Application of Orthogonal Experimental Design in Overlay Welding Roll Heat Treatment Process Optimization

Literature Overview

The paper by Li Zhaohui, published in Ansteel Technology in 2004, addresses the optimization of heat treatment processes for overlay-welded support rolls used in continuous rolling mills. The author employed orthogonal experimental design methodology to systematically evaluate the effects of multiple heat treatment parameters on the performance of these critical rolling mill components. This work bridges metallurgical science and industrial application, demonstrating how statistical experimental design can accelerate process development.

Background and Technical Context

Overlay Welded Support Rolls in Rolling Mills

Support rolls (also called backup rolls) in continuous rolling mills bear the primary load transmitted through the work rolls. These components are subject to:

Extremely high contact pressures (2–5 GPa at the roll-to-roll interface)
Severe thermal cycling during hot rolling operations
Abrasive wear from scale and oxide particles
Potential spalling and fatigue cracking

The overlay welding process deposits a wear-resistant alloy layer on the base roll body, providing enhanced surface properties without requiring the entire roll to be made from expensive alloy steel. Common overlay materials include high-chromium cast iron, martensitic stainless steels, and nickel-based alloys.

Heat Treatment Requirements

The heat treatment of overlay welded rolls must achieve several objectives simultaneously:

Relieve welding residual stresses in the base material and overlay layer
Optimize hardness and toughness of the overlay layer for wear resistance
Reduce HAZ hardness to prevent cold cracking during subsequent service
Maintain dimensional stability to preserve roll geometry tolerances
Promote carbide precipitation in the overlay layer for enhanced wear resistance

Orthogonal Experimental Design Methodology

Experimental Parameters

The orthogonal design approach allows evaluation of multiple factors with a reduced number of experiments compared to full factorial designs. For overlay welded roll heat treatment, the key process parameters typically include:

Factor	Symbol	Levels	Unit
Heating temperature	T₁	550, 600, 650, 700	°C
Holding time	t₁	2, 4, 6, 8	h
Cooling method	C₁	Air, Furnace, Oil	—
Number of cycles	N	1, 2, 3	—
Post-quench tempering temperature	T₂	200, 300, 400	°C

Response Variables

The performance evaluation likely includes:

Surface hardness (HV or HRC) of the overlay layer
Impact toughness (J) at various temperatures
Residual stress (MPa) after heat treatment
Wear resistance (mm³/Wh or equivalent)
Dimensional change (mm or percentage)

Data Analysis

The orthogonal analysis involves:

Calculating the range (R) for each factor to determine its influence magnitude
Identifying the optimal combination of parameters
Conducting verification experiments to confirm the predicted optimum
Analyzing interaction effects if significant

Technical Analysis and Process Windows

Optimal Heat Treatment Regime

Based on the principles of overlay welding heat treatment, the optimal regime for support rolls typically involves:

Stage	Temperature	Time	Cooling	Purpose
Stress relief	550–650°C	4–8 h	Furnace cool	Reduce residual stress in base metal
Solution treatment	950–1050°C	2–4 h	Air cool	Dissolve carbides in overlay layer
Tempering	200–400°C	2–4 h	Air cool	Optimize hardness/toughness balance

Critical Technical Considerations

Hardness vs. Toughness Trade-off: The overlay layer must balance wear resistance (requiring high hardness, typically 45–60 HRC for high-chromium overlays) with impact resistance (requiring adequate toughness). The tempering temperature is the primary lever for this balance:

At 200°C: maximum hardness (~60 HRC) but reduced toughness
At 300°C: good balance (~55 HRC with improved toughness)
At 400°C: reduced hardness (~50 HRC) but significantly improved toughness

Residual Stress Reduction: The stress relief stage must be carefully controlled. Temperatures above 650°C risk softening the base roll material, while temperatures below 550°C are insufficient for stress relief. The optimal range of 550–600°C allows significant stress reduction without compromising base metal strength.

Carbide Morphology Control: In high-chromium overlay alloys, the morphology and distribution of M₇C₃ and M₂₃C₆ carbides critically influence wear resistance. The solution treatment temperature and subsequent cooling rate determine whether carbides form in a continuous network (brittle, detrimental) or as discrete particles (tough, beneficial).

FMEA Analysis of Heat Treatment Process

Failure Mode	Effect	Severity	Cause	Occurrence	Detection	RPN	Countermeasure
Excessive HAZ softening	Reduced roll strength	9	Overheating during stress relief	3	2	54	Thermocouple monitoring, limit T < 650°C
Overlay spalling	Roll surface damage	8	Thermal expansion mismatch	4	1	32	Gradual cooling, interlayer optimization
Carbide network formation	Brittle overlay	7	Too fast cooling from solution temp	5	2	70	Controlled furnace cooling rate
Dimensional distortion	Roll geometry out of tolerance	6	Uneven heating/cooling	4	3	72	Symmetric loading, pre-machining allowance
Cracking during quench	Component failure	10	High thermal gradient	3	1	30	Avoid water quench, use oil or air

Engineering Practice Integration

Implementation in Rolling Mill Maintenance

The optimized heat treatment process should be integrated into the rolling mill maintenance schedule as follows:

Pre-welding preparation: Grind the roll surface to remove decarburized layer and ensure proper contact for overlay welding
Overlay welding: Apply the overlay alloy using the qualified welding procedure (typically SMAW with high-chromium electrodes or submerged arc welding)
Post-weld heat treatment: Execute the optimized heat treatment cycle in a properly instrumented furnace
Post-treatment inspection: Verify hardness distribution, residual stress levels, and dimensional accuracy
Machining and finishing: Restore final roll geometry and surface finish

Quality Control Points

The following quality control checkpoints are essential:

Pre-heat treatment: Measure baseline hardness and residual stress
During heat treatment: Monitor furnace temperature uniformity (±10°C) and ramp rates
Post-heat treatment: Conduct full hardness profile measurement from surface to base metal
Final acceptance: Verify overlay thickness, bond strength, and surface integrity

Key Questions and Reflections

The application of orthogonal experimental design to heat treatment optimization offers several advantages over traditional trial-and-error approaches:

Efficiency: Reduces the number of experiments needed from potentially dozens to a manageable set (typically 9–25 runs)
Systematic factor evaluation: Identifies the most influential parameters and their optimal levels
Reproducibility: Provides a clear, documented process that can be replicated

However, several limitations deserve attention:

Orthogonal designs assume additive effects and may miss significant interaction effects
The number of levels per factor is limited, potentially missing the true optimum
Statistical significance testing requires adequate replication, which may be constrained by the high cost of roll testing

For future work, response surface methodology (RSM) or Taguchi methods with more levels might provide superior optimization results, particularly when interaction effects are significant.

Study Insights and Implications

This paper demonstrates the value of applying statistical experimental design to metallurgical process optimization. For engineers managing rolling mill maintenance and overlay welding programs, the key insight is that systematic optimization of heat treatment parameters can significantly extend component life and reduce unplanned downtime.

The methodology is directly transferable to other overlay welding applications in the pipe and fitting industry, such as optimizing the post-weld heat treatment of corrosion-resistant overlay welds on pipeline elbows or hardfaced reducers. The fundamental principles of balancing competing performance objectives through controlled thermal cycles remain universal across applications.

In conclusion, Li Zhaohui's work exemplifies how statistical experimental design can bridge the gap between metallurgical theory and industrial practice, providing a rigorous framework for process optimization that is both scientifically sound and practically implementable in demanding industrial environments.