Study on the mechanical properties and critical temperature of FeNiCrMn alloy using MD-ML-MA framework

J Liu and JY Mao and B Wang and QK Wang and N Zhang and SY Pan, JOURNAL OF MOLECULAR MODELING, 31, 340 (2025).

DOI: 10.1007/s00894-025-06575-6

Context and resultsThe FeNiCrMn alloy gasket is vital for the sealing performance of the engine cylinder head-block interface and thus engine reliability. The transition temperature at which the plastic region disappears in the FeNiCrMn alloy gasket remains ambiguous. Molecular dynamics (MD) simulations show that lowering temperature suppresses plastic deformation under tension but improves compressive performance, while strain rate has negligible effects on elastic and strength properties. Based on MD data, a machine learning (ML) model achieved high prediction accuracy (MAE=0.0072,R2=0.9949\documentclass12ptminimal \usepackageamsmath \usepackagewasysym \usepackageamsfonts \usepackageamssymb \usepackageamsbsy \usepackagemathrsfs \usepackageupgreek \setlength\oddsidemargin-69pt \begindocument$$\text MAE = 0.0072, R<^>2 = 0.9949$$\enddocument). Mathematical analysis (MA) further identified critical temperatures of Tc=509K\documentclass12ptminimal \usepackageamsmath \usepackagewasysym \usepackageamsfonts \usepackageamssymb \usepackageamsbsy \usepackagemathrsfs \usepackageupgreek \setlength\oddsidemargin-69pt \begindocument$$T_c = 509\,\text K$$\enddocument (tension) and 526 K (compression), beyond which tensile plasticity vanishes and compressive behavior exhibits the opposite trend.MethodsA combined MD-ML-MA framework was employed to investigate the mechanical properties and critical temperature of the FeNiCrMn alloy gasket. MD simulations assessed tensile and compressive responses across temperatures and strain rates. The resulting dataset was used to train an ML neural network with a backpropagation algorithm for predictive modeling, while MA quantified the plastic region m(T), enabling determination of critical temperature thresholds.

Return to Publications page