TY - CHAP A1 - Birkenmaier, Clemens A1 - Krenkel, Lars ED - Chinesta, F. ED - Abgrall, R. ED - Allix, O. ED - Kalistke, M T1 - Convolutional Neural Networks for Approximation of Internal Non-Newtonian Multiphase Flow Fields T2 - 14th World Congress on Computational Mechanics (WCCM), ECCOMAS Congress 2020: 19–24 July 2020, Paris, France N2 - Neural networks (NNs) as an alternative method for universal approximation of differential equations have proven to be computationally efficient and still sufficiently accurate compared to established methods such as the finite volume method (FVM). Additionally, analysing weights and biases can give insights into the underlying physical laws. FVM and NNs are both based upon spacial discretisation. Since a Cartesian and equidistant grid is a raster graphics, image-to-image regression techniques can be used to predict phase velocity fields as well as particle and pressure distributions from simple mass flow boundary conditions. The impact of convolution layer depth and number of channels of a ConvolutionDeconvolution Regression Network (CDRN), on prediction performance of internal non-Newtownian multiphase flows is investigated. Parametric training data with 2055 sets is computed using FVM. To capture significant non-Newtownian effects of a particle-laden fluid (e.g. blood) flowing through small and non-straight channels, an Euler-Euler multiphase approach is used. The FVM results are normalized and mapped onto an equidistant grid as supervised learning target. The investigated NNs consist of n= {3, 5, 7} corresponding encoding/decoding blocks and different skip connections. Regardless of the convolution depth (i.e. number of blocks), the deepest spacial down-sampling via strided convolution is adjusted to result in a 1 × 1 × f · 2nfeature map, with f = {8, 16, 32}. The prediction performance expressed is as channel-averaged normalized root mean squared error (NRMSE). With a NRMSE of < 2 · 10-3, the best preforming NN has f = 32 initial feature maps, a kernel size of k = 4, n = 5 blocks and dense skip connections. Average inference time from this NN takes < 7 · 10-3s. Worst accuracy at NRMSE of approx 9 · 10-3is achieved without any skips, at k = 2, f = 16 and n = 3, but deployment takes only < 2 · 10-3s Given an adequate training, the prediction accuracy improves with convolution depth, where more features have higher impact on deeper NNs. Due to skip connections and batch normalisation, training is similarly efficient, regardless of the depth. This is further improved by blocks with dense connections, but at the price of a drastically larger model. Depending on geometrical complexity, spacial resolution is critical, as it increases the number of learnables and memory requirements massively. KW - Deep Learning KW - Convolutional neural networks KW - Non-Newtonian multiphase flow Y1 - 2021 U6 - https://doi.org/10.23967/wccm-eccomas.2020.107 PB - CIMNE ER - TY - CHAP A1 - Krenkel, Lars A1 - Pennecot, Julien A1 - Lenz, Christian A1 - Feldmann, Daniel A1 - Wagner, Claus ED - Füssel, Jens ED - Koch, Edmund ED - Malberg, Hagen T1 - Optimierung der Hochfrequenz-Oszillationsventilation mittels strömungsmechanischer Methoden und kontrastgasgestützter Magnetresonanztomografie - Teilprojekt: Rekonstruktion von Geometrien aus medizinischen Bilddaten und Erstellung von Modellen für experimentelle Strömungsuntersuchungen T2 - Laseranwendung in der Medizin, Erfassen und Verarbeiten kardiovaskulärer Signale, protektive Beatmungskonzepte : 3. Dresdner Medizintechnik-Symposium - mit DFG-Forschungsschwerpunkt Protektive Beatmungskonzepte, 6. bis 9. Dezember 2010, Dresden N2 - Der vorliegende Artikel skizziert die neu entwickelte Prozesskette für detailgenaue Rekonstruktion von Geometrien anhand medizinischer Bilddaten sowie die Herstellung von transparenten Modellen für experimentelle Untersuchungen mit bildgebenden Strömungsmessverfahren. Die Geometriedaten und Modelle werden für systematische Untersuchungen der komplexen Transportvorgänge in den Atemwegen bei künstlicher Beatmung mittels Hochfrequenzoszillationsventilation (HFOV) verwendet. Y1 - 2010 UR - https://elib.dlr.de/66863/ SN - 978-3-942710-02-2 SP - 107 EP - 114 PB - TUDpress CY - Dresden ER - TY - CHAP A1 - Birkenmaier, Clemens A1 - Steiger, Tamara A1 - Philipp, Alois A1 - Lehle, Karla A1 - Krenkel, Lars T1 - Flow-induced accumulations of von Willebrand factor inside oxygenators during extracorporeal life support therapy T2 - Proceedings of 12th International Conference BIOMDLORE 2018, June 28–30, 2018, Białystok, Poland N2 - BACKGROUND: Shear-induced conformational changes of von Willebrand factor (vWF) may be responsible for coagulation disorder and clot formation inside membrane oxygenators (MOs) during extracorporeal membrane oxygenation (ECMO) therapy. OBJECTIVE: The aim was to identify vWF structures inside clinically used MOs and employ computational fluid dynamics to verify the corresponding flow conditions. METHODS: Samples from gas exchange membranes (GEM) from MOs were analysed for accumulations of vWF and P-selectin-positive platelets using immunofluorescence techniques. Streamlines and shear rates of the flow around GEMs were computed using a laminar steady Reynolds-Averaged-Navier-Stokes approach. RESULTS: Most samples were colonized with equally distributed leukocytes, integrated in thin cobweb-like vWF-structures. Only 25 % of the samples showed extended accumulations of vWF. Computed streamlines showed considerable cross flow between interconnected neighbouring channels. Stagnation points were non-symmetric and contact faces were washed around closely. The occurring maximum shear rates ranged from 2,500 to 3,000 1/s. CONCLUSIONS: If pronounced vWF structures are present, shape and extent match the flow computations well. Computed shear rates bear a critical degree of uncertainty due to the improper viscosity model. If flow conditions inside the MO were sufficient to affect vWF, a more consistent distribution of vWF across the samples should be present. KW - Blood Viscosity KW - Shear Rate Induced Coagulation KW - Hemodynamics KW - Membrane Oxygenator KW - von Willebrand factor Y1 - 2018 SN - 978-1-5386-2396-1 U6 - https://doi.org/10.1109/BIOMDLORE.2018.8467205 PB - IEEE CY - Piscataway, NJ ER - TY - CHAP A1 - Markus Rütten, A1 - Krenkel, Lars A1 - Kessler, Roland T1 - Secondary Flow Effects as Physical Mechanism of Molecular Species Transport in Highly Oscillating Generic-Trachea Flows T2 - 83rd Annual Scientific Conference of the International Association of Applied Mathematics and Mechanics, 26.-30. März 2012, Darmstadt, Germany N2 - The high frequency oscillation artificial respiration technique is often the last hope for patients to survive highly damaged lung tissue. The mortality can significantly be reduced. In comparison to conventional artificial respiration the applied volume flow rate and pressure is significantly lowered in order to avoid further damaging of lung tissue and remaining intact alveolae. However, the physical mechanism of transport of oxygen to the aeriols under high frequency oscillation is not well understood. In the upper part of the lung convection is dominant, in contrast, the gas exchange in the lower parts of the lung is mainly driven by diffusion. It is not clear how associated gradients of concentrations of different molecular species are then achieved. Highly oscillating fluid flows has been a long research topic in fluid dynamics. It is known that oscillating pressure fluctuations are able to induce secondary flows, in particular, in curved ducts and pipes. The question is, whether the trachea enforces the generation of secondary flow by its kidney like cross section geometry. The influence of molecular species of different densities onto the formation of secondary flows and the convectional transport within the trachea is investigated. In order to clarify the physical mechanisms behind flow simulations have been conducted by using state of the art CFD techniques. KW - CFD KW - Generic-Trachea Flows Y1 - 2012 UR - https://elib.dlr.de/75733/ ER - TY - CHAP A1 - Friedrich, Janet A1 - Feldmann, Daniel A1 - Krenkel, Lars A1 - Wagner, Claus A1 - Schreiber, Laura Maria T1 - 19F Gas Flow Measurement of C3F7H During Constant Flow and High Frequency Oscillatory Ventilation T2 - Discovery, innovation & application - advancing mr for improved health : ISMRM 21st Annual Meeting & Exhibition ; SMRT 22nd Annual Meeting Salt Lake City, Utah, USA 20-26 April 2013 N2 - The aim of the current study is the development of MRI methods that enable the investigation of gas flow mechanisms during high frequency oscillatory ventilation. This work includes flow measurements during three constant flows (19.9, 30.6 and 41.4 L min-1) and the comparison to direct numerical simulations (DNS) using a second-order-acurate finite-volume method and to data measured with a volume flow meter. 19F-MRI, DNS and flow meter data are in good agreement. Flow measurements during HFOV of 4 Hz were successfully performed and velocity profiles could be recorded at different phases of the ventilation cycle. Y1 - 2013 UR - https://archive.ismrm.org/2013/1482.html VL - 21 ER - TY - CHAP A1 - Birkenmaier, Clemens A1 - Krenkel, Lars ED - Dillmann, Andreas ED - Heller, Gerd ED - Krämer, Ewald ED - Wagner, Claus T1 - Convolutional Neural Networks for Approximation of Blood Flow in Artificial Lungs T2 - New Results in Numerical and Experimental Fluid Mechanics XIII: Contributions to the 22nd STAB/DGLR Symposium N2 - Blood flow in channels of varying diameters <500μm exhibits strong non-linear effects. Multiphase finite volume approaches are feasible, but still computationally costly. Here, the feasibility of applying convolutional neural networks for blood flow prediction in artificial lungs is investigated. Training targets are precomputed using an Eulerian two-phase approach. To match with experimental data, the interphase drag and lift, as well as intraphase shear-thinning are adapted. A recursively branching regression network and convolution/deconvolution networks with plain skip connections and densely connected skips are investigated. A priori knowledge is incorporated in the loss functional to prevent the network from learning non-physical solutions. Inference from neural networks is approximately six orders of magnitude faster than the classical finite volume approach. Even if resulting in comparably coarse flow fields, the neural network predictions can be used as close to convergence initial solutions greatly accelerating classical flow computations. KW - Deep learning fluid mechanics KW - Multiphase blood flow Y1 - 2021 SN - 978-3-030-79560-3 U6 - https://doi.org/10.1007/978-3-030-79561-0_43 IS - 1. Auflage SP - 451 EP - 460 PB - Springer International Publishing CY - Cham ER - TY - CHAP A1 - Tauwald, Sandra Melina A1 - Quadrio, Maurizio A1 - Rütten, Markus A1 - Stemmer, Christian A1 - Krenkel, Lars T1 - High Spatial Resolution Tomo-PIV of the Trachea Focussing on the Physiological Breathing Cycle T2 - New Results in Numerical and Experimental Fluid Mechanics XIV - Contributions to the 23nd STAB/DGLR Symposium N2 - Investigations of complex patient-specific flow in the nasopharynx requires high resolution numerical calculations validated by reliable experiments. When building the validation base and the benchmark of computational fluid dynamics, an experimental setup of the nasal airways was developed. The applied optical measurement technique of tomo-PIV supplies information on the governing flow field in three dimensions. This paper presents tomo-PIV measurements of the highly complex patient-specific geometry of the human trachea. A computertomographic scan of a person’s head builds the basis of the experimental silicone model of the nasal airways. An optimised approach for precise refractive index matching avoids optical distortions even in highly complex non-free-of-sight 3D geometries. A linear-motor-driven pump generates breathing scenarios, based on measured breathing cycles. Adjusting of the CCD cameras‘ double-frame-rate PIV-Δt enables the detailed analysis of flow structures during different cycle phases. Merging regions of interest enables high spatial resolution acquisition of the flow field. KW - Tomographic PIV KW - Flow visualisation KW - Breathing cycle KW - Nasal airflow Y1 - 2023 N1 - Accepted for publication, not yet published PB - Springer ER - TY - CHAP A1 - Tschurtschenthaler, Karl A1 - Krenkel, Lars A1 - Schreiner, Rupert T1 - Mechano-optical micro pillar sensor for biofluidmechanic wall shear stress measurements T2 - 25th Congress of the European Society of Biomechanics (ESB), July 7-10, 2019, Vienna, Austria Y1 - 2019 UR - https://esbiomech.org/conference/archive/2019vienna/Contribution_608.pdf ER - TY - CHAP A1 - Rütten, Markus A1 - Krenkel, Lars A1 - Quadrio, Maurizio T1 - Simulation and Analyis of the Unsteady Flow within Nasal Airways T2 - 9th European Congress on Computational Methods in Applied Sciences and Engineering - ECCOMAS Congress, 3-7 June 2024, Lisbon, Portugal Y1 - 2024 UR - https://re.public.polimi.it/handle/11311/1269952 ER - TY - CHAP A1 - Krenkel, Lars A1 - Wagner, Claus A1 - Wolf, Ursula A1 - Scholz, Alexander-Wigbert K. A1 - Terekhov, Maxim A1 - Rivoire, Julien A1 - Schreiber, W. ED - Hirschel, Ernst Heinrich ED - Schröder, Wolfgang ED - Fujii, Kozo ED - Haase, Werner ED - Leer, Bram ED - Leschziner, Michael A. ED - Pandolfi, Maurizio ED - Periaux, Jacques ED - Rizzi, Arthur ED - Roux, Bernard ED - Shokin, Yurii I. ED - Dillmann, Andreas ED - Heller, Gerd ED - Klaas, Michael ED - Kreplin, Hans-Peter ED - Nitsche, Wolfgang T1 - Protective Artificial Lung Ventilation: Impact of an Endotracheal Tube on the Flow in a Generic Trachea T2 - New Results in Numerical and Experimental Fluid Mechanics VII : Contributions to the 16th STAB/DGLR Symposium Aachen, Germany 2008 N2 - Computational Fluid Dynamics (CFD) and experimental investigations on a generic model of the trachea have been carried out focusing on the impact of an endotracheal tube (ETT) on the resulting flow regime. It could be shown that detailed modelling of the airway management devices is essential for proper flow prediction, but secondary details as Murphy Eyes can be neglected. Models with bending and connector promote the formation of stronger secondary flows and disturbances which persist for a longer time. KW - computational fluid dynamics KW - Computational Fluid Dynamics Simulation KW - Endotracheal Tube KW - Particle Image Velocimetry KW - Turbulent Kinetic Energy Y1 - 2010 SN - 978-3-642-14242-0 U6 - https://doi.org/10.1007/978-3-642-14243-7_62 SP - 505 EP - 512 PB - Springer Berlin Heidelberg CY - Berlin, Heidelberg ER -