A local frame formulation of dual stress-strain pairs and time derivatives (Q1921990)

The justification for this complex study is to be found in a referenced work authored by \textit{P. Haupt} and \textit{Ch. Tsakmakis} (loc. cit.). If \(T^0\) and \(T'\) are respectfully the Jaumann and Oldroyd objective time rates of (Cauchy) stress tensor \(T\), a common form of constitutive equation is either \(T^0=f(T,D)\), or \(T'=f(T,D)\), where \(f(T,D)\) is an isotropic, tensor valued function of its arguments. Since terms bilinear in \(T\) and \(D=1/2(L+L^*)\), where \(L\) is the velocity gradient, appear in both sides of either equation, it has long been held that it is immaterial which time rate one chooses to place in the left hand side of these constitutive equations, because the difference between them may be allowed for in the values of the material constants. If one insists on retaining bilinear terms on the left hand sides, while retaining only terms linear in \(T\) and \(D\) on the right, as is the case in the objective differential form of the Maxwell model of linear viscoelasticity, there is a dilemma: the choice of the Jaumann derivative yields undesirable physical conclusions. Haupt and Tsakmakis produced arguments, based on variational considerations involving `stress powers', for their choice of dual rates of change of stress and strain tensor. Svendsen gives an alternative geometric formulation for these pairings based on Noll's concept of a body element (loc. cit.).