Open Access. Powered by Scholars. Published by Universities.®
- Keyword
-
- Machine learning (2)
- Aerospace engineering (1)
- Algorithms (1)
- CNN auto encoder model (1)
- Colleges & universities (1)
-
- Computer architecture (1)
- Convolutional neural networks (1)
- CubeSat (1)
- Design specifications (1)
- Documents (1)
- Encoding airfoil geometry (1)
- Fluid dynamics (1)
- Hyperparameter optimization (1)
- Improvements (1)
- MBSE (1)
- Model-based systems (1)
- National Aeronautics & Space Administration (1)
- Patch-wise model training (1)
- Payloads (1)
- Programming languages (1)
- Satellites (1)
- Sequential learning (1)
- Software (1)
- Structural optimization (1)
- Subsonic flow (1)
- Success (1)
- Systems engineering (1)
- United States (1)
Articles 1 - 3 of 3
Full-Text Articles in Aerospace Engineering
Patch-Wise Training With Convolutional Neural Networks To Synthetically Upscale Cfd Simulations, John P. Romano, Alec C. Brodeur, Oktay Baysal
Patch-Wise Training With Convolutional Neural Networks To Synthetically Upscale Cfd Simulations, John P. Romano, Alec C. Brodeur, Oktay Baysal
Mechanical & Aerospace Engineering Faculty Publications
This paper expands the authors’ prior work[1], which focuses on developing a convolutional neural network (CNN) model capable of mapping time-averaged, unsteady Reynold’s-averaged Navier-Stokes (URANS) simulations to higher resolution results informed by time-averaged detached eddy simulations (DES). The authors present improvements over the prior CNN autoencoder model that result from hyperparameter optimization, increased data set augmentation through the adoption of a patch-wise training approach, and the predictions of primitive variables rather than vorticity magnitude. The training of the CNN model developed in this study uses the same URANS and DES simulations of a transonic flow around several NACA 4-digit airfoils …
Convolutional-Neural-Network-Based Des-Level Aerodynamic Flow Field Generation From Urans Data, John P. Romano, Oktay Baysal, Alec C. Brodeur
Convolutional-Neural-Network-Based Des-Level Aerodynamic Flow Field Generation From Urans Data, John P. Romano, Oktay Baysal, Alec C. Brodeur
Mechanical & Aerospace Engineering Faculty Publications
The present paper culminates several investigations into the use of convolutional neural networks (CNNs) as a post-processing step to improve the accuracy of unsteady Reynolds-averaged Navier–Stokes (URANS) simulations for subsonic flows over airfoils at low angles of attack. Time-averaged detached eddy simulation (DES)-generated flow fields serve as the target data for creating and training CNN models. CNN post-processing generates flow-field data comparable to DES resolution, but after using only URANS-level resources and properly training CNN models. This document outlines the underlying theory and progress toward the goal of improving URANS simulations by looking at flow predictions for a class of …
Model Based Systems Engineering With A Docs-As-Code Approach For The Sealion Cubesat Project, Kevin Chiu, Sean Marquez, Sharanabasaweshwara Asundi
Model Based Systems Engineering With A Docs-As-Code Approach For The Sealion Cubesat Project, Kevin Chiu, Sean Marquez, Sharanabasaweshwara Asundi
Mechanical & Aerospace Engineering Faculty Publications
The SeaLion mission architecture team sought to create a model-based systems engineering approach to assist improving CubeSat success rates as well as for the SeaLion CubeSat project to guide an implementation for the flight software. This is important because university CubeSat teams are growing in number but often have untrained students as their core personnel. This was done using a document-as-code, or docs-as-code, approach. With this the team created tools for the systems architecture with the Mach 30 Modeling Language to create an architecture that is easy to learn and use even for newly admitted team members with little to …