Rapid Aero Analysis through Deep Learning: FY22 Engineering Research Technical Investment Program

Abstract

MIT Lincoln Laboratory frequently designs and fabricates state-of-the-art airborne sensors to be flown on Flight Test Facility aircraft. During the design process of these sensors, aerodynamic performance must be evaluated to determine the best design and to establish the airworthiness of the aircraft modification. Quantities of interest often derive from in-flight aerodynamic forces such as: lift, drag, moments, shedding frequencies, and vibrational profiles. Of these quantities, aerodynamic drag is a commonly-evaluated design metric because of its impact on fuel burn, time on station, and flight speed and its correlation to undesirable vortex shedding. With present-day methodologies, high-fidelity analysis requires three major steps as shown in Figure 1:simplifying the geometric model to an approximate outer mold shape, building the computational mesh, and running the high-fidelity fluid-flow simulation to extract the flow quantity of interest. The entire process requires days or weeks of work and careful expertise from the analyst. These time and resource commitments are a significant hurdle during the conceptual design phase when design parameters are rapidly changing and the opportunity for design impact is largest. The goal is to apply modern advances in machine learning in the prediction of aerodynamics to accelerate the burdensome analysis process.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Feb 23, 2023
Accession Number
AD1196293

Entities

People

  • Anthony Mccourt
  • Matthew C. Jones
  • Suhas Kodali

Organizations

  • Massachusetts Institute of Technology

Tags

Communities of Interest

  • Air Platforms
  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Aerodynamic Characteristics
  • Aerodynamic Drag
  • Aircrafts
  • Boundary Layer
  • Computational Fluid Dynamics
  • Deep Learning
  • Drag
  • Engineering
  • Flow
  • Fluid Dynamics
  • Fluid Flow
  • Geometry
  • Mach Number
  • Machine Learning
  • Physics Laboratories
  • Reliability
  • Simulations
  • Software Testing
  • Test Facilities

Readers

  • Distributed Systems and Data Platform Development
  • Fluid Mechanics and Fluid Dynamics.
  • Software Engineering

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - DoD AI Strategy