Vision-Based Weld Pool Width Detection in TIG Welding of Copper-Clad Aluminum Cable
Vision-based weld pool width detection represents a critical advancement in real-time process monitoring for TIG (Tungsten Inert Gas) welding applications, particularly in the manufacturing of copper-clad aluminum (CCA) cable. This technology leverages machine vision systems to capture, process, and analyze the geometry of the weld pool during active welding, enabling closed-loop control of weld parameters to ensure consistent cladding quality, dimensional accuracy, and metallurgical integrity. For Cladding Technology Shanxi Co., Ltd., this capability bridges the gap between traditional manual TIG weld overlay and automated, data-driven production systems capable of meeting the stringent quality demands of electrical cable manufacturers, power utilities, and aerospace supply chains.
1. Definition and Fundamental Principles
1.1 Core Concept
Vision-based weld pool width detection refers to the use of optical imaging systems—typically high-speed cameras equipped with narrow-band filters or structured light sources—to capture the morphology of the molten weld pool during TIG welding. The system extracts the weld pool width (and often length, area, and surface temperature distribution) from the captured images in real time, converting raw pixel data into quantitative dimensional measurements that can be fed back into the welding control system.
1.2 Physical Basis of Weld Pool Imaging
The weld pool emits intense thermal radiation across the visible and near-infrared spectrum due to temperatures exceeding 2,000°C at the arc contact point. Key optical phenomena exploited by vision systems include:
- Thermal radiation intensity: The brightness of the weld pool correlates with local temperature, allowing thermal mapping from intensity profiles.
- Surface tension boundary: The free surface of the molten pool is delineated by the transition from liquid metal to solidified weld bead, visible as a sharp intensity gradient.
- Plasma arc reflection: The arc itself creates specular highlights that can interfere with pool boundary detection and must be filtered or compensated.
1.3 Image Processing Pipeline
- Image acquisition: High-speed camera (typically 25–120 fps) captures images of the weld zone through a protective filter (e.g., OD 10–14 neutral density or band-pass filter in the 800–1100 nm range).
- Preprocessing: Noise reduction, background subtraction, and arc glare suppression are applied to isolate the weld pool region.
- Segmentation: Thresholding, edge detection (Canny, Sobel), or machine learning classifiers (e.g., U-Net, YOLO) segment the weld pool from the surrounding base material.
- Dimensional extraction: The pool width is measured as the maximum lateral extent of the segmented region, calibrated to real-world units (mm) using a known reference or camera calibration matrix.
- Feedback control: The measured width is compared against a target setpoint, and deviations trigger adjustments to welding current, travel speed, arc length, or wire feed rate.
2. Category and Business Positioning
2.1 Technology Classification
Within Cladding Technology Shanxi Co., Ltd.'s capability portfolio, vision-based weld pool monitoring falls under Intelligent Process Monitoring and Quality Assurance. It is not a standalone cladding process but rather an enabling technology that enhances the precision, repeatability, and auditability of TIG weld overlay operations—particularly for thin-clad or dissimilar-metal joints such as copper-clad aluminum cable.
2.2 Strategic Positioning
- Process intelligence layer: Adds real-time metrology to existing TIG/MIG weld overlay equipment without requiring fundamental changes to the welding power source or torch design.
- Quality assurance enabler: Provides objective, traceable process data that supports qualification records, lot traceability, and customer audits.
- Differentiation driver: Distinguishes the company's automated production lines from competitors relying solely on manual visual inspection and post-weld dimensional checks.
3. Technical Purpose and Value
3.1 Primary Technical Objectives
- Weld geometry control: Maintain weld pool width within specified tolerances (typically ±0.5 mm for cable cladding applications) to ensure uniform clad thickness and bonding quality.
- Defect prevention: Detect anomalies such as pool oscillation (indicating arc wander), excessive pool depression (risk of burn-through), or asymmetric pool shape (indicating torch misalignment).
- Parameter optimization: Accumulate process data to refine WPS (Welding Procedure Specification) parameters and reduce trial-and-error during new product qualification.
- Operator training and standardization: Provide visual feedback to operators during manual TIG welding, reducing skill-dependency and improving consistency across shifts.
3.2 Quantifiable Value Metrics
| Value Dimension | Without Vision Monitoring | With Vision Monitoring |
|---|---|---|
| Weld width tolerance | ±1.5–2.0 mm (manual) | ±0.3–0.5 mm (closed-loop) |
| Scrap/rework rate | 3–8% (typical) | <1.5% |
| Inspection time per lot | Manual sampling, 30–60 min | Continuous, automated logging |
| Operator skill requirement | High (5+ years experienced) | Moderate (guided by real-time feedback) |
| Process traceability | Post-hoc dimensional checks only | Full parameter history per meter |
4. Key Process and Implementation Points
4.1 System Architecture
A complete vision-based weld pool monitoring system for CCA cable TIG welding comprises the following subsystems:
| Subsystem | Key Specification | Function |
|---|---|---|
| Camera | Monochrome CMOS, ≥1024×768, ≥30 fps, global shutter | High-speed image capture of weld zone |
| Optics | Telecentric or macro lens, DOF ≥5 mm at working distance | Minimize perspective distortion; maintain focus across pool depth |
| Filter | Band-pass 800–1100 nm or OD 10+ ND filter | Suppress arc UV/visible glare; enhance pool thermal signal |
| Lighting | Structured LED (850 nm) or laser line scan | Provide contrast for pool boundary definition |
| Processing unit | Industrial PC or embedded GPU (e.g., NVIDIA Jetson) | Real-time image processing and control signal generation |
| Communication | RS-485 / EtherCAT / OPC-UA | Interface with welding power source and motion controller |
4.2 Critical Parameters for CCA Cable TIG Welding
The following parameters govern the TIG welding process for copper-clad aluminum cable and must be monitored and controlled via the vision system:
| Parameter | Typical Range (CCA Cable) | Effect on Weld Pool Width | Monitoring Priority |
|---|---|---|---|
| Welding current | 40–120 A (AC/DC depending on joint design) | Directly proportional; higher current → wider pool | Critical |
| Travel speed | 50–200 mm/min | Inversely proportional; higher speed → narrower pool | Critical |
| Arc length | 2–4 mm | Longer arc → wider, flatter pool with reduced penetration | High |
| Torch angle | 0°–15° (traveling direction) | Asymmetric pool if misaligned | High |
| Shielding gas flow | 8–15 L/min (Ar or Ar/He mix) | Indirect; affects arc stability and pool surface calmness | Moderate |
| Filler wire diameter | 1.0–2.4 mm (Cu or Cu-Al alloy) | Larger wire → wider pool; affects bead profile | Moderate |
4.3 Closed-Loop Control Strategy
The vision system operates within a feedback control loop:
- Measurement: Weld pool width (Wmeasured) is extracted from each camera frame.
- Comparison: Wmeasured is compared against the target width (Wtarget) defined in the WPS.
- Error calculation: ΔW = Wmeasured − Wtarget.
- Control action:
- If ΔW > +tolerance (pool too wide): Increase travel speed by 5–10% or decrease current by 2–5%.
- If ΔW < −tolerance (pool too narrow): Decrease travel speed or increase current.
- Logging: All measurements and control actions are timestamped and stored for traceability.
4.4 Calibration and Validation
- Geometric calibration: Use a calibrated target (e.g., checkerboard or known-width reference) to establish pixel-to-mm conversion. Accuracy target: ±0.1 mm at working distance.
- Thermal calibration: If temperature mapping is performed, calibrate intensity-to-temperature using blackbody references or emissivity-corrected models for copper and aluminum surfaces.
- Periodic verification: Validate system accuracy weekly using a known-width test weld or calibration gauge.
5. Applicable Standards and Acceptance Criteria
5.1 Welding Procedure and Qualification Standards
- GB/T 985.1-2008 (Welding procedure specification — General rules): Governs the preparation and approval of WPS documents including monitoring parameters.
- GB/T 15058-2008 (Rules for qualification of welding procedures): Requires documented evidence of weld quality; vision data serves as objective process monitoring records.
- ASME Section IX: If the CCA cable welds are part of pressure vessel or piping applications, the WPS and PQR must comply with Section IX qualification requirements.
- ASTM E165-18 (Standard Guide for Radiographic Examination): While primarily for RT, the traceability principles are analogous to vision-based monitoring documentation.
5.2 Cable and Cladding Product Standards
- GB/T 3955-2009 (Copper-clad aluminum wire and cable): Specifies dimensional tolerances, tensile strength, and conductivity requirements for CCA cable.
- ASTM B180-18 (Standard Specification for Copper-Clad Aluminum Wire): Defines clad thickness ratio (typically 10–25% by cross-sectional area) and bonding quality requirements.
- IEC 61159 (Copper-clad aluminium wire and cable): International standard for CCA wire with dimensional and mechanical acceptance criteria.
- NB/T 47013.2-2015 (Non-destructive testing — Radiographic testing): If RT verification is required post-weld, this standard governs the acceptance criteria for weld defects.
5.3 Acceptance Criteria for Vision Monitoring System
| Criterion | Acceptance Threshold | Verification Method |
|---|---|---|
| Measurement accuracy | ±0.3 mm (1σ) on weld pool width | Comparison with calibrated micrometer on test coupons |
| Measurement repeatability | CV ≤ 2% over 100 consecutive readings | Statistical analysis of continuous monitoring data |
| System latency | ≤ 100 ms from image capture to control signal | Timing test with oscilloscope on I/O signals |
| Defect detection sensitivity | Detect pool width deviation ≥ 1.0 mm | Simulated defect injection during test weld |
| Data logging completeness | 100% of frames logged with timestamp and parameters | Post-weld data audit against expected frame count |
6. Common Risks and Controls
6.1 Technical Risks
| Risk | Cause | Impact | Mitigation |
|---|---|---|---|
| Arc glare obscuring pool boundary | Insufficient optical filtering; high arc intensity | False width readings; control instability | Use OD 12+ ND filter or band-pass filter; add arc suppression algorithm |
| Spatter contamination of lens | Metal spatter from arc or filler wire | Progressive image degradation; measurement drift | Install protective glass window; implement automated cleaning cycle; use air purge around lens |
| Camera misalignment during production | Mechanical vibration; thermal expansion of mount | Systematic measurement bias | Use rigid mounting; implement auto-calibration routine at shift start |
| False positive defect alarms | Noise in image processing; threshold too tight | Unnecessary weld interruptions; reduced throughput | Tune detection thresholds based on historical data; implement moving-average filtering |
| Electromagnetic interference | High welding currents inducing noise in camera signal | Image corruption; data loss | Shield camera cable with braided shield; use isolated power supply; ground camera chassis |
6.2 Process Risks in CCA Cable Welding
- Intermetallic compound formation: Excessive weld pool width (indicating excessive heat input) promotes formation of brittle Cu-Al intermetallic phases (CuAl₂, Cu₅Al₈), reducing joint ductility. Control: Limit pool width to specified maximum; monitor thermal input via pool area estimation.
- Incomplete bonding: Insufficient pool width (low heat input) results in poor metallurgical bonding between copper and aluminum layers. Control: Set minimum pool width threshold; alarm if width drops below specification.
- Weld porosity: Unstable pool geometry (oscillation, irregular shape) can trap gas inclusions. Control: Monitor pool shape factor (aspect ratio) in addition to width; alarm on shape irregularity.
- Crack initiation: Thermal stress from uneven pool geometry can initiate cracks in the brittle intermetallic zone. Control: Ensure uniform pool width along the weld length; flag any deviation exceeding 15% of nominal.
7. Application Scenarios Across Company Technology Routes
7.1 TIG/MIG Weld Overlay Route
For the company's primary TIG and MIG weld overlay operations, vision-based weld pool width detection provides the following specific value:
- Transition layer welding: When welding a 309L or 2205 transition layer between dissimilar metals (e.g., carbon steel and stainless steel), pool width directly affects dilution ratio. Vision monitoring ensures the dilution remains within the specified range (typically 20–40% for 309L on carbon steel), preventing under-alloying (loss of corrosion resistance) or over-alloying (excessive cost and potential cracking).
- Multi-pass overlay build-up: In multi-pass cladding, each pass must maintain consistent width to achieve uniform clad thickness. Vision monitoring detects pass-to-pass variation and triggers automatic compensation.
- CCA cable production: The direct application described in the technical entry—monitoring pool width during TIG welding of copper-clad aluminum cable to ensure uniform clad thickness and bonding quality.
- WPS qualification welding: During procedure qualification trials (per GB/T 15058 or ASME Section IX), vision monitoring provides objective data demonstrating process control, strengthening the qualification case.
7.2 Hydraulic Explosive Bonding (HEB) Route
While HEB does not involve welding, vision-based monitoring contributes in supporting roles:
- Pre-bond TIG welding of backing plates: HEB stacks often require TIG-welded backing plates or containment fixtures. Vision monitoring ensures these auxiliary welds meet quality standards.
- Post-bond inspection support: The same vision system can be repurposed for automated visual inspection of bonded interfaces, detecting surface defects, wrinkles, or contamination on HEB clad plates.
- Fixture fabrication quality: HEB requires precision-machined and welded fixtures. Vision monitoring during fixture fabrication ensures dimensional accuracy critical to bonding quality.
7.3 Explosion Welding Route
Similar to HEB, explosion welding benefits from vision monitoring in ancillary processes:
- TIG welding of explosion welding backing plates: Backing plates must be welded to specific geometric tolerances. Pool width monitoring ensures consistent weld geometry.
- Post-explosion weld seam inspection: After explosion welding, seam welds joining clad panels can be inspected using the vision system for surface quality and dimensional compliance.
- Production line integration: In integrated explosion welding + TIG finishing lines, the vision system provides continuous quality data across both processes, enabling holistic process optimization.
8. Contribution to Qualification Building, Product Delivery, and Customer Value
8.1 Qualification Building
- WPS data enrichment: Vision monitoring data (pool width, travel speed, current, arc length over time) provides the objective process parameter records required for WPS qualification per GB/T 15058-2008 and ASME Section IX. This transforms subjective "operator judgment" into quantifiable, auditable process evidence.
- PQR support: Performance qualification records benefit from documented process monitoring data demonstrating consistent weld geometry throughout the test coupon, reducing the risk of qualification rejection.
- Operator certification: Real-time vision feedback enables systematic operator training and certification, with performance metrics tracked over time.
- ISO 3834 compliance: The vision monitoring system supports the documented process control requirements of ISO 3834 (Quality requirements for fusion welding of metallic materials), particularly Clause 5.5 (Welding procedures) and Clause 5.6 (Welding operator qualification).
8.2 Product Delivery Enhancement
- Reduced rework: Real-time defect detection prevents the production of non-conforming welds, reducing rework rates by an estimated 60–80%.
- Increased throughput: Automated monitoring eliminates the need for frequent manual inspection stops, increasing effective production rate by 15–25%.
- Lot traceability: Every meter of welded cable or every square meter of clad plate has an associated process data record, enabling rapid root-cause analysis in the event of a quality issue.
- Consistent quality: Elimination of operator-dependent variation ensures uniform product quality across shifts, operators, and production runs.
8.3 Customer Value Proposition
- Quality assurance documentation: Customers receive comprehensive process monitoring data with each delivery, providing confidence in product quality and simplifying their incoming inspection procedures.
- Reduced warranty claims: Consistent process control reduces the incidence of field failures, protecting the company's reputation and reducing warranty costs.
- Competitive differentiation: The ability to provide real-time process monitoring data distinguishes the company from competitors who rely solely on post-production inspection.
- Regulatory compliance support: For customers in regulated industries (nuclear, aerospace, oil and gas), the vision monitoring data supports compliance with NACE, API, and ASME requirements for documented process control.
9. Implementation Roadmap and Recommendations
9.1 Phased Implementation
| Phase | Timeline | Activities | Deliverables |
|---|---|---|---|
| Phase 1: Feasibility | Month 1–2 | Literature review; prototype camera setup; image capture trials on CCA cable test welds | Feasibility report; preliminary accuracy data |
| Phase 2: Development | Month 3–6 | Image processing algorithm development; closed-loop control integration; calibration procedure | Functional prototype; calibration procedure document |
| Phase 3: Validation | Month 7–9 | Systematic accuracy testing; comparison with manual measurement; WPS parameter correlation | Validation report; updated WPS with monitoring parameters |
| Phase 4: Deployment | Month 10–12 | Production line integration; operator training; SOP development; quality system integration | Production-ready system; trained operators; updated quality manual |
9.2 Key Success Factors
- Cross-functional team: Assemble a team comprising welding engineers, vision system specialists, quality engineers, and production operators.
- WPS integration: Incorporate pool width as a monitored parameter in all relevant WPS documents from the outset.
- Operator engagement: Involve operators in the development phase to ensure the system is user-friendly and builds trust.
- Continuous improvement: Establish a feedback loop where production data informs algorithm refinement and WPS optimization.
- Standards alignment: Ensure all documentation and acceptance criteria align with applicable standards (GB/T 15058, ASME Section IX, ISO 3834, ASTM B180).
10. Conclusion
Vision-based weld pool width detection in TIG welding of copper-clad aluminum cable represents a transformative capability for Cladding Technology Shanxi Co., Ltd. By converting the weld pool from an invisible, rapidly changing phenomenon into a measurable, controllable process parameter, this technology enables unprecedented levels of quality assurance, process traceability, and production efficiency. When integrated across the company's three technology routes—TIG/MIG weld overlay, hydraulic explosive bonding, and explosion welding—it provides a unified quality monitoring framework that strengthens qualification records, enhances product delivery reliability, and delivers measurable value to customers in demanding industrial markets. The investment in this capability positions the company at the forefront of intelligent cladding manufacturing, supporting the transition from experience-based production to data-driven, standards-compliant manufacturing excellence.