The calculation of the aerodynamic coefficients of an airfoil with established methods is a time consuming and computationally expensive process. This work takes a deep learning approach to determine the aerodynamic coefficients of a given airfoil much faster. It uses an already existing data base and tries three different platforms (Keras and SMT package in python and Monolith AI) to predict the coefficients as accurately as possible. The newly obtained surrogate model can then be used to give scalar and graph predictions for certain Mach numbers and angles of attack or in optimization algorithms to calculate the aerodynamic coefficients in the optimization steps.
-
Notifications
You must be signed in to change notification settings - Fork 1
mid2SUPAERO/PIR_Mario_Leupolt
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
Learning Aerodynamics Through Data to Improve Optimization Algorithms
Topics
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published