A mass-conserved formulation for the Ffowcs-Williams–Hawkings (FW–H) integral is proposed to suppress contributions of spurious mass flux to the far-field sound at very low Mach numbers. The far-field condition and compact-source region assumptions are employed. By using higher-order derivatives of Green’s function, an expansion of the integrand in the monopole term is performed. This expansion transforms the mass-flux like monopole term into a series including different orders of velocity moment. At very low Mach numbers, the zero-order term is exactly the contribution from the spurious mass flux. The proposed mass-conserved formulation is confirmed by using an unsteady dipole, a two-dimensional (2D) incompressible convecting vortex, a circular-cylinder flow, and a co-rotating vortex pair. Additional spurious mass flux is added to the unsteady dipole, 2D incompressible convecting vortex, and flows over a circular cylinder; and the spurious mass flux of the co-rotating vortex pair comes from the residual of an incompressible-flow simulation. The far-field sound is found to be sensitive to spurious mass flux in the unsteady dipole and 2D incompressible convecting vortex cases. Then, the computation of the monopole-term expansion with the flow over a circular cylinder is presented. Fast convergence performance was observed, suggesting that the expansion requires little extra computational resources. Finally, FW–H boundary dependence is observed in the co-rotating vortex-pair case and eliminated by using the proposed mass-conserved formulation.

We develop a finite-domain force element theory to improve the calculation of body forces from local flow structures within a finite domain. By applying Gauss's theorem, the contributions of vorticity outside the integral region are transformed into boundary contributions of pressure, kinetic energy, and surface vorticity. The Bernoulli principle is combined with a virtual auxiliary potential function to eliminate dependence on pressure. This method establishes an accurate quantitative relationship between the local flow structures near a body and the forces acting on the body.

This paper presents a neural network-based turbulence modeling approach for transonic flows based on the ensemble Kalman method. The approach adopts a tensor basis neural network for the Reynolds stress representation, with modified inputs to consider fluid compressibility. The normalization of input features is also investigated to avoid feature collapsing in the presence of shock waves. Moreover, the turbulent heat flux is accordingly estimated with the neural network-based turbulence model based on the gradient diffusion hypothesis. The ensemble Kalman method is used to train the neural network with the experimental data in velocity and wall pressure due to its derivative-free nature. The proposed framework is tested in two canonical configurations, i.e., 2D transonic flows over the RAE2822 airfoils and 3D transonic flows over the ONERA M6 wings. Numerical results demonstrate the capability of the proposed method in learning accurate turbulence models for external transonic flows.