Automatic Control Method for Roller Surfacing System

Literature Overview

Li Zhenying, Dai Liping, Hou Ming, and He Qiong from Beijing Information Science and Technology University published this research in the Transactions of the China Welding Institute in 2008, presenting an automatic control method for open-arc surfacing of coal mill rollers. The study addresses the challenge of achieving consistent surfacing quality on large cylindrical surfaces through precise trajectory control of the welding torch.

System Architecture and Control Strategy

The coal mill roller surfacing system requires precise multi-axis coordination to deposit uniform weld beads on a large cylindrical surface. The authors established a coordinate system based on the mechanical structure of the surfacing machine and developed mathematical formulations for torch trajectory control.

The surfacing process is divided into two distinct phases:

Phase	Description	Control Objective
Same-arc segment	Continuous welding along a circumferential path	Maintain constant arc parameters and bead geometry
Transition arc segment	Torch movement between adjacent circumferential passes	Ensure smooth transition without weld defects

Mathematical Model Development

The core innovation of this work is the mathematical formulation of the torch trajectory control. The roller surface is approximated as a spherical surface based on wear patterns, and the control method accounts for:

Roller geometry: The cylindrical surface with variable radius due to wear creates a non-uniform deposition profile that must be compensated.
Multi-axis coordination: Simultaneous control of traverse, rotation, and vertical axes to maintain the torch at a constant standoff distance.
Arc parameter maintenance: Constant arc voltage and current throughout the surfacing cycle to ensure uniform bead properties.

The control formulas derived for same-arc segments ensure:

Constant linear welding speed synchronized with roller rotation
Vertical axis compensation for surface irregularities
Consistent overlap between adjacent passes (typically 30–40% overlap for full coverage)

Error Analysis and Validation

Using MATLAB simulation, the authors quantified the trajectory control error and demonstrated that the proposed method meets surfacing process requirements. The error analysis considers:

Error Source	Magnitude	Impact
Encoder resolution	±0.01 mm	Negligible for most applications
Mechanical backlash	±0.05 mm	Requires compensation algorithm
Thermal deformation of machine	±0.1 mm	Time-dependent, requires monitoring
Surface contour deviation	±0.5 mm	Compensated by spherical approximation
Arc force variation	±10%	Affects standoff distance

Practical Implementation Considerations

The automatic surfacing of coal mill rollers presents specific challenges:

Large diameter: Typical coal mill rollers have diameters of 600–1200 mm, requiring substantial machine capacity and stable support.
Heavy section thickness: Roller thickness of 150–300 mm creates high thermal mass, affecting interpass temperature control.
Wear pattern variation: Each roller has a unique wear profile requiring individual scanning and contour mapping.
Material requirements: The surfacing alloy must provide abrasion resistance against coal and slag while maintaining toughness for impact loading.

The control method must also account for:

Preheating requirements (typically 150–250 °C for low-alloy steel rollers)
Interpass temperature limitation (≤250 °C) to prevent excessive grain growth
Post-weld heat treatment for stress relief
Surface preparation (grinding to remove existing wear layer and oxide)

Study Insights and Engineering Value

This work represents an important contribution to the automation of heavy industrial surfacing operations. The mathematical framework provides a systematic approach to trajectory planning that can be adapted to other cylindrical surfacing applications, including cement mill rollers, mining equipment, and power generation components. The key insight is that successful automation requires not only mechanical precision but also process-aware control algorithms that account for welding physics — arc force, thermal input, and bead geometry — in real-time trajectory adjustments.

For practitioners implementing similar systems, the study recommends investing in surface scanning capability (laser triangulation or structured light) to capture the actual roller contour, rather than relying solely on nominal dimensions. This enables the control algorithm to compensate for wear irregularities and produce uniform coverage with consistent bead geometry throughout the surfacing cycle.