FLOW-3D-Based Numerical Analysis Model for GMAW Weld Pool Behavior
1. Definition and Fundamental Principles
The FLOW-3D-based numerical analysis model for GMAW (Gas Metal Arc Welding) weld pool behavior is a computational fluid dynamics (CFD) simulation approach that leverages the proprietary Volume-of-Fluid (VOF) method within the FLOW-3D solver to predict and analyze the complex multiphysics phenomena occurring within the weld pool during gas metal arc welding overlay processes. This model integrates fluid dynamics, heat transfer, electromagnetic forces, and mass transfer into a unified computational framework to simulate the transient evolution of weld pool geometry, temperature field distribution, velocity vector fields, and solidification patterns.
The governing equations underlying this model include:
- Navier-Stokes equations for momentum conservation with Boussinesq approximation for buoyancy-driven natural convection
- Energy equation with arc heat source models (Gaussian double-ellipsoidal or modified Gaussian) to represent the moving heat input
- VOF equation for tracking the liquid-solid interface and free surface of the weld pool
- Electromagnetic force equations incorporating Lorentz force (J × B) and electromagnetic stirring effects
- Marangoni convection driven by surface tension gradients induced by temperature variations
- Mass transfer equations accounting for wire feed deposition, dilution rates, and alloying element diffusion
The FLOW-3D solver employs a sharp-interface VOF method on a Cartesian computational grid, providing superior resolution of the weld pool boundary compared to diffuse-interface approaches. This enables accurate prediction of weld bead geometry, penetration profiles, and fusion boundary morphology that are critical for weld overlay qualification and process optimization.
2. Category and Business Positioning
Within the technical capability framework of Cladding Technology Shanxi Co., Ltd., the FLOW-3D-based GMAW weld pool numerical analysis model is categorized as a process simulation and digital engineering tool that supports all three primary manufacturing routes: TIG/MIG weld overlay, hydraulic explosive bonding, and explosion welding. Its primary business positioning is as follows:
2.1 Role in Weld Overlay Technology (TIG/MIG)
The model serves as the digital backbone for GMAW/MIG weld overlay process development and qualification. It enables virtual WPS (Welding Procedure Specification) optimization before physical trial welding, significantly reducing the cost and time associated with traditional trial-and-error qualification procedures. The model supports prediction of dilution rates, bead overlap ratios, layer build-up geometry, and residual stress initiation zones — all critical parameters for meeting ASME Section IX, AWS D10.9, and NB/T 25072 requirements.
2.2 Role in Hydraulic Explosive Bonding
While the primary bonding mechanism in hydraulic explosive bonding is high-velocity jet impact, the numerical model contributes to understanding the thermal effects at the interface during post-bonding stabilization and to the design of any transition weld layers that may be required at clad plate edges or repair areas.
2.3 Role in Explosion Welding
For explosion welding applications, the model assists in simulating the thermal boundary conditions at the explosive weld interface and predicting the effects of post-explosion welding repairs or transition layers that are often required to achieve metallurgical continuity at clad plate edges.
3. Technical Purpose and Value
3.1 Process Qualification Acceleration
The primary value of the FLOW-3D-based GMAW weld pool model lies in its ability to dramatically accelerate the WPS qualification cycle. Traditional qualification of a new weld overlay procedure may require 3–8 physical trial runs, each involving material procurement, welding execution, NDT testing, mechanical testing, and metallographic examination. The numerical model reduces this to 1–2 physical trials by providing pre-qualification predictions of:
- Weld pool geometry (width, depth, penetration profile)
- Dilution rate prediction for clad material into base metal
- Interpass temperature control strategy
- Layer build-up geometry and cumulative deformation
- Solidification defect susceptibility (hot cracking, porosity, lack of fusion)
3.2 Product Delivery Quality Assurance
For production delivery, the model provides a digital reference for process monitoring. When physical process parameters deviate from the qualified WPS, the model can predict the likely impact on weld quality, enabling immediate corrective action. This is particularly valuable for complex multi-layer clad pipe and large-diameter clad plate production where real-time process adjustment is challenging.
3.3 Customer Value Enhancement
The numerical analysis model elevates the company's technical proposition to customers by demonstrating:
- Engineering rigor — data-driven process development rather than empirical guesswork
- Traceability — simulation records provide documented justification for process parameters
- Predictive capability — ability to forecast weld performance before physical production
- Cost optimization — reduction of trial welding costs, material waste, and rework
- Customization — rapid adaptation of process parameters for novel clad material combinations
4. Key Process and Implementation Points
4.1 Model Configuration and Parameter Setup
| Parameter Category | Key Variable | Typical Range/Value | Source/Standard Reference |
|---|---|---|---|
| Heat Input | Arc power (P) | 3–15 kW | WPS-specific |
| Heat Input | Travel speed (v) | 50–300 mm/min | WPS-specific |
| Heat Input | Heat input (q = P/v) | 0.2–0.8 kJ/mm | ASME Section IX |
| Wire Feed | Wire diameter (d) | 1.0–1.6 mm | AWS A5.18 / A5.22 |
| Wire Feed | Feed rate (WFR) | 1.0–6.0 m/min | WPS-specific |
| Heat Source | Arc radius (r) | 0.5–2.0 mm | FLOW-3D Gaussian model |
| Heat Source | Energy fraction front (η) | 0.4–0.6 | Double-ellipsoidal model |
| Thermal Properties | Thermal conductivity (k) | Material-dependent (20–40 W/m·K for steels) | ASTM E1461 / supplier data |
| Thermal Properties | Specific heat (c) | Material-dependent (450–650 J/kg·K) | ASTM E1461 |
| Thermal Properties | Latent heat of fusion (L) | 200–270 kJ/kg | Material-specific |
| Surface Tension | Surface tension gradient (dγ/dT) | -0.03 to -0.08 N/m·K | Experimental measurement |
| Electromagnetic | Current density (J) | 10⁶–10⁸ A/m² | Model calculation |
| Shielding Gas | Gas composition | Ar/CO₂, Ar/He, Ar/O₂ | AWS A5.1 / WPS |
4.2 Computational Domain and Mesh Design
The computational domain must encompass the weld pool region with sufficient margin to capture heat conduction into the base material. Key mesh design considerations include:
- Domain size: Typically 20–30 mm width × 15–20 mm height × 10–15 mm length for single-pass simulation
- Mesh resolution: Critical refinement (0.05–0.1 mm) within the weld pool and fusion boundary; coarser mesh (0.5–1.0 mm) in the far-field base material
- Total cell count: 500,000–3,000,000 cells for production-grade accuracy
- Grid type: Adaptive mesh refinement (AMR) near the weld pool boundary for computational efficiency
- Boundary conditions: Adiabatic on lateral faces (except heat flux input); convective-radiative on top surface; fixed temperature or adiabatic on bottom surface
4.3 Heat Source Model Selection
The accuracy of weld pool prediction is highly sensitive to the heat source model employed. The following comparison guides model selection:
| Heat Source Model | Applicable Weld Pool Aspect Ratio | Advantages | Limitations | Recommended Use |
|---|---|---|---|---|
| Gaussian Surface | Wide, shallow (a < 1) | Simple implementation; fast convergence | Poor penetration prediction | High-speed, low-penetration GMAW |
| Single Ellipsoidal (Rosenthal) | Moderate (a ≈ 1) | Better penetration shape | Cannot represent front/back asymmetry | General-purpose simulation |
| Double Ellipsoidal (Goldak) | Deep, narrow (a > 1) | Captures front/back energy distribution | More parameters to calibrate | High-penetration GMAW overlay |
| Hybrid Gaussian-Ellipsoidal | Variable | Flexibility across parameter ranges | Complex calibration | Multi-configuration qualification |
4.4 Multi-Physics Coupling Implementation
The FLOW-3D model for GMAW weld pool analysis requires coupling of the following physics modules:
- Thermal module: Transient heat conduction-convection with moving heat source; enthalpy-porosity method or VOF for solidification tracking
- Fluid dynamics module: Incompressible Navier-Stokes with Boussinesq buoyancy; k-ε or k-ω turbulence model for high-Reynolds-number flow
- Electromagnetic module: Induced current density calculation from welding current; Lorentz force source term in momentum equation
- Surface tension module: Temperature-dependent surface tension with Marangoni shear stress on free surface
- Mass transfer module: Wire feed deposition as volumetric source; dilution calculation from mixing zone
4.5 Validation Protocol
Every FLOW-3D model configuration must be validated against experimental data before use in production qualification. The validation protocol includes:
- Weld pool shape validation: Comparison of predicted vs. measured (via sectioning or CT scanning) weld bead cross-section geometry
- Penetration depth validation: Within ±15% of experimental measurement
- Weld width validation: Within ±10% of experimental measurement
- Dilution rate validation: Predicted dilution within ±3 percentage points of spectrographic measurement
- Temperature field validation: Comparison with thermocouple or infrared pyrometer measurements at defined positions
- Velocity field validation: Comparison with particle image velocimetry (PIV) or particle tracking data where available
5. Applicable Standards and Acceptance Criteria
5.1 Standards Governing GMAW Weld Overlay Processes
| Standard | Title/Scope | Relevance to Numerical Model |
|---|---|---|
| ASME Section IX | Welding, Brazing, and Fusing Qualifications | WPS qualification parameters must be within model-predicted acceptable ranges |
| AWS D10.9 | Recommended Practices for Welding in Piping | Process parameters for clad pipe GMAW overlay |
| NB/T 25072 | Welding Procedure Qualification for Overlay Welding | Chinese nuclear industry WPS qualification requirements |
| GB/T 985.1 | Welding Procedure Test Method | Test specimen preparation and evaluation methodology |
| GB/T 3375 | Basic Terms in Welding | Terminology and definitions for weld geometry parameters |
| ASTM A388 | Standard Specification for Clad Steel Plate | Performance requirements for clad plate produced with GMAW overlay |
| ASTM A240 | Standard Specification for Chromium and Chromium-Nickel Stainless Steel Plate | Clad material specification for stainless overlay |
| NACE MR0175/ISO 15156 | Materials for Use in H₂S-Containing Environments | Material qualification for sour service clad components |
| ASME B31.3 | Process Piping | Design and fabrication requirements for clad piping |
| API 5L | Specification for Line Pipe | Clad line pipe requirements for oil and gas applications |
| GB/T 11266 | Corrosion-Resistant Clad Steel Plate | Chinese standard for clad plate specifications and testing |
5.2 Acceptance Criteria for Model Predictions
The numerical model predictions must satisfy the following acceptance criteria to be considered valid for process qualification support:
- Weld geometry prediction accuracy: Weld width within ±10%, penetration depth within ±15%, reinforcement height within ±20% of experimental measurements
- Dilution prediction accuracy: Predicted dilution rate within ±3 percentage points of experimental spectrographic analysis
- Defect prediction: Model must correctly identify susceptibility to hot cracking (via solidification cracking index), porosity (via gas evolution model), and lack of fusion (via fusion boundary temperature criterion)
- Reproducibility: Model predictions must be consistent across multiple simulation runs (coefficient of variation < 5% for key output parameters)
- Parameter sensitivity: Model must demonstrate reasonable sensitivity to input parameter variations, with predictions changing directionally consistent with physical expectations
6. Common Risks and Controls
6.1 Model-Specific Risks
| Risk Category | Description | Potential Impact | Mitigation Control |
|---|---|---|---|
| Heat source model mismatch | Inappropriate heat source geometry for the actual welding configuration | Systematic error in weld pool shape prediction; incorrect dilution rate | Calibrate against at least 3 experimental welds spanning the intended parameter range; document model applicability limits |
| Thermal property inaccuracy | Use of generic or room-temperature property values instead of temperature-dependent data | Incorrect heat distribution; poor solidification prediction | Use experimentally measured temperature-dependent properties; validate with ASTM E1461 or equivalent testing |
| Boundary condition oversimplification | Inappropriate boundary conditions for the actual welding setup (e.g., ignoring back-side cooling) | Overprediction of penetration; incorrect cooling rate estimates | Include realistic boundary conditions; use conjugate heat transfer for multi-material systems |
| Numerical instability | Mesh-too-coarse or time-step-too-large leading to numerical oscillations or divergence | Unreliable results; wasted computational resources | Perform mesh convergence study; use adaptive time stepping; monitor Courant number |
| Over-reliance on model | Using model predictions without experimental validation for critical decisions | Unqualified WPS; product nonconformance | Mandate experimental validation before production use; maintain model validation database |
| Software version drift | Different FLOW-3D versions producing different results for identical inputs | Inconsistent qualification records; audit traceability issues | Lock software version for each qualification project; document version and solver settings |
6.2 Process-Specific Risks Addressed by the Model
- Hot cracking: Model identifies critical solidification temperature range and strain rate; recommends wire composition modifications or parameter adjustments to reduce susceptibility
- Excessive dilution: Model predicts dilution for each pass; identifies parameter combinations that maintain clad integrity above minimum required thickness
- Welding defects: Model identifies porosity-prone conditions (high hydrogen, low pressure), lack of fusion risk (low heat input, high travel speed), and undercut susceptibility
- Residual stress initiation: Model provides thermal cycle data that feeds into residual stress analysis for distortion prediction and stress relief planning
- Microstructure prediction: Thermal cycle data from model enables prediction of grain growth, phase transformation, and hardness distribution in the weld and HAZ
7. Application Across the Company's Three Technology Routes
7.1 TIG/MIG Weld Overlay Applications
The FLOW-3D GMAW weld pool model is most directly applicable to the MIG/GMAW weld overlay route, providing comprehensive process development support:
- Multi-layer clad plate development: Optimization of interpass temperature, layer thickness, and bead overlap for 3–8 layer stainless steel or Ni-based overlay on carbon steel base plates per ASTM A388 and GB/T 11266
- Clad pipe production: Prediction of internal and external weld overlay bead geometry for large-diameter pipes (DN500–DN3000) requiring rotational GMAW welding
- Transition layer design: Simulation of dilution behavior for transition welds between dissimilar materials (e.g., carbon steel to 309L to 316L sequence) ensuring adequate metallurgical compatibility
- High-alloy overlay qualification: WPS development for Hastelloy, Inconel, and Stellite overlay where dilution control is critical for corrosion resistance per NACE MR0175/ISO 15156
- Robotized welding parameter optimization: Model-based parameter selection for automated GMAW systems with wire feed tracking, ensuring consistent weld quality across long production runs
7.2 Hydraulic Explosive Bonding Applications
While hydraulic explosive bonding relies on kinetic energy transfer rather than thermal fusion, the GMAW weld pool model contributes in supporting applications:
- Edge transition welding: Prediction of weld pool behavior for GMAW transition welds at clad plate edges where the bonded interface terminates, ensuring proper fusion and dilution control
- Repair welding simulation: Modeling of repair welds on hydraulic explosive bonded clad surfaces where local damage requires GMAW overlay repair without compromising the bonded interface
- Post-bonding thermal analysis: Understanding of thermal effects during any post-bonding stabilization welding on the base material side
- WPS development for hybrid joints: Qualification of welded-bonded hybrid connections where GMAW welds join to or integrate with explosively bonded interfaces
7.3 Explosion Welding Applications
For explosion welding (air-gap and water-gap), the model supports complementary welding operations:
- Edge weld qualification: Simulation of GMAW welds used to seal or transition explosion-welded clad plate edges, predicting dilution and weld geometry for qualification testing
- Clad pipe end preparation: Modeling of weld overlay beads applied to explosion-welded pipe ends for subsequent joining operations per API 5L and ASME B31.3
- Crater area repair: Prediction of weld pool behavior for GMAW repair of crater areas (local defects) on explosion-welded surfaces
- Multi-layer repair builds: Design of multi-pass GMAW builds for repair of significant defects in explosion-welded cladding, with dilution control for each subsequent pass
8. Integration with Qualification and Certification Systems
8.1 WPS Qualification Support
The FLOW-3D model integrates into the WPS qualification workflow as follows:
- Phase 1 — Virtual screening: Model screening of candidate parameter combinations to identify the most promising 2–3 configurations from a broader parameter space
- Phase 2 — Predictive modeling: Detailed simulation of selected configurations to predict weld geometry, dilution, and defect susceptibility
- Phase 3 — Experimental validation: Physical welding of model-selected parameters; comparison of results with predictions; model calibration if discrepancies exceed acceptance criteria
- Phase 4 — Qualification confirmation: Final WPS parameters confirmed by both model prediction and experimental testing; documented in qualification records
- Phase 5 — Production deployment: Model retained as reference for production monitoring and troubleshooting
8.2 Certification System Integration
The numerical analysis model supports the company's certification system by providing:
- Technical documentation for quality management system audits (ISO 9001, ISO 3834) demonstrating systematic process development
- Evidence of process control for ASME "U" stamp and "S" stamp qualification, showing data-driven parameter selection
- Customer technical submissions providing simulation-based justification for proposed welding procedures in bid proposals
- Regulatory compliance documentation for nuclear (NB/T 25072) and pressure equipment (TSG 21) applications requiring rigorous process qualification
9. Continuous Improvement and Future Development
The FLOW-3D-based GMAW weld pool model is not a static tool but a continuously evolving asset. Planned development directions include:
- Multi-pass coupling: Extension from single-pass to multi-pass sequential simulation with realistic interpass cooling, capturing the cumulative effects of multi-layer build-up
- Microstructure prediction integration: Coupling thermal cycle data with cellular automata or phase-field models to predict grain morphology and phase distribution
- Residual stress coupling: Integration with thermo-elastic-plastic finite element models for residual stress and distortion prediction
- Machine learning acceleration: Development of surrogate models trained on FLOW-3D simulation databases for rapid parameter optimization
- Real-time process monitoring: Development of simplified models for real-time comparison with actual process conditions, enabling closed-loop control
- Multi-physics extension: Incorporation of electromagnetic arc modeling, plasma flow simulation, and spatter prediction for comprehensive process understanding
10. Conclusion
The FLOW-3D-based numerical analysis model for GMAW weld pool behavior represents a critical digital engineering capability that enhances the technical foundation of Cladding Technology Shanxi Co., Ltd.'s weld overlay operations. By providing predictive, validated, and standards-aligned process analysis, the model accelerates WPS qualification, improves product consistency, reduces development costs, and elevates the company's technical credibility with customers and regulatory authorities. Its application extends across all three technology routes — directly supporting TIG/MIG weld overlay development and providing complementary support for hybrid welding operations associated with hydraulic explosive bonding and explosion welding. The model's ongoing refinement through validation, calibration, and feature extension ensures its continued relevance as the company's digital engineering platform evolves alongside advancing computational capabilities and industry requirements.