In modern thin slab casting and rolling (TSCR) production lines, the operational continuity and efficiency are paramount. These lines are characterized by their strong capability for rolling thin-gauge products and low comprehensive energy consumption. A critical component ensuring this smooth operation is the rolling mill work roll, which is subjected to extreme conditions including high rotational speeds, severe thermal cycling, and significant mechanical loads. In our specific 1880mm TSCR line, the roughing mill work rolls are primarily fabricated from high-chromium composite cast steel, a material chosen for its balance of wear resistance, toughness, and thermal fatigue properties inherent to the steel casting process. However, the occurrence of two catastrophic drive-side neck fractures within a single year presented a severe challenge, disrupting production continuity and necessitating a thorough root cause investigation and the development of effective control measures.
The first incident occurred in February 2019. During the processing of a Q345B steel grade, a cobble event took place at the finishing mill. The roughing mill stand, designated R1, remained with the slab inside. After resolving the cobble and attempting to rotate the rolls, the upper work roll was found stationary. Subsequent inspection confirmed a clean fracture at its drive-side neck. The second, nearly identical failure happened in August 2019 while rolling Q245B, where the upper work roll in the same R1 stand ceased rotation, again due to a drive-side neck fracture. Both events led to unplanned downtime, breaking the production sequence and highlighting a critical reliability issue.
A detailed investigation into the operational circumstances preceding both failures revealed a common factor: both rolls had experienced a cobble or similar jam event shortly before the fracture. Process data logs were scrutinized. No abnormal parameters were recorded during the light reversal maneuvers used to clear the jam. Standard procedure was followed, including the immediate shutdown of roll cooling water to prevent thermal shock. The fracture itself was deduced to have occurred at the precise moment the R1 mill attempted to bite into the next slab, as indicated by a sudden, sharp drop in rolling force from the expected level (see data representation below).
| Stand | Main Motor Power (kW) | Rolling Speed (m/min) | Max. Rolling Force (kN) | Max. Bending Force (kN) |
|---|---|---|---|---|
| R1 | 6600 | 45 / 79 (Base/High) | 39200 | 210 |
| R2 | 6600 | 65 / 114 (Base/High) | 39200 | 221 |
| Property | Outer Layer | Core |
|---|---|---|
| Tensile Strength (MPa) | 700 – 900 | 400 – 800 |
| Yield Strength (MPa) | 700 – 900 | – |
| Impact Toughness (J/cm²) | 1 – 5 | – |
| Elongation (%) | ≤ 2.0 | – |
Visual inspection of the fracture surfaces from both incidents provided immediate clues. The breaks were macroscopically flat and granular, characteristic of brittle fracture with minimal plastic deformation. Notably, on the circumferential surface of the fracture, approximately 400-540 mm from the roll end, clusters of a porous, honeycomb-like structure were observed. This was not a feature of a clean, sound metal fracture. The location and appearance pointed towards an inherent material discontinuity.
Further laboratory analysis was conducted on samples from the fracture zone of the first failed roll. Chemical composition analysis at three points within the fractured neck area revealed an average chromium content slightly exceeding the specified core material limits, as detailed below. More critically, metallographic examination of the fracture surface revealed a substandard graphite morphology. The spheroidization degree and nodularity of the graphite were poor, indicating a deviation from the optimal microstructure required for good toughness in this grade of cast steel.
| Measurement Point | Mn | Cr | Ni | Mo |
|---|---|---|---|---|
| Point 1 | 0.409 | 0.514 | 0.302 | 0.065 |
| Point 2 | 0.442 | 0.526 | 0.697 | – |
| Point 3 | 0.443 | 0.636 | 0.358 | 0.494 |
| Average at Fracture | 0.431 | 0.559 | 0.358 | 0.498 |
| Specified Core Limit | 0.100-0.500 | < 0.400 | < 1.800 | – |
The most definitive diagnostic tool employed was non-destructive testing (NDT). Ultrasonic testing (UT) was performed on the fractured roll necks using both 1 MHz and 2 MHz straight-beam probes. In a sound material, the ultrasonic wave travels through the metal and reflects off the back surface, producing a clear back-wall echo on the display. In both fractured rolls, a distinct zone of significant back-wall echo attenuation was detected at a circumferential location corresponding to the observed honeycomb clusters. This attenuation indicated an area of coarse, non-uniform grain structure that scattered and absorbed the ultrasonic energy—a classic indicator of a casting defect known as a “hot spot” or “thermal node.”
This finding led to the conclusive root cause analysis. The “hot spot” is a fundamental concern in steel casting. It refers to a volume within a casting that remains molten longer than the surrounding material during solidification, becoming the last region to freeze. The prolonged solidification time in this zone leads to several detrimental effects:
- Coarse Microstructure: Larger grain sizes and irregular phase formations, such as the poor graphite morphology observed.
- Micro-shrinkage and Porosity: As the final liquid solidifies, it contracts. If fed inadequately, it draws liquid from within itself, creating microscopic shrinkage pores—the likely source of the honeycomb appearance.
- Element Segregation: Alloying elements like chromium can become concentrated in the last-to-freeze liquid, explaining the locally elevated Cr levels.
The combination of these factors severely degrades the local mechanical properties. The material’s strength, toughness (impact energy), and fatigue resistance in the hot spot zone fall well below the design specifications. The neck of a work roll is a critically stressed area, primarily subjected to torsional shear stress from the driving torque and bending stress. The stress in a cylindrical shaft under torsion is given by:
$$ \tau = \frac{T \cdot r}{J} $$
where $\tau$ is the shear stress, $T$ is the applied torque, $r$ is the radius, and $J$ is the polar moment of inertia. A defect like a hot spot creates a local stress concentration factor ($K_t$), drastically increasing the effective stress ($\sigma_{eff}$) at that point:
$$ \sigma_{eff} = K_t \cdot \sigma_{nominal} $$
Furthermore, the material’s ability to resist crack propagation, its fracture toughness ($K_{IC}$), is diminished. When the stress intensity factor ($K_I$) at the tip of a flaw (like a shrinkage pore) exceeds the material’s toughness, catastrophic brittle fracture occurs:
$$ K_I \geq K_{IC} $$
The cobble events served as the final trigger. The sudden, high-impact loading during the bite of a new slab, potentially combined with residual stresses from the thermal shock of the previous cobble, provided the sufficient driving force for a pre-existing flaw cluster within the hot spot to propagate instantaneously through the weakened cross-section, resulting in the observed clean, brittle fractures.
The image above illustrates the complexity of the steel casting manufacturing process, where meticulous control of solidification is required to prevent defects like hot spots. To prevent recurrence, a multi-pronged strategy was implemented, focusing on detection, prevention, and procedural controls.
1. Revised Incoming Inspection Standard Based on UT: A new, stringent acceptance criterion for the drive-side neck of all incoming high-chromium composite cast steel work rolls was established. This was quantified based on back-wall echo attenuation using a 1 MHz ultrasonic probe:
| UT Condition (1 MHz Probe) | Assessment | Disposition |
|---|---|---|
| No back-wall echo obtained (Complete attenuation) | Severe hot spot/discontinuity | Reject and return to supplier |
| Back-wall echo attenuation > 50% of reference level | Significant hot spot | Reject and return to supplier |
| Back-wall echo attenuation ≤ 50% of reference level | Minor anomaly, acceptable with caution | Accept for use, subject to on-site monitoring |
This standard was formally incorporated into the roll procurement technical agreement. Existing rolls in inventory were screened against this new standard, and suppliers were required to provide written quality certifications.
2. Casting Process Optimization by the Supplier: Collaborating with the roll manufacturer was essential. They conducted their own failure analysis and implemented corrective actions in their steel casting process. The primary adjustment involved modifying the gating and risering system to shift the inevitable thermal center (hot spot) away from the high-stress neck region and into a less critically stressed area of the roll body. This leverages principles of directional solidification, where the solidification front is controlled to move towards a riser that feeds the shrinking metal. The solidification time ($t_f$) for a simple shape can be estimated by Chvorinov’s rule:
$$ t_f = B \cdot \left( \frac{V}{A} \right)^n $$
where $V$ is volume, $A$ is surface area, $B$ is a mold constant, and $n$ is an exponent (~2). By changing the geometry of the mold or using chills, the $(V/A)$ ratio—and thus the solidification sequence—is altered to move the hot spot.
3. Operational Procedure Enhancement: A clear procedure was mandated for cobble or jam events. Any roll involved in a significant cobble, especially in the roughing mill, must be replaced immediately upon clearing the event, regardless of its apparent condition or remaining tonnage life. This eliminates the risk of subjecting a potentially flawed roll to further high-impact loads. Furthermore, the磨辊间 (roll grinding shop) must be notified immediately about any rolls removed due to cobbles for a detailed inspection, including UT, before any decision on regrinding and reuse is made.
The effectiveness of these measures has been demonstrated over subsequent years. Since the implementation of the new UT screening standard in late 2019, no further drive-side neck fractures have occurred on the production line. The revised inspection protocol successfully identified and led to the rejection of at least two new rolls that exhibited severe ultrasonic attenuation, preventing them from ever entering service. Rolls supplied after the manufacturer’s casting process adjustments now consistently show clean, full back-wall echoes in the neck region during UT, indicating the successful mitigation of the hot spot defect. The post-cobble roll replacement policy has become a standard, non-negotiable operating practice, further reducing systemic risk.
In conclusion, the investigation into the repeated neck fractures of high-chromium composite cast steel work rolls pinpointed a critical steel casting quality defect—the formation of a “hot spot” in the high-stress neck region. This defect acted as a severe stress concentrator and initiated brittle fracture under operational loads. The problem was systematically addressed through a combination of rigorous non-destructive testing to screen out defective rolls, collaboration with the supplier to optimize the fundamental steel casting and solidification process, and enhancements to mill operating procedures to manage high-risk events. This case underscores that the reliability of critical cast components in heavy industry is not solely a function of material specification but is profoundly dependent on the precision and control exercised during the steel casting manufacturing process itself. Continuous monitoring and a partnership-based approach with suppliers are essential for sustainable operational integrity.