Session U29: CFD: Applications

Mode-C transition to three-dimensionality in the wake behind a circular cylinder undergoing 2-DOF VIV

We have performed the three-dimensional numerical simulations using the open-source Finite Volume-based solver of OpenFOAM. In the simulations, the 2-DOF VIV system is modelled by solving the spring-damper equations allowing the movement of the cylinder in both the inline and cross-flow directions. The cylinder is 9.6 times the cylinder diameter (D) in length and corresponds to a low mass ratio of m^* = 2.546. The Reynolds number is varied from 100 (corresponding to fully 2D wake) to 600 (corresponding to 3D/chaotic wake) to assess the transition to three-dimensionality in the wake. The damping ratio (ζ) is 0 (to maximize the oscillation amplitude), and the velocity ratio (U_r) is chosen to be U_r = 6, which corresponds to the lock-in regime. The data are parametrized using the frequency ratio (f^* = f_nx/f_ny), which is defined as the ratio of inline frequency (f_nx) to cross-flow frequency (f_ny). We have systematically varied this ratio from zero (corresponding to a 1-DOF VIV system with transverse oscillation only) to eight. Further, to represent the 3-D vortex patterns in the cylinder wake, we have used the iso-surfaces of the second eigenvalue (λ₂) of the tensor S² + Ω², where S and Ω are the symmetric and antisymmetric parts of the velocity gradient tensor ▽u, is used to identify the vortex core. An elastically mounted circular cylinder, with the varying inline to transverse oscillation frequency ratios (f^* = f_nx/f_ny), shows a drastic change in the transition Reynolds number and the wake-mode transition to the three-dimensional wake from a fully two-dimensional wake in the lock-in regime. The 2-DOF VIV system follows Mode-C instability for the transition from 2D to 3D wake instead of Mode-B transition observed in the 1-DOF VIV system. 

    

Large eddy simulation of backdraft phenomena on a 2/5th scale compartment

Among building fires, backdraft is a phenomenon that poses great danger to occupants and first responders. It may occur when a heated and oxygen depleted closed compartment holds significant unburned fuel concentrations. When an opening to the exterior is established, a cold air current can introduce oxygen back in the enclosure. If ignition conditions are adequate, the flammable mixture can ignite producing a fireball that travels across the opening. It is of great interest to understand and simulate the detailed fire physics behind backdraft. The Fire Dynamics Simulator (FDS) is an engineering tool for simulation of fire scenarios in forensic studies and fire protection systems design. FDS solves the Low Mach approximation equations for thermally driven buoyant flows by means of Large Eddy Simulation (LES), including combustion and radiation physics.

In this talk, we will describe the work being done at NIST on FDS simulation of backdraft phenomena. We target model validation against the detailed thermal and mixture composition measurements done at the National Fire Research Laboratory on a compartment with 2/5th the dimensions of the standard ASTM fire test room (NIST Technical Note 2185). A fire model assessment on different experimental cases will be provided.

    

Blockage detection in pipeline by estimating reflected pressure signal velocity in a dissipative medium

The removal of a blockage in long oil pipelines requires accurate detection of the blockage location.  The applied pulse propagates with the acoustic speed in a non-dissipative system. However, waxy crude oil at low temperatures shows high viscosity, and hence pressure propagation speed reduces considerably. Thus a dissipative model is required to estimate the true pressure propagation speed. Once the accurate pressure signal speed is known we can find the exact location of the blockage. For estimating accurate speed, the present work utilizes a finite volume method along with volume of fluid method to simulate a multiphase system consisting of a 1 km long oil pipeline with blockage. The blockage mostly consisting of paraffin wax is considered either shear thinning or elasto-viscoplastic with a higher compressibility than free-flowing oil. The transient equations are written in a 2D cylindrical coordinate system for an axisymmetric pipeline, which resolves acoustic time scale. The discretized equations are solved iteratively for both oil and blockage. The time taken by the pulse to reflect and reach the inlet is recorded and is used to calculate the blockage location. The model is able to record accurate blockage location, with errors in calculation ranging from 0.79% to 1.65%.

    

Aggregate Loss Data Assimilation (ALDA) for Supersonic BOS

Experimental measurements of practical flows are needed to develop and validate numerical models for design. However, real fluid measurements exhibit limited resolution and often provide path-integrated or bulk information. We explore a new aggregate loss (AL) approach to data assimilation (DA): ALDA. The method is applied to high-speed flow using background-oriented schlieren data. CFD solvers often incorporate measurements using computationally-intensive DA techniques such as Kalman filter, state observer, and adjoint–variational methods. Physics-informed neural networks (PINNs) were proposed as a low-cost DA tool, but PINNs implicitly filter the flow and considerable effort is required to optimize their architecture. For complex flows, an expensive domain decomposition technique must be employed to capture the dynamically relevant scales. ALDA uses the same optimization tools as a PINN to quickly solve for flow fields that roughly satisfy the governing equations and match experimental data. However, the flow is parameterized as integral quantities in finite volumes (voxels) and the physics loss is evaluated using discrete numerical operators. The result is a more scalable and interpretable model with well-known and controllable numerical filtering properties.   

Validating the design optimisation of ultrasound-based flow meters using computational fluid dynamics and surrogate modelling

Small-diameter ultrasonic flow meters, with an intrusive two-stand configuration, present an interesting industrial internal-flow problem due to their unique geometry and complex interactions with fluid flow. In order to efficiently optimise these systems, their flow physics and operation must be accurately predicted. In this study, Design and Analysis of Computer Experiments (DACE) by computational fluid dynamics is used to predict the turbulent flow and to perform robust design optimisation of the flow meter. The optimisation is performed by surrogate modelling based on Kringing, Latin Hypercube Sampling (LHS), and a Multi-Objective Evolutionary Algorithm (MOEA), where minimisation of pressure drop and maximisation of the flow meter accuracy, are taken as objective functions. The optimisation results are shown and compared numerically and experimentally against a baseline geometry, displaying performance gains and geometrical changes in the 3D space. The applied methodology provides a robust and time-efficient framework to analyse and optimise internal-flow problems with similar features.   

Numerical investigation on heat transfer enhancement using ionic wind in a perforated micro-pin fin heat sink

Overheating of the electronic components is a major cause of device failure. Heat sinks are used to mitigate this issue which relies on the thermodynamics of conduction, convection, and radiation. Further, the heat transfer ability of heatsinks can be increased by adding fins in heat sinks, increasing the overall surface area for convection and radiation. In this modern era of miniaturization, the micro pin fins (MPFs) heatsink can be one of the solutions for enhancing heat transfer in such small-scale devices. Recently, researchers have demonstrated that using perforated MPFs enhances the heat transfer performance compared to the solid micro pin-fins due to the larger surface area to volume ratio. To further increase the performance, ionic wind generated by corona discharges has received much attention because it has distinct advantages over traditional methods. This work aims to examine the electrode configuration study to determine the effect of longitudinal electrode arrangement with a different number on the heat transfer enhancement through numerical simulation and multi-physical characteristics, including the discharge process and charged particle movements. The multi-physics simulations incorporate the interplay between the flow physics and electrodynamics of charged particles.   

A three-dimensional solver for multiphase flow with soluble surfactant beyond the critical micelle concentration

Surfactant-laden multiphase flows are of paramount importance to industrial applications in energy, manufacturing, and healthcare. Despite the attention that such flows have received, there is a dearth of computational studies that focus on systems with surfactant loading above the critical micelle concentration (CMC), beyond which surfactant monomers aggregate to form micelles, highly relevant to industrial applications. Beyond the CMC, predictive models of the system must account for flow and interfacial dynamics (including surface-tension-gradient-driven Marangoni stresses), surfactant transport in the form of monomers and aggregates in the bulk and on the interface, in addition to sorptive fluxes, and micelles mass-action kinetics. In this talk, we will present a new numerical framework able to handle surfactant-laden flows with concentrations above the CMC using our high-fidelity hybrid front-tracking and level-set approach.   

The Effect of Boundary Conditions on Numeric Simulations of Wind Tunnel Experiments

Computational simulations are often performed in conjunction with wind tunnel experiments. The choice of boundary conditions in simulations are the choice of the user. The boundary conditions in a wind tunnel, however, are prescribed by the experimental facility. In this work we look at a study aimed at reducing drag on Olympic luge sleds. This study combines computation simulation and optimization using numeric tools with experimental drag measurements to design luge sleds. We compare the computational results, and more importantly the trends in the drag values, for several different computational domains. The domains included an "open" computational domain with a ground plane, an enclosed domain matching the test section of the wind tunnel, and a domain defined by the wind tunnel contraction, test section and diffuser. These results are also then compared to the results from the wind tunnel experiments. The results indicate that the choice of computational domain and boundary conditions affect the trends in the drag value. In some cases the choice in computational domain showed decreases in drag in the computational domain while the experiments resulted in drag increases. The results highlight the importance of simulating the wind tunnel domain when comparing computation and experimental results.   

The Use of Computational Adjoint Design Processes in Optimization of Olympic Luge Sleds

The combination of CFD with adjoint design processes offers the prospect of automated optimization of aerodynamic shapes. In this work we investigate the application of an adjoint design process to the minimization of aerodynamic drag on Olympic luge sleds. We compare two approaches: 1. Using the adjoint process to automate changes in the aerodynamic shape; and 2. Using the adjoint process to inform human choice in the changes in the aerodynamic shapes. The computational results are compared to wind tunnel drag measurements. The results indicate that the automated adjoint process resulted in shapes that increased, rather than decreased, the drag in all cases. This appeared to be due to the inability of the automated process to add or remove material from the shape. Specifically, the total amount of the physical material was conserved resulting in warped versions of the original shape. The trends in the automated shape were qualitatively similar to the human designed shapes, which were shown to decrease aerodynamic drag. This indicates that automated computational adjoint design processes are currently capable of informing by not automating the design process.   

Numerical study of the flow through and around mangrove-inspired structures

This work aims to study the flow and coherent structures developed through and around different configurations of a circular array of mangrove-inspired structures, modeled as rigid cylinders, with computational fluid dynamics (CFD). We developed two-dimensional (2D) and three-dimensional (3D) simulations to predict hydrodynamic forces and to understand the role of relevant mechanisms of the flow dynamics on the near and far wake of the circular array. We perform a detailed study of the wake to analyze the evolution of the vorticity and vortex shedding in the downstream direction of the mangrove. We validate the results of the simulations with the PIV experiments performed by [1,2]. Finally, we discuss the impact of the mangrove-inspired structure layout on sediment transport and shoreline erosion. Moreover, we highlight the role and benefits of numerical simulations of simplified models in the early design stage of engineered mangroves for coastal protection.
References:
1. Kazemi, A., Castillo, L., & Curet, O. M. (2021). Mangrove roots model suggest an optimal porosity to prevent erosion. Scientific Reports, 11(1), 1-14.
2. Kazemi, A., Van de Riet, K., & Curet, O. M. (2017). Hydrodynamics of mangrove-type root models: the effect of porosity, spacing ratio and flexibility. Bioinspiration & biomimetics, 12(5), 056003.   

Design-by-Morphing: A Shape Design Methodology for Fluid Machinery

Even with years of experience, the intuition of skilled engineers often falls short when designing around fluid flows. The non-intuitive behavior of fluids can mean the optimal geometry of fluid machinery is surprising or even extreme (consider, for instance, the bulbous bow of a ship). Conventional design techniques, such as parametric or control point methods, optimize small segments of geometry and do not easily allow for the discovery of such out-of-the-box designs. These techniques are also hampered by designer bias and a high number of design parameters. Here we present Design-by-Morphing (DbM), a novel methodology for shape design, that overcomes the limitations of traditional strategies. DbM works by weighting homeomorphic (i.e., topologically equivalent) shapes with positive weights (exploration) and negative weights (extrapolation) and combining them together to create a continuous search space that can produce radical geometries from a few design parameters and without designer bias. We present our application of DbM for the shape optimization of airfoils, draft-tubes for hydrokinetic turbines, and vertical-axis wind turbines. In all cases, we show significantly improved and radical designs.   

An eXtended Discontinuous Galerkin method for three-dimensional two-phase flows: Application to large amplitude oscillations of viscous drops

We are going to present a high-order eXtended Discontinuous Galerkin (XDG) method for transient two-phase flows. The XDG method adapts the local ansatz functions to conform with the position of the interface and provides separate degrees of freedom for each phase [1,2]. This allows a sub-cell accurate approximation of the incompressible Navier-Stokes equations in their sharp-interface formulation. The interface is defined as the zero-set of a signed-distance level-set function. 

The numerical investigations focus on nonlinear axisymmetric shape oscillations of a drop in a dynamically nearly inert ambient phase. The initial deformation is given by a Legendre polynomial. We compare the numerical results with the analytical results of the weakly nonlinear theory [3]. The properties to be compared include the droplet aspect ratio over time and mode decomposition of the droplet shape. Further, we present the kinetic and surface energy over time for the numerical simulations.