The scaling form of the free energy near a critical point allows for the definition of various thermodynamical amplitudes and the determination of their dependence on the microscopic nonuniversal scales. Universal quantities can be obtained by considering special combinations of the amplitudes. Together with the critical exponents they characterize the universality classes and may be useful quantities for their experimental identification. We compute the universal amplitude ratios for the tricritical Ising model in two dimensions by using several theoretical methods from perturbed conformal field theory and scattering integrable quantum field theory. The theoretical approaches are further supported and integrated by results coming from a numerical determination of the energy eigenvalues and eigenvectors of off-critical systems in an infinite cylinder.
We compute current correlators of the
CP${}^{N\ensuremath{-}1}$ field theory in 2+1 dimensions, both at the critical point and in the phase with spontaneously broken
SU(N) symmetry. Universal constants are obtained to next-to-leading order in the
1/N expansion. Implications are noted for quantum critical points of antiferromagnets and their vicinity.
Power laws are distinctive features of critical phenomena and can be ascribed to diverging length-scales. One of the most important contributions of the studies of critical phenomena is the shift in point of view to a length-scale based analysis and classifying terms of a Hamiltonian as relevant or irrelevant rather than classifying them as numerically weak or strong. The idea of universality transcends the domain of critical phenomena. Critical points are classified by the number of relevant variables required to describe them. A universality class would be described by the exponents and also the amplitude ratios of the various singulax quantities on both sides of the critical point. In a correlation based approach, we need to compute the critical correlations of all possible combinations of the degrees of freedom. Scaling is the rule for criticality, originating from a diverging length scale, while anomalous dimension is seen for fluctuation dominated cases.
Using modern estimates and exact values of critical points and exponents for Ising systems, existing series expansions for various lattices are analyzed to estimate the universal amplitude ratio
R0\ensuremath{\equiv}(
C4+)2/
C+C6+, where ${\mathit{C}}^{\ifmmode\pm\else\textpm\fi{}}$ are the amplitudes of divergence of the susceptibility \ensuremath{\chi}, above and below
Tc, while ${\mathit{C}}_{\mathit{k}}^{\ifmmode\pm\else\textpm\fi{}}$ are, similarly, the amplitudes for (${\mathrm{\ensuremath{\partial}}}^{\mathit{k}\mathrm{\ensuremath{-}}2}$\ensuremath{\chi}/\ensuremath{\partial}${\mathit{h}}^{\mathit{k}\mathrm{\ensuremath{-}}2}$
)h=0; we obtain
R0=0.1582\ifmmode\pm\else\textpm\fi{}0.0002 for d=2 dimensions and 0.1275\ifmmode\pm\else\textpm\fi{}0.0003 for d=3. Similarly, we find the universal ratio
C4+/${\mathit{C}}_{4}^{\mathrm{\ensuremath{-}}}$=-9.0\ifmmode\pm\else\textpm\fi{}0.3 and -121.5\ifmmode\pm\else\textpm\fi{}0.5 for d=3 and 2, respectively. On using existing estimates for
C+, for the second-moment correlation-length amplitude
f1+, and for the universal ratios
C+/${\mathit{C}}^{\mathrm{\ensuremath{-}}}$ and
f1+/${\mathit{f}}_{1}^{\mathrm{\ensuremath{-}}}$, we estimate the renormalized coupling constants to be
g+∗=
G1+\ensuremath{\equiv}-
C4+/(
C+)2(
f1+a
)d/
v0=24.45\ifmmode\pm\else\textpm\fi{}0. 15 for d=3 and 14.700\ifmmode\pm\else\textpm\fi{}0.017 for d=2. Beneath
Tc we find ${\mathit{G}}_{1}^{\mathrm{\ensuremath{-}}}$=-501\ifmmode\pm\else\textpm\fi{}60, -1768\ifmmode\pm\else\textpm\fi{}80 and -${\mathit{C}}_{3}^{\mathrm{\ensuremath{-}}}$B/(${\mathit{C}}^{\mathrm{\ensuremath{-}}}$
)2=6.
47\ifmmode\pm\else\textpm\fi{}0.2, 33.
06\ifmmode\pm\else\textpm\fi{}0.1, where B is the spontaneous-magnetization amplitude, and thence ${\mathit{g}}_{\mathrm{\ensuremath{-}}}^{\mathrm{*}}$\ensuremath{\simeq}85 and 23 for d=3 and 2, respectively. \textcopyright{} 1996 The American Physical Society.