Visual Detection of Copper Weld Pool Images in Non-Melting-Depth Weld Overlay

1. Definition and Technical Principles

The research titled "Visual Detection Method for Copper Weld Pool Images in Non-Melting-Depth Weld Overlay" addresses a critical process monitoring technology for cold weld overlay applications. Non-melting-depth weld overlay—also referred to as cold weld transfer or non-penetrating weld overlay—is a specialized cladding technique in which the deposited metal (in this case, copper or copper alloys) is transferred onto the substrate surface through a controlled welding arc without achieving full penetration into the base material. The weld pool exists exclusively within the deposited copper layer, maintaining a distinct metallurgical boundary with the underlying substrate.

The visual detection methodology employs high-speed imaging sensors, typically equipped with narrow-band optical filters, to capture real-time weld pool morphology during the overlay process. The system analyzes weld pool width, length, temperature gradient distribution, and surface topology to determine whether the deposition process remains within the non-melting-depth regime. Key optical parameters include:

The fundamental principle relies on the fact that when the welding arc energy is insufficient to penetrate through the copper deposit into the base metal, the weld pool exhibits characteristic thermal signatures and geometric profiles that are visually distinguishable from penetrating welds. The visual detection system identifies these signatures through image processing algorithms that analyze pool boundary sharpness, temperature gradient uniformity, and surface spatter patterns.

2. Category and Business Positioning

This technology falls under the advanced process monitoring and quality assurance category within Cladding Technology Shanxi Co., Ltd.'s technical capability portfolio. It represents a research-driven innovation that bridges the gap between conventional weld overlay manufacturing and intelligent process control systems. The study positions the company at the forefront of:

Within the company's three primary technology routes—TIG/MIG weld overlay, hydraulic explosive bonding, and explosion welding—this visual detection technology primarily supports the TIG/MIG weld overlay route but also provides complementary quality verification capabilities for explosion welding cladding interfaces where copper transition layers are deposited.

3. Technical Purpose and Value

3.1 Process Control Objectives

The primary technical purpose of the visual detection system is to ensure that the copper weld overlay process maintains non-penetrating conditions throughout the entire deposition sequence. This is critical for several engineering reasons:

3.2 Quality Assurance Value

The visual detection system provides continuous, non-contact process verification that complements post-weld destructive and non-destructive testing. Its value proposition includes:

4. Key Process and Implementation Points

4.1 Non-Melting-Depth Weld Overlay Process Parameters

Parameter Typical Range Non-Penetrating Condition Penetrating Condition
Welding Current (DC) 80–180 A Lower 40% of range Upper 60% of range
Travel Speed 80–250 mm/min Higher speeds Lower speeds
Wire Feed Rate (MIG) 1.5–4.0 m/min Proportional to current Excessive relative to current
Shielding Gas Ar / Ar+CO₂ Pure Ar preferred Variable
Deposition Height 2–8 mm ≥2 mm maintained <2 mm (risk of penetration)
Interpass Temperature ≤150°C Maintained below threshold Exceeds threshold

4.2 Visual Detection System Configuration

Component Specification Function
Industrial Camera Monochrome, 5MP, global shutter High-resolution pool imaging
Optical Lens Macro lens, 65 mm focal length Close-up pool capture
Light Source LED array, 620 nm center wavelength Controlled illumination
Optical Filter Bandpass, 620±10 nm Arc interference rejection
Image Processor Real-time FPGA or GPU Frame-by-frame analysis
Control Interface PLC/IPC communication Process parameter feedback

4.3 Image Analysis Algorithms

The visual detection system employs a multi-stage image processing pipeline:

  1. Pre-processing: Background subtraction, noise filtering (median filter, 3×3 kernel), and contrast enhancement to isolate the weld pool from the surrounding workpiece surface
  2. Pool Boundary Extraction: Edge detection using Canny algorithm with adaptive thresholds calibrated to the thermal gradient at the pool periphery
  3. Morphological Analysis: Calculation of pool area, major/minor axis dimensions, eccentricity, and aspect ratio to characterize pool geometry
  4. Thermal Signature Classification: Machine learning classification (trained on labeled datasets) to distinguish non-penetrating pool signatures from penetrating pool signatures based on boundary sharpness, internal temperature uniformity, and surface spatter patterns
  5. Decision Logic: Binary classification output (non-penetrating/penetrating) with confidence level, triggering process parameter adjustment or alarm if penetration risk is detected

4.4 Critical Weld Pool Image Features

Feature Non-Penetrating Pool Penetrating Pool Detection Method
Pool Boundary Sharpness Well-defined, high contrast Diffuse, low contrast Edge gradient magnitude
Pool Aspect Ratio Elongated (2.0–3.5) Near-circular (1.0–1.8) Axis ratio calculation
Surface Spatter Minimal, localized Excessive, widespread Particle detection algorithm
Pool Temperature Uniformity High uniformity Lower uniformity Pixel intensity variance
Leading Edge Shape Rounded, smooth Irregular, jagged Fourier descriptor analysis

5. Applicable Standards and Acceptance Criteria

5.1 Process Standards

5.2 Copper Cladding Standards

5.3 Acceptance Criteria for Non-Melting-Depth Copper Overlay

Acceptance Parameter Criteria Verification Method
Base Metal Penetration Zero penetration (no fusion to substrate) Visual detection system + macrograph examination
Copper Deposit Thickness ±0.5 mm tolerance on specified thickness Ultrasonic thickness measurement (GB/T 11344)
Interface Bond Strength ≥70% of base metal tensile strength (minimum) Shear bond test (ASTM E8/E8M)
Surface Quality No cracks, porosity >1 mm, or undercut Visual examination (ASTM E1657)
Visual Detection System Accuracy ≥95% classification accuracy for non-penetrating vs. penetrating Calibration with known process conditions

6. Common Risks and Controls

6.1 Process Risks

Risk Cause Consequence Control Measure
Accidental penetration into base metal Excessive welding current or travel speed reduction Base metal dilution, loss of mechanical properties Real-time visual monitoring with automatic parameter adjustment
Incomplete bonding at copper-base interface Insufficient heat input or contamination at interface Delamination, reduced bond strength Pre-weld surface preparation verification; interpass temperature monitoring
Porosity in copper deposit Inadequate shielding gas coverage Reduced conductivity, potential crack initiation Gas flow rate verification; trailing shield configuration
Cracking in copper weld metal Excessive cooling rate or hydrogen absorption Structural failure Pre-heat control; low-hydrogen consumables; post-weld stress relief

6.2 Visual Detection System Risks

7. Application Scenarios Across Company Technology Routes

7.1 TIG/MIG Weld Overlay Applications

The visual detection technology is most directly applicable to the company's TIG and MIG weld overlay operations, particularly for copper cladding on electrical and thermal management components:

7.2 Hydraulic Explosive Bonding Applications

While hydraulic explosive bonding (hydrodynamic metal forming) does not involve traditional welding, the visual detection technology supports quality verification of copper transition layers deposited by weld overlay prior to bonding:

7.3 Explosion Welding Applications

In explosion welding operations, the visual detection technology provides complementary quality assurance for copper transition layers:

8. Contribution to Qualification Building, Product Delivery, and Customer Value

8.1 Qualification Building

The visual detection technology directly supports the company's qualification programs under ASME BPVC Section IX and NB/T standards by providing:

8.2 Product Delivery Enhancement

The technology enhances product delivery through:

8.3 Customer Value Creation

The visual detection technology creates measurable customer value by:

9. Implementation Roadmap and Future Development

9.1 Near-Term Implementation (0–12 Months)

9.2 Medium-Term Development (12–24 Months)

9.3 Long-Term Vision (24–36 Months)

10. Conclusion

The research on visual detection of copper weld pool images in non-melting-depth weld overlay represents a significant advancement in process monitoring technology for cold weld overlay applications. By providing real-time, non-contact verification of weld pool conditions, this technology enables the company to deliver copper cladding products with verified non-penetrating interfaces, enhanced quality documentation, and reduced production risk. The methodology aligns with international standards for visual testing and non-destructive examination while establishing proprietary capabilities that differentiate the company in the competitive cladding technology market.

As the company continues to expand its technical capabilities across TIG/MIG weld overlay, hydraulic explosive bonding, and explosion welding routes, the visual detection technology will serve as a foundational element for intelligent manufacturing systems, supporting qualification programs, enhancing product delivery reliability, and creating measurable customer value through verifiable process control and quality assurance.