Development of a custom supervised learning network to determine the recession rate of ablative thermal protection systems materials

Vijay B.Mohan Ramu, Savio J. Poovathingal

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Carbon-carbon composites and carbon-phenolic ablators are used as thermal protection system (TPS) materials to regulate the surface temperature of hypersonic vehicles. Material response codes that model ablation of the material do not resolve the effects of microscale changes because of the disparate length scales. In this work, we have developed a framework based on supervised learning to account for material microstructure while capturing the recession of ablative materials. The proposed supervised learning model is inspired by conventional neural networks used for temporal image-based prediction. The training images consist of snapshots depicting the physical changes of the microstructure that are obtained through coupled flow-material simulations. During the simulation, snapshots of the oxidized microstructure are recorded. A custom surface function is used to capture the profile of the evolving microstructure. The surface function acts as the encoder and decoder of the supervised learning model. Support Vector Regression (SVR) forms the heart of the model by relating the surface function parameters with input variables (porosity, time, and microscopic probabilities). The developed supervised learning model is capable of resolving the effects of needling and thinning of carbon fibers on overall recession rates. Once trained, the supervised learning model is able to predict the recession rate in a very short period (few seconds) as opposed to expensive DSMC simulations. The final model consists of analytical functions directly dependent on the input variables, which facilitates the smooth integration of the supervised learning model with the macroscale solvers. The supervised learning model provides an effective pathway to account for the effect of microstructural changes of carbon ablators on macroscale recession rate.

Original languageEnglish
Title of host publicationAIAA Aviation Forum and ASCEND, 2024
DOIs
StatePublished - 2024
EventAIAA Aviation Forum and ASCEND, 2024 - Las Vegas, United States
Duration: Jul 29 2024Aug 2 2024

Publication series

NameAIAA Aviation Forum and ASCEND, 2024

Conference

ConferenceAIAA Aviation Forum and ASCEND, 2024
Country/TerritoryUnited States
CityLas Vegas
Period7/29/248/2/24

Bibliographical note

Publisher Copyright:
© 2024, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved.

ASJC Scopus subject areas

  • Energy Engineering and Power Technology
  • Nuclear Energy and Engineering
  • Aerospace Engineering
  • Space and Planetary Science

Fingerprint

Dive into the research topics of 'Development of a custom supervised learning network to determine the recession rate of ablative thermal protection systems materials'. Together they form a unique fingerprint.

Cite this