Information Technology Reference
In-Depth Information
ε
cr
.
linear fits and avoid the problem ofthe uncertainty ofthe position ofthe
Fig. 3 shows the data and linear fits for all functions listed above.
Fig. 3.
Estimates for correlation-length exponents
ν
. Log-log plots of derivatives
∂
ε
log
|m|
L
,
∂
ε
log
m
L
,
∂
ε
log
m
L
(left scale) and log-plots of
V
2
,
V
4
and
V
6
are pre-
sented (right scale). The solid lines correspond to linear fits for L=40,60,80,100.
4.4 Estimates of Other Critical Exponents
Estimates for
β
and
γ
are found from relations (12). Again, we work with linear
approximations of
|m|
L
,
m
L
,
m
L
and
χ
L
for
ε
in critical regimes. We find that
values of
β/ν
and
γ/ν
are very close to the Ising model ones, i.e.,
β
Ising
=0
.
125,
γ
Ising
=1
.
75. Moreover, we could easily calculate
ε
cr
by chosing such a value
of
ε
for which the corresponding
β/ν
and
γ/ν
take values that are nearest to
the Ising ones. This way we can provide the following description of the critical
regimes in all considered systems:
