Zhuojin Pipe Fitting Co., LtdLaser Vision Detection System for TIG Welding Rapid Manufacturing
Literature Overview
The research by Luo Yong, Zhang Hua, Li Yuehua, Xiao Min, and Xu Jianning, published in Laser Technology in 2007 (Vol. 31, Issue 4, pp. 367-369), presents the development and validation of a laser vision detection system for monitoring metal structure formation during TIG welding-based rapid manufacturing. The system employs laser triangulation to acquire three-dimensional surface profiles of additively built components, enabling real-time assessment of layer height accuracy and workpiece deformation. The study demonstrates that the detection error falls within the natural forming height variation range of single-pass welds, validating the approach for process monitoring applications.
Laser Vision Sensing Principle and System Configuration
Laser vision detection for welding process monitoring relies on the principle of laser triangulation. A laser line or laser sheet is projected onto the workpiece surface, and a camera positioned at a known offset angle captures the reflected laser pattern. As the surface height varies, the apparent position of the laser line in the camera image shifts, allowing three-dimensional surface profiling through triangulation geometry.
| System Parameter | Description | Engineering Significance |
|---|---|---|
| Laser source | Line or sheet laser projection | Defines scan width and spatial resolution |
| Camera | CCD or CMOS imaging sensor | Captures laser pattern; resolution determines measurement precision |
| Triangulation angle | Angle between laser axis and camera axis | Larger angles improve height sensitivity but reduce field of view |
| Sampling rate | Frame rate of image acquisition | Must match welding travel speed for adequate data density |
| Data processing | Image analysis algorithm for height extraction | Converts pixel displacement to physical height values |
The system described in the paper acquires images during the TIG welding process and analyzes them to extract metal structure height data. The key innovation is the application of this sensing technology to rapid manufacturing, where multiple layers of weld metal are deposited sequentially to build up three-dimensional components. In this context, monitoring each layer's height and position relative to the previous layer is critical for ensuring dimensional accuracy and geometric fidelity.
Process Monitoring and Dimensional Accuracy Assessment
In TIG welding rapid manufacturing, each deposited weld bead forms a layer of the final component. The height of each bead is determined by the welding parameters (current, voltage, travel speed, wire feed rate) and the interaction between the arc and the previously deposited material. Natural variation in bead height occurs due to fluctuations in arc characteristics, gas shielding effectiveness, and thermal interaction with adjacent material.
The laser vision system measures the actual bead height and compares it with the programmed target height. The study reports that the measurement error is within the natural forming height error range of single-pass welds, which is a significant finding. This means that the measurement uncertainty of the laser system does not mask real process variations, and the system can reliably detect deviations that would require corrective action.
The detection of workpiece deformation is equally important. As layers are deposited, accumulated thermal stresses cause the base material to warp or distort. The laser vision system captures the surface profile of the base material before and during welding, providing quantitative data on deformation magnitude and pattern. This information can feed into process optimization strategies, such as adjusting welding sequence, implementing interlayer cooling, or modifying support fixture design.
Engineering Practice and Integration with Process Control
For practical implementation in a production environment, the laser vision detection system must be integrated with the welding control system to enable closed-loop process adjustment. The following integration architecture is recommended:
- Real-time height measurement: The laser system continuously measures bead height during welding and transmits data to the welding controller.
- Deviation detection: The controller compares measured height with target values and identifies deviations exceeding tolerance limits.
- Parameter adjustment: If deviation is detected, the controller modifies welding parameters (typically travel speed or wire feed rate) to compensate for height error in subsequent passes.
- Deformation compensation: Measured base material deformation data is used to adjust the programmed welding path for subsequent layers, ensuring that deposited material follows the intended geometry.
The accuracy of the laser vision system is sufficient for monitoring purposes, as confirmed by the study's finding that measurement error is within natural process variation. However, for closed-loop control applications, additional considerations include measurement latency, the time required for parameter changes to take effect, and the risk of over-correction due to process dynamics. A well-tuned control algorithm, such as a proportional-integral controller with appropriate gain settings, is essential to achieve stable closed-loop performance.
Study Insights and Reflections
This work represents an early but important contribution to the field of process monitoring for additive manufacturing by welding. The use of laser vision detection provides a non-contact, high-resolution measurement method that does not interfere with the welding process. The finding that measurement accuracy is commensurate with natural process variation is particularly significant, as it establishes a benchmark for the minimum required measurement precision in welding-based rapid manufacturing. For engineers developing process monitoring systems, the key lesson is that measurement accuracy must be evaluated relative to the process variability it is intended to detect, not in absolute terms. A highly accurate sensor that is more precise than the process noise floor provides no additional value and may introduce unnecessary complexity and cost. The laser vision approach described here strikes an appropriate balance between measurement precision, system complexity, and practical applicability.