Five-Axis Continuous Sweep Scanning Method Based on Geodesic Distance Field for Multi-Entrance Inwall Surface Inspection

Background Introduction

In industrial applications, Multi-Entrance Inwall (MEI) surfaces pose significant challenges for precise inspection due to their complex topological structures and potential collision risks. Traditional point-by-point inspection methods are inefficient, while the recently developed Five-Axis Continuous Sweep Scanning Technology has significantly improved inspection efficiency, offering new possibilities for large-area and complex surface inspection. However, the current path planning for five-axis continuous scanning still heavily relies on manual intervention, especially for MEI surfaces, where automated path generation is particularly difficult due to complex collision scenarios and multi-entrance structures. To address this, this paper proposes a novel method based on the Geodesic Distance Field (GDF), aiming to automatically and efficiently generate five-axis continuous sweep paths to solve the path planning and collision avoidance issues in MEI surface inspection.

Source of the Paper

This paper was co-authored by Yuzhu Ding, Zhaoyu Li, Dong He, Kai Tang, and Pengcheng Hu, affiliated with the Department of Mechanical and Aerospace Engineering, The Hong Kong University of Science and Technology, and the Thrust of Smart Manufacturing, The Hong Kong University of Science and Technology (Guangzhou). The paper was published in the September 2020 issue of IEEE Transactions on Automation Science and Engineering.

Research Process and Main Content

1. Research Background and Problem

The five-axis Coordinate Measuring Machine (CMM) is one of the primary tools for surface geometry inspection. Traditional three-axis CMMs are less efficient, while the five-axis continuous scanning technology significantly improves inspection efficiency and accuracy through the synchronized movement of five axes. However, the complex multi-entrance structures and potential collision risks of MEI surfaces make automated path planning a significant challenge. This paper aims to address this by introducing the GDF to generate five-axis continuous sweep paths for the automatic and efficient inspection of MEI surfaces.

2. Basic Principles of Five-Axis Scan Path Generation

The five-axis CMM consists of a three-axis translation arm and a two-axis rotational probe. The collaborative movement of the five axes enables the probe to perform continuous scanning on the surface. The generation of the scan path primarily involves three steps: 1. Nominal Scan Path (Nominal SP) Generation: The nominal scan path is generated based on the guiding curve and probe orientation. 2. Nominal Scan Path Expansion: More nominal scan paths are iteratively generated to cover the entire surface. 3. Final Scan Path Generation: The final continuous scan path is generated by bridging adjacent nominal scan paths.

3. GDF-Based Guiding Curve Generation

To generate scan paths suitable for MEI surfaces, this paper proposes a GDF-based guiding curve generation method: 1. GDF Generation: By treating the entrances of the MEI surface as heat sources, the geodesic distance field is calculated using the heat conduction equation. The gradient field of the GDF is used to guide the generation of the guiding curve. 2. Guiding Curve Generation: Starting from a seed point, the guiding curve is traced along the gradient field direction of the GDF. The generated guiding curve aligns with the topological structure of the MEI surface, ensuring the continuity and efficiency of the scan path.

4. Probe Orientation Planning and Surface Partitioning

After generating the guiding curve, this paper further proposes an optimized probe orientation planning method to ensure collision-free and efficient scanning. Simultaneously, based on the GDF, the MEI surface is partitioned into multiple accessible regions, each corresponding to a guiding curve and a set of scan paths. This partitioning method minimizes overlap in the scan paths, improving inspection efficiency.

5. Experimental Results and Discussion

The effectiveness of the proposed method is validated through physical inspection experiments and computer simulations. The experimental results show that the GDF-based scan path generation method outperforms two existing benchmark methods in terms of inspection efficiency and surface coverage. Specifically: - Inspection Efficiency: The total inspection time of the proposed method is 253 seconds, reducing the time by 14.81% to 52.97% compared to other methods. - Surface Coverage: The proposed method covers 84.66% of the accessible area, improving coverage by 0.37% to 21.56% compared to benchmark methods.

Research Conclusion and Significance

This paper proposes a GDF-based five-axis continuous sweep scan path generation method, successfully addressing the path planning and collision avoidance issues in MEI surface inspection. By introducing the GDF, the generated long and smooth guiding curves significantly improve inspection efficiency. Additionally, the GDF-based surface partitioning method considers the accessibility of inspection points, further enhancing surface coverage. This method not only achieves automated path generation, reducing reliance on manual intervention, but also provides new technical means for efficient inspection of complex surfaces.

Research Highlights

1. Introduction of GDF for Guiding Curve Generation: The guiding curves generated from the geodesic distance field align with the topological structure of the MEI surface, significantly improving the continuity and efficiency of the scan paths.
2. Optimized Probe Orientation Planning: By considering collision avoidance and probe contact angles, the generated scan paths ensure collision-free and highly efficient inspection processes.
3. Surface Partitioning Method: The GDF-based surface partitioning method minimizes path overlap, improving inspection efficiency.

Future Research Directions

Future research will focus on optimizing the inspection of inaccessible regions on MEI surfaces and exploring the relationship between scan path characteristics and inspection accuracy. Additionally, more precise scan path generation methods will be developed to further enhance the automation level of five-axis CMMs.