Code Development of Linearized Euler Equation on Block-Structured Cartesian Mesh for Complicated Geometries

Abstract

The Linearized Euler Equation code for aeroacoustic analysis has been developed on block-structured Cartesian mesh to treat the complicated geometries robustly and accurately. In the present method, spatial derivation and time integration are conducted by high order schemes and the immersed boundary method using ghost cell approach is employed at the wall boundary. At the block boundary, higher-order data exchange is conducted by Lagrange interpolation and near the outer boundary outgoing wave is damped by absorbing boundary domain. The present code is validated for two-dimensional and three-dimensional test cases. The accuracy and effectiveness of the present method are demonstrated in two-dimensional acoustic scattering problems. In the three-dimensional cases, acoustic scattering around a sphere and noise propagation from JT15D nacelle configuration were simulated and it showed the good agreement with the analytical and experimental results. Finally, noise shielding effect of an Over-the-Wing-mounted-Nacelle (OWN) configuration was estimated as an application. The results show that far-field SPL of OWN configuration is lower than that of conventional Under-the-Wing-mounted-Nacelle configuration by about 10dB due to the shielding effect of OWN configuration.

Wavelet‐based data compression for flow simulation on block‐structured Cartesian mesh

“…Here, the vorticity magnitude is considered as the characteristic value i in Equation (2). The large value of gives small quantization step size in Equation (6), and small error as shown in Figure 7(b). The contour lines from the compressed data and the original data represent similar features, which means that the compressed data can preserve a complex wake flow structure behind the sphere.…”

“…Here, it is shown that the value of in Equation (2) is large in the wake region (Figure 7(a)), where strong vortices exist. The large value of gives small quantization step size in Equation (6), and small error as shown in Figure 7(b). This results in a good preservation property for the characteristic vortex structure.…”

“…These problems emerged in large-scale simulations implemented by the BCM, that is, the incompressible unsteady flow simulation around a landing gear model with approximately 100m mesh points [5], and the aeroacoustic analysis around an over-the-wing-mounted-nacelle configuration with approximately 220m mesh points [6]. The performance development of supercomputers enables high-fidelity flow simulations with a large number of mesh points, which give a large amount of flow simulation data.…”

“…A large amount of data are difficult to treat in the post-processing, and it also requires much time for data transfer and much space for data storage. These problems emerged in large-scale simulations implemented by the BCM, that is, the incompressible unsteady flow simulation around a landing gear model with approximately 100m mesh points [5], and the aeroacoustic analysis around an over-the-wing-mounted-nacelle configuration with approximately 220m mesh points [6].…”