A Coupled Thermal/Material Flow Model of Friction Stir Welding Applied to Sc-Modified Aluminum Alloys
- First Online:
- Cite this article as:
- Hamilton, C., Kopyściański, M., Senkov, O. et al. Metall and Mat Trans A (2013) 44: 1730. doi:10.1007/s11661-012-1512-y
- 1.5k Downloads
A coupled thermal/material flow model of friction stir welding was developed and applied to the joining of Sc-modified aluminum alloy (7042-T6) extrusions. The model reveals that surface material is pulled from the retreating side into the weld zone where it is interleaved with in situ material. Due to frictional contact with the shoulder, the surface material is hotter than the in situ material, so that the final weld microstructure is composed of bands of material with different temperature histories. For this alloy and the associated FSW heating rates, secondary phase dissolution/precipitation temperatures are in proximity to the welding temperatures. Therefore, depending on the surface and in situ material temperatures in relation to these transformation temperatures, disparate precipitate distributions can develop in the bands of material comprising the weld nugget. Based on the numerical simulation and on thermal analysis data from differential scanning calorimetry, a mechanism for the formation of onion rings within the weld zone is presented.
Over the last 20 years, numerous investigations have sought to characterize the principles of friction stir welding (FSW) and to model the material flow behavior, the temperature distribution, and the microstructural evolution within the weld. The review papers of Nandan et al. and Threadgill et al. provide an excellent account of past and current FSW research. Early numerical simulations, such as those of Colegrove and Shercliff or Khandkar et al., focused on the temperature distribution during welding and studied its potential influence on the weld microstructure and precipitation kinetics. More recently, researchers have been able to model both the material flow behavior and temperature characteristics during FSW despite the complex material flow associated with the process. Robson and Campbell developed a grain growth and recrystallization model of FSW that successfully predicted the weld nugget size during the joining of 2524 aluminum alloy plates. Colegrove et al. created a numerical model that combined material hot deformation and thermal properties to predict temperature, flow stress, and strain rate in age hardenable aluminum alloys 2024, 7449, and 6013. The current investigation presents a coupled thermal/flow model of friction stir welding applied to Sc-modified Al-Zn-Mg-Cu extrusions (Al alloy 7042-T6).
Additions of scandium (Sc) and zirconium (Zr) to 7000 series alloys stabilize the microstructure at temperatures greater than 423 K (150 °C) through the formation of fine, secondary strengthening phases such as Al3(Sc,Zr).[7,8] The nanometer-sized Al3(Sc,Zr) particles also stabilize the microstructure formed during hot working operations and inhibit recrystallization during heat treatment, thus potentially enhancing the residual properties after joining operations such as FSW. These additions also affect the kinetics of precipitation and growth of the primary strengthening precipitates (GP zones, η′), thus modifying heat treatment conditions for enhancing the mechanical properties of these alloys. The numerical simulation proposed here gives insight into the material flow and temperature distribution of the weld zone during the joining of 7042-T6 extrusions. Combined with thermal analysis data from differential scanning calorimetry (DSC), the precipitation behavior within the weld is discussed in terms of the volume fraction of the metastable (GP zones and η′) and equilibrium [η (MgZn2) and/or T (Al2Mg3Zn3)] strengthening particles found in the 7042 aluminum alloy. It is assumed that FSW does not change the size and volume fraction of the Al3(Sc,Zr) precipitates due to their high thermal stability.
DSC is a powerful technique for the investigation of precipitation and dissolution processes in Al alloys.[10,13] By detecting the heat variations due to the phase transformations, the technique is able to identify the temperature ranges in which they occur. For example, Dixit et al.[14,15] utilized DSC to study the nucleation of precipitates within the nugget of friction stir welded aluminum 2024 and to correlate the weld microstructure to mechanical properties. In the present work, results of the DSC thermal analysis of the FSW regions of the 7042-T6 Al alloy, together with a developed coupled thermal/material flow model of FSW, were used to propose a mechanism of onion ring formation within the weld zone. The model reveals that surface material is pulled from the retreating side into the weld zone where it is interleaved with in situ material.
2 Experimental Procedure
2.1 Alloy Chemistry and Heat Treatment
Chemical Composition of 7042
7042, This Work
2.2 Friction Stir Welding
After heat treatment, the bar was cut into twelve, 305-mm-long pieces and sent to the Edison Welding Institute (EWI, Columbus, OH) to produce six longitudinal friction stir welds. The diameter of the FSW tool shoulder was 17.8 mm, the pin diameter tapered linearly from 10.3 mm at the tool shoulder to 7.7 mm at the tip, and the pin depth was 6.1 mm. With a constant weld velocity of 2.1 mm s−1 and a constant applied force of 22 kN, unique welds were produced at the following pin rotation speeds (PRS): 175, 225, 250, 300, 350, and 400 rev min−1. The temperature profile across the weld surface was experimentally recorded for each condition using a Mikron M7815 Infrared Thermal Imaging Camera during welding. These data were used to verify the temperature predictions of the coupled thermal/flow simulation developed during this investigation. The uncertainty in these measurements was ±2 pct (or approximately ±9 K). The thermal emissivity for the infrared data was calibrated by imaging an extrusion length heated to 733 K (460 °C) and adjusting the emissivity value until the recorded temperature of the camera matched the reference temperature. The appropriate thermal emissivity value was determined to be 0.285.
2.3 Post-Weld Investigation
Subsequent to joining, the welded panels were stored at room temperature and allowed to naturally age for at least 30 days prior to testing and investigation. Small samples (approximately 20 to 50 mg) were extracted from the T6-tempered baseline material and from the weld center of each welded sample for thermal analysis. The samples were sealed in Al pans and analyzed in a Perkin Elmer Jade differential scanning calorimeter, using an argon atmosphere. Depending on the data desired, samples were heated from room temperature to 673 K (400 °C) at a constant heating rate that ranged from 10 to 100 K min−1. A polarized optical microscope was used to study the microstructure of the welds. To enhance the appearance of precipitate distributions and grains, the studied surfaces of the weld samples were polished and anodized in an electrolytic solution of 1.8 pct fluoroboric acid in water at room temperature and an electric current of 0.15 A. The anodizing time was 2.5 to 3 minutes.
3 Coupled Model for Flow and Temperature Behavior
3.1 Materials Properties and Boundary Conditions for Flow
The maximum strain rates within the flow region occur adjacent to the weld tool, i.e., under the tool shoulder and along the pin, where the velocity gradients are the greatest. The strain rates decrease rapidly away from the tool since the material flow velocities also decrease quickly away from the tool. In their work on aluminum, Frigaard et al. estimated the maximum effective strain rate under the tool shoulder to be 20 s−1, while Nandan et al. calculated the maximum strain rate as 100 s−1 near the tool shoulder and as 30 s−1 approximately 4 mm below the shoulder. More recently, Arora et al. computed the maximum strain rate in aluminum 2524 as 9 s−1 for their FSW parameters and tools. Colegrove et al. used constant strain rate values ranging from 0.001 to 1000 s−1 in their thermal/flow simulations of aluminum alloys.
For the process parameters utilized in this investigation, the maximum effective strain rate varies from 20.9 s−1 at 175 rev min−1 to 47.9 s−1 at 400 rev min−1. These numbers certainly fall within the range of effective strain rate values calculated and utilized by other researchers.
Material Constants for the Sheppard and Wright Flow Stress Equation and the Zener–Hollomon Parameter
1.26 × 108
3.2 Materials Properties and Boundary Conditions for Temperature
A thermal insulation constraint is applied at each interface of the flow-capable region with the non-flow areas of the model, i.e., with the retreating side, with the advancing side and with the backing spar. These constraints assure temperature continuity across the flow-capable boundaries into the other areas of the model. Thermal insulation constraints are also applied to the tool shoulder/workpiece and pin bottom/workpiece interfaces to assure heat continuity across these boundaries as well. For the boundaries exposed to ambient conditions, i.e., the workpiece top, workpiece side, and tool side, the convective heat transfer coefficient is set to 15 W/m2 K to approximate free convection on these surfaces. As suggested in References 22 and 23, convection coefficients of 200 and 250 W/m2 K are applied to the tool top and spar bottom, respectively. For the underside of the workpiece and the sides of the backing spar, a convective coefficient of 100 W/m2 K is used to represent the dissipation of heat into the backing plate. Heat dissipation due to radiation is ignored in this model.
3.3 Model Verification
4 Results and Discussion
4.1 Material Flow and Temperature Distribution
These numerical results of material flow are consistent with the experimental observations of other researchers. Hamilton et al. studied the friction stir welding of 6101-T6 extrusions plated with tin (0.05-mm thick). After welding, the unique presence of tin within the weld nugget and the lack of tin within the TMAZ led them to conclude that the weld nugget forms as surface material is extruded from the retreating side into a region of plasticized material around the FSW pin and under the tool shoulder. Colligan, in his study of material flow during the FSW of 6061-T6, used steel shot tracers and concluded that surface material extrudes from the retreating side of the pin and deposits in the wake of the tool. Guerra et al., in their study of FSW of 6061, utilized a faying surface tracer and similarly hypothesized that material from the front of the retreating side of the pin extrudes between the deformed surface material rotating with the tool and the parent material into the area behind the pin.
Perhaps this result is more clearly illustrated by the inset to Figure 6. Here, a two-dimensional coupled thermal/flow model for aluminum flowing around a rotating, heat-generating circular boundary is presented. The streamlines in the inset show that, like in the three-dimensional FSW model for 7042-T6, cooler material ahead of the rotating boundary is swept to the retreating side and material heated by the boundary is deposited on the advancing side. As a result, the temperature distribution is skewed, and higher temperatures develop on the advancing side.
4.2 Formation of Onion Rings
Endothermic and Exothermic Peak Transformation Temperatures Estimated (From DSC Data) for Different Weld Conditions at the Weld Edge (Retreating Side) and Weld Center.
Welding Condition (rev min−1)
Endothermic Transformation Temperature [K (°C)]
Exothermic Transformation Temperature [K (°C)]
Weld Edge (Retreating Side)
The potential for onion ring formation in the 7042-T6 alloy therefore depends on the surface and in situ material temperatures relative to the endothermic and exothermic reaction temperatures in these regions for the given weld conditions. If both the surface and in situ material temperatures are less or greater than the exothermic temperature, then the appearance of onion rings is minimized. In the first case, no coarse particles will form in both regions, while in the second case, similar coarse particles will form in both regions; thus, low optical contrast between the regions will develop. If, however, one of the surface and in situ material temperatures is below and another is above the peak exothermic transformation temperature, then onion ring formation is maximized since each material zone will exhibit unique precipitation characteristics.
At 225 rev min−1, the in situ material temperature is greater than the endothermic reaction temperature, but about 45 K smaller than the peak exothermic temperature. On the other hand, the surface material temperature corresponds to the peak exothermic reaction temperature. The in situ material will again be free of the η/T particles. However, the rate of formation of these particles now reaches a maximum value in the surface material before it is pulled from the retreating side down into the nugget. These stable secondary phases are retained in the microstructure upon cooling, thus creating the potential for onion ring formation due to the difference in precipitation behavior between the material zones. Examination of the metallographic image of the 225 rev min−1weld (Figure 10) reveals the emergence of a more prominent onion ring pattern.
The onion rings are most pronounced for the 250 and 300 rev min−1 welds, as illustrated in Figure 10. As previously discussed, at 250 rev min−1, the surface material slightly exceeds the peak exothermic transformation temperature, while the in situ material reaches a temperature in between the respective endothermic and exothermic temperatures. Before being introduced into the weld zone, the surface material will experience a strong precipitation of stable η and/or T phases, and the in situ material will dissolve fine GP and η′, which will re-precipitate upon cooling and holding at room temperature. The resulting microstructure will have defined bands of particle-rich surface material and particle-poor in situ material comprising the weld. Similarly at 300 rev min−1, the surface material now exceeds the exothermic temperature by 52 K, and the in situ material temperature is effectively equal to the peak exothermic temperature. As such, η and/or T phases in the surface material will coarsen and begin to dissolve while they approach the maximum rate of precipitation in the in situ material. Suggesting that the volume fraction of these phases is the same for these two conditions, the surface material would have a smaller particle number density and a larger average particle size than the in situ material. Mixing the bands of surface material containing coarser η/T particles with the bands of in situ material containing finer η/T particles again imparts a strong onion ring pattern due to the optical contrast in precipitation.
For pin rotation speeds greater than 300 rev min−1, both the in situ and surface material temperatures exceed the respective peak exothermic phase transformation temperatures. For the surface material, the temperature of which now exceeds 673 K (400 °C), the trend will be toward complete dissolution of the η and/or T phases and GP zone formation upon cooling and holding at room temperature. The in situ material will experience limited dissolution of the η and/or T phases as well, but since the temperature lags behind that of the surface material, some overaged η and/or T phases can be retained at room temperature. As a result, the onion ring pattern will become fainter for pin rotation speeds beyond 300 rev min−1. The microstructural images for the 350 and 400 rev min−1 welds in Figure 10 show faint onion patterns, but they are certainly less distinct than those in the 250 and 300 rev min−1 welds.
The appearance of onion rings within the weld nugget is a well-established, but not fully understood, phenomenon in FSW. The terms “particle-rich” and “particle-poor” were originally introduced by Sutton et al. who observed the onion ring phenomenon in FSW 2024 and determined the rings were interleaved bands of material both rich and depleted in secondary phases. Not all researchers, however, credit the formation of onion rings to disparate particle distributions. Rather, some researchers have concluded that the weld nugget structure is actually an interleaving of layers of fine-equiaxed, recrystallized grains with coarse recrystallized grains.[30, 31] Perhaps both hypotheses are actually correct and the mode of onion ring formation is simply dependent on both the workpiece material and the welding conditions. For aluminum 7042-T6, at least, the formation of onion rings depends on the temperature of the surface material and in situ material relative to the endothermic and exothermic precipitate transformation temperatures, which are driven by the heating rates during friction stir welding. As the fine Al3(Sc,Zr) particles presented in the 7042-T6 alloy effectively suppress recrystallization, the onion rings are evidently not related to recrystallized grains in this material.
A coupled thermal/material flow model of friction stir welding was developed and utilized to simulate the joining of Sc-modified aluminum extrusions (7042-T6). The model successfully predicts the temperatures in and near the weld zone for all weld conditions. For higher pin rotation speeds (300 rev min−1 and greater), the model shows good agreement with the experimental temperature distribution within the workpiece away from the weld zone. For slower pin rotation speeds (250 rev min−1 and less), however, the model overpredicts the workpiece temperatures away from the weld.
Within the weld zone, the model demonstrates that surface material approaching the rotating tool is swept to the retreating side. The material then (1) rotates with the tool under the shoulder and is deposited behind the tool toward the advancing side, (2) rotates under the shoulder and is captured by the tool, or (3) rotates under the shoulder and is pulled into the weld zone by the threaded pin. Mid-plane and bottom plane materials rotate with the tool and show some downward migration. The extrusion of “hotter” surface material into the weld nugget where it interleaves with “cooler” in situ material gives rise to the formation of onion rings in the 7042-T6 alloy. If the surface material temperature is greater than the exothermic transformation temperature of the alloy, then the surface becomes enriched in stable η and/or T phases before being pulled into the weld nugget. If, at the same time, the in situ material temperature is greater than the endothermic reaction temperature, but less than the exothermic reaction temperature, the in situ material becomes depleted in the non-equilibrium η′ phase and supersaturated. Upon cooling, GP zones form in the in situ material, such that bands of particle-rich surface material interleaved with particle-poor in situ material comprise the weld nugget and impart the characteristic onion ring appearance.
The authors would like to acknowledge the Polish Ministry of Science and Higher Education (Grant No. N507 446337) and UES, Inc. for their support of this work. ONS acknowledges financial support through the United States Air Force Research Laboratory Contract No. FA8650-10-D-52226.
This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.