https://hal-mines-paristech.archives-ouvertes.fr/hal-01809350Arora, ShitijShitijAroraCEMEF - Centre de Mise en Forme des Matériaux - Mines Paris - PSL (École nationale supérieure des mines de Paris) - PSL - Université Paris sciences et lettres - CNRS - Centre National de la Recherche ScientifiqueFourment, LionelLionelFourmentCEMEF - Centre de Mise en Forme des Matériaux - Mines Paris - PSL (École nationale supérieure des mines de Paris) - PSL - Université Paris sciences et lettres - CNRS - Centre National de la Recherche ScientifiqueA nodally condensed SUPG formulation for free-surface computation of steady-state flows constrained by unilateral contact - Application to rollingHAL CCSD2018[SPI.MAT] Engineering Sciences [physics]/MaterialsPrudon, MagalieBuffa G.,Fratini L.,Ingarao G.,Di Lorenzo R.2018-06-06 16:10:402023-03-24 14:53:072018-06-06 16:10:40enConference papers10.1063/1.50349291In the context of the simulation of industrial hot forming processes, the resultant time-dependent thermo-mechanical multi-field problem (v→,p,σ,ε) can be sped up by 10-50 times using the steady-state methods while compared to the conventional incremental methods. Though the steady-state techniques have been used in the past, but only on simple configurations and with structured meshes, and the modern-days problems are in the framework of complex configurations, unstructured meshes and parallel computing. These methods remove time dependency from the equations, but introduce an additional unknown into the problem: the steady-state shape. This steady-state shape x→ can be computed as a geometric correction t→ on the domain X→ by solving the weak form of the steady-state equation v→.n→(t→)=0 using a Streamline Upwind Petrov Galerkin (SUPG) formulation. There exists a strong coupling between the domain shape and the material flow, hence, a two-step fixed point iterative resolution algorithm was proposed that involves (1) the computation of flow field from the resolution of thermo-mechanical equations on a prescribed domain shape and (2) the computation of steady-state shape for an assumed velocity field. The contact equations are introduced in the penalty form both during the flow computation as well as during the free-surface correction. The fact that the contact description is inhomogeneous, i.e., it is defined in the nodal form in the former, and in the weighted residual form in the latter, is assumed to be critical to the convergence of certain problems. Thus, the notion of nodal collocation is invoked in the weak form of the surface correction equation to homogenize the contact coupling. The surface correction algorithm is tested on certain analytical test cases and the contact coupling is tested with some hot rolling problems.