Exploration of physics-related latent vectors in hot working of Inconel 718 superalloy using autoencoder

The thermo-mechanical processing (TMP) behavior of nickel-based superalloys during hot working exhibits highly non-linear and complex characteristics, necessitating an understanding of deformation mechanisms for industrial applications. In this study, we developed an autoencoder-prediction network (...

Full description

Saved in:
Bibliographic Details
Main Authors: Min Jik Kim, Seon Yeong Yang, Woo Seok Yang, Sehyeok Oh, Sang Min Park, Da Seul Shin
Format: Article
Language:English
Published: Elsevier 2025-03-01
Series:Journal of Materials Research and Technology
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2238785425005502
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849762090007396352
author Min Jik Kim
Seon Yeong Yang
Woo Seok Yang
Sehyeok Oh
Sang Min Park
Da Seul Shin
author_facet Min Jik Kim
Seon Yeong Yang
Woo Seok Yang
Sehyeok Oh
Sang Min Park
Da Seul Shin
author_sort Min Jik Kim
collection DOAJ
description The thermo-mechanical processing (TMP) behavior of nickel-based superalloys during hot working exhibits highly non-linear and complex characteristics, necessitating an understanding of deformation mechanisms for industrial applications. In this study, we developed an autoencoder-prediction network (AE-PN) to model the TMP behavior of the Inconel 718 superalloy in an unsupervised way within the temperature range of 900–1200 °C and strain rates of 0.001–10 s−1. The AE-PN effectively reduces the dimensionality of stress-strain curves in a non-linear manner while preserving critical flow features, enabling accurate reconstruction of stress-strain curves. The AE-PN established strong correlations between processing parameters and the latent space, achieving high reconstruction accuracy for continuous flow curves (testing RMSE = 4.19). Latent features derived from stress-strain curves were linked to key characteristics of hot deformation behavior, including peak stress, strain hardening, and flow softening. Notably, one major latent component was strongly correlated with peak stress (R2 = 0.9971), highlighting its role as a physics-related variable connected to TMP behavior. For additional datasets within the processing window, the AE outperformed a conventional artificial neural network (ANN) model, achieving an RMSE of 7.98 compared to the ANN's RMSE of 60. Furthermore, hot workability characteristics, such as flow softening and dynamic recrystallization (DRX), were systematically analyzed using a processing map and microstructural analysis. This AE-PN framework provides interpretable insights beyond traditional ‘black-box’ systems, offering accurate stress-strain curve predictions and deeper interpretation of processing parameters through latent space exploration.
format Article
id doaj-art-18a1272d6ac14d9e9ce84d7a8acaa4f0
institution DOAJ
issn 2238-7854
language English
publishDate 2025-03-01
publisher Elsevier
record_format Article
series Journal of Materials Research and Technology
spelling doaj-art-18a1272d6ac14d9e9ce84d7a8acaa4f02025-08-20T03:05:50ZengElsevierJournal of Materials Research and Technology2238-78542025-03-01356749676210.1016/j.jmrt.2025.03.040Exploration of physics-related latent vectors in hot working of Inconel 718 superalloy using autoencoderMin Jik Kim0Seon Yeong Yang1Woo Seok Yang2Sehyeok Oh3Sang Min Park4Da Seul Shin5Materials Processing Research Division, Korea Institute of Materials Science, 797 Changwon-Daero, Seongsan-Gu, Changwon, Gyeongnam, 51508, South Korea; School of Mechanical Engineering, Pusan National University, 2 Busandaehak-ro 63beon-gil, Geumjeong-gu, Busan, 46241, South KoreaMaterials Processing Research Division, Korea Institute of Materials Science, 797 Changwon-Daero, Seongsan-Gu, Changwon, Gyeongnam, 51508, South KoreaMaterials Processing Research Division, Korea Institute of Materials Science, 797 Changwon-Daero, Seongsan-Gu, Changwon, Gyeongnam, 51508, South KoreaMaterials Data & Analysis Research Division, Korea Institute of Materials Science, 797 Changwon-Daero, Seongsan-Gu, Changwon, Gyeongnam, 51508, South KoreaSchool of Mechanical Engineering, Pusan National University, 2 Busandaehak-ro 63beon-gil, Geumjeong-gu, Busan, 46241, South Korea; Corresponding author.Materials Processing Research Division, Korea Institute of Materials Science, 797 Changwon-Daero, Seongsan-Gu, Changwon, Gyeongnam, 51508, South Korea; Corresponding author.The thermo-mechanical processing (TMP) behavior of nickel-based superalloys during hot working exhibits highly non-linear and complex characteristics, necessitating an understanding of deformation mechanisms for industrial applications. In this study, we developed an autoencoder-prediction network (AE-PN) to model the TMP behavior of the Inconel 718 superalloy in an unsupervised way within the temperature range of 900–1200 °C and strain rates of 0.001–10 s−1. The AE-PN effectively reduces the dimensionality of stress-strain curves in a non-linear manner while preserving critical flow features, enabling accurate reconstruction of stress-strain curves. The AE-PN established strong correlations between processing parameters and the latent space, achieving high reconstruction accuracy for continuous flow curves (testing RMSE = 4.19). Latent features derived from stress-strain curves were linked to key characteristics of hot deformation behavior, including peak stress, strain hardening, and flow softening. Notably, one major latent component was strongly correlated with peak stress (R2 = 0.9971), highlighting its role as a physics-related variable connected to TMP behavior. For additional datasets within the processing window, the AE outperformed a conventional artificial neural network (ANN) model, achieving an RMSE of 7.98 compared to the ANN's RMSE of 60. Furthermore, hot workability characteristics, such as flow softening and dynamic recrystallization (DRX), were systematically analyzed using a processing map and microstructural analysis. This AE-PN framework provides interpretable insights beyond traditional ‘black-box’ systems, offering accurate stress-strain curve predictions and deeper interpretation of processing parameters through latent space exploration.http://www.sciencedirect.com/science/article/pii/S2238785425005502Inconel 718Thermo-mechanical processingHot deformationDeep learningAutoencoderHot workability
spellingShingle Min Jik Kim
Seon Yeong Yang
Woo Seok Yang
Sehyeok Oh
Sang Min Park
Da Seul Shin
Exploration of physics-related latent vectors in hot working of Inconel 718 superalloy using autoencoder
Journal of Materials Research and Technology
Inconel 718
Thermo-mechanical processing
Hot deformation
Deep learning
Autoencoder
Hot workability
title Exploration of physics-related latent vectors in hot working of Inconel 718 superalloy using autoencoder
title_full Exploration of physics-related latent vectors in hot working of Inconel 718 superalloy using autoencoder
title_fullStr Exploration of physics-related latent vectors in hot working of Inconel 718 superalloy using autoencoder
title_full_unstemmed Exploration of physics-related latent vectors in hot working of Inconel 718 superalloy using autoencoder
title_short Exploration of physics-related latent vectors in hot working of Inconel 718 superalloy using autoencoder
title_sort exploration of physics related latent vectors in hot working of inconel 718 superalloy using autoencoder
topic Inconel 718
Thermo-mechanical processing
Hot deformation
Deep learning
Autoencoder
Hot workability
url http://www.sciencedirect.com/science/article/pii/S2238785425005502
work_keys_str_mv AT minjikkim explorationofphysicsrelatedlatentvectorsinhotworkingofinconel718superalloyusingautoencoder
AT seonyeongyang explorationofphysicsrelatedlatentvectorsinhotworkingofinconel718superalloyusingautoencoder
AT wooseokyang explorationofphysicsrelatedlatentvectorsinhotworkingofinconel718superalloyusingautoencoder
AT sehyeokoh explorationofphysicsrelatedlatentvectorsinhotworkingofinconel718superalloyusingautoencoder
AT sangminpark explorationofphysicsrelatedlatentvectorsinhotworkingofinconel718superalloyusingautoencoder
AT daseulshin explorationofphysicsrelatedlatentvectorsinhotworkingofinconel718superalloyusingautoencoder