Figure 4 - uploaded by Dmitry A. Lysenko
Content may be subject to copyright.
Unstructured triangular mesh (a) and 40 equally spaced contours of log density (b). 2D computational domain is chosen in such a way to replicate the conditions of TST-27 wind tunnel (Figure 4a). The length and the width of domain are: 1 m and 0.27 m, respectively. The wedge-plate bluff body is placed at the distance of 0.3 m from the inlet (left) boundary. The unstructured triangular mesh with total number cells of 171 K (~ 90K nodes) is used. The size of elements for bluff-body walls is chosen as 0.5 mm. Outer domain boundaries are meshed with the element's size of ~7 mm with the purpose to provide the smooth transition in the grid cell sizes. The viscous structured sublayer is applied also to the walls of the wedge-plate body (first row size-0.01 mm, growth factor-1.5 and number of rows-6). Thus the maximum Y+ value is received as 8.83. The general view of the computational domain with triangular grid is presented on Figure 4a.

Unstructured triangular mesh (a) and 40 equally spaced contours of log density (b). 2D computational domain is chosen in such a way to replicate the conditions of TST-27 wind tunnel (Figure 4a). The length and the width of domain are: 1 m and 0.27 m, respectively. The wedge-plate bluff body is placed at the distance of 0.3 m from the inlet (left) boundary. The unstructured triangular mesh with total number cells of 171 K (~ 90K nodes) is used. The size of elements for bluff-body walls is chosen as 0.5 mm. Outer domain boundaries are meshed with the element's size of ~7 mm with the purpose to provide the smooth transition in the grid cell sizes. The viscous structured sublayer is applied also to the walls of the wedge-plate body (first row size-0.01 mm, growth factor-1.5 and number of rows-6). Thus the maximum Y+ value is received as 8.83. The general view of the computational domain with triangular grid is presented on Figure 4a.

Source publication
Article
Full-text available
Four state-of-the-art applications of unstructured tetrahedral meshes (both static and dynamic- adaptive) in numerical simulations of different fluid dynamics phenomena are presented. The numerical methods are varied form simple Euler equations through unsteady Reynolds averaged equations to filtered Navie-Stokes equations. Meshes practically of al...

Context in source publication

Context 1
... of the gas, such as heat capacity and viscosity are modeled using functional dependences of temperature 8 (polynomial and Sutherland laws, respectively). Figure 4b. The more detailed flow visualization near the bluff- body is presented in Figure 5. Schlieren image obtained in experiment (Figure 5a) is compared to numerically predicted density field (Figure 5b). ...

Similar publications

Conference Paper
Full-text available
Increasing demands on the capabilities of engine thermo-fluid dynamic simulation and the ability to accurately predict both performance and acoustics has lead to the development of several approaches, ranging from fully 3D to simplified 1D models. The 3D cell approach is proposed as a compromise between the time demanding 3D CFD analysis and the fa...
Article
Full-text available
The gradient calculation for the reconstruction of dependent variables is one of the critical issues for the accuracy and robustness of computational fluid dynamics (CFD) methods. If this condition is not satisfied, unphysical new extrema can be generated, leading to the blowup of the flow computation. The geometrical monotonicity condition, which...