Laser shock peening in friction stir welded joints with lack of penetration defects

Abstract

This experimental work was conducted in order to assess the influence of the application of Laser Shock Peening (LSP) to Friction Stir Welded (FSW) joints. LSP has the ability to potentially recover the reduction in joint mechanical properties that arose due to the presence of a common FSW defect known as Lack of Penetration (LOP). The material used throughout this study was 3 mm thick AA6082-T6 Aluminium. This specific material was selected due to its weldability and common use in the manufacturing of aircraft structures.

A 20 mm diameter, spiralled profile shoulder and 5 mm diameter, tapered, three facetted pin was used in the manufacturing of the FSW joints. Experimental assessment and optimisation of the FSW parameters window consisted of the varying the tool rotational speed from 630 to 1600 RPM, at five increments and the feed rate from 200 to 600 mm/min, 200 mm/min intervals. The joints were assessed on the overall quality, microstructure, ductility and static strength. The parameter combinations produced joints with ultimate tensile joint efficiencies which ranged from 48% to 74% that of the base material. The initial study showed that higher welding rates, typically associated with low feed rates, resulted in the highest quality joints. This was attributed to the sufficient thermal softening of the material during welding. The increased welding temperatures improved the joint formation, material flow and mechanical properties. Due to the elevated welding temperatures and material flow, substantial flash formation was observed on all joints manufactured with a welding rate of 5 rev/mm and higher. These results formed the foundation of the multi-objective optimisation in order to determine the most suitable parameters for this welding configuration. The optimisation simulation determined the optimum parameters to be, a tool rotational speed of 1433 RPM, feed rate of 196 mm/min and a welding rate of 7.3 rev/mm. Due to the fixed gearing of the CNC FSW machine the required tool rotational speed could not be achieved, thus, all FSW completed after the optimisation was completed at 1600 RPM, 200 mm/min and 8.00 rev/mm. This parameter combination produced a joint of high structural integrity, high ductility and with no visual sign of internal voids, defects or lack of penetration.

The performance of the defect free joint formed the foundation of the characterisation of the influence of intentionally introduced LOP defects. LOP was defined as a pre-initiated crack which formed at the root surface during the fabrication of a FSW joint. Controlled and consistent LOP was introduced into the joints, manufactured with welding parameters of 1600 RPM and 200 mm/min, by offsetting the welding tool in combinations of the normal and lateral directions relative to the joint line. The defects originated at the root surface and extended at various lengths through the thickness of the joints due to the numerous offsets. The defects lengths ranged from a few microns to as much as 954.5 μm. A number of defects affected as much as 43% of the joint thickness. The presence of the defects negatively affected the joints structural static strength by as much as 9% to 27% (dependent on the size of the defect) and dynamic fatigue life of as much as 36%.

Laser Shock Peening (LSP) is a novel post manufacturing technique, which has been used to introduce compressive residual stresses within the near surface of the metallic components. A LSP processing was completed without a protective ablative coating (LSPwC), at a wavelength of 1064 nm and pulsed nano-second laser at 10 Hz. The characterisation of varying the laser power intensity and processed coverage to the base of FSW AA6082-T6 Aluminium was completed through extensive parameter window exploration. Factors such as the quality of the energy delivery, sample deflection, strain hardening, penetration of effects through the FSW joint thickness and the improvement of the fatigue life of the base material were used to define the appropriate parameters. A multi-objective optimisation strategy was implemented in the attempt to fully explore the regions between tested parameter combinations; to provide an optimum set of LSP parameters which would be used in combination of the optimum FSW parameters. The simulation predicted two optimum sets of parameters dependent on the desired outcome of either maximising component fatigue or LSP depth of penetration effect. Due to the nature of this research requiring both fatigue and penetration depth, a parameter set was selected based on parameters that would theoretically provide the maximum for both desired outcomes. The optimum power intensity and coverage was specified as 3.33 GW/cm2 and 1067 spots/cm2.

The optimum parameters of each process was combined in an attempt to recover the drop in fatigue life of the joints due to the presence of the LOP defect. LSP was capable of altering the near surface residual stress states of approximately 100 MPa tensile to -150 to -200 MPa compressive across the three measurement ranges. It was found that LSP had minimal effect on the fatigue life of the components in the low cycle fatigue due to the applied stress relaxing the introduced stress thus having minimal effect on the life of the joints. LSP was found to increase the fatigue life of the non-defective joints by as much as 68%. LSP showed a life improvement of approximately 20% in a joint which had a defect length of roughly 175μm. After the application of LSP the samples in the low cycle fatigue tended to fail at a closed cycles to failure as the non-defective unpeened samples. Application of a LSP to a FSW was found to shift the fracture position of the flawed components from the region of the defect to that of the Heat Affected Zone (HAZ) on the advancing side of the weld. It is suggested that, the shift in fracture position was due to critical relocation of the tensile stress during LSP into the HAZ on the advancing side. The results did not conclusively show that LSP was capable of recovering the effects of the LOP but was plausible that it could possible. This has been said due to some samples exhibiting an increase and due to the change in location of the fracture position.