MPE-lab

Computational science and engineering

Bertil Nilsson

At present three research projects are pursued:

New models for cavitation has been developed (numerical as well as analytical) and implemented in different finite element models for the flow equations of Reynolds and of Stokes, together with a stable method for the coupling of these models across an artificial interface. For these models we have also developed new adaptive finite element methods, aiming at the automatic control of discretization errors through local mesh refinement. In particular, these adaptive procedures are goal-oriented, i.e., the aim of the computation (computation of pointwise errors in pressure or velocity, error in lift, etc.) is taken into account in the adaptive process.

Extended abstract for "Tribology" (pdf, 92 kB)

A new numerical model for contact mechanics with friction and wear, based on the discontinuous Galerkin method, has been developed. In this approach, continuity in the displacement field across the contact zone is taken into account in a weak sense. The approach allows for more sophisticated elasto-plastic models of interface friction than those based only on shear stresses (which is the traditional approach). We have also developed an alternative model using stabilized Lagrange multipliers to enforce the non-penetration condition in the contact zone. The stabilization approach allows for arbitrary multipliers, which can thus be defined on a topologically rectangular mesh for a numerically efficient implementation. We aim to proceed with lubricated contact using our models derived for tribology.

Extended abstract for "Contact mechanics" (pdf, 120 kB)

We are here studying skeletal muscle models aiming at a deeper understanding of their internal mechanics. The goal is to be able to use computer simulations for predictions of the effects of training as well as surgical procedures. Focus has been put on improving on the basic continuum models derived from experiments in the 1930´s that are still predominant in simulations today.

Extended abstract for "Muscle modelling" (pdf, 220 kB)