My research interests lie in the area of fluid mechanics, particularly with respect to computational fluid dynamics (CFD). The objective of my research is to develop and implement the state-of-the-art CFD technologies to solve both fundamental and practical engineering problems, with specially focus on energy sustainability and new methodology and procedure for quantitative verification and validation for numerical simulations. My research involves large-scale scientific parallel computations that require up to thousands of processors. Thanks to the support from University of Idaho (UI) and Idaho National Laboratory (INL), sufficient computing resources have been supplied to me both locally and remotely. Local computing resources include a Dell Precision T7500 with 12 cores and 48 GB RAM and an adaptive computation server that has 160 cores and 4TB RAM (purchased using the NSF MRI grant). The server has 160 cores and 4 Terabyte RAM. I also used the startup fund from UI to purchase state-of-the-art CFD software for my graduate students and myself such as Pointwise, Tecplot, and ANSYS FLUENT. Remote resource includes supercomputer at INL provided to us for free. The INL machines enable us to run very large-scale CFD computations using thousands of processors. My research has been and will continue to be boosted by these resources.
My research activities have been closely linked to research projects funded by the U.S. National Science Foundation, Office of Naval Research, Vorsana Inc., and University of Idaho. A few completed (2009-2011 before I joined UI) or ongoing (2011 – present) areas of research in which I have contributed are:
CFD for Onshore Wind Turbines: wind energy provides a clean alternative to fossil fuel electricity production. However, for the design and development of more efficient and reliable wind turbines, accurate prediction of aerodynamic behavior is of critical significance, since the interaction of the wind with the blades influences the efficiency. It also has a significant effect on the loads on bearings and gearbox, ultimately affecting the lifespan and reliability of the machine. The Computational methods are appropriate tools to resolve these challenges, but developments and improvements to current capabilities are needed. Most approaches omit fluid viscosity, simplify blade and tower geometries, and lack sufficient grid resolutions, severely limiting the utilization for offshore machines. Since I joined Tuskegee University in 2009, I started thinking of extend CFD to the area of onshore wind turbines. With collaboration with The University of Iowa, we developed a high-performance CFD tool capable of simulating the onshore wind turbine flows. The tool is built on top of the general-purpose CFD code CFDShip-Iowa, which uses state-of-the-art technology and has been validated for many applications in ship hydrodynamics. It has been applied for onshore wind turbines for a wide range of conditions with good agreement between predictions and experimental data.
(produced by Yuwei Li)
CFD for Offshore Wind Turbines: right before I joined UI in 2011, I was awarded the NSF award on “Collaborative Research: Simulation Based Design for Deep Water Offshore Wind Turbines Including Wave Loads and Motions.” Majority of the grant has been sub-awarded to UI after I joined UI, which is used to support a full-time Ph.D. student. Offshore wind turbines have unique advantages over onshore wind turbines such as steadier and higher air velocity, and larger sizes. Although deepwater wind resources in the United States are abundant, there are no commercial wind facilities operating off the coasts. Technologies used in Europe are restricted to fixed-bottom in shallow water and thus inapplicable in the US. Several floating turbine concepts exist, but methodologies accounting for wind inflow, aerodynamics, elasticity and control of the turbine, along with the incident waves, sea currents, hydrodynamics, and platform and mooring dynamics of the floater are needed to determine technical and economic feasibility. This project aims to extend the CFD tool validated for onshore wind turbine flows to study the aero-hydro-elastic problem of offshore wind turbines. So far, we have developed a new mooring line (crowfoot) model with a system simulation using a coarse grid for a floating wind turbine in regular waves.
(Produced by Sean Quallen)
Verification & Validation: quantitative verification of numerical methods and validation of models are very important to evaluate the quality of a simulation. However, the best V&V methodology and procedure is not achieved mainly due to the assumptions made by the methods and lack of statistical evidence. Recently, I developed a new factor of safety method (Xing and Stern, 2010, 2011) based on statistical analysis, which has shown advantages over the Grid Convergence Index (GCI) method that has been widely accepted and recommended by ASME and AIAA. Since the publication of this method in 2010, it has been widely applied in ship hydrodynamics and used by researchers from the U.S., Italy, and Swed
CFD Modeling of the Vorsana McCutchen Processor: This was a pilot study sponsored by the Vorsana Inc. The McCutchen processor can be used to separate mixture flows. My long term objective is to use CFD to simulate three-dimensional flow in the Vorsana processors (system simulations). Nonetheless, the flows are very complicated, i.e., two-phase or multi-phase between two counter-rotating disks and one stationary tank. To completely solve the problem will require the use of the most advanced state-of-the-art CFD technologies including multiple reference frames, two-/multi- phase models, dynamic/sliding meshes, and turbulence models, etc. To reduce the risk for Vorsana Inc.’s investments, it was suggested that a pilot study be conducted by myself in the summer of 2012. The objective of the pilot study is to implement and validate CFD models for flows between two counter-rotating disks, including both qualitative and quantitative validation for individual CFD model using available experimental data, which will prepare my team for the real system simulation in the future. Gradually adding complexity is most likely robust and dependable strategy to build a simulation tool that can effectively be employed for industrial applications.
Numerical Simulation of Magnetic Nanoparticles Using Electromagnetic Separation Device: This study is supported by the seed grant from UI. This project aims to use computational fluid dynamics (CFD) to design and optimize the magnetic separation system used in spent nuclear fuel recycling. The project will develop advanced models for both the fluid and particles such that the accuracy of predicting the removal efficiency of particles will be significantly improved. The model will be validated using data measured by a research team led by Dr. You Qiang at the Department of Physics, University of Idaho, and other experimental data when available. The validated model will be used to elucidate the best combination of flow rate, concentration, magnetic field gradient, and types of particles to achieve the highest removal efficiency of particles. The research results/findings will be published in refereed Journals and conference proceedings. The model can be easily transformed to be used for many disciplines such as nuclear engineering, chemical engineering, and mineral engineering.
CFD Modeling for Vortical and Turbulent Structures for Ships with Large Drift Angles: DES on a 13M grid is conducted for a KVLCC2 tanker hull form at drift angle 30 degrees. The objective is to identify and analyze the vortical structures, instabilities, and turbulent structures with analogy to the vortex breakdown and helical instability analysis for delta wing at large angles of attack. The simulation results were used to guide the experimental study performed at Hamburg University of Technology. Their very recent measurements confirmed my CFD results.
CFD Modeling for Entropy Generation in Bypass Transitional Boundary Layer Flows (funded by DOE EPSCOR): The primary objective of this study is to evaluate the accuracy of using computational fluid dynamics (CFD) turbulence models to predict entropy generation rates in bypass transitional boundary layers flows under zero and adverse pressure gradients. Entropy generation rates in such flows are evaluated employing the commercial CFD software, ANSYS FLUENT. Various turbulence and transitional models are assessed by comparing their results with the direct numerical simulation (DNS) data and two recent CFD studies. A solution verification study is conducted on three systematically refined meshes. The factor of safety method is used to estimate the numerical error and grid uncertainties. Monotonic convergence is achieved for all simulations. The Reynolds number based on momentum thickness, ReΘ, skin-friction coefficient, Cf, approximate entropy generation rates, S''' , dissipation coefficient, Cd, and the intermittency, γ, are calculated for bypass transition simulations. All Reynolds averaged Navier-Stokes (RANS) turbulence and transitional models show improvement over previous CFD results in predicting onset of transition. The transition SST k -ω 4 equation model shows closest agreement with DNS data for all flow conditions in this study due to a much finer grid and more accurate inlet boundary conditions. The other RANS models predict an early onset of transition and higher boundary layer entropy generation rates than the DNS shows.