Zhuojin Pipe Fitting Co., LtdMachine Vision-Based Robot Overlay Welding System for Roller Press Sleeve
Literature Overview
This paper by Chen Zhonghua, published in Cement (2024, No. 5, pp. 35-37), presents the design and implementation of an intelligent robot overlay welding system for roller press sleeves in cement manufacturing. The author is affiliated with CNBM (Hefei) Powder Technology Equipment Co., Ltd., and the Anhui Provincial Key Laboratory of Green Low-Carbon Technology for Cement Manufacturing. The system addresses the well-known challenges of roller press sleeve overlay welding: inconsistent weld quality, difficulty in process monitoring, harsh operating environment, and high labor costs.
Core Technical Content
Roller presses are critical equipment in cement grinding circuits, where the rollers undergo severe abrasive wear from grinding cement clinker and raw meal. The roller sleeves require periodic overlay welding to restore surface hardness and geometry. Traditional manual overlay welding of roller sleeves is labor-intensive, produces inconsistent results, and exposes operators to high temperatures, metal fumes, and noise.
System Architecture
The proposed system integrates several key components:
- Industrial robot: A 6-axis articulated robot with sufficient reach to access the entire roller sleeve surface. The robot is equipped with a GTAW (Tungsten Inert Gas) or GMAW (Gas Metal Arc Welding) welding torch.
- Machine vision system: A camera-based inspection system that captures the roller sleeve surface before and during welding to determine weld bead positioning, surface condition, and weld quality.
- Process monitoring system: Sensors that monitor welding current, voltage, arc length, and travel speed in real time, enabling immediate detection of process deviations.
- Quality database: A traceability system that records all welding parameters and inspection results for each weld pass, enabling quality traceability and process optimization.
Machine Vision Functions
The machine vision system serves multiple purposes:
| Function | Description | Benefit |
|---|---|---|
| Pre-weld inspection | Detects surface defects, wear pattern, and geometry | Enables adaptive welding strategy |
| Bead tracking | Monitors weld bead position relative to programmed path | Corrects for roller rotation and positioning errors |
| Real-time quality monitoring | Detects porosity, lack of fusion, and undercuts during welding | Enables immediate correction |
| Post-weld inspection | Measures weld bead height, width, and profile | Verifies dimensional compliance |
| Surface roughness measurement | Quantifies surface finish after grinding | Ensures functional surface quality |
Welding Process Parameters
| Parameter | Typical Value | Control Method |
|---|---|---|
| Welding process | GTAW or GMAW | Robot-controlled |
| Consumable | Hardfacing wire (e.g., Cr-Cr₃C₂, Co-based) | Wire feed motor |
| Current | 180-250 A (GTAW) or 200-300 A (GMAW) | Closed-loop control |
| Voltage | 15-22 V | Arc length regulator |
| Travel speed | 200-400 mm/min | Robot servo control |
| Wire feed speed | 4-8 m/min | Synchronized with travel |
| Shielding gas | Ar (GTAW) or Ar + CO₂ (GMAW) | Flow controller |
| Interpass temperature | < 150 °C | IR thermometer monitoring |
Quality Control and Traceability
The system's quality assurance approach follows a comprehensive PDCA (Plan-Do-Check-Act) cycle:
- Plan: Pre-weld inspection determines the wear pattern and establishes the welding strategy, including number of passes, wire selection, and parameter settings.
- Do: The robot executes the welding program with real-time parameter monitoring and vision-based bead tracking.
- Check: Post-weld inspection verifies weld quality through visual examination, dimensional measurement, and hardness testing.
- Act: Results are fed back into the process database to optimize parameters for future welds.
The traceability system records the following data for each weld pass:
- Robot path coordinates and timestamps
- Welding current and voltage waveforms
- Wire feed speed and travel speed
- Shielding gas flow rate
- Preheat and interpass temperatures
- Vision system images and measurements
- Operator identification and authorization
Engineering Practice Insights
The integration of machine vision with robot welding represents a significant advancement over traditional manual or semi-automated overlay welding. The key benefits include:
- Consistency: Robot-controlled parameters eliminate operator variability, producing uniform weld beads with consistent mechanical properties.
- Safety: Operators are removed from the welding environment, reducing exposure to heat, fumes, and noise.
- Productivity: Continuous operation without fatigue enables higher throughput and shorter maintenance windows.
- Quality traceability: Complete parameter recording enables root cause analysis of any quality issues and supports continuous improvement.
However, the system also presents challenges:
- Cost: The initial investment in robot, vision system, and control software is substantial.
- Flexibility: The system is optimized for specific roller sleeve geometries and may require reprogramming for different equipment.
- Maintenance: The vision system requires regular calibration and cleaning to maintain accuracy.
- Skill requirements: Operation and maintenance require specialized training in robotics and welding automation.
Key Reflections
This paper represents the state-of-the-art in intelligent welding systems for heavy equipment maintenance. The integration of machine vision for both pre-weld assessment and real-time process monitoring is particularly noteworthy, as it transforms the welding process from a purely parameter-driven operation to an adaptive, feedback-controlled process.
The traceability aspect of the system is especially valuable for quality management in cement manufacturing, where roller press sleeve failure can result in significant production losses. Having complete records of every welding parameter and inspection result enables rapid identification of the root cause of any premature sleeve failure.
The paper also highlights the importance of process quality monitoring during welding, not just post-weld inspection. Real-time detection of process deviations allows immediate correction, preventing the accumulation of defects that would require costly rework.