Abstract
Unsteady, 3D particle tracking velocimetry (PTV) data are applied as an inlet boundary condition in a direct numerical simulation (DNS). The considered flow case is a zero pressure gradient (ZPG) turbulent boundary layer (TBL) flow over a flat plate. The study investigates the agreement between the experimentally measured flow field and its simulated counterpart with a hybrid 3D inlet region. The DNS field inherits a diminishing contribution from the experimental field within the 3D inlet region, after which it is free to spatially evolve. Since the measurement does not necessarily provide a spectrally complete description of the turbulent field, the spectral recovery of the flow field is analyzed as the TBL evolves. The study summarizes the pre-processing methodology used to bring the experimental data into a form usable by the DNS as well as the numerical method used for simulation. Spectral and mean flow analysis of the DNS results show that turbulent structures with a characteristic length on the order of one average tracer particle nearest neighbor radius {bar{r}}_{text {NN}} or greater are well reproduced and stay correlated to the experimental field downstream of the hybrid inlet. For turbulent scales smaller than {bar{r}}_{text {NN}}, where experimental data are sparse, a relatively quick redevelopment of previously unresolved turbulent energy is seen. The results of the study indicate applicability of the approach to future DNS studies in which specific upstream or far field boundary conditions (BCs) are required and may provide the utility of decreasing high initialization costs associated with conventional inlet BCs.Graphic abstract
Highlights
The relative advantages and disadvantages of computational fluid dynamics (CFD) and experimental fluid dynamics (EFD) as they pertain to the study of turbulent flows are well understood
The test section was seeded with Helium filled soap bubbles (HFSBs), which were illuminated by 10 LED array panels.Images were captured by a total of 12 high-speed cameras, 4 of which were focused primarily on the aft-most half of the zero pressure gradient (ZPG) region, having a resolution of 4096 × 2160 pixels per camera and capable of producing 1000 exposures per second
The dimensional and non-dimensional parameters of the experiment and direct numerical simulation (DNS) are listed in Table 1.The defined inlet hydrodynamic boundary layer thickness Reynolds number Reδ99 ≈43e3 is kept constant, the freestream Mach number M∞ of the dimensionless problem is increased to M∞=0.85 to reduce the simulation cost.Compared to an equivalent simulation at M∞=0.3, the computational effort at M∞=0.85 is reduced by a factor of roughly three due to the less strict stability constraints on the timestep Δtmax related to the discretization of the energy equation, see Babucke (2009)
Summary
The relative advantages and disadvantages of computational fluid dynamics (CFD) and experimental fluid dynamics (EFD) as they pertain to the study of turbulent flows are well understood. Hybrid methods refer to techniques seeking to leverage the relative advantages of EFD and CFD for combinations of various purposes, including noise reduction through assimilation to governing equations, the derivation of unknown field quantities, resolution enhancement, and the application of highly specific boundary conditions (Suzuki and Yamamoto 2015). Such goals are by no means separate, and the pursuit of one of these objectives often implicitly involves others. The hybrid studies referenced employ advanced coupling schemes that are at least partially active within a complete subdomain of interest
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
