One neural network approach for the surrogate turbulence model in transonic flows

With the rapid development of artificial intelligence techniques such as neural networks, data-driven machine learning methods are popular in improving and constructing turbulence models. For high Reynolds number turbulence in aerodynamics, our previous work built a data-driven model applicable to subsonic airfoil flows with different free stream conditions. The results calculated by the proposed model are encouraging. In this work, we aim to model the turbulence of transonic wing flows with fully connected deep neural networks, where there is less research at present. The proposed model is driven by two flow cases of the ONERA (Office National d'Etudes et de Recherches Aerospatiales) wing and coupled with the Navier-Stokes equation solver. Four subcritical and transonic benchmark cases of different wings are used to evaluate the model performance. The iteration process is stable, and final convergence is achieved. The proposed model can be used to surrogate the traditional Reynolds averaged Navier-Stokes turbulence model. Compared with the data calculated by the Spallart-Allmaras model, the results show that the proposed model can be well generalized to the test cases. The mean relative error of the drag coefficient at different sections is below 4% for each case. This work demonstrates that modeling turbulence by data-driven methods is feasible and that our modeling pattern is effective.

Automated summary

This study applies artificial intelligence techniques and data-driven machine learning methods to model turbulence in transonic wing flows. By constructing a fully connected deep neural network model, the researchers achieve successful modeling results and demonstrate good generalization capabilities in test cases. This research is significant for improving turbulence modeling techniques.