Kevin J. Maki

RANS Simulation Versus Experiments of Flow Induced Motion of Circular Cylinder With Passive Turbulence Control at 35,000 < RE < 130,000

Two-dimensional (2D) Unsteady Reynolds-Averaged Navier–Stokes equations (URANS) equations with the Spalart–Allmaras turbulence model are used to simulate the flow and body kinematics of the transverse motion of spring-mounted circular cylinder. The flow is in the high-lift TrSL3 regime of a Reynolds number in the range 35,000 < Re < 130,000. Passive turbulence control (PTC) in the form of selectively distributed surface roughness is used to alter the cylinder flow induced motion (FIM). Simulation is performed using a solver based on the open source Computational Fluid Dynamics (CFD) tool OpenFOAM, which solves continuum mechanics problems with a finite-volume discretization method. Roughness parameters of PTC are chosen based on tests conducted in the Marine Renewable Energy Lab (MRELab) of the University of Michigan. The numerical tool is first tested on smooth cylinder in vortex-induced vibration (VIV) and results are compared with available experimental measurements and URANS simulations. For the cylinder with PTC cases, the sandpaper grit on the cylinder wall is modeled as a rough-wall boundary condition. Two sets of cases with different system parameters (spring, damping) are simulated and the results are compared with experimental data measured in the MRELab. The amplitude ratio curve shows clearly three different branches, including the VIV initial and upper branches, and a galloping branch. The numerical branches are similar to those observed experimentally. Frequency ratio, vortex patterns, transitional behavior, and lift are also predicted well for PTC cylinders at such high Reynolds numbers.

Transom-stern Flow for High-speed Craft

Abstract Two series of experiments have been conducted, one at the University of Michigan (U-M), and one by The University of New South Wales (UNSW), with a focus to characterize the flow in the transom region of a high-speed vessel. At U-M, we have tested a destroyer type model, with and without a stern flap, while measuring pressures in the aft region of the hull and on the flap. The model was tested in both the free-to-sinkand-trim condition and the fixed condition. At UNSW, a series of geosimilar models was tested while measuring the free-surface elevation behind the vessel. The non-dimensional free-surface elevation was found to be primarily a function of the calm-water-transom-draft Froude number. To this end, an empirical formula that estimates the unwetting of the transom has been developed. This formula can be employed in a resistance prediction computer program which will provide an accurate calculation of the hydrostatic force on the transom. As a consequence, the total resistance of the vessel can now be computed accurately, even in the low-Froude-number region.

Flow separation over a backward-facing ramp with and without a vortex generator

Large Eddy Simulation (LES) and Unsteady Reynolds-Averaged Navier-Stokes (URANS) calculations are performed for flow over a 25◦ backward-facing ramp with and without a passive vortex generator (VG). The objective of this work is to determine the influence of the VG on the separated flow region. The LES study is performed with a synthetic time-varying turbulent inflow boundary condition whereas a mean turbulent velocity inlet is used for URANS. The wall layer is resolved in the LES and these results are used as a baseline to evaluate the accuracy of the URANS calculations. A dynamic one equation model is used for LES while the k − ω SST model is used for URANS. Upstream of the ramp edge the URANS simulations lack the near wall turbulence structures observed in LES due to the synthetic inflow. Therefore URANS calculations overpredict the reattachment length as compared to LES with and without the VG. Based on the turbulent kinetic energy profiles the separation tends to induce more turbulence away from the wall and decrease it close to the wall. The effect of the VG is to entrain momentum from the mean flow to the near wall region, therefore reducing the reattachment length. Proper Orthogonal Decomposition (POD) is used to identify the dominant modes in separation region with and without the VG.

Dynamic Response of a Marine Vessel Due to Wave-Induced Slamming

This paper presents simulation results of the structural response of a surface vessel advancing in head seas. In this particular fluid-structure interaction problem, the discretized geometry in the fluid domain is significantly different than that of the structuralmodel. It is therefore necessary to use interpolation to transfer information between the two domains. In this paper we discuss a numerical procedure to obtain the dynamic response of a marine vessel based on fluid-structure interaction as applied to the test case of the S175 container ship. The vessel is advancing in head seas, and the sea conditions result in bow-flare and bottom impact slamming. The fluid and structure interact in a one-way coupled scheme where the fluid stresses are applied to the structural modal model. The simulation results are compared to previously published experimental data.

Flow control using passive vortex generators

Passive vortex generators (VGs) can be employed to control boundary layer separation effectively. As the boundary layer encounters an adverse pressure gradient, it tends to separate from the surface due to a momentum deficit. Passive VGs entrain high momentum fluid from the mean flow and enhance mixing which energizes the near-wall flow. In this study the performance of a single cube used as a passive VG to control flow separation on a backward-facing ramp is evaluated using wall-resolved large-eddy simulations. The Reynolds number of the flow is 19,600. Since our interests lie in low-profile VGs within a turbulent boundary layer, we evaluate cube height to boundary layer thickness ratio of 0.6 to understand the effect of the physical VG on the separation region.

A Velocity Decomposition Approach for Unsteady External Flow

In this work we address the development of the velocity decomposition algorithm, a numerical flow solution method that incorporates both velocity potential and Navier-Stokes-based solution procedures. The motivation for this is so that the field discretization required by the Navier-Stokes solver can be reduced to the region of the flow domain in which the flow is vortical. Specific advantages are that the computational cost is reduced, it is easier to discretize the flow domain, and difficult problems such as the simulation of ships maneuvering in a seaway are closer to being within reach. The target applications are broad, ranging from vortex shedding on bluff objects such as risers, to the wave induced loads on a platform in a current and irregular seas. In previous work, the algorithm has been successfully applied to address steady flows of 3-D non-lifting bodies without water waves, or 2-D bodies that can have lift and be near a water surface. In this paper, the velocity decomposition approach is extended to numerically solve for the unsteady flow of single-phase viscous flows. The velocity vector is decomposed into irrotational and vortical components. A boundary element method is used to solve for the irrotational component (designated as the viscous potential) by applying a viscous boundary condition to the body boundary. A field method is used to solve for the total velocity on a reduced domain where the flow is vortical. The new algorithm investigates two approaches to solve the unsteady problem based on different types of time-dependence exhibited by the solution. The unsteady velocity decomposition method is demonstrated on two cases, and the solutions are compared to those generated by a conventional viscous flow solver. The results by the new algorithm agree well with the benchmark solutions and exhibit a reduction in time.Copyright

Joint High Speed Sealift (JHSS) Segmented Model Test Results

The Joint High Speed Sealift segmented model (Model 5663) tests performed in 2007, in the Maneuvering and Seakeeping Basin at the Naval Surface Warfare Center, Carderock Division, were designed to provide a large data set for validation of numerical simulations. Model 5663 is a segmented structural ship model that has scaled longitudinal bending and torsional stiffness. The scaled stiffness is obtained by building a backspline into the model that accounts for the bending stiffness and cutting the shell in several places, segmenting it to isolate the stiffness to the backspline. The alternative way to obtain structural loads would be to build a model with scaled plates and stiffeners; however, this would be very difficult and expensive. The backspline allows the stiffness to be scaled properly while using reasonable materials and simple construction. The hull segments of the model are connected with silicone to maintain a watertight connection. The model is self propelled and steered during data collection. The test matrix spans a wide set of wave conditions, including regular and irregular seas, with heading angles spanning the possible range. A wide range of speeds are also included, with Froude numbers ranging from 0 to 0.43. This test matrix, which includes about 2000 runs, allows for validation of codes from still water test, through operational conditions, to extreme design load determination. Different aspects of the data have been studied, but much is still left to be considered. An aspect of the model tests that has not previously been considered in detail is the hydroelastic response of the vessel. Hydroelastic phenomena are a subset of fluid-structure interaction problems where the elasticity of the structure is important. The vibrational characteristics of the model are determined. The main phenomena of interest are springing and whipping, and an analysis of the springing response and the whipping response in head seas is also discussed.Copyright

Aerothermal Optimization of Internal Cooling Passages Using a Discrete Adjoint Method

Aerothermal optimization is a powerful technique for the design of internal cooling passages because it maximizes heat transfer and simultaneously minimizes pressure loss. Moreover, the optimization is fully automatic, which reduces the duration of design process compared with a human-supervised design approach. Existing optimization studies commonly rely on gradient-free methods, which can only handle a small number of design variables. However, cooling passage designs use complex geometry configurations (e.g., serpentine channels with rib-roughened surfaces) to enhance heat transfer; what is needed is to parameterize the passage using a large number of design variables. To address this need, we perform aerothermal optimization using a gradient-based optimization algorithm along with the discrete adjoint method to compute derivatives. The benefit of using the adjoint method is that its computational cost is independent of the number of design variables. In this paper, we focus on optimizing a U-bend duct, which is representative of a simplified, rib-free turbine internal cooling passage. The duct geometry is parameterized using 135 design variables, which gives us sufficient design freedom for geometric modification. We construct a Pareto front for heat transfer enhancement and total pressure loss reduction by running multi-objective optimizations. We also compare our optimization results with those from the gradient-free methods and demonstrate that we achieve better pressure loss reduction and heat transfer enhancement. The above results show that our gradient-based optimization framework functions as desired and has the potential to be a useful tool for turbine aerothermal designs with full internal cooling configurations.

Understanding the effect of cube size on the near wake characteristics in a turbulent boundary layer

Wall-resolved large-eddy simulation of flow over a wall-mounted cube-shaped obstacle placed in a spatially evolving boundary layer is performed. The Reynolds number of the flow based on the mean velocity and the inlet boundary layer thickness is 19,600. Our interests lie in understanding how variations in the cube height h modify the flow dynamics for situation where the cube is within the boundary layer. For this purpose, we conduct simulations with height to boundary layer thickness ratios of h/δo = 0.2, 0.6 and 1.0, where δo is the boundary layer thickness at the cube location. The size of the vortices formed around the cube, the Reynolds stresses and the turbulent kinetic energy in the boundary layer all depend on the cube height relative to the boundary layer thickness. The drag coefficient of the cube is found to increase with increasing h/δo.

Structural vibration of an elastically supported plate due to excitation of a turbulent boundary layer

High-Reynolds number turbulent boundary layers are an important source for noise and structural vibration. Indeed small features of a structure can have important influence on the resulting noise and vibration. In this work we develop a new method to couple a high-fidelity fluid solver with a dynamic-finite-element solver for the structure. The flow solver is based on the OpenFOAM opensource CFD toolkit. The fluid solver uses the Large-Eddy Simulation closure for the unresolved turbulence. Specifically, a local and dynamic one-equation eddy viscosity model is employed. The fluid pressure fluctuation on the structure is mapped to the dynamic finite element model with a suitable interpolation routine. A modal decomposition of the dynamic finite-element model is used to reduce the number of degrees-of-freedom of the numerical structural model. The numerical method is validated for the turbulent flow over a flat plate. The plate is elastically supported along its edges, and the turbulence is excited upstream ...