Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
DY: Fachverband Dynamik und Statistische Physik
DY 17: Modeling and Data Analysis
DY 17.6: Vortrag
Mittwoch, 2. April 2014, 11:00–11:15, ZEU 146
Kriging-supported Adaptive Sampling for Non-Oscillatory Central Schemes — •D. G. Roehm1, R. S. Pavel2, T. C. Germann3, and A. L. McPherson4 — 1ICP, Universität Stuttgart, Stuttgart, Germany — 2ECE, University of Delaware, Newark, USA — 3T1, LANL, Los Alamos, USA — 4CCS, LANL, Los Alamos, USA
We propose an enhanced adaptive sampling method for heterogeneous multi-scale simulations with stochastic data, based on a non-oscillatory high-resolution scheme for 2D hyperbolic conservation laws. This general framework is used to compute the evolution of a mechanical shockwave in a perfect copper crystal on the macro-scale by evaluating stress and energy fluxes on the micro-scale. A finite-volume method was used as the macro-scale solver, which launches for every volume element a light-weighted MD simulation (called CoMD) to incorporate details from the micro scale. Since the execution of an MD simulation is rather costly, we reduced the number of actual MD simulations through the use of an adaptive sampling scheme. Our adaptive scheme utilizes a key-value database for ordinary Kriging and a gradient analysis to reduce the number of finer-scale response functions. Kriging estimates an unknown value at a certain location by using weighted averages of the neighboring points. It also provides an error estimate, which we use as trigger for our adaptive scheme. In this contribution we will focus on how the accuracy of the physical values is affected by several thresholds in our adaptive scheme and their connection to the overall performance. The presented adaptive scheme allows for the future inclusion of details present in real materials. (LA-UR-13-29087)