \(\epsilon _{AB} = \sqrt{-G}\varepsilon _{AB}\) with \(\varepsilon _{AB}\) being the Levi-Civita symbol and \(\epsilon ^{AB} = G^{AC}G^{BD}\epsilon _{CD}\). Working in the zero shift gauge gives
The tensor \(K^A\) is extended to a 4-vector \(K^\mu \) by posing \(K^\theta = K^\phi = 0\). The quantity \(S^\mu = T^\mu _\nu K^\nu \) is thus a conserved current (see [48] and [47] for explanations) and its integral, the Kodama mass, is a conserved quantity. The Kodama mass within a sphere of radius r at time t is thus defined by
By developing \(S^t\), we find
The expression (9) gives, in the case of a universe filled with one fluid of matter (other cases do not change much things),
In terms of BSSN variables, we have
where we have use (11) and (121314) for the last equality. In conclusion, the expression for the Kodama mass in BSSN variables is
The corresponding quantity for the Friedmann universe used as background is thus
Note that in the definition of the compaction function (55) we need to compute it at the same areal radius than the local Kodama mass, that is to say
where we made the change of variable \(x = \psi (t,y)^2\sqrt{\hat{b}(t,y)}y\) for the last equality. The last expression, though less simple, can be useful because of its similarity with the first term of (128). For example, in a comoving gauge it gives the relation
only in term of the energy-density contrast \(\delta = \frac{e-\overline{e}}{\overline{e}}\).
Concerning the mean energy-density contrast, recall that it is defined in the following way:
By using BSSN equations, it gives
where the denominator can be replaced by \(\frac{R^3}{3} = \frac{\psi ^6a^3\sqrt{\frac{\hat{b}}{\hat{a}}}r^3}{3}\). We thus see the direct relation between the mean energy-density contrast and the compaction in the comoving gauge by looking at the relation (130).
Appendix D: Derivation of the long-wavelength solution in the BSSN variables
We assume that, at fixed time, the universe becomes locally flat homogeneous and isotropic in the limit \(\epsilon \rightarrow 0\). We thus assume, still following [50], that \(\hat{\gamma }_{ij} = O(\epsilon ^2)\). As described in details in [47], we then obtain
where we used the following notations
Moreover, the conformal factor can be decomposed in such a way:
where \(\Psi = O(\epsilon ^0)\) and \(\xi = O(\epsilon ^2)\). In fact, as also shown in [50] and [47], all slicings coincide up to \(O(\epsilon )\), which allows to apply the long-wavelength scheme in any slicing. After having done it in the special case where \(p = \omega e\), the evolution equations become
where \(\overline{D}\) and \(\overline{\Delta }\) are the covariant derivative and the Laplacian operators related to the flat metric written in the coordinates \(\lbrace x^i\rbrace \). In spherical coordinates, we have that this metric is given by \(\overline{\hat{\gamma }}_{ij} = \text {diag}(1,r^2,r^2\sin ^2\theta )\). The resulting equations have been solved in [47] for the constant mean curvature (CMC) slicing (for which \(K = \overline{K}\)), the comoving slicing, the uniform-density slicing (for which \(\delta = 0\)) and the geodesic slicing. The paper also gives comparison of the solutions between these four slicings. In our case, we will solve them for a general Bona-Masso slicing (5354). This solution has, to our knowledge, never been derived in the literature.
We start with the computation of the variables \(\hat{A}_{ij}\) and \(h_{ij}\) which, as explained in [47], do not depend on the slicing to \(O(\epsilon ^2)\). If we use the intermediate variable
the explicit expressions of \(\hat{A}_{ij}\) and \(h_{ij}\) are given, solving (149) and (150), by
If we want to use spherical symmetry and the BSSN variables presented in Sect. 2, we must use the quantities
With these definitions, we can write
where we have \(p_a+2p_b = 0\), implying the desired conditions \(A_a + 2A_b = 0\) and \(\hat{a}\hat{b}^2 = 1 + O(\epsilon ^4)\). The variable \(\hat{\Delta }^r\) can be determined thanks to equation (5).
We will now determine the expressions for \(\delta \), \(\xi \), \(\kappa \) and \(\chi \) by solving equations (144), (145), (147), (148) and (5354). The combination of the first two ones gives
The Bona-Masso equation (5354) reads, after being expanded around \(\alpha = 1\),
With equation (148) and the change of variable \(s = \ln a\), we obtain the differential linear system
where the matrix \(M(\omega )\) is given by
This matrix admits the following eigenvalues:
The general solution of the system (159) is thus given by
where \(\mathbf {v}_*\), \(\mathbf {v}_+\) and \(\mathbf {v}_-\) are the eigenvectors associated respectively to \(\lambda _*\), \(\lambda _+\) and \(\lambda _-\) and \(C_*\), \(C_+\) and \(C_-\) are integration constants.
For the geodesic slicing, \(\lambda _+ = \displaystyle -\frac{3(1+\omega )}{4} <0\) and \(\lambda _- = 0\). In the other Bona–Masso slicings of Sect. 3.1.3, \(\lambda _+\) and \(\lambda _-\) are complex conjugates with a negative real part. Since we only take pure growing modes, we set \(C_+ = C_- = 0\). We have that the eigenvector of \(\lambda _*\) is given by
We can then deduce the two useful relations
By inserting (165) in (147), we finally obtain
where we have defined
We immediately deduce the values of \(\kappa \), \(\chi \) and \(\xi \) thanks to (165), (166) and (145):
where the integration constant C can be absorbed into \(\Psi \) (see [47] in a similar case).
The last equation that remains to solve is (146) to find the velocity \(v_i\). By using the expressions of \(\delta \) and \(\chi \), we find
where the integration constant C must be set to zero to keep only growing modes (see [47]). We deduce the value for \(v_i\):
To summarize the results in terms of BSSN variables in spherical symmetry, the long-wavelength solution is given by
Rights and permissions
About this article
Cite this article
Staelens, F., Rekier, J. & Füzfa, A. Universality of the spherical collapse with respect to the matter type: the case of a barotropic fluid with linear equation of state. Gen Relativ Gravit 53, 38 (2021). https://doi.org/10.1007/s10714-021-02804-4
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10714-021-02804-4