Concepto procesal de tercero y de tercero legitimado

negligible near collapsed quasi-static systems, as is the pressure term. The function 1_{− F}′ is a dielectric-like scalar free function of the covariant scalar quantity_|E_|=_|∇Φ_|/a0. This

dielectric-like function is thus the analogue of the µ-function of MOND (Bekenstein & Milgrom 1984). To recover the MOND phenomenology, a specific choice of the function can be F′≡ _ddF |E_|2 = 1 +|E| n −n . (1.36)

Zhao (2007) prefers the parametern_∼3 to best explain the amplitude of the cosmological constant. Here we rather choosen= 1 corresponding to the ‘simple’µ-function of MOND (Famaey & Binney 2005, Zhao & Famaey 2006, Famaey et al. 2007b, Sanders & Noor- dermeer 2007). Note finally that, when a galaxy is embedded in an external field gext,

we have that thetotal gravitational force_−∇Φ_≡gext− ∇Φint, where Φint is the internal

potential of the galaxy.

1.2

External Field Effects

The CDM and MOND interpretations of the galaxy kinematics can sometimes be viewed as degenerate, at the price of accepting some mysterious conspiracy between the distribution of baryons and dark matter at all radii in the CDM context (see e.g. McGaugh 2005). There are, however, some fundamental differences, one of which is the role of dynamical friction (e.g. S´anchez-Salcedo et al. 2006, Nipoti et al. 2008). Another one is the fact that the Strong Equivalence Principle is violated in MOND when considering the external field in which a system is embedded (e.g. Bekenstein & Milgrom 1984, Zhao & Tian 2006, Famaey et al. 2007a, S´anchez-Salcedo et al. 2008), even if the external field is constant and uniform. Hence, it is also important to study the effects of the environment in MOND: unlike in CDM, the rotation curves and the morphology of galaxies depend on both the background and self-gravity acceleration (Wu et al. 2007). The escape velocity at any location in a galaxy is also dependent on this external field (Famaey et al. 2007a, Wu et al. 2007). In addition to the escape and circular speeds, the influence of external fields has been studied for a variety of different situations, including the motion of probes in the inner solar system (Milgrom 2009b), the Roche lobe of binary stars (Zhao & Tian 2006), the kinetics of stars in globular clusters (Milgrom 1983c, Perets et al. 2009), and satellites

Chapter 1. The MOdified Newtonian Dynamics

Figure 1.5: The gravity contour of the “bullet cluster” 1E0657-56, in units ofa0.

surrounding a host galaxy (Brada & Milgrom 2000, Tiret et al. 2007, Angus 2008). MOND-based theories generate different degrees of dark matter-like effects depending on the absolute acceleration. Most galaxies are, like the Milky Way, in the field, where they accelerate slowly with respect to the Cosmic Microwave Background, typically at a rate of 0.01a0 to 0.03a0 (Famaey et al. 2007a, Milgrom 2002, Angus & McGaugh 2008, Wu

et al. 2007). For Milky Way, it is impossible to precisely calculate the value of the external field at the location of the Milky Way without describing the Local Group formation in the context of a MOND N-body simulation. Nevertheless, one can estimate its order of magnitude from the acceleration endured during a Hubble time in order to attain a peculiar velocity of 600 km s−1with respect to the CMB,gext ≃H0×600 km s−1 ≃0.01a0.

The Milky Way is accelerated toward the Great Attractor, which is lying at _∼ 46 Mpc far away, in the direction of the Hydra Centaurus supercluster. This value is also roughly the one produced by the M31 galaxy, as well as the one produced by the Virgo and Coma clusters (Famaey et al. 2007a). But in X-ray clusters, galaxies accelerate much faster, from 0.3a0 to 3a0 (Pointecouteau et al. 2005, Wu et al. 2007). We show here the gravitational

acceleration contours inside the “bullet cluster” (Fig. 1.5) which represents the MONDian potential of Angus et al. (2007b).

This external gravitational field has wider and more subtle implications for the internal system in MOND than in Newton-Einstein gravity, for the very reason that MOND breaks

1.2. External Field Effects the Strong Equivalence Principle.

1.2.1 Binding energy of an accelerating Milky Way

Galaxies free-fall, but with slowly-changing systemic (center-of-mass) velocityvm(t). Their

present non-zero systemic velocity is mainly the accumulation of the acceleration by the gravity from neighbouring galaxies over a Hubble time. Consider, as a first approximation, that a galaxy is stationary in a non-inertial frame (in the Galilean sense), which free-falls with a “uniform” systemic acceleration ˙vm =gextXˆ =cstdue to an external linear poten-

tial, say, ₋gextX along the X-direction, where the over-dot means time-derivatives. Let

˙

vint= ( ¨X,Y ,¨ Z¨) be the peculiar acceleration of a star-like test particle in the coordinates

relative to the center of a non-evolving galaxy internal mass densityρ(X, Y, Z), then

˙

vint=g−v˙m =−∇Φint(X, Y, Z), (1.37)

whereg is the absolute acceleration g satisfying the MOND Poisson’s equation − ∇.[µ(x)g] = 4πGρ(X, Y, Z), x_≡ |g|

a0

. (1.38)

One can define an ”effective potential” Φint(X, Y, Z) (called ”internal” potential) and an

”effective energy” Eeff ≡ v

2 int

2 + Φint(X, Y, Z), where Eef f is conserved along the orbit of

the test particle effectively moving in a force field_−∇Φint(X, Y, Z), which is curl-free and

time-independent because the absolute gravity g is curl-free, center-of-mass acceleration ˙

vmis assumed a constant, and the galaxy densityρ(X, Y, Z) is assumed time-independent.

Far away from the center of the free-falling system, we have _|v˙int| ≪ |v˙m|, hence

µ_→ µm ≡µ( ˙vm/a0) = cst, and the equation 1.38 reads (BM1984; Milgrom1986; Zhao &

Tian 2005; Zhao & Famaey 2006):

∇2Φint+ ∆

∂2

∂X2Φint→4πGρ/µm, (1.39)

where Y, Z denote the directions perpendicular to the external field X-direction, and ∆ = [dlnµ/dlnx]_x=|v˙m|/a0 is a dilation factor (note that 0 _≤ ∆ _≤ 1). So at large radii where the external field dominates and the equation is linearizable, the potential satisfies

Chapter 1. The MOdified Newtonian Dynamics

a mildly anisotropic Poisson equation and the solution at large radii 2 goes to

Φ∞_int(X, Y, Z) =₋ GMint

µm

p

(1 + ∆)(Y2_+Z2_{) +X}2_+s2, (1.40)

where we included a softening radius s, comparable to the half-light radius of a galaxy. Hence the internal potential Φint is finite, and approaches zero at large radii.

The escape speed at any location rin the system can then be meaningfully defined by 0 =Eeff =

v2_esc(X, Y, Z)

2 + Φint(X, Y, Z). (1.41)

Such escape speed is a scalar independent of ”the path to escape” because the “effective energy”Eeff is conserved, and a particle withEeff equal zero (the maximum value of Φint)

will reach infinite distance from the system, and never return, hence will be lost into the MOND potential of the background (from which it cannot escape). However, equal escape speed contours across a disk galaxy are generally not axisymmetric, meaning that the escape speed differs on opposite symmetric locations of the Galaxy (Wu et al. 2007).

In summary, MOND potentials are logarithmic for isolated distributions of finite mass, and consequently infinitely deep, but that the internal potential becomes ”polarized” Keplerian with a dilation factor 1 + ∆ at large distances (see Eq. 1.40) when an external field is applied.