Beyond CFD: Emerging methodologies for predictive simulation in cardiovascular health and disease.

Casting computational fluid mechanics into a convex quadratic optimization framework

In recent years, computational simulations have emerged as valuable tools for the evaluation of atherosclerosis progression in coronary anatomies, although only a few studies have utilized more realistic Fluid-Structure Interaction (FSI) simulations. This work aims to compare the results of Computational Fluid Dynamics (CFD), Structural Finite Element Analysis (structural FEA) and FSI simulations in order to assess differences in plaque progression indices estimation. We performed structural FEA, CFD and FSI on five patient-specific epicardial coronary anatomies using the commercial software LS-Dyna. To account for the vessel pre-stress, the zero-pressure configuration was calculated for each anatomy with an inverse elastostatic algorithm. CFD, structural FEA and FSI simulations were performed applying boundary conditions based on physiological values. The comparison between structural FEA and FSI showed similar stress distribution and vessel expansions, with differences found only in the distal parts of the coronaries, where pressure reduction due to pressure loss affects the vessel walls. The elastic walls of the coronaries impact blood flow, resulting in a more disturbed flow. However, time averaged wall shear stress (TAWSS) and oscillatory shear index (OSI) distributions are similar across each coronary between CFD and FSI; TAWSS is slightly higher in CFD while OSI peaks are higher in FSI. In conclusion, given the significantly higher computational costs of FSI, we believe that CFD and structural FEA offer a more practical and cost-effective approach, providing results comparable to those of FSI, making them preferable options.

Indoor airflow field reconstruction using physics-informed neural network

The objective of this work is to gain the mechanical understanding of the water management in the proton exchange membrane fuel cell (PEMFC). Water is a by-product of the fuel cell reaction and its amount is proportion to the current of fuel cell output. Water is used to retain the proper hydration level in the PEM. At the same time, condensed water in the flowfield and the gas diffusion layer (GDL) reduces oxygen transport to the oxygen reduction reaction (ORR) area. Therefore, excess accumulation of condensed water causes to lower the performance, particularly at the higher current densities. Detailed simulation of two-phase water in the GDL is significantly important to understand the water management.This work shows the successful in development of a multi-scale calculation technique that incorporates various scales of numerical model in a fuel cell and simultaneously performed a prediction. The outcomes of this work are, (i) development of two-phase fluid model in the micro-scale porous structure of GDL, and (ii) integration of this micro-scale GDL fluid model with a state-of-the-art macro-scale flowfield fluid model to predict two-phase water in the fuel cell. Figure 1 gives the concept of modeling approach at each component of the fuel cell. The flowfield in the bipolar plate and MEA models are calculated using traditional computational fluid dynamics (CFD) method with existing PEMFC model [1-3]. The two-phase fluid in the micro-scale porous structure of GDL is numerically predicted by Lattice Boltzmann Method (LBM) [4-5]. The solution of each iteration from CFD and LBM are required to simultaneously exchange for the next iteration until all solutions are converged, especially at the interfaces. In this figure, it shows the flowfield is transformed into the macro-scale CFD model and the image of detail structured GDL is converted into micro-scale LBM model. Actual porous structure imaging of GDL is obtained by the nano-scale high special resolution x-ray computed tomography. Figure 2 shows an example of the predictions from the simulation when the macroscopic flowfield model is combined with the microscopic GDL model while performing the calculation. The predictions give the detail distributions of variables in every components of the multiscale model. In this figure, the current density distribution on MEA surface, the liquid water transport inside the GDL, and the temperature of solid and fluid phases inside the GDL are presented.

Distributed-parallel CFD computation for all fuel assemblies in PWR core

Promena paradigme za stabilnu koronarnu bolest u hronični koronarni sindromi - novine u vodiču Evropskog udruženja kardiologa iz 2019. godine

Objective To explore the clinical usefulness of low-dose dual-source CT (DSCT) angiography in detecting coronary artery anomaly (CAA) in patients with complex congenital heart disease (CHD). Methods A total of 615 consecutive patients with complex CHD who underwent DSCT angiography between 2008 and 2012 were retrospectively reviewed. According to the ultrasonic results, a total of 312 patients with tetralogy of Fallot (n=176 cases) or double outlet right ventricle (n=72 cases) or pulmonary artery atresia (n=64 cases) were included. Scans were divided into 3 groups: retrospective electrocardiogram (ECG)-triggered spiral scanning in 75 cases, prospective ECG-triggered spiral scanning in 158 cases, and prospective ECG-triggered high-pitch spiral scanning in 79 cases. The subjective image quality was evaluated with a four-point scale by two radiologists. Radiation dose values were calculated. Interobserver agreement in subjective image quality grading was assessed by using Kappa statistics. The scanning groups were compared with the Kruskal-Wallis test. Results The incidence of CAA was 8.0% (25/312) . CAA was detected in 7.9% (14/176) patients with tetralogy of Fallot, 8.3% (6/72) patients with double outlet right ventricle, 7.8% (5/64) patients with pulmonary artery atresia. In 76% (19/25) of the patients, the anomalous vessels were crossing the right ventricular outflow tract (RVOT). Thirty of 312 CT examinations were nondiagnostic (9.6%). Two radiologists acquired good agreement (Kappa=0.72, P<0.01). The median subjective image quality score of three scan groups was 3 (range 1—4). The score in high-pitch spiral scanning group was significantly lower than those in other two groups (H=29.1, P<0.01) . Retrospective ECG-triggered scans had a median dose of 1.19 mSv (range 0.52—3.29 mSv) , prospectively ECG-triggered scans had a median dose of 0.51 mSv (range 0.27—2.13 mSv) , and prospectively ECG-triggered high-pitch spiral scans had a median effective dose of 0.30 mSv (range 0.18—0.62 mSv) . The difference between groups was statistically significant (H=160.0, P<0.01) . Conclusions DSCT angiography is a reliable diagnostic method for the assessment of CAA with complex CHD and low-dose prospectively ECG-triggered DSCT scan is the best choice. Key words: Coronary artery disease; Heart defects, congenital; Tomography, X-ray computed

The design of new multiphase separation plants or the replacement of vessels represents an opportunity to review and update the separation systems, to ensure the surface facilities lifetime with maturing production conditions. Potentially adverse operating scenarios involving slug flow could occur in the production header, leading to potentially damaging fluid forces acting on the high-pressure production trap internals. The internals design is based on the fluid forces occurring during a slug flow condition. This paper presents the application of 3-D Computational Fluid Dynamics (CFD) to simulate the realistic fluid forces acting on a novel separator inlet device, and to assess the structural integrity of the mechanical design. To generate slug conditions resulting from several adverse, but low probability, operating scenarios, OLGA, a 1-D transient multiphase flow hydraulic simulation application, was used to simulate the transient multiphase flow conditions in the production header entering the production separator using the 20-year production forecast. In a one-way coupling, the output from OLGA is used as input to the CFD model. The scenarios were carefully developed, thoroughly evaluated and agreed on by the different project stakeholders — technical services, operations and project management teams. The OLGA simulations produced the transient phase velocities and hold-ups that were applied as inlet boundary conditions to the CFD models of the separator. The transient CFD turbulent multiphase simulations used the Volume-of-Fluid model to track the gas-oil-water mixture in the inlet device and the inlet section of the separator. A discussion on the appropriate CFD multiphase model selection is presented. The hydrodynamic forces computed in the CFD model are integrated over the internal surfaces of the inlet device and mapped as a distributed pressure load for a Finite Element Analysis (FEA) structural model, to determine the mechanical reactions of the structure, a fluid-structure interaction (FSI) analysis. The methodology developed is intended for use in new internal designs for two- and three-phase separators, and slug catchers in gas/oil separation plants (GOSPs).

Computational fluid dynamics (CFD) simulations are essential for studying cardiovascular diseases, particularly atherosclerosis. The left coronary artery (LCA) and its bifurcations, which supply blood to a large portion of the myocardium, are recognized as high-risk sites for plaque formation. This systematic review addresses the question: “What are the main CFD approaches, modeling strategies, and hemodynamic parameters used to study LCA bifurcation over the last decade, and how have these advanced the understanding of atherosclerosis progression?” A comprehensive search was conducted in PubMed, Web of Science, and Scopus databases. Following preferred reporting items for systematic reviews and meta-analyses guidelines, the search focused on original studies published within the last decade employing CFD to analyze LCA bifurcation hemodynamics. The review includes 108 CFD studies, mostly three-dimensional (3D) simulations (104 studies). Forty-seven used idealized 3D LCA models and 48 used patient-specific models focusing on bifurcation. Recent studies increasingly focus on hemodynamic parameters linked to atherosclerosis and the impact of stenosis and post-stenting restenosis on local blood flow. Simulations incorporating physiological features, such as fluid–structure interaction, cardiac motion, and blood rheology, are also included. Additionally, artificial intelligence is being integrated to efficiently predict hemodynamic parameters, reducing simulation time and cost. There is a clear trend toward using high-resolution imaging and advanced modeling for patient-specific CFD simulations. Despite methodological differences, disturbed flow patterns, low wall shear stress, and high oscillatory shear index are studied in the LCA bifurcation, supporting their role in atherogenesis. Standardizing protocols and integrating clinical data are essential for translating CFD findings into practice.

&lt;div class="section abstract"&gt;&lt;div class="htmlview paragraph"&gt;The number of computational fluid dynamics (CFD) simulations performed during the vehicle aerodynamic development process continues to expand at a rapid rate. One key contributor to this trend is the number of analytically based designed experiments performed to support vehicle aerodynamic shape development. A second contributor is the number of aerodynamic optimization studies performed for vehicle exterior components such as mirrors, underbody shields, spoilers, etc. A third contributor is the increasing number of “what if” exploratory studies performed early in the design process when the design is relatively fluid. Licensing costs for commercial CFD solutions can become a significant constraint as the number of simulations expands. A number of alternative products (e.g., independently developed, supported and documented forks of the popular open-source OpenFOAM® toolbox [&lt;span class="xref"&gt;1&lt;/span&gt;]) have become available in recent years, offering a lower cost alternative to traditional commercial CFD products. This paper summarizes results from a broad and deep evaluation of the capability of iconCFD® to substitute for the more traditional commercial CFD solutions currently used to support vehicle aerodynamic development early in the program development cycle. Included in this study were detailed B-car, sedan, SUV and truck shapes as well as multiple variants of each shape. The study investigated both static and moving ground boundary conditions as well as alternative turbulence models. Trends of the predicted aerodynamic drag coefficients (&lt;i&gt;C&lt;sub&gt;d&lt;/sub&gt;&lt;/i&gt;) are compared against experimental data. Both transient and steady state simulation ranking accuracy of total vehicle &lt;i&gt;C&lt;sub&gt;d&lt;/sub&gt;&lt;/i&gt; were found to be equivalent to that historically observed with more traditional commercial solver results. A consistent upward bias in absolute &lt;i&gt;C&lt;sub&gt;d&lt;/sub&gt;&lt;/i&gt; values was observed in the transient results.&lt;/div&gt;&lt;/div&gt;

This research investigates the wake dynamics of low-Reynolds-number flow past a circular cylinder by combining high-fidelity Computational Fluid Dynamics simulations with modern data-driven analysis. The flow is simulated with the incompressible solver icoFoam in the OpenFOAM 11.0, faithfully capturing the laminar regime and its transition to unsteady vortex shedding at a Reynolds number of one hundred. These simulations provide a clear picture of wake development, the emergence of flow instabilities, and the onset of vortex shedding. Central to the study is Singular Value Decomposition, which decomposes the complex velocity field into its most energetic spatial structures. By retaining the dominant modes, Singular Value Decomposition reduces the dimensionality of the data while preserving essential flow characteristics. The velocity field reconstructed from these modes offers a compact yet accurate description of vortex formation, energy distribution, and the temporal evolution of shedding patterns. The synergy between traditional Computational Fluid Dynamics and contemporary machine-learning-inspired techniques yields a versatile and computationally efficient framework for exploring intricate fluid phenomena. Beyond accelerating calculations, this approach deepens our physical insight into the mechanisms that govern wake dynamics. By quantitatively predicting vortex-shedding onset and elucidating the underlying flow structures, the study delivers actionable knowledge for applications ranging from aerodynamic design to environmental flow management, where precise modeling of complex flow behavior is critical.

The high performance of nature’s creations and biological assemblies has inspired the development of engineered counter parts that may outperform or provide new capabilities to conventional systems. In particular, the wings of bats contain distributed arrays of micro-scaled flow sensitive hair receptors over their surface, which inspires artificial hair sensors (AHS) development in aerodynamic feedback control designs using the micro-electro-mechanical systems (MEMS). One approach investigates the possibility of installing AHS on the leading edges of the wings of small-scaled unmanned aerial vehicles (UAVs) to improve the aerodynamic control. Our major motivation for the present study is that current mathematical models have limited relevance to aerodynamic situations because they are analyzed in steady or purely oscillatory flows. Our overall objective is to understand AHS fluid-structure interaction (FSI) in flow regimes relevant to small-scaled UAVs, for which we speculate a steady baseline flow perturbed by an oscillatory component is an appropriate flow reference condition. Towards understanding the AHS in this situation, we investigate the dynamic response of a hair receptor in a creeping flow environment with a steady and oscillatory component. We present time varying deflection and bending moment of the artificial hair sensors at different freestream velocities. For this, a three-dimensional FSI model is developed for the flexible hair-structure in the airflow, which is coupled with a finite element model using the computational fluid dynamics (CFD). The Navier-Stokes equations including continuity equation are solved numerically for the CFD model. To describe the dynamic response of the hair receptors, the natural frequencies and mode shapes of the hair receptors, computed from the FSI model, are compared with the excitation frequencies of the surrounding airflow. This model also describes both the boundary layer effects and effects of inertial forces due to FSI of the hair receptors. For supporting the FSI model, the dynamic response of the hair receptor is also validated considering the Euler-Bernoulli beam theory including the steady and unsteady airflow.

This paper deal with structural deformation of one rectangular, closed, thin walled, steel made, thermally loaded channel and influence of this deformation on the air flow through the channel, recognised as coupled fluid-structure interaction problem. The geometrically non-linear deformation of thermally loaded channel was solved using commercial finite-element analysis software ANSYS. The unsteady Navier-Stokes equations in their conservation form defining fluid flow are solved using commercial computer fluid dynamics software ANSYS CFX. A fuzzy logic-based expert system has been developed to predict some results of structural and fluid flow numerical simulation instead of directly evaluating it by a time-consuming software coupling of finite-element and computer fluid dynamics numerical analysis. Three input parameters, namely temperature of channel inner walls, difference in temperature of inner and outer surface of channel wall, and factor of mesh size, are varied while solving the fluid-structure interaction problem. A pressure drop downstream the flow was considered as a computer fluid dynamics output parameter of importance, while Von Misses maximum stresses were an finite-element analysis output parameter observed. The developed FLES can predict the results of finite-element and computer fluid dynamics analysis, a pressure drop and Von Misses maximum stresses respectively, within a reasonable accuracy limit and at lower computation cost compared to the finite-element, and CFD computer fluid dynamics packages combined into series of multi-physics analyses.

Research and development on digital twins of nuclear power systems has focused on high-precision real-time simulation and the prediction of local complex three-dimensional fluid dynamics. Traditional computational fluid dynamics (CFD) methods cannot take into consideration the efficiency and accuracy of fluid dynamics. In this study, a fast-flow field-prediction framework based on proper orthogonal decomposition (POD) and deep learning is proposed. Compressed data containing the original flow field information are obtained using POD and deep neural network (DNN) is used to construct the POD-DNN flow field reduction model to achieve fast flow field prediction. The calculation accuracy and speed of the reduced-order model are analyzed in detail, considering the flow field of the nuclear compressor and key flow equipment of the nuclear power system as objects. The results show that the average relative deviation of the POD-DNN is &amp;lt;10% and calculation time is &amp;lt;1% when compared to those of CFD. This research shows that the high-fidelity model constructed using model reduction and deep learning is a feasible method for the realization of digital twins of the nuclear power system in engineering.

This paper presents a novel approach to address computational challenges in predicting flow features by employing a Graph Neural Network (GNN), which is proficient in predicting flow domain values. Traditional Computational Fluid Dynamics (CFD) simulations, although effective, often require substantial computational resources and time, limiting their applicability in time-sensitive scenarios and optimization studies necessitating extensive case studies. The main objective was to evaluate the feasibility of employing node classification on a graph generated from a 2D baffle flow system to segment the domain based on relative fluid age. A second objective was to compare the computational time required for CFD simulations with the inference time of the network to quantify the efficiency gains achieved by utilizing the network. Results demonstrate the potential of utilizing graph convolutional networks for domain segmentation to predict regions of holdup and bypass. The GNN achieved 97% and 92% accuracy in predicting recirculation regions in single and double baffle cases, particularly excelling in high Reynolds number scenarios. Importantly, the proposed GNN-based approach reduces computation time by over 2100%, showcasing significant efficiency gains. Results here highlight the promise of employing graph convolutional networks for flow feature prediction, offering substantial computational efficiency improvements over traditional CFD simulations.