Hydrodynamical models for semiconductors

Monte Carlo and hydrodynamic simulationof a one dimensional n+ ‐ n ‐ n+ silicon diode

“…The AP model differs from the BW because: i) more moment equations are taken into account and a rigorous closure is obtained; ii) there are no free parameters; iii) we compare our results with the same parabolic MC simulations. The simulation shows that [7], at the drain junction, the heat flow evaluated with our HM overestimates the MC one and that, as consequence, the hydrodynamic energy is lower than the corresponding quantity evaluated with the MC model and the velocity obtained by the HM exhibits the wellknown "spurious" peak We ascribe this discrepancy to the limitations of the Maxwellian iterative approximation. This can be overcome by: a) including viscous effects; b) employing a flux-limiter heat equation which can be obtained rigorously from the full system [8].…”

“…This can be overcome by: a) including viscous effects; b) employing a flux-limiter heat equation which can be obtained rigorously from the full system [8]. In this article we investigate the first route by extending the previous system to incorporate Navier-Stokes like viscous effects nktT ( lg uj 2 Uk sij ) (7) O<ij> m-----;-'co xi -)xi 3…”