85
2.
Before examining the real galaxy interaction wemust recall that SE for MF,
∂ ∂
fmt mdmdm
(,)
=
∫
12
[
Uff
m
1212
δ
−
− −
cyclebicycle
] (1)where
( )
ffmt
≡
,
,
( )
ffmt
11
≡
,,
etc, are MFs,
( )
δ δ
m
mmm
≡ − −
12
,
δ
— Dirac’s delta-function,
( )
UUmm
1212
≡
,
, which describes the merging,permits two nontrivial (here we do not touch the case of
U = const
) exact solutions describing the MF evolutionfrom its initial state localized on small masses [8]. In thesecond and third terms we have
δ
(m
2
-m-m
1
)
and
δ
(m
1
-m
2
-m)
correspondingly.For the merging probability proportional to the productof colliding masses the power index of IA is s
1
=-5/2.Expressed through the uniformity power u of coagulation coefficient
( ) ( )
( )
UamaUm
u
=
it equals
s
1
= – (u+3)/2 (u=2).
As known (see also below) suchindex corresponds to the constant mass flux on thespectrum. In the case of
U = c(m
1
+m
2
)
the power of IAis
s
0
= – 3/2
or
s
0
= – (u+2)/2 (u=1).
The lattercorresponds to the constant flux of the number of massive objects. Though at first sight such aconservation law fails with the mergers it is realized inthe form of approximate integral in the case of predominant interaction of large objects with the smallones. Such a "nonlocal" situation fits the last solution.The condition of locality, i.e. that of convergence of thecollision integral in SE for s
1
- solution is in the form of
|u
2
- u
1
|<1
[9], where the indexes
u
1
and
u
2
are definedby the expression for
U
provided the masses differstrongly
Umm
uu
∝
12
12
( )
mm
12
<<
. Obviously, inthe first case
( )
uu
12
1
= =
the locality criterion isfulfilled and in the second the marginal case occurs (
u
1
=0, u
2
=1
). That is, the galaxy interaction with the mostdistinguished scales prevails in the latter case and, thus,the conservation law of the "number of particles" isrealized.
3.
For gravitational galaxy interaction the cross-sectionof coagulations is usually taken as product of co-factors,which describe, respectively, the geometrical cross-section, gravitational focusing and conditional mergingprobability at the frontal collision of galaxies (seereferences in 3]):
( ) ( )
σ π γ ϕ γ γ
= + ≡
rvv
g
222
1,,
vGmrmmmrrr
g
21212
2
= = + = +
,,
.The homogeneity index differs for "large" and "small"masses. Further we will focus on this particular region,regarding a small mass region as contracted to zero.This scheme can be attributed with a more accurateformal meaning. On the assumption that
ϕ
decreases assquare of relative velocity we can take it in the form of
( )
ϕ γ
= +
−
11
1
. The resulting cross-section will be auniform function in all of the mass-changing interval:
σ π γ
=
r
2
. By averaging over velocities we come tothe coagulation coefficient
Uv
=< >
σ
in the form
( )
( )
Ummmm
∝ + +
1212
β β
where the radius-massdependence is chosen as
rm
∝
β
. Below we employonly the fact that U is the uniform function of masseswith
uuuu
= + = =
10
12
β
,,
( )
β
= ÷
1312
For u>1, as known from general theory of SE
, theevolution of MF has an explosive character and a quasipower asymptotics is established in a wide mass intervalbetween the region of initial mass localization
mm
≈
*
and the coagulation front which is turned to the infinitymass for the finite time [11,12]. Our goal is to find thepower IA of the spatial homogeneous
2
solutions of theSE (1) with the considered kernel
U
discussed above.
( )
JmdmmI
st m
1
=−
∫
,
( )
∂ ∂ ∂ ∂
mf t Jmm
+ =
1
0
(2)Here the
I
st
is the right part of SE (1).
4
. Both the numerical solution of SE and modelling byMonte Carlo method show that the power index of IA
α
lies between
s
0
= – (
u+2)/2
and
s
1
= – (u+3)/2
(seefor example [10] Fig. 2b; [13]). In order to understandwhat it means consider the symmetry properties of thecollision integral of SE in the case of exact uniformity:
UaU
amamamummm
1212
=
.To utilize the similarity of
U
we must changesimultaneously the scale of all three arguments
mm
12
,
and m. But, as one of them (m) is fixed in SE, from thecontinuous group transformation only two discretetransformations remain (except a trivial one):
G
1
,transforming
mm
1
→
, and
G
2
, transforming
mm
2
→
. These Zakharov transformations are
1
In this case the initial distribution localized on small masseswithin a finite time forms a power "tail" spreading on theregion of formally infinite masses [8]. This the so-calledkinetic phase transition was first discovered and studied indetail by Stockmayer for the above-mentioned model with
U = cm
1
m
2
and was utilized for describing polymerization, inparticular, zol–gel transition (in addition to the above-mentioned see also the references in [10]). In the case of thegravitating systems we are interested in, the new phase whichemerges at the transition and corresponds to the "infinite"mass is juxtaposed with cD-galaxies in the center of thecluster.
2
This surely leads to the loss of a number of distributionfeatures, including spatial stratification of galaxy clusters, witha more compact central and less dense periphery, etc. At thesame time the chaotization in the systems considered confirmsthe made assumption.