Friction stir welding
Fundamentals
Principle of Operation
Friction stir welding (FSW) is a solid-state joining technique that employs a non-consumable rotating tool to generate frictional heat and plastic deformation, softening the base materials without reaching their melting points to form a high-strength joint.[5] This process enables the metallurgical bonding of similar or dissimilar materials, primarily metals like aluminum alloys, through intense localized thermomechanical working in the solid phase.[6] The operation commences with the tool, comprising a cylindrical shoulder and an attached pin, being positioned at the butt joint interface of two clamped workpieces. The pin is plunged into the joint until the shoulder contacts the surface, establishing initial frictional contact.[7] The tool then rotates at high speed, generating heat primarily through friction between the shoulder and the workpiece surface, as well as through viscous dissipation from the plastic deformation induced by the pin's stirring action within the material.[7] This softens a zone around the pin into a plasticized state, creating a dynamic equilibrium where the material ahead of the tool is heated and deformed, while the material behind consolidates under the forging pressure from the shoulder.[8] As the tool traverses along the joint line, the rotating pin stirs and mixes the plasticized material from both workpieces, transporting it in a swirling flow pattern to the trailing edge of the tool.[6] The shoulder confines the deformed material, applying downward force to forge it into a seamless weld nugget as it cools below the recrystallization temperature, resulting in a fine-grained microstructure without solidification defects.[7] Upon completion, the tool is withdrawn, leaving a keyhole at the start or end, depending on the setup.[6] The shoulder of the tool primarily handles surface-level frictional heating and provides the axial forging force to consolidate the stirred material, while the pin penetrates to a predetermined depth to enable subsurface stirring and material intermixing.[6] Heat input in FSW arises from two main sources: sliding friction at the tool-workpiece interfaces, which accounts for a significant portion of the thermal energy, and adiabatic heating due to the high strain rates in the deforming material volume.[7] As a thermomechanical solid-phase process, FSW avoids the melting associated with fusion welding, thereby eliminating issues such as liquation cracking, porosity, and hydrogen-induced defects that arise from solidification.[5] This results in joints with superior mechanical properties, including reduced distortion and no need for filler materials or shielding gases.[6]Historical Development
The origins of friction stir welding (FSW) trace back to a friction-based welding method patented in the Soviet Union by Yu. V. Klimenko in 1967, which described a process for joining metals through frictional heating without melting, though it did not lead to immediate commercial development.[9] The modern FSW process was invented and patented by researchers at The Welding Institute (TWI) in the United Kingdom in 1991, with the foundational patent (GB Patent Application No. 9125978.9, granted as US Patent 5,460,317) outlining the use of a rotating tool to generate frictional heat and plasticize material for solid-state joining.[10] This innovation addressed challenges in fusion welding aluminum alloys, such as porosity and cracking, by enabling defect-free joints at lower temperatures.[6] Initial demonstrations of FSW focused on aluminum alloys in the early 1990s, shortly after TWI's invention, where the process successfully joined series 2000 and 7000 aluminum alloys used in aerospace applications, demonstrating superior mechanical properties compared to traditional arc welding.[10] By the late 1990s, research expanded to higher-strength materials, including feasibility studies on low-carbon and 12% chromium steels, marking the beginning of FSW's adaptation beyond aluminum to ferrous alloys despite challenges like tool wear.[11] A key milestone was the first industrial application in shipbuilding during the mid-1990s, where FSW was used to manufacture hollow aluminum panels for fish-freezing compartments on fishing vessels, enabling long, distortion-free welds up to 16 meters in length with high reproducibility.[12] In the aerospace sector during the 2000s, NASA adopted FSW for repairing and fabricating components of the Space Shuttle's external tank, including the super lightweight aluminum-lithium alloy tanks, which improved joint strength and reduced weight by eliminating fusion weld defects.[13] By the 2010s, FSW evolved to incorporate hybrid variants, such as laser-assisted and electrically-enhanced processes, to extend applicability to dissimilar materials and thicker sections while minimizing defects.[14] Automation advancements, including torque and force control systems, enabled precise manufacturing integration, particularly for aluminum in automotive and marine industries.[15] This period saw rapid patent growth, with over 1,800 FSW-related filings worldwide by 2007, reflecting its widespread adoption across sectors.[10]Process Mechanics
Tool Design
The friction stir welding (FSW) tool consists of two primary components: the rotating shoulder and the pin, also known as the probe. The shoulder generates frictional heat through contact with the workpiece surface and forges the softened material to consolidate the weld, while the pin penetrates the material to stir and key it, promoting plastic flow and intermixing across the joint interface.[16][17] Common tool geometries are designed to optimize material flow and heat distribution. Pins typically feature cylindrical, tapered, or threaded profiles, with threaded or tapered designs enhancing stirring action by facilitating better material transport and reducing defects. Shoulders may be flat, concave, or ridged, where concave shapes help contain the deformed material and ridged patterns improve flow control by promoting upward extrusion.[16][17] Material selection for the tool prioritizes high strength, thermal stability, and wear resistance, varying by workpiece. For welding aluminum alloys, hot work tool steels such as H13 are commonly used due to their toughness and ability to maintain hardness up to 600°C. In high-temperature applications like steel welding, polycrystalline cubic boron nitride (PCBN) or tungsten-rhenium (W-Re) alloys are preferred, as they withstand temperatures exceeding 1000°C and exhibit superior abrasion resistance.[16][17][18] Key design considerations include geometric proportions and durability features to ensure effective welding. The shoulder diameter is typically 2 to 3 times the pin diameter to balance heat input and forging pressure, while the pin length is matched to the plate thickness, often slightly shorter to avoid excessive plunge depth. Wear resistance is critical, as the tool endures high shear stresses and temperatures; coatings or composite materials like PCBN inserts are often incorporated in the pin to extend usability. Tool life generally exceeds 1000 meters for aluminum welding but is significantly shorter for steels, limited to around 60 meters with PCBN tools due to temperatures up to 900–1200°C causing rapid degradation.[16][17][18]Key Welding Parameters
In friction stir welding (FSW), the key adjustable parameters govern the heat input, material flow, and overall weld integrity, allowing optimization for specific materials and thicknesses. The primary parameters include tool rotational speed, traverse speed, tool tilt angle, and plunge depth, each influencing the balance between frictional heating and mechanical stirring.[19] Tool rotational speed, denoted as ω and typically ranging from 400 to 2000 rpm for aluminum alloys, primarily controls the rate of frictional and deformational heat generation. Higher rotational speeds increase the heat input through intensified material deformation and friction at the tool-workpiece interface, promoting better material softening and flow, but excessive speeds can lead to defects such as tunneling or voids due to overheating and inadequate consolidation.[19][20] Traverse speed, denoted as v and commonly set between 50 and 2000 mm/min depending on material thickness, determines the rate at which the tool moves along the weld line. Slower traverse speeds allow greater heat accumulation, which is beneficial for thicker sections to ensure sufficient softening without melting, while faster speeds enhance process efficiency but may result in insufficient heating and defects like incomplete penetration if not balanced properly.[19][21] The tool tilt angle, usually maintained at 1–3° backward from the vertical, facilitates enhanced forging action on the plasticized material trailing the tool. This slight tilt increases the contact pressure from the tool shoulder, improving material consolidation and reducing surface defects by promoting downward flow and expulsion of stirred material.[19][22] Plunge depth refers to the extent of pin insertion into the workpiece, which directly controls the volume of the stirred zone and the depth of material interaction. Proper plunge depth ensures complete penetration and adequate mixing without excessive flash formation or thinning, with values typically adjusted to 0.1–0.5 mm below the plate surface for thin aluminum sheets to avoid incomplete stir zones.[23][24] These parameters are interdependent, with heat input Q qualitatively approximated by the frictional power relation $ Q = \mu P \omega r $, where is the friction coefficient, is the axial force, is the rotational speed, and is the effective contact radius; this highlights how increases in or amplify heating, while variations affect weld properties like strength and microstructure uniformity. For aluminum alloys, an optimal ratio of 10–50 rev·min/mm balances heat generation and material flow to achieve defect-free welds with maximal mechanical performance.[25][26]Welding Forces
In friction stir welding (FSW), three primary forces act on the tool and workpiece: the axial (or plunge) force, which applies downward pressure to maintain tool penetration; the traverse force, which opposes the tool's forward movement along the weld line; and the torque, which resists the tool's rotational motion. For aluminum alloys, typical axial forces range from 3 to 10 kN during welding, depending on plate thickness and tool geometry, while traverse forces are generally lower at 0.5 to 5 kN, and torque values fall between 10 and 60 Nm.[27] These forces arise from the frictional and deformational interactions between the rotating tool and the workpiece material, essential for generating the necessary heat and material flow without melting.[28] Force profiles vary across the welding stages. During the initial plunge phase, the axial force rises sharply to a peak—often 4 to 6 kN for 5-6 mm thick aluminum plates—as the tool shoulder contacts the surface, then decreases by up to 40% once full penetration is achieved. In the steady-state traverse phase, the axial force stabilizes at a lower level with minor fluctuations, while the traverse force remains relatively constant. Torque exhibits a pronounced peak at the start due to the cold, unsoftened material, reaching maxima around 60 Nm before settling to 20-40 Nm during traversal.[27] These profiles provide indicators of process stability, with deviations signaling potential defects like incomplete penetration or excessive flashing.[28] Forces in FSW are measured using dynamometers or load cells integrated into the welding machine's fixture or spindle. Multi-component dynamometers, such as piezoelectric types with capacities up to 60 kN for axial loads and 15 kN for traverse, capture real-time data at high sampling rates, often paired with torque sensors on the spindle. Load cells offer a cost-effective alternative for uniaxial measurements, though they may require calibration for multi-axis accuracy.[27] Accurate monitoring enables process control, as high forces necessitate robust clamping fixtures to prevent workpiece distortion, and excessive traverse force can indicate suboptimal lubrication, high travel speeds, or tool wear.[28] The magnitude of welding forces scales with the workpiece material's strength and thickness; for instance, steels generate 2-5 times higher axial and traverse forces than aluminum alloys under comparable conditions, often exceeding 20 kN axially for 6 mm thick low-carbon steel plates. This scaling arises from the greater resistance to plastic deformation in harder materials, influencing tool design and machine requirements. Traverse forces, in particular, increase with thickness due to larger deformed volumes.[27]Material Behavior During Welding
Microstructural Evolution
Friction stir welding (FSW) induces a characteristic zonal microstructure in the weld region due to the thermomechanical processing, consisting of the stir zone (SZ), thermo-mechanically affected zone (TMAZ), and heat-affected zone (HAZ). The SZ, also known as the nugget zone, is the intensely stirred central area where severe plastic deformation and frictional heating lead to full dynamic recrystallization, producing fine equiaxed grains typically 1-10 μm in size.[29] This zone often exhibits banded structures, including onion rings or flow arms, which result from the periodic extrusion and deposition of material layers as the rotating tool advances, creating alternating light and dark bands visible in cross-sections.[30] The TMAZ surrounds the SZ and experiences partial plastic deformation coupled with thermal exposure, resulting in distorted, elongated grains that are not fully recrystallized and serve as a transitional region between the recrystallized SZ and the unaffected base metal.[31] In contrast, the HAZ is subjected only to heat conduction from the weld, without mechanical stirring, leading to grain coarsening or over-aging in precipitation-hardened alloys, while maintaining a structure similar to the parent material but with altered properties.[32] The grain refinement in the SZ and TMAZ is governed by continuous dynamic recrystallization mechanisms, driven by high strain rates and temperatures reaching 80-90% of the material's absolute melting point, which promote nucleation and growth of new grains without melting.[33] Material-specific microstructural changes are prominent in FSW. In aluminum alloys, such as 6063 or 6061, the SZ develops a uniform distribution of fine, equiaxed recrystallized grains; in precipitation-hardened alloys like AA6061-T6, this typically results in reduced hardness (e.g., 80-95 HV) compared to the base metal (typically 95-110 HV), due to dissolution of strengthening precipitates despite grain refinement effects via the Hall-Petch relationship.[34] In steels, the solid-state process suppresses the formation of brittle martensite phases common in fusion welding, instead yielding fine ferrite, bainite, or tempered martensite structures that improve ductility and toughness.[35] Consequently, the SZ often displays altered hardness relative to the base metal, with material-specific behaviors influenced by the refined microstructure.Material Flow
In friction stir welding (FSW), material flow is driven by the combined action of the rotating tool's pin and shoulder, resulting in complex plastic deformation patterns within the softened workpiece material. The primary flow mechanisms include an extrusion-like flow induced by the leading shoulder edge, which pushes material ahead of the tool; vigorous stirring by the pin that generates a vortex-like motion, entraining and recirculating material around its periphery; and forging action from the trailing shoulder, which consolidates the deformed material behind the pin. These mechanisms ensure thorough mixing without melting, distinguishing FSW from fusion-based processes.[36][37] Material paths exhibit asymmetry relative to the tool's rotation and travel directions. On the advancing side—where the tool rotation and traverse motion align—material experiences higher strain rates and layered, rotational flow patterns, with particles undergoing multiple loops around the pin due to enhanced shear. In contrast, the retreating side features simpler, predominantly shear-dominated flow, where material is more directly displaced and less intensely recirculated. This asymmetry arises from the rotational direction, leading to differential velocities and deformation across the weld zone. Peak flow velocities occur near the pin tip, reaching up to 60% of the tool's peripheral speed close to the pin surface, with the highest rates on the advancing side.[37][38][39] Visualization of these flow patterns has been achieved through experimental techniques such as marker insert experiments and X-ray radiography. In marker insert studies, thin strips or prismatic markers of contrasting material (e.g., AA5454 in AA2195) are embedded along the joint line, revealing post-weld deformation that indicates recirculation paths, including vortex formation and layered deposition. X-ray radiography provides in-situ observation, capturing real-time material motion within a small field (e.g., 2 mm × 2 mm) and confirming three-dimensional recirculation, such as downward flow ahead of the pin and upward return behind it. These methods highlight the dynamic, non-uniform nature of the flow, with material from both sides contributing to the stir zone.[40][41][42] Inadequate material flow can lead to defect formation, while excessive flow may cause other issues. Insufficient stirring or shear, often due to low rotational speeds or high traverse rates, results in voids or kissing bonds—unbonded interfaces resembling lap joints—where material fails to fully intermix. Conversely, overly vigorous flow from high rotation or tool pressure expels excess material as flash, forming surface irregularities that may require post-processing. Optimizing flow balance is critical for defect-free welds, as these flaws compromise joint integrity.[43][44]Heat Generation and Distribution
In friction stir welding (FSW), heat is generated primarily through two mechanisms: frictional heating at the interface between the tool shoulder and the workpiece, which is the dominant source, and plastic deformation within the stir zone, providing a secondary contribution.[45][46] The frictional heating arises from the relative motion and contact pressure between the rotating tool and the softened material, while plastic deformation heating results from the intense shearing and flow of the material around the tool pin. These sources create a localized thermal field that softens the workpiece without reaching the melting point, enabling solid-state joining.[47] Temperature profiles in FSW exhibit a characteristic distribution, with peak temperatures occurring near the tool centerline in the stir zone. For aluminum alloys, these peaks typically range from 400-500°C, representing 80-90% of the material's melting point, while for higher-melting-point materials like steels, peaks reach 800-1000°C.[48][49] The temperature decreases radially outward through the thermo-mechanically affected zone (TMAZ) to the base metal, forming a steep gradient that influences material properties across the weld. This asymmetric profile, often higher on the advancing side, results from the combined effects of tool rotation and translation.[50] Heat flow during FSW is predominantly governed by conduction through the workpiece and tool, with advective contributions from the tool's linear traverse that transports hot material along the weld path. Cooling rates post-peak temperature typically fall between 10-100°C/s, depending on welding parameters and workpiece thickness, which affects recrystallization and residual stress development.[51][46] Analytical and numerical modeling of heat generation emphasizes the frictional component, often represented by the heat flux equation
where is the friction coefficient, is the axial force, is the tool rotational speed, and is the radial distance from the centerline; however, models prioritize qualitative insights into the spatiotemporal distribution rather than precise flux calculations.[47] The overall heat input is commonly parameterized by , where is the traverse speed, which strongly correlates with peak temperature and the width of the heat-affected zone. Higher values increase both temperature and zone extent, guiding parameter selection for defect-free welds.[52]