@article{GoubergritsOsmanMevertetal.2016, author = {Goubergrits, Leonid and Osman, Jan and Mevert, Ricardo and Kertzscher, Ulrich and P{\"o}thkow, Kai and Hege, Hans-Christian}, title = {Turbulence in blood damage modeling}, volume = {39}, journal = {The International Journal of Artificial Organs}, number = {4}, doi = {10.5301/ijao.5000476}, pages = {147 -- 210}, year = {2016}, abstract = {Purpose: To account for the impact of turbulence in blood damage modeling, a novel approach based on the generation of instantaneous flow fields from RANS simulations is proposed. Methods: Turbulent flow in a bileaflet mechanical heart valve was simulated using RANS-based (SST k-ω) flow solver using FLUENT 14.5. The calculated Reynolds shear stress (RSS) field is transformed into a set of divergence-free random vector fields representing turbulent velocity fluctuations using procedural noise functions. To consider the random path of the blood cells, instantaneous flow fields were computed for each time step by summation of RSS-based divergence-free random and mean velocity fields. Using those instantaneous flow fields, instantaneous pathlines and corresponding point-wise instantaneous shear stresses were calculated. For a comparison, averaged pathlines based on mean velocity field and respective viscous shear stresses together with RSS values were calculated. Finally, the blood damage index (hemolysis) was integrated along the averaged and instantaneous pathlines using a power law approach and then compared. Results: Using RSS in blood damage modeling without a correction factor overestimates damaging stress and thus the blood damage (hemolysis). Blood damage histograms based on both presented approaches differ. Conclusions: A novel approach to calculate blood damage without using RSS as a damaging parameter is established. The results of our numerical experiment support the hypothesis that the use of RSS as a damaging parameter should be avoided.}, language = {en} } @misc{SagnolHegeWeiser2016, author = {Sagnol, Guillaume and Hege, Hans-Christian and Weiser, Martin}, title = {Using sparse kernels to design computer experiments with tunable precision}, issn = {1438-0064}, url = {http://nbn-resolving.de/urn:nbn:de:0297-zib-59605}, year = {2016}, abstract = {Statistical methods to design computer experiments usually rely on a Gaussian process (GP) surrogate model, and typically aim at selecting design points (combinations of algorithmic and model parameters) that minimize the average prediction variance, or maximize the prediction accuracy for the hyperparameters of the GP surrogate. In many applications, experiments have a tunable precision, in the sense that one software parameter controls the tradeoff between accuracy and computing time (e.g., mesh size in FEM simulations or number of Monte-Carlo samples). We formulate the problem of allocating a budget of computing time over a finite set of candidate points for the goals mentioned above. This is a continuous optimization problem, which is moreover convex whenever the tradeoff function accuracy vs. computing time is concave. On the other hand, using non-concave weight functions can help to identify sparse designs. In addition, using sparse kernel approximations drastically reduce the cost per iteration of the multiplicative weights updates that can be used to solve this problem.}, language = {en} }