Barlat's anisotropic yield potential is based on linear transformations of
stress. The stress potential function (Yld2004-18p) proposed by Barlat et al.
(2005) is
where
is the exponent and
,
(
=1, 2, 3) are principal values of stress tensors
and
, respectively.
The principal values of a stress tensor
(
or )
are the roots of the characteristic equation
where the first, second, and third stress invariants are
The principal values are
where the deviatoric polar angle is
and the stress invarients are
The tensors
and
are defined by two linear transformations of the deviatoric
stress
,
where the tensors
and
are transformation tensors contains anisotropy coefficients
and the tensor
transforms the Caushy stress,
, to the deviatoric stress. The transformations can be
expressed in matrix form:
and
The 18 anisotropy coefficients
,
can be calibrated from experiments.
If
and ,
the yield function reduces to the Hosford isotropic yield function. In
addition, when
or ,
the Mises isotropic plasticity model is recovered.
The flow rule is
where, from the definition of
above,
Barlat Plasticity (Yld91)
When the Barlat plasticity considers only one linear transformation (that
is,
), the yield surface reduces to the Barlat Yld91 yield
surface (Barlat et al., 1991) as
where
is the exponent; and
,
, and
are principal values of the stress tensor
.
The linear transformation is defined as
or expressed in matrix form,
The six anisotropy coefficients
,
,
,
,
, and
can be calibrated from experiments. If all six coefficients
are set to unity, the yield function reduces to the Hosford isotropic yield
function. In addition, when
or
, the Mises isotropic plasticity model is recovered.