Chapter Four · failure evidence
What Hydrodynamic & Fluid Simulation got wrong, from 84 dissertations
Computational fluid dynamics and hydrodynamic simulations frequently encounter bottlenecks spanning excessive computational demands, turbulence model failures, and boundary instability. Across these investigations, practitioners frequently abandon complex direct solvers or uncalibrated empirical approximations when simpler representations or alternative physical formulations prove more accurate and numerically robust. These records come from PhD theses at 23 institutions, 2021 to 2026. Each links to its thesis. They were extracted by language models reading the full text, so treat each as a lead to read, not a verdict.
Prohibitive computational expense and excessive runtime in high-fidelity simulations
Full three-dimensional Navier-Stokes solvers, direct numerical simulations, smoothed particle hydrodynamics, and aggressive spatial mesh refinements repeatedly become computationally intractable or incompatible with real-time requirements. As a consequence, researchers often reject fine-scale multiscale models and particle methods in favor of analytical solutions, discrete element methods, or lumped-parameter models.
Considered and rejected
Considered and rejected: Rejected Computational Fluid Dynamics (CFD) in favor of small-scale salt bath physical experiments due to high computational expense, turbulence modeling uncertainties, boundary condition sensitivity, and lack of validation data.
The fluid mechanics of horizontal exchange flows in naturally ventilated buildings · Imperial
Considered and rejected
Considered and rejected: Rejected computational fluid dynamics (CFD) modeling due to excessive computational demands incompatible with real-time forward simulation.
Adapting Random Tree Search to Fixed Wing Aerial Vehicles with Closed-Loop Prediction and Hybrid Control · Georgia Tech
Considered and rejected
Considered and rejected: Directly solving 3-D Navier-Stokes equations for external particle flow, rejected because computational expense prevents real-time interactive rendering and continuity assumptions fail when fluid separates into particles
Interactive computer graphics environment for particle simulation · Iowa State
Considered and rejected
Considered and rejected: Rejected full 3D time-dependent multi-phase Navier-Stokes fluid flow modeling due to excessive complexity of geometry and boundary conditions.
Thermodynamic analysis of single-screw oil-flooded refrigerant compressors · Virginia Tech
Considered and rejected
Considered and rejected: Rejected fluid-dynamics and fluid-empirical air brake models because solving 1D fluid equations required co-simulation or parallel computing, compromising real-time simulation speed.
Development of numerical and experimental tools for the simulation of train braking operations · IRIS - POLITO - prod
Tried and failed
smoothed particle hydrodynamics applied to large-scale 3D dynamic impact simulations. Outcome: infeasible cost. Reason: computationally impractical and numerically unstable when scaling up to full-scale 3D models
Computational modelling of traumatic cerebral vasculature injury · Imperial
Tried and failed
adaptive mesh refinement algorithm applied to collapsing multiphase fluid bubble simulations. Outcome: too slow. Reason: excessive over-refining of the grid during bubble collapse drastically increased computational time
Numerical Study of Energy Loss Mechanisms in Oscillating Underwater Explosion (UNDEX) Bubbles · Virginia Tech
Tried and failed
isotropic spatial mesh refinement applied to shock tube detonation hydrodynamic simulation. Outcome: too slow. Reason: runtime exceeded acceptable limits without providing sufficient computational practicality
Numerical Modeling of the Energy Release of Aluminized Oxyacetylene Detonations · Virginia Tech
Tried and failed
uniform fine mesh grid discretization applied to large-scale 2D hydrodynamic floodplain simulation. Outcome: too slow. Reason: Massive computational runtime increase without producing measurable differences in inundation extent
Modeling the Impact of Flood Pulses on Disease Outbreaks in Large Water Basins with Scarce Data · Virginia Tech
Tried and failed
full concurrent multiscale direct numerical simulation applied to multiscale fluid-structure interface dynamics. Outcome: infeasible cost. Reason: simulating fine-scale microdomains at every macroscale boundary point each timestep is computationally prohibitive
Modeling Droplet Impingement Dynamics on Micropillar-Arrayed Viscoelastic Substrates Through Microgeometry-Free and Multiscale Methods · Virginia Tech
Considered and rejected
Considered and rejected: Rejected SPH (Smoothed Particle Hydrodynamics) due to extreme computational particle cost and difficulty setting viscous absorbing boundary conditions.
Numerical and Experimental Studies in Evaluating Engineering Effectiveness of Nature Based and Eco-Engineered Sea Defences · Research Repository UCD
Considered and rejected
Considered and rejected: Rejected Smoothed Particle Hydrodynamics (SPH) due to lack of constitutive material models handling simultaneous fracture/flow/cohesion and slower computation than DEM.
Considered and rejected
Considered and rejected: Rejected Smoothed Particle Hydrodynamics (SPH) and other Lagrangian mesh-free methods due to explicit computational cost, pressure field instability, and difficulties in implementing Dirichlet/Neumann boundary conditions.
Numerical modelling of aeration and hydroelasticity in slamming loads and responses of marine structures · Leibniz Universität Hannover Repository
Considered and rejected
Considered and rejected: Rejected subsurface volumetric modelling and fluid dynamics for hydrogeology/sedimentology due to high computational overhead and interface complexity
Exploring the Potential of an Interactive Geoscience Teaching Application for Dynamically Visualizing 3D Terrain in Undergraduate Classrooms · Queens University Institutional Repository
Considered and rejected
Considered and rejected: Rejected pure 3D numerical hydrodynamic plume simulation over decadal scales due to excessive computational requirements and artificial numerical diffusion, selecting analytical solutions instead.
Sediment plumes and financial modeling in the context of deep-sea polymetallic nodule mining · MIT
Considered and rejected
Considered and rejected: Rejected using numerical quadrature rules for integrating rapidly varying regularized slender-body kernels due to prohibitive computational runtime during elastohydrodynamic simulations.
Mathematical modelling and computational simulation studies of flagellate motility · Oxford
Considered and rejected
Considered and rejected: Rejected convolution integral approach for hydrodynamics during optimization due to excessive computational time required by step-by-step for-loop ODE solver execution in MATLAB
Vibration Reduction of a Semisubmersible Floating Wind Turbine using Optimized Tuned Mass and Tuned Inerter Dampers · Virginia Tech
Considered and rejected
Considered and rejected: Rejected computationally expensive fluid dynamics modeling in favor of a lumped-capacitance steady-state thermal model.
Characterization of Internal Formaldehyde Production within The Pandora Spectrometer Instrument · Virginia Tech
Considered and rejected
Considered and rejected: Rejected Lagrangian Particle Tracking multiphase model because it is computationally inefficient and unsuitable for high particle volume fractions with turbulence.
TRANSIENT COMPUTATIONAL FLUID DYNAMICS MODELING OF WET STEAM IN A PIPING SYSTEM · Calhoun
Turbulence and eddy viscosity model breakdowns in non-ideal or separated flows
Standard RANS closures, isotropic eddy viscosity assumptions, and classical Smagorinsky formulations fail to capture separated flow dynamics, turbulent kinetic energy distributions, and thermal stratification. These closures frequently mispredict near-wall turbulent dissipation or severely diverge when encountering strong non-ideal thermodynamic regimes, vortex interactions, or viscoelastic effects.
Tried and failed
algebraic and one-equation turbulence models applied to recirculating and separated fluid flows. Outcome: did not generalise. Reason: unable to capture separated flow dynamics without extensive empirical tuning
An experimental and theoretical study of flow in a vortex controlled diffuser · Cranfield
Considered and rejected
Considered and rejected: Rejected the Macroscopic Fluctuation Theory (MFT) hydrodynamic scaling limit because fixing viscosity while vanishing noise is physically invalid for turbulence.
Spontaneous Stochasticity and the Space-Time Statistical Solutions of Landau–Lifshitz Hydrodynamics · JScholarship
Tried and failed
variable turbulent Prandtl number closure models applied to turbulent non-ideal compressible fluid flow. Outcome: worse than baseline. Reason: models severely underestimated heat transfer under strong non-ideal thermodynamic effects compared to constant Pr_t
A study of the internal flow of dense vapours used in Organic Rankine Cycle (ORC) turbines · Imperial
Tried and failed
Standard two-equation RANS turbulence models applied to Natural convection in stagnant enclosed vapor. Outcome: did not generalise. Reason: Models overpredicted turbulent kinetic energy and thermal diffusivity, underpredicting thermal stratification and vapor superheating
Modelling the evaporation of cryogenic liquids in storage tanks · Imperial
Tried and failed
Newtonian eddy-viscosity turbulence modeling applied to viscoelastic fluid flow drag reduction. Outcome: did not generalise. Reason: Failed to capture near-wall turbulent kinetic energy peaks and anisotropic viscoelastic stress contributions.
Drag Reduction by Polymeric Additive Solutions · Virginia Tech
Tried and failed
Strong Reynolds Analogy turbulence modeling applied to non-ideal compressible fluid flows. Outcome: did not generalise. Reason: large discrepancies between turbulent velocity and enthalpy mixing length scales under strong non-ideal thermodynamic regimes
A study of the internal flow of dense vapours used in Organic Rankine Cycle (ORC) turbines · Imperial
Tried and failed
hybrid RANS-LES turbulence modeling applied to differentially heated cavity natural convection flows. Reason: flow laminarization suppressed modeled turbulent kinetic energy, collapsing the model to coarse direct numerical simulation
Tried and failed
baseline k-omega two-equation turbulence model applied to unsteady aerodynamic vortex interaction simulation. Outcome: unstable. Reason: overpredicted turbulent eddy viscosity by several orders of magnitude, causing numerical divergence
Computational Assessment of the Interactional Proprotor-Wing Aerodynamics during Tiltrotor Conversion · Georgia Tech
Tried and failed
classical Smagorinsky Large Eddy Simulation model applied to oscillating wall-bounded fluid flows. Outcome: worse than baseline. Reason: overestimated near-wall eddy viscosity and turbulent dissipation, skewing oscillation amplitudes and natural periods
Nonlinear forcing, response and damping of oscillating water columns · Imperial
Lost to a baseline
SA model with default and varied coefficients over-predicted lift coefficient CL for NACA 0012 at angles of attack > 15 deg compared to other baseline turbulence models.
Data-driven RANS turbulence modeling for compressor stall · Imperial
Considered and rejected
Considered and rejected: Rayleigh-Plesset cavitation model and k-epsilon turbulence model were rejected in favor of Schnerr-Sauer and k-omega SST models which aligned better with reference data.
Health monitoring of cavitation erosion on airframe fuel pumps · Cranfield
Considered and rejected
Considered and rejected: Rejected turbulence modeling (e.g., RANS) in favor of a laminar solver with mesh refinement to avoid empirical constants and artificial dissipation
Considered and rejected
Considered and rejected: Rejected standard k-ω and SST k-ω turbulence models in favor of Realizable k-ε with enhanced wall treatment based on error analyses and R² comparisons of velocity and cooling effectiveness.
Analysis on the effect of trench geometry on film cooling effectiveness of shaped holes using rans simulation · Institutional Repository University of Moratuwa
Considered and rejected
Considered and rejected: Rejected freezing the Reynolds stress field during UQ realizations (as done in prior methods), instead fully solving and transporting transport equations through the turbulence model.
A First Complete Approach to Address Model Error in Computational Turbulent Heat Transfer · MIT
Tried and failed
Standard Smagorinsky large eddy simulation model applied to turbulent jet expansion and mixing. Outcome: did not generalise. Reason: Inability of constant eddy viscosity to capture spatial variations in jet spreading and near-wall turbulence decay
TURBULENT MIXING BETWEEN LIQUIDS OF DISPARATE VISCOSITY ANDTHEIR EFFECT ON REACTION · Georgia Tech
Considered and rejected
Considered and rejected: Rejected algebraic URANS turbulence models due to incomplete formulations with empirical coefficients that do not generalize to complex 3D separated rotary flows.
Computational Investigation of Separated Flow and Stall Events on Rotating Systems · Georgia Tech
Failure of dimensional reductions and empirical correlations under complex flow physics
Reduced-dimension models and empirical correlations calibrated under simple operating states fail when applied to complex conditions such as cryogenic boiling, downward multiphase flows, and stall conditions. Omitting three-dimensional effects, spatial crosswind variations, or nonlinear drag profiles leads to significant underpredictions of power savings, meniscus stability, and flow rates.
Tried and failed
computational fluid dynamics aerodynamic modeling applied to airfoils at high angles of attack. Reason: fails to realistically simulate flow separation and aerodynamic behavior under stall conditions
Anomaly Detection in the Vibration of Wind Turbine Blades using Gaussian Process Regression · Carleton University Institutional Repository
Tried and failed
empirical flow boiling correlation applied to cryogenic liquid boiling heat transfer. Outcome: did not generalise. Reason: failed to provide satisfactory predictive accuracy across cryogenic fluid flow regimes
Flow and Phase Change Phenomena of Cryogens and Cryogenic Fuels · Georgia Tech
Tried and failed
2D axisymmetric computational fluid dynamics modeling applied to boundary layer ingestion aerodynamic performance. Reason: neglecting 3D flow effects led to systematic underprediction of power savings compared to full 3D models
Propulsion integration design and evaluation for novel aircraft configurations · Cranfield
Tried and failed
empirical two-phase pressure drop correlations applied to vertical downflow fluid boiling. Outcome: did not generalise. Reason: Models calibrated on horizontal or upward flows fail to capture downward multiphase flow physics accurately.
Lost to a baseline
Fundamentally-based turbulent flame propagation models performed less satisfactorily than zero-dimensional thermodynamic correlations in the absence of detailed in-cylinder turbulence data
A zero-dimensional model of turbulent combustion in a homogeneous charge spark ignition engine. · Cranfield
Tried and failed
lumped equivalent density parameter estimation applied to non-Newtonian fluid with thermal viscosity inversion. Outcome: did not generalise. Reason: non-linear spatial pressure losses under temperature gradients cannot be captured by a single effective density parameter
Rheological characterization of complex drilling fluid and the effects on drilling hydraulics · UT Austin
Lost to a baseline
2D BEM outperformed all URANSE turbulence/cavitation models in predicting CD and following the experimental CD trend across cavitation numbers at AoA=10 deg.
Numerical Simulations of Supercavitating Propellers and Hydrofoils · Virginia Tech
Considered and rejected
Considered and rejected: Rejected turbulent fluid flow models in heatsink channel analysis in favor of laminar flow models, justifying it as a valid worst-case analysis while reducing analytical complexity
Designing Power Converter-Based Energy Management Systems with a Hierarchical Optimization Method · Virginia Tech
Considered and rejected
Considered and rejected: Rejected linear increasing wind drag coefficient in the sigma-layer hydrodynamic model, replacing it with a piecewise linear function to match observed velocities and temperatures.
Considered and rejected
Considered and rejected: Pure linear or pure quadratic velocity damping hydrodynamic models during least-squares parameter identification of impulsive maneuvers (rejected in favor of a combined linear and quadratic damping model due to consistently higher L2 norm error)
Nonlinear Free Surface and Viscous Effects on Underwater Vehicle Maneuvering and Seakeeping · Virginia Tech
Tried and failed
reduced-scale physical flume modeling applied to sediment transport in river pools. Outcome: did not generalise. Reason: Wall boundary effects and missing 3D secondary currents distorted local hydrodynamics
Hydromorphological assessment of sediment augmentation measures in gravel-bed rivers · EPFL
Considered and rejected
Considered and rejected: Rejected 1D and 2D shallow water hydrodynamic modeling for complex fishways due to inability to resolve vertical velocities, anisotropic turbulence, and slot flow parameters.
Upstream passage of small sized fish species - effects of light and hydrodynamics · IRIS - POLITO - prod
Tried and failed
constant crosswind assumption in trajectory modeling applied to projectile ballistic drift prediction. Outcome: did not generalise. Reason: Atmospheric field wind distributions rarely match uniform velocity assumptions due to spatial and temporal turbulence.
Effect of wind turbulence on wind ballistic trajectory of a medium calibre weapon system · Cranfield
Lost to a baseline
The EWOD pump's droplet velocity (126 mm/s) and flow rate fell roughly one order of magnitude below theoretical slug-flow model predictions due to unaccounted interface turbulence and contact angle hysteresis.
Electrowetting on a Dielectric (EWOD) for soft robotics · EPFL
Tried and failed
1D Darcy's law for macroscopic fluid flow applied to slot die coating over porous substrates. Outcome: did not generalise. Reason: 1D porous flow oversimplified transverse infiltration dynamics, mispredicting meniscus stability limits and coating thickness by over 57%
Scalable Approaches to Manufacturing Nanocellulose-Glass-Fiber Reinforced Composites · Georgia Tech
Lost to a baseline
Dittus-Boelter correlation (underpredicted/shifted down calculated fluid average velocities compared to measured values because the thermal boundary layer develops within the laminar viscous sublayer near the wall)
Intelligent Non-Invasive Thermal Energy Flow Rate Sensor for Laminar and Turbulent Pipe Flows · Virginia Tech
Breakdown of continuum and transport assumptions in porous and particulate systems
Standard continuum formulations including Darcy flow, static drag laws, and pore network approximations break down when applied to mobile granular beds, dilute packings, or microscopic cavitation. Neglecting capillary pressures or relying on literature-derived dispersivity metrics causes severe mispredictions of fluid front propagation and mixing zones.
Tried and failed
Literature-derived high dispersivity values in transport modeling applied to subsurface solute transport simulation. Reason: Excessive hydrodynamic dispersion caused unrealistically broad mixing zones compared to observed salinity gradients
Controls on groundwater and surface water salinity in coastal Bangladesh · Imperial
Tried and failed
predicting sediment transport from shear stress applied to turbulent flows with emergent obstructions. Outcome: did not generalise. Reason: transport rates varied by two orders of magnitude at identical shear stress due to uncaptured turbulence
The Impact of Vegetation Morphology on Turbulence and Bedload Transport · MIT
Tried and failed
multiphase displacement using non-equilibrated working fluids applied to pore-scale fluid displacement experiments. Outcome: unstable. Reason: the non-wetting phase completely dissolved into the non-equilibrated wetting fluid across repeated cycles
Pore-scale investigation of two-phase flow hysteresis in hydrogen storage using micro-ct imaging · Imperial
Tried and failed
multiphase reservoir simulation neglecting capillary pressure applied to thermal porous media drainage modeling. Reason: neglecting capillary forces significantly overestimated fluid front propagation rate and displacement chamber dimensions
Considered and rejected
Considered and rejected: Rejected using standard dried oil film (DOF) upwelling dilution alone to represent hyporheic contamination because pre-coating dry gravel fails to reproduce liquid droplet hydrodynamics, trapping, and spatial heterogeneity.
Bioavailability and Toxicity of Oil Trapped in River Sediments by Hyporheic Flows · Queens University Institutional Repository
Tried and failed
Eulerian contour thresholding for coherent vortex detection applied to fluid flow transport boundary identification. Outcome: did not generalise. Reason: Eulerian boundaries lack material invariance and fail to track Lagrangian coherent transport over monthly timescales
A Lagrangian perspective of mesoscale biophysical interactions in the subtropical ocean · MIT
Tried and failed
static drag laws in mobile granular beds applied to fluid-driven granular shear flows. Outcome: did not generalise. Reason: fixed-particle assumptions severely under-predict drag under mobile, shearing bed conditions
Tried and failed
semi-analytical colloid filtration correlations applied to charged particle deposition in porous media. Outcome: did not generalise. Reason: correlations fail at low fluid velocities under unfavourable electrostatic surface interaction conditions
Mathematical modelling of nanoparticle transport through tumours for application in radiotherapy · Imperial
Tried and failed
hydrodynamic transport modeling via Boltzmann equation applied to phonon transport in defective nanomaterials. Outcome: no signal. Reason: low point defect concentrations strongly scatter momentum and destroy collective drift motion
Tried and failed
pore network modeling of fluid flow applied to high-porosity dilute particulate packings. Outcome: did not generalise. Reason: throat conductance and network topology assumptions break down in dilute, unconfined pore spaces without well-defined throats
Particle-scale numerical simulation of the thermal behaviour of granular materials · Imperial
Considered and rejected
Considered and rejected: Rejected standard Darcy multiphase poroelastic extensions because non-wetting fluid completely evacuates solid during macroscopic cavitation, invalidating standard pore-flow assumptions.
Formation and collapse of non-wetting cavities in soft porous media · Oxford
Boundary condition and numerical interface formulation defects causing instabilities
Standard open outflow conditions, truncated domains without buffer zones, and cut-cell immersed boundary methods induce artificial pressure reflections, backflow entrainment, and flow leakage. Similarly, sharp patient pressure inputs and Rhie-Chow formulations with drag terms trigger numerical divergence and prevent proper valve closure.
Tried and failed
lumped parameter boundary conditions without branch flow correction applied to computational fluid dynamics in vascular networks. Outcome: unstable. Reason: omitted side branches without flow adjustment caused severe artificial volume overload and unphysiological pressure spikes
Computational analysis of hemodynamics and thrombosis in aortic dissection for clinical applications · Imperial
Tried and failed
standard open outflow boundary conditions applied to swirling and recirculating fluid flows. Outcome: unstable. Reason: nonphysical artificial pressure gradients and backflow entrainment cause numerical ill-posedness
Dynamics of swirling jets and flames · Georgia Tech
Tried and failed
immersed boundary method on raw point clouds applied to computational fluid dynamics simulation. Outcome: unstable. Reason: cut elements lacking boundary points caused severe velocity oscillations and flow leakage
NIMBUS: A novel immersed boundary framework for scanned geometries · Iowa State
Tried and failed
direct patient pressure waveforms as boundary conditions applied to heart valve fluid dynamics simulation. Reason: the thinness and sharpness of pressure peaks prevented proper valve closure in CFD simulations
Towards a Tissue Engineered Heart Valve Bioreactor and a Coptic Ethic for Tissue Engineering · Harvard
Tried and failed
SPH-FE coupling with ghost particles applied to coupled fluid-structure interaction simulations. Outcome: unstable. Reason: excessive computational cost and increased numerical instability at the interface
Computational modelling of traumatic cerebral vasculature injury · Imperial
Tried and failed
increasing spatial resolution in particle hydrodynamics applied to semi-analytical boundary fluid-solid coupling. Outcome: unstable. Reason: amplified boundary inconsistencies causing unphysical fluid jets and noisy pressure fields
Smoothed particle hydrodynamics for fluid-solid coupling: modelling fixed and mobile boundaries · Imperial
Tried and failed
short computational domain without buffer zones applied to unsteady fluid dynamics simulations. Reason: spurious boundary wave reflections constructively interfered, overpredicting modal forces
Multi-row Aeromechanical and Aeroelastic Aspects of Embedded Gas Turbine Compressor Rotors · DukeSpace
Tried and failed
harmonic balance unsteady fluid simulations applied to turbomachinery stator vane aerodynamics. Outcome: did not converge. Reason: numerical solver failed to converge at specific operating condition
Aeroacoustic Scaling of Turbomachinery Blade Rows · Leibniz Universität Hannover Repository
Tried and failed
Rhie-Chow interpolation with drag terms in Navier-Stokes applied to porous media fluid flow simulation. Outcome: unstable. Reason: numerical instabilities prevented steady state convergence in complex 3D geometry
Nuclear thermal hydraulics analysis using devito’s code generation framework · Imperial
Neglect of structural compliance and fluid-solid coupling in biological flow simulations
Simulating pulsatile flow through biological vessels using rigid walls or plug flow assumptions overestimates peak velocities, wall shear stresses, and viscous energy losses. Purely Eulerian wall boundaries also fail to capture dynamic wall compliance and longitudinal strain, whereas unaccelerated iterative coupling methods require excessive iterations.
Tried and failed
rigid wall computational fluid dynamics applied to pulsatile arterial blood flow. Reason: neglecting wall compliance significantly overestimates peak flow velocity and wall shear stress
Tried and failed
low-resolution laminar flow simulation applied to pulsatile cardiovascular fluid dynamics. Outcome: worse than baseline. Reason: neglecting turbulent dissipation and phase-averaging severely overpredicts derived viscous energy loss metrics
Assessing turbulence in patient-specific aortas with aortic valve disease and valve prostheses · Imperial
Tried and failed
rigid-wall computational fluid dynamics simulation applied to compliant multi-lumen vessel flow. Outcome: did not generalise. Reason: neglecting dynamic wall compliance prevents accurate prediction of cross-luminal pressure gradients
Computational analysis of hemodynamics and thrombosis in aortic dissection for clinical applications · Imperial
Considered and rejected
Considered and rejected: Rejected using purely Eulerian velocity boundary conditions (V_wall(x,t)) for deforming elastic channels because they fail to capture finite longitudinal strain, material periodicity, and fluid-structure coupling.
Lost to a baseline
Block-Gauß-Seidel iterative method required significantly more iterations than inexact block-Newton and Quasi-Newton (BFGS) methods in coupled fluid-structure interactions.
Cooperative and multirate simulation: analysis, classification and new hierarchical approaches · DSpace-CRIS at TU Wien
Considered and rejected
Considered and rejected: Rejected plug flow assumption in large vessels (arteries/veins) because fluid flow follows Navier-Stokes equations closer to parabolic flow, which produced 22-fold lower error.
Left open by the authors
Problems the authors named and did not get to.
Left open
Validate computational fluid dynamics simulations of mechanical circulatory support using an optimization function for boundary conditions. Blocker: Requires unpublished experimental flow and pressure data from the thesis's physical mock loop and porcine models
Optimization of Hybrid Hemodynamics from Mechanical Support Devices in Cardiogenic Shock · MIT
Left open
Integrate a dynamic growth model with finite element fluid-structure interaction simulations to model cardiovascular tissue remodeling dynamics. Blocker: Lacks specific target growth equations, constitutive parameters, and validation data from the thesis
Soft Robotic Platforms for the Simulation of Cardiovascular Disease and Device Development · MIT
Left open
Determine porosity coefficients and improve pressure drop calculation methods for generalized brain stent fluid dynamics simulations. Blocker: Lacks specific methodology or criteria for calibrating generalized porosity coefficients without empirical validation data
Fluid-structure interaction simulations for medical applications · Texas Tech
Left open
Evaluate unsteady aerodynamic load distributions under distributed bleed using the Shih et al. vorticity balance formulation with measured or simulated flow fields. Blocker: Requires experimental wind tunnel flow field data (PIV) or high-fidelity fluid-structure interaction simulations specific to the flexible wing bleed setup.
Flow-Interactive Control of a Flexible Wing Using Aerodynamic Distributed Bleed Actuation · Georgia Tech
Left open
Develop a fluid-structure interaction model for coronary artery biomechanics to simulate wall compliance beyond rigid-wall computational fluid dynamics. Blocker: Requires specific arterial geometry, tissue mechanical properties, and hemodynamic boundary condition data from the pig study
Left open
Validate the numerical solver accuracy by comparing simulation results against additional experimental data under complex hydrodynamic conditions. Blocker: No specific experimental benchmark datasets or experimental test facilities specified
Left open
Implement fluid-structure interaction simulations on the tricuspid valve finite element model to quantify regurgitant flow volume. Blocker: Requires the private subject-specific tricuspid valve finite element mesh and experimental material data from the thesis
Towards understanding tricuspid valve mechanics & function in health, disease, and repair · UT Austin
Left open
Validate multiphase structured packing CFD simulations against experimental multiphase hydrodynamics and mass transfer data. Blocker: Requires experimental multiphase mass transfer/hydrodynamic laboratory data from structured packing rigs
Predictions of the mass transfer in structured packings using computational fluid dynamics · UT Austin
Left open
Implement two-phase fluid-solid-interaction simulations incorporating blood cells and plasma for embryonic zebrafish cardiac mechanics. Blocker: Requires private 4D light-sheet/confocal microscopy geometry and tissue motion tracking data from the thesis lab
Left open
Model and test vertical capillary imbibition with graphene-oxide and sickled blood cell suspensions coupled to numerical multi-phase flow simulations. Blocker: Requires wet lab experiments with graphene-oxide, sickled blood cells, and microfluidics apparatus for validation
Flow of colloidal and living suspensions in confined geometries · Harvard
Checking a claim in this area?
We can run the same search on any method or claim. If nothing turns up, we will say so, and that proves nothing on its own.