Coupled solutions are implemented in a general way, using a stepped algorithm and data-transfer database. Variables can be exported to the transfer database as integration point data or node point data. These fields can then be imported into the sub problems as external parameters allowing a variety of influences on the calculation such as imposed geometry displacements in thermal calculations, or in coefficient dependancies.
Note that any number of sub-problems can be combined, and the import/export data is totally user defined. Also, the data transfer does not require that the meshes be the same between sub-problems.
the input for this problem uses a mechanical and thermal sub-problem,
and a number of exported variables. The material has variable coefficients,
and is in finite strain. An option is used to save the coefficient
values (such as initial yield R0 plotted at right), which can be
very useful in coupled analysis. We use the same course mesh here as
given in the reference.
****calcul weak_coupling ***resolution **sequence *time 4.0 8.0 *increment 10 10 *ratio 1.e-4 ***coupled_resolution **iteration 2 ***sub_problem fem MechTherm/plastic **transfer integ_nodeparam *variable q_dot *file MechTherm/heat_out **transfer node_kinematic *file MechTherm/kine_out ***sub_problem fem MechTherm/thermal **transfer node_nodeparam *variable TP *file MechTherm/temp_out ****return % % Material from Simo and Miehe % Comp. Meth Appl. Mech Eng 98 (1992) 41-104 % ***behavior gen_evp lagrange_polar **save_coefficients R0 H Q b **heat_generation f 0.9 **elasticity young 240581.0 poisson 0.499 **thermal_strain isotropic alpha 1.e-5 ref_temperature 0.0 **potential gen_evp ep *flow plasticity *isotropic linear_nonlinear R0 function 450.00*(1.0-2.e-3*temperature); H function 129.24*(1.0-2.e-3*temperature); Q function 265.00*(1.0-2.e-3*temperature); b 16.93 **coefficient masvol 1.e-3 ***return ***behavior thermal **conductivity isotropic k 45. % N/sK ***coefficient capacity 3.588 % m^2/s^2K ***return