Tiny Cracks, Big Losses: Simulation Pinpoints When Steel Blooms Begin to Tear

Corner cracks in large-section micro-alloyed steel blooms have haunted steelmakers for decades, quietly turning otherwise perfect casts into scrap and driving up costs in one of the world’s most fundamental industries. A new study published in the Journal of Materials Science by Yutang Li, Jianxun Fu, and Wensheng Fu of Shanghai University’s State Key Laboratory of Advanced Metallurgy now offers one of the clearest pictures yet of exactly when and why these cracks form. By combining physical hot-tearing experiments with a coupled thermomechanical numerical model, the team traced the fatal window in which the corner of a solidifying bloom becomes vulnerable to fracture, and the numbers they found are strikingly small: at the moment of failure, the steel can only bear loads of a few megapascals, roughly the mechanical strength of a soft polymer rather than the material we think of as the backbone of modern civilization.

The research focuses on continuous casting, the process by which molten steel is poured into a water-cooled mold and drawn downward as a solidifying strand. In large-section blooms, the thick square or rectangular semi-finished products used to make heavy bars and forgings, the corners cool fastest and solidify first, making them natural stress concentrators. As the shell of solidifying steel descends below the meniscus, the liquid-metal pool inside shrinks, the solidifying skin contracts, and thermal stresses build up at the edges. If those stresses exceed what the semi-solid material can withstand, a hot tear initiates. What has always been difficult is pinning down the precise temperature, position, and stress state at which that happens for a specific steel grade, because the relevant mechanics unfold over meters of strand travel and a narrow band of temperatures just below solidification.

To capture that window experimentally, the researchers performed hot-tearing simulation experiments designed to mimic the stresses that a solidifying bloom corner experiences in the caster. Specimens of the micro-alloyed steel were heated to 1509 degrees Celsius, above the solidification range, and then cooled at controlled rates between 0.2 and 1.0 degrees Celsius per second, spanning the cooling regimes typical of continuous casting. Once each specimen reached the zero-strength temperature, or ZST, the point at which a coherent solid skeleton first appears and can transmit force, an external tensile load was applied at pulling rates of 0.25 to 1.00 millimeters per minute. This physical simulation approach, which builds on decades of hot-tearing methodology developed for steels and aluminum alloys, allowed the team to measure directly the high-temperature strength of the material in its most fragile state and to establish fracture criteria for use in numerical models.

The experimental results were unambiguous. Fracture occurred in a narrow temperature band between 1363 and 1415 degrees Celsius, and at those temperatures the hot tensile strength of the steel was a mere 1.9 to 2.8 megapascals. That number is the heart of the story. During solidification, the mushy zone, the region where solid dendrites are still bathed in interdendritic liquid, has almost no load-bearing capacity. Thin liquid films segregated with solutes such as sulfur and manganese coat the dendrite boundaries, and when tensile strain concentrates there, the semi-solid skeleton simply pulls apart. Classic criteria in the literature, from the Clyne-Davies cracking susceptibility framework through Rappaz’s mass-feed criterion and Kou’s solidification cracking model, all describe versions of this same vulnerability, but quantitative data for a specific micro-alloyed steel grade at caster-relevant conditions are precious, and this study provides them.

Micro-alloyed steels add a further complication. These grades deliberately contain small additions of elements such as vanadium and niobium, often with controlled sulfur for machinability, to precipitate fine carbonitrides that strengthen the final product. But during solidification and cooling, those same alloying choices affect microsegregation, the uneven partitioning of solutes between solid and liquid described by Scheil-type solidification models and refined approaches such as the Won-Thomas microsegregation model. Segregated liquid enriched in sulfur can form low-melting films, and precipitates such as manganese sulfide in the interdendritic regions further weaken the boundaries at high temperature. The result is a material whose mechanical identity at 1400 degrees Celsius bears little resemblance to its room-temperature personality, and whose cracking behavior cannot be predicted from conventional high-temperature strength data.

With the experimental fracture criteria in hand, the team built a coupled thermomechanical numerical simulation of the continuous casting process itself. Coupled models of this kind solve simultaneously for heat transfer during solidification and for the mechanical response of the growing shell, in the tradition of visco-plastic damage models used to simulate hot tear formation in steel castings. The model tracks how temperature and stress evolve at every point of the bloom as it descends through the caster, from the intense cooling of the mold through the secondary cooling zones, and it can therefore answer the question that laboratory specimens alone cannot: where along the strand does the corner of a real bloom actually reach the conditions measured in the hot-tearing tests?

The simulation’s answer was precise. When the bloom had traveled 1.4 to 1.6 meters below the meniscus, the temperature at a location 40 millimeters from the corner reached 1403 to 1413 degrees Celsius, while the thermal stresses there amounted to 1.0 to 2.6 megapascas. Set those numbers against the measured high-temperature strength of 1.9 to 2.8 megapascals and the conclusion is inescapable: the thermal stress at the bloom corner overlaps and, at the upper end, meets or exceeds the material’s resistance to fracture. The corner of the strand, at that specific position in the caster, sits exactly at the critical condition for hot tearing. This is the mechanical explanation for why corner cracks in large-section micro-alloyed blooms appear so persistently, and it converts a long-standing empirical problem into a quantitatively defined failure event with a known location and temperature range.

The practical implications reach deep into steel plant operations. Knowing that the danger zone lies 1.4 to 1.6 meters below the meniscus gives operators a concrete target: the cooling regime in that region, and the bulging, bending, and unbending strains imposed by the rolls and guides there, can be tuned so that the strain and stress experienced by the corner never push past the measured fracture envelope. The findings also connect to a wider body of work on transverse and surface corner cracks in peritectic and medium-carbon steels, where sulfur content, secondary cooling intensity, and mold behavior all modulate the same underlying mechanics. More broadly, the paired approach of physical simulation to establish material fracture criteria and coupled numerical simulation to locate the critical event in the machine offers a template that other plants and other grades can follow, replacing trial-and-error mitigation with prediction.

There is also a scientific elegance in the result. Hot tearing has been studied since the earliest in-situ tensile tests of solidifying aluminum shells in the 1980s, and theoretical criteria have proliferated, yet the steel industry still loses enormous value to cracks whose onset could not previously be timed. By measuring the strength of a real micro-alloyed steel in its mushy state and matching those measurements against the computed stress history of an industrial bloom, the Shanghai University team has shown how narrow the margin of safety truly is, often less than a single megapascal between intact product and fracture. That a material capable of carrying thousands of tonnes in service survives its birth with a strength of two megapascals is a vivid reminder that the most dangerous moments in metallurgy occur in the first meters of a strand’s life, where liquid and solid coexist and physics gives almost no room for error. For steelmakers battling corner cracks, the crack clock now has a readable face, and the numbers to set it by.

Subject of Research: Hot tearing and thermomechanical behavior of micro-alloyed steel during continuous casting

Article Title: Thermomechanical behavior of micro-alloyed steel: physical and numerical simulation

Article References: Li, Y., Fu, J., & Fu, W. (2026). Thermomechanical behavior of micro-alloyed steel: physical and numerical simulation. Journal of Materials Science. https://doi.org/10.1007/s10853-026-13888-1

Image Credits: AI Generated

DOI: 10.1007/s10853-026-13888-1

Keywords: continuous casting, micro-alloyed steel, hot tearing, thermomechanical simulation, corner cracks, zero-strength temperature, thermal stress, solidification, bloom casting, fracture criterion, high-temperature strength, materials science