Source-linked AI summary
Numerical modeling of friction stir welding process: a literature review
Diogo Mariano Neto, Pedro Neto
TL;DR
FSW modeling remains challenging because the process combines heat flow, high-temperature plastic deformation, and material flow, while some process behavior is poorly understood. This survey reviews contact-based heat-generation models, heat-loss formulations, and material-flow approaches, concluding that calibrated numerical models can support parameter selection but still cannot optimize the process.
Problem
FSW modeling is challenging because the multiphysics process is not fully understood and requires difficult contact, heat-flow, and material-flow representations.
Method
The survey reviews analytical and numerical approaches for contact-dependent heat generation, heat losses, and material-flow modeling, including their advantages and drawbacks.
Results
Heat loss into the tool is about 5%.
Takeaways & Limitations
Numerical FSW modeling can help tune process and material parameters with less effort and economic advantages than costly experimental trials.
Takeaways & Limitations
Current FSW models still cannot be used to optimize the process.
Abstract
from arXiv · showhide
This survey presents a literature review on friction stir welding (FSW) modeling with a special focus on the heat generation due to the contact conditions between the FSW tool and the workpiece. The physical process is described and the main process parameters that are relevant to its modeling are highlighted. The contact conditions (sliding/sticking) are presented as well as an analytical model that allows estimating the associated heat generation. The modeling of the FSW process requires the knowledge of the heat loss mechanisms, which are discussed mainly considering the more commonly adopted formulations. Different approaches that have been used to investigate the material flow are presented and their advantages/drawbacks are discussed. A reliable FSW process modeling depends on the fine tuning of some process and material parameters. Usually, these parameters are achieved with base on experimental data. The numerical modeling of the FSW process can help to achieve such parameters with less effort and with economic advantages.
1. INTRODUCTION
FSW is a solid-state joining process in which a rotating, translating tool generates heat, softens material, and forges it into a joint. Modeling must account for process parameters, thermomechanical effects, material flow, and difficult-to-measure temperature histories.
- Process description: FSW joins materials without melting by moving a non-consumable rotating tool through the joint line.The tool penetrates the joint, rotates, and translates along the welding direction.
- Process phases: The process comprises plunge, dwell, and welding phases, with the tool or workpiece moving along the joint during welding.During dwell, the tool remains stationary relative to the workpiece while continuing to rotate.
- Process description: Frictional work and plastic deformation generate heat, which softens material and enables severe deformation and flow around the tool.Material is transported from the tool front to its trailing edge, where it is forged into the joint.
- Process parameters: Tool rotation, traverse speed, axial force, tilt angle, and tool geometry influence heating, material flow, weld quality, and joint uniformity.Excessive pressure can cause overheating and thinning, whereas insufficient pressure can cause inadequate heating and voids.
- Weld microstructure: FSW produces distinct microstructural regions, including the stir zone, thermomechanically affected zone, and heat affected zone.The stir-zone grains are often an order of magnitude smaller than those in the base material, while the HAZ experiences thermal cycling without deformation.
- Numerical modeling: Numerical modeling is useful because several FSW aspects remain poorly understood and nugget temperature measurements are difficult.Models can help visualize flow, temperature, stresses, and strains and adjust process parameters and tool design.
2. HEAT GENERATION
FSW heat generation is governed by viscous dissipation and high shear at the tool–workpiece interface, making contact behavior central to modeling. The reviewed analytical treatment distinguishes sticking, sliding, and mixed conditions while recognizing complex boundary conditions.
- Power input: The tool shoulder and probe have much higher peripheral speeds than the tool’s translational speed, affecting the power input introduced into the weld.The heat generated during welding is treated as equivalent to the introduced power input, subject to losses.
- Heat-generation basis: FSW heat generation primarily involves viscous dissipation in the workpiece driven by high shear stresses at the tool–workpiece interface.Modeling therefore requires representations of both interface behavior and material viscous dissipation.
- Contact conditions: Interface material may stick to the rotating tool with the same local velocity or slip with a lower velocity.These alternatives define the principal contact conditions considered in the analytical heat-generation model.
- Contact conditions: A mixed contact state combines sliding and sticking, with sliding promoting frictional heat generation and sticking promoting heat generation through the stated contact condition.The survey presents an analytical model based on different contact assumptions.
2.1. Contact Condition
FSW contact modeling distinguishes sliding, sticking, and mixed states according to contact shear stress, internal yield shear stress, and relative velocity. A dimensionless slip-rate variable links these conditions to tool and workpiece motion.
- Contact-state framework: Coulomb friction estimates contact shear stress from the friction coefficient and contact pressure, but rigid contact-pair assumptions are not sufficiently representative for FSW.The review therefore introduces three tool/workpiece interface states.
- Sliding condition: Sliding occurs when contact shear stress is below the matrix yield shear stress, producing slight elastic deformation and relative motion.The sliding condition promotes heat generation through friction.
- Sticking condition: Sticking occurs when friction shear stress exceeds matrix yield shear stress, accelerating the matrix surface toward the tool velocity until equilibrium is established.Sticking promotes heat generation through plastic deformation.
- Partial sliding/sticking condition: Partial sliding/sticking combines both states: the matrix moves below tool-surface velocity, while contact shear stress equals internal yield shear stress.This mixed state corresponds to quasi-stationary plastic deformation.
- Contact state variable: The dimensionless slip rate relates workpiece-surface velocity to tool-surface velocity and can impose a slip boundary condition in CFD material-flow models.The tool velocity is calculated from angular velocity and radius, while workpiece velocity is local interface velocity.
- Contact-state framework: The modeling assumption that transverse welding speed does not affect slip or deformation rate treats workpiece velocities as tangential to the rotation axis.The relationships among contact conditions are summarized in Table I.
2.2. Analytical Estimation of Heat Generation
The review develops analytical heat-generation estimates for simplified FSW tool geometries by integrating contributions from the shoulder, probe side, and probe tip. The estimates distinguish sliding, sticking, and mixed contact conditions and are compared with experiments.
- Tool geometry: The analytical model represents a simplified tool with conical or horizontal shoulder, vertical cylindrical probe, and horizontal probe-tip surfaces.The cone angle is zero for a flat shoulder.
- Heat-generation formulation: Total heat generation combines contributions from the shoulder, probe side, and probe tip at the tool/workpiece contact interface.The model neglects mechanical power from traverse movement because it is considered negligible relative to rotational power.
- Heat-generation formulation: The heat contributions are estimated by integrating contact heat flux over shoulder, probe-side, and flat probe-tip surfaces.The formulation uses contact area, cylindrical radius, force, and moment terms.
- Heat-generation formulation: For a flat shoulder, the total heat-generation expression simplifies to a geometry-dependent form involving shoulder radius, probe radius, and probe height.The three surface contributions are combined before applying the flat-shoulder simplification.
- Contact-condition influence: Sliding heat generation uses frictional shear stress, whereas sticking heat generation uses the material yield shear stress and plastic deformation.The mixed-condition estimate combines sliding and sticking solutions with a weighting function.
- Contact-condition influence: For partial sliding/sticking, slip is a fraction of ωr, reducing frictional heat while adding plastic dissipation from material deformation.The resulting expression linearly combines the sliding and sticking contributions, with δ spanning the contact states.
- Heat-generation ratios: Probe heat is negligible for thin plates but typically 10% or more for thick plates, with the probe radius influencing shoulder-to-probe heat ratios.Figure 7 presents these ratios as a function of probe radius and highlights the shoulder-to-probe radius ratio.
- Experimental comparison: Analytical heat-generation estimates correlate with experimental heat generation under either sliding or sticking assumptions.Sliding comparisons use a reasonable metal-to-metal friction coefficient, while sticking comparisons use elevated-temperature yield shear stress.
2.3. Heat Generation Mechanism
FSW heat generation depends on uncertain tool–workpiece contact behavior and includes frictional surface heating and plastic or viscous dissipation. Literature models differ in their assumptions, while a volume-to-surface heat parameter can improve agreement with experimental thermal data.
- The nature of tool–workpiece contact, particularly at the shoulder, remains unclear and affects heat-generation modeling.
- Reported models attribute heating variously to shoulder friction, plastic deformation near the pin, or complete sticking between material and tool.A purely frictional model using a low friction coefficient was judged probably inadequate, while high temperatures near the pin suggest plastic deformation is important.
- Plastic dissipation heat sources use the plastic strain-rate tensor, Cauchy stress tensor, and Taylor–Quinney coefficient β, typically ranging from 0.8 to 0.99.
- Heat input is divided into surface frictional heating and volume viscous or plastic-dissipation heating.
- The parameter γ represents the relative importance of surface and volume heat contributions in thermal-flow models.
- γ = 1 produced the best agreement with experimental thermal data in computational models that included material fluid flow.
2.4. Heat Input Estimation using the Torque
Torque and rotational speed provide an experimental route for estimating FSW tool input power, while translational power is typically neglected. A power-efficiency factor accounts for the fraction converted directly into weld heat and is usually high.
- Translational movement consumes approximately 1% of total power and is typically neglected in total heat-input estimates.
- Tool input power can be obtained experimentally from the measured torque and angular rotation speed.The power expression also includes traverse force and velocity, although their contribution is typically small.
- The power-efficiency factor is the fraction of tool-generated power directly converted into heat in the weld material.
- The power-efficiency factor is usually between 0.9 and 1.0 and is calculated from heat loss into the tool.
3. HEAT DISSIPATION
FSW heat must be balanced between generation and dissipation to reproduce experimental temperatures. Losses occur through the workpiece, tool, backing plate, and surroundings, with backing-plate contact and thickness strongly affecting modeled heat flow.
- Heat generation and heat dissipation must be adjusted and balanced to agree with experimental temperature values.
- Heat is generated by friction and plastic deformation, then dissipated into the workpiece, tool, backing plate, and surrounding atmosphere.The workpiece thermal conductivity influences the resulting thermomechanically affected and heat-affected zones.
- Heat loss into the tool is about 5%, based on modeled tool-temperature distributions compared with experiments.
- Adiabatic, perfect-contact, and tool-region-only perfect-contact assumptions are among the options for modeling the workpiece–backing-plate interface.An adiabatic model achieved reasonable agreement but consistently over-predicted measured temperatures.
- Most dissipated heat flows from the workpiece to the backing plate, especially near the tool where applied pressure reduces the conductance gap and locally maximizes heat flow.
- A backing spar reduces the number of equations and processing time while capturing the essential heat flow between workpiece and backing plate.Its width is usually approximately equal to the tool diameter.
- Increasing backing-plate thickness increases heat dissipation.Reported thicknesses include 12 mm, 25.4 mm, and 60 mm in different studies.
- An excessively high bottom convection coefficient underestimates the maximum temperature.
4. METAL FLOW
FSW metal-flow modeling must represent coupled thermal and mechanical behavior despite complex flow, steep velocity gradients, free surfaces, and constitutive sensitivity. Approaches range from efficient 2-D models to detailed but computationally expensive 3-D analyses.
- Flow models should capture thermal and mechanical aspects sufficiently for flow visualization, heat-flow evaluation, and tool-design optimization.
- Material flow depends on tool geometry and process parameters and is a determinant of FSW success and defect susceptibility.
- Analyzing predominantly in-plane flow in 2-D at mid-thickness provides significant computational-efficiency benefits compared with full 3-D flow.
- A kinematic model decomposes FSW into rotation, translation, and ring-vortex incompressible flow fields combined into distinct currents.
- CFD and finite-element methods, including Eulerian, Lagrangian, and hybrid mesh formulations, are used to model material flow.
- CFD models neglect elasticity, cannot predict absolute forces, and exclude some mechanical effects such as varying downforce.
- Steep velocity gradients near the tool motivate zoned meshes with a rotating, refined region containing the deformation zone.
- Three-dimensional elastic-plastic ALE analyses provide physical insight and handle concave shoulders, tool tilt, and threaded pins, but require very long computation times.
5. NUMERICAL SIMULATION OF FSW
Numerical FSW studies use models capable of representing severe deformation, frictional contact, material behavior, and complex geometries. Simulations examine how welding parameters affect temperatures and power dissipation, including an inverse trend between total power and maximum temperature with welding velocity.
- Modeling requirements: FSW analysis codes require rotational boundary conditions, frictional contact algorithms, high-deformation support, suitable material models, and complex-geometry capability.Elastic-plastic or elastic-viscoplastic formulations are identified among the required material models.
- Modeling approaches: Three-dimensional thermomechanical CFD models describe material flow around the tool during welding.One cited study modeled an AA2024 sheet with 3.2 mm thickness.
- Temperature effects: Maximum weld temperature decreases as welding velocity increases and increases as tool rotational velocity increases.The temperatures are evaluated in the workpiece close to the tool shoulder.
- Power dissipation: Plastic power partition is estimated through the sliding ratio associated with tool-workpiece contact.Predicted and measured power evolution are compared as functions of welding parameters.
- Power dissipation: Although total power generated in the weld increases with welding velocity, maximum temperature decreases.The model separates predicted power dissipation into plastic and surface contributions.
6. CONCLUSIONS
FSW modeling helps visualize welded-material behavior and assess process parameters without costly experiments, but its multiphysics complexity currently limits numerical optimization. Future improvements may support parameter selection and reduce reliance on experimental trials.
- FSW modeling visualizes welded-material behavior and analyzes weld parameters, including tool design, and boundary conditions without costly experiments.
- FSW simulation is challenging because the process combines heat flow, high-temperature plastic deformation, and microstructure and property evolution.
- Numerical FSW simulation still cannot be used to optimize the process.The conclusion links this limitation to the current state of process knowledge and computer resources.
- Increasing process knowledge and computer resources may enable simulations to predict suitable process-parameter combinations and replace current experimental trials.The anticipated use is described as helping expand FSW to a wider range of applications.