Multi-Field Coupled Numerical Simulation of Submerged Arc Cladding Process on Steel Rolls

Literature Overview

This study focuses on the multi-field coupled numerical simulation of the submerged arc welding (SAW) cladding process applied to steel rolls, which are critical components in hot rolling mills subject to severe wear, thermal shock, and mechanical loading. The research addresses the inherent complexity of the SAW cladding process by simultaneously modeling the thermal field, stress field, and magnetic field interactions during deposition. The core motivation is to predict and optimize cladding quality—particularly residual stress distribution, dilution rate, and bonding integrity—before committing to expensive physical trials on large-diameter rolls. The study bridges computational modeling with practical process parameter selection, offering engineers a systematic framework for cladding process design on cylindrical geometries.

Core Technical Points

The multi-field coupling approach captures three interdependent physical phenomena during SAW cladding:

• Thermal field: Governs the melt pool geometry, solidification rate, and dilution between the cladding metal and the roll substrate. Key parameters include arc voltage (typically 28–35 V), current (400–600 A), travel speed (150–350 mm/min), and preheating temperature (150–250 °C for carbon steel rolls).
• Stress field: Residual stresses arise from non-uniform cooling and differential thermal expansion between the overlay layer and the substrate. For steel rolls with wall thickness exceeding 80 mm, longitudinal residual stresses can reach 200–350 MPa near the weld root, posing a risk of cracking in high-strength materials.
• Magnetic field: The electromagnetic force generated by the welding current influences melt pool convection and droplet transfer behavior. In SAW cladding, the magnetic field also affects the arc stability when welding on curved surfaces, where the electrode-to-workpiece distance varies along the circumference.

Process Parameter Windows for Roll Cladding

The numerical model employs a moving heat source with a double-ellipsoid or Gaussian distribution to represent the arc energy input. The thermal boundary conditions account for radiative and convective heat loss on the roll surface, which is critical because the cylindrical geometry creates non-uniform heat dissipation compared to flat plates.

Interpretation of Key Findings

The simulation results reveal that the peak temperature in the cladding layer reaches approximately 1650–1850 °C during deposition, while the base metal temperature at the fusion boundary fluctuates between 850–1100 °C depending on the number of passes and interpass temperature control. The residual stress distribution shows a characteristic pattern: compressive stresses dominate in the weld cap region, while tensile stresses concentrate at the root and in the heat-affected zone (HAZ). For multi-pass cladding on thick rolls, the interaction between successive passes causes stress relaxation in previously deposited layers, reducing the peak tensile stress by 15–25% compared to single-pass conditions.

A particularly important finding is the sensitivity of dilution rate to travel speed and current. When the travel speed drops below 180 mm/min with currents above 550 A, dilution can exceed 40%, leading to excessive softening of the overlay layer and loss of desired wear resistance. Conversely, high travel speeds above 320 mm/min with lower currents produce incomplete fusion at the root, compromising bond strength. The optimal window identified through simulation balances these competing factors, recommending travel speeds of 220–280 mm/min and currents of 450–520 A for typical 304 stainless steel overlay on C45 steel rolls.

Integration with Engineering Practice

In industrial roll cladding, the numerical predictions must be validated against physical measurements. A recommended verification protocol includes:

1. Thermal measurement: Infrared thermography or embedded thermocouples at multiple positions along the roll circumference to verify the predicted temperature profile.
2. Stress measurement: X-ray diffraction or hole-drilling strain gauge methods to map residual stresses in the cladding layer and HAZ.
3. Dilution verification: Optical emission spectroscopy (OES) or wet chemical analysis of cross-sections at the fusion boundary to quantify actual dilution rates.
4. Microstructural examination: Metallographic analysis of the fusion line to assess bonding quality, microsegregation, and the presence of defects such as lack of fusion or porosity.

The PDCA cycle is particularly applicable here: the Plan phase uses numerical simulation to select initial parameters; the Do phase executes the cladding with those parameters; the Check phase validates results through NDT and mechanical testing; and the Act phase feeds measured deviations back into the simulation model for refinement. This iterative approach significantly reduces the number of physical trial runs required for qualification, saving both time and material costs.

Key Questions and Reflections

One critical question arising from this study is how accurately the multi-field coupled model captures the magnetic field effects on curved surfaces. While the thermal and stress fields are well-established in the literature, the electromagnetic force distribution on a rotating or stationary cylinder differs substantially from flat-plate assumptions. The arc force vector may tilt away from the normal direction as the electrode traverses the roll circumference, potentially causing asymmetric bead profiles and uneven penetration. Future work should incorporate measured arc force data from actual roll cladding operations to refine the electromagnetic boundary conditions.

Another reflection concerns the scale-up challenge. Most validated models are calibrated on small coupon specimens or short test rolls (diameter 200–400 mm), whereas production rolls can exceed 1500 mm in diameter and 3000 mm in length. The thermal mass of large rolls creates a fundamentally different heat sink condition, and the accumulated thermal history over hundreds of passes can lead to microstructural degradation that simple linear models may not capture. Engineers should exercise caution when extrapolating simulation results from small-scale validation to full-scale production, and should always include a margin of safety in parameter selection.

Study Insights and Implications

This research demonstrates that multi-field coupled simulation is a powerful tool for optimizing SAW cladding processes on steel rolls, but its value is maximized only when integrated with rigorous experimental validation and iterative process improvement. The key takeaway for practicing engineers is that numerical models should be treated as decision-support tools rather than absolute predictors. The residual stress predictions, dilution estimates, and temperature distributions provide valuable guidance for initial parameter selection, but final qualification must always rely on physical testing including bond strength tests, hardness profiling across the cladding layer, and non-destructive examination of the full cladding surface. When properly applied within a structured quality management framework, this approach can reduce cladding defect rates by 30–50% and significantly shorten the qualification cycle for new roll designs or overlay material combinations.