Fast pressure prediction along the NACA airfoil using the convolution neural network

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F49777513%3A23520%2F19%3A43957400" target="_blank" >RIV/49777513:23520/19:43957400 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Fast pressure prediction along the NACA airfoil using the convolution neural network

Popis výsledku v původním jazyce

The knowledge of the pressure coefficient distribution around the airfoil body is the most important airfoil characteristic. Using the pressure distribution, lift and drag coefficients could be established. When a new profile is being designed, some of the optimisation techniques are usually used, for example, to have a maximal lift. This requires the calculation of the pressure field for many different variations of the geometry, which is very time-consuming. This contribution aims to utilize a convolution neural network (CNN) for the fast prediction of the pressure distribution around NACA airfoil. The CNN input contains a mesh point coordinates. At the CNN output the flow field, which includes density, velocity and pressure, is generated. The CNN was created using opensource software Keras and TensorFlow through the python interface. The CNN was trained on the set of 785 random profiles generated using an opensource CFD solver FlowPro.

Název v anglickém jazyce

Fast pressure prediction along the NACA airfoil using the convolution neural network

Popis výsledku anglicky

The knowledge of the pressure coefficient distribution around the airfoil body is the most important airfoil characteristic. Using the pressure distribution, lift and drag coefficients could be established. When a new profile is being designed, some of the optimisation techniques are usually used, for example, to have a maximal lift. This requires the calculation of the pressure field for many different variations of the geometry, which is very time-consuming. This contribution aims to utilize a convolution neural network (CNN) for the fast prediction of the pressure distribution around NACA airfoil. The CNN input contains a mesh point coordinates. At the CNN output the flow field, which includes density, velocity and pressure, is generated. The CNN was created using opensource software Keras and TensorFlow through the python interface. The CNN was trained on the set of 785 random profiles generated using an opensource CFD solver FlowPro.

Druh

O - Ostatní výsledky

CEP obor

—

OECD FORD obor

20302 - Applied mechanics

Projekt

<a href="/cs/project/TE01020068" target="_blank" >TE01020068: Centrum výzkumu a experimentálního vývoje spolehlivé energetiky</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2019

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů