A Computational Re-evaluation of the Grossmann-Asimow Model and Its Implications for Heat Treatment Failures.

Hardenability is the foundational property governing the depth and distribution of hardness in steel components during quenching. For decades, the industry has relied on the Grossmann-Asimow analytical model to link intrinsic steel properties to extrinsic quench severity (H-values). However, modern heat treatment failures—such as insufficient hardening depths or unpredicted unhardened cores—can often be traced back to an over-reliance on the simplifying assumptions embedded within these 80-year-old models.

This study presents a comprehensive re-evaluation of the Grossmann-Asimow framework using modern MATLAB computational tools. By reconstructing the 1D heat equation from first principles, this work systematically tests the robustness of the model's core assumptions. Specifically, we address the limitations of assuming constant thermal diffusivity and Newtonian cooling, which neglect the non-linear heat transfer dynamics caused by vapor blankets and nucleate boiling. Furthermore, our re-evaluation identifies potential inaccuracies in the universally utilized historical master charts, specifically regarding the placement of H=1 and H=0.3 curves. Such discrepancies can lead engineers to miscalculate quench severity and fail to achieve the critical 50% martensite criterion.

To provide a more physically representative predictive tool, this work compares hardenability predictions derived from the original "half-temperature time" (t_0.5) criterion against modern alternatives, such as cooling rates at transformation-relevant temperatures (e.g., 704°C per ASTM E23) and average cooling rates. By illuminating these historical limitations, this analysis equips metallurgists with a deeper understanding of model breakdown points, providing a crucial framework for the root-cause analysis of quenching failures.