Browsing by Subject "cfd"
Now showing 1 - 3 of 3
Results Per Page
Sort Options
Item A numerical method for fully nonlinear aeroelastic analysis(2009-05-15) Gargoloff, Joaquin IvanThis work presents a numerical method for the analysis of fully nonlinear aeroelastic problems. The aeroelastic model consisted of a Navier-Stokes flow solver, a nonlinear structural model, and a solution methodology that assured synchronous interaction between the nonlinear structure and the fluid flow. The flow around the deforming wing was modeled as unsteady, compressible and viscous using the Reynolds-averaged Navier-Stokes (RANS) equations. To reduce the computational time, a three-level multigrid algorithm was implemented and the flow solver was parallelized. The message-passing interface (MPI) standard libraries were used for the parallel interprocessor communication. The computational domain was divided into topologically identical layers that spanned from the root to past the tip of the wing. A novel mesh deformation algorithm was developed to deform the mesh as the structure of the wing was being displaced. The mesh deformation algorithm was able to handle wing tip deformations of up to 60 % of the wing semi-span. Besides being robust, the mesh deforming algorithm was computationally more efficient than regriding, since deforming an existing mesh was computationally less expensive than generating a new mesh for each wing position. Results are presented for the validation and verification of both the flow solver and the aeroelastic solver. The flow solver was validated using: (1) the flow over a flat plate, to validate the turbulent model implementation, and (2) the flow over the NACA 0012 airfoil and over the F-5 wing, to validate the implementation of the convective and viscous fluxes, the time integration algorithm, and the boundary conditions. The aeroelastic solver was validated using: (1) the unsteady F-5 wing undergoing forced pitch motion, and (2) the Nonlinear Aeroelastic Test Apparatus (NATA) wing. In addition, aeroelastic results were generated for the Goland wing. The aeroelastic solver developed herein allows the analysis of aeroelastic phenomena using a fully nonlinear approach. Limit cycle oscillations, which are highly nonlinear phenomena, were captured by the nonlinearities of the flow solver and the structural solver. The impact of the nonlinearities was assessed for the Goland wing, where nonlinear terms changed dramatically the aeroelastic behavior of the wing.Item Direct Numerical Simulation of the Flow in a Pebble Bed(2014-06-24) Ward, PaulThe flow in a tightly packed array of spheres is important to various engineering fields. In nuclear engineering applications, for instance, researchers have proposed core geometries of the pebble bed reactor (PBR) type cooled by gas or molten salt. Proper core cooling, both at operation and during accident conditions, is a key issue that must be addressed in any reactor design; and the limited amount of data available for the complicated geometry of PBR cores makes this task even more complex. A detailed understanding of coolant flow patterns and properties must be developed in order to meet safety requirements and ensure core longevity. We addressed this issue by using the spectral-element computational fluid dynamics code Nek5000, developed at Argonne National Laboratory, to conduct both large eddy simulation (LES) and direct numerical simulation (DNS) of fluid flow through a single face-centered cubic sphere lattice with periodic boundary conditions. Multiple LES were conducted with varying Reynolds numbers in an effort to determine how the Reynolds number affects the development of asymmetries within the flow patterns. The DNS focused on the development of turbulence and were used to compute the turbulent kinetic energy budgets. A set of statistical analyses were also conducted to support the validity of the results.Item Performance Evaluation and CFD Simulation of Multiphase Twin-Screw Pumps(2013-05-16) Patil, AbhayTwin-screw pumps are economical alternatives to the conventional multiphase system and are increasingly used in the oil and gas industry due to their versatility in transferring the multiphase mixture with varying Gas Void Fraction (GVF). Present work focuses on the experimental and numerical analysis of twin-screw pumps for different operating conditions. Experimental evaluation aims to understand steady state and transient behavior of twin-screw pumps. Detailed steady state evaluation helped form better understanding of twin-screw pumps under different operating conditions. A comparative study of twin-screw pumps and compressors contradicted the common belief that compressor efficiency is better than the efficiency of twin-screw pumps. Transient analysis at high GVF helped incorporate necessary changes in the design of sealflush recirculation loop to improve the efficiency of the pump. The effect of viscosity of the sealflush fluid at high GVF on pump performance was studied. Volumetric efficiency was found to be decreased with increase in viscosity. Flow visualization was aimed to characterize phase distribution along cavities and clearances at low to high GVF. Dynamic pressure variation was studied along the axis of the screw which helped correlate the GVF, velocity and pressure distribution. Complicated fluid flow behavior due to enclosed fluid pockets and interconnecting clearances makes it difficult to numerically simulate the pump. Hence design optimization and performance prediction incorporates only analytical approach and experimental evaluation. Current work represents an attempt to numerically simulate a multiphase twin-screw pump as a whole. Single phase 3D CFD simulation was performed for different pressure rise. The pressure and velocity profile agreed well with previous studies. Results are validated using an analytical approach as well as experimental data. A two-phase CFD simulation was performed for 50% GVF. An Eulerian approach was employed to evaluate multiphase flow behavior. Pressure, velocity, temperature and GVF distributions were successfully predicted using CFD simulation. Bubble size was found to be most dominant parameter, significantly affecting phase separation and leakage flow rate. Better phase separation was realized with increased bubble size, which resulted in decrease in leakage flow rate. CFD results agreed well with experimental data for the bubble size higher than 0.08 mm.