Data Driven Aerodynamics | Dr. Emre Yılmaz

A Convolutional Neural Network Approach for Training Predictors for Airfoil Performance

Medium-fidelity predictors were trained for predicting airfoil performance using deep convolutional neural networks. The framework was implemented in Tensorflow.
Published a conference paper: AIAA Aviation’17 (Yilmaz & German, 2017).

Why to use Convolutional Neural Networks

The Application of Deep Convolutional Neural Networks to Airfoil Performance Prediction

A Deep Learning Approach to an Airfoil Inverse Desing Problem

Airfoil inverse problem was reformulated and solved using a framework constructed with deep convolutional neural networks.
Published a conference paper: AIAA Aviation’18 (Yilmaz & German, 2018).

The Application of Deep Convolutional Neural Networks to Airfoil Inverse Design

Conditional Generative Adversarial Network Framework for Airfoil Inverse Design

A game theoretic framework was investigated by training conditional generative adversarial networks (CGANs) to generate airfoil shapes given the desired airfoil performance criteria.
Published a conference paper: AIAA Aviation’20 (Yilmaz & German, 2020).
Funding Proposal Experience
-ARPA-E: DIFFERENTIATE (Design Intelligence Fostering Formidable Energy Reduction and Enabling Novel Totally Impactful Advanced Technology Enhancements) Program.

Airfoil Shape Generation by the Generator in GAN Framework Shown at Different Epochs

References

2020

C6.AVIATION

Conditional Generative Adversarial Network Framework for Airfoil Inverse Design

Emre Yilmaz, and Brian J. German

In AIAA AVIATION Forum, 2020

DOI Bib HTML

@inproceedings{yilmaz20,
  author = {Yilmaz, Emre and German, Brian J.},
  title = {Conditional Generative Adversarial Network Framework for Airfoil Inverse Design},
  booktitle = {AIAA AVIATION Forum},
  year = {2020},
  publisher = {aiaa},
  doi = {10.2514/6.2020-3185},
  url = {https://arc.aiaa.org/doi/abs/10.2514/6.2020-3185},
}

2018

C4.AVIATION

A Deep Learning Approach to an Airfoil Inverse Design Problem

Emre Yilmaz, and Brian J. German

In 2018 Multidisciplinary Analysis and Optimization Conference, AIAA AVIATION Forum, 2018

DOI Bib HTML

@inproceedings{yilmaz18,
  author = {Yilmaz, Emre and German, Brian J.},
  title = {A Deep Learning Approach to an Airfoil Inverse Design Problem},
  booktitle = {2018 Multidisciplinary Analysis and Optimization Conference, AIAA AVIATION Forum},
  year = {2018},
  publisher = {aiaa},
  doi = {10.2514/6.2018-3420},
  url = {https://arc.aiaa.org/doi/abs/10.2514/6.2018-3420},
}

2017

C3.AVIATION

A Convolutional Neural Network Approach to Training Predictors for Airfoil Performance

Emre Yilmaz, and Brian J. German

In 18th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference, AIAA AVIATION Forum, 2017

DOI Bib HTML

@inproceedings{yilmaz17,
  author = {Yilmaz, Emre and German, Brian J.},
  title = {A Convolutional Neural Network Approach to Training Predictors for Airfoil Performance},
  booktitle = {18th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference, AIAA AVIATION Forum},
  year = {2017},
  publisher = {aiaa},
  doi = {10.2514/6.2017-3660},
  url = {https://arc.aiaa.org/doi/abs/10.2514/6.2017-3660},
}