Fig. 1 — Transverse view of a conventional friction stir weld with regions of interest labeled. A B C 19, 30, 33–35) report a range of complex precipitate morphologies in the SZ with coarse particles dissolving providing solute for postweld natural aging. In contrast, studies on 2219 report either particle coarsening (Refs. 36–38, 40) and/or the dissolution of the Al2Cu phase in the SZ (Refs. 15, 20). Nonhomogeneities observed at the macroscale have been attributed to banding of large constituent particles, which correspond to tool rotation variations in 2024 (Refs. 41, 42), different tempers of 2219 (Ref. 36), or overpass repair welds in 2219 (Ref. 38). While differences in the microstructural characterization of 2195 in two different studies were attributed to variations in phase transformations kinetics as influenced by FSW process parameters (Refs. 14, 31), no systematic study has been conducted to verify. Conductivity measurements provide a well-established nondestructive evaluation (NDE) technique for determining the temper of a metal. However, its sensitivity is affected by variations in alloy uniformity due to heat treatment condition, the degree of cold work, presence of residual stresses, or effect of thermal exposure (Refs. 43–45). Thus, a combination of NDE techniques are often used to evaluate the temper of an alloy such as combining eddy current with hardness testing. While these standard techniques are typically used at the macroscale where homogeneity of the thermomechanical processing is assumed, characterization at the microscale can provide insight into nonhomogenous variations. This study evaluated the combined use of conductivity measurements with hardness testing at the macro and micro length scales to evaluate the resulting microstructure in a FSW SZ formed by varying the tool rotation. The range of tool rotations in the study was selected based on earlier studies where a large change in the resulting SZ strength was observed (Ref. 46). Microstructural features were correlated with conductivity and hardness measurements. The results in this study were also compared with temperature calculations based on conversion of weld power to thermal energy (Ref. 27). Experimental Procedure Friction stir welds were made in rolled panels of aluminum alloy 2219-T87 approximately 610 mm long, 152 mm wide, and 6.4 mm thick that were butted together. Nominal composition of the 2219 alloy (wt-%) is Cu 6.30%, Mn 0.30%, Zr 0.17%, V 0.10%, Ti 0.06%, Fe 0.15%, Si 0.10%, and balance Al. The FSW tool consisted of a 12.7-mmdiameter UNF left-handed pin, a 30.5-mmdiameter scrolled shoulder, and a pin length of approximately 6.2 mm. All FSWs were performed with a zero degree lead angle and in-position control. A RM-1 model FSW machine from Manufacturing Technology, Inc. (MTI), was used to produce the welds with the data recorded using a highspeed National Instruments Data Acquisition system. Metallographic specimens were taken of the transverse section of each FSW segment. The specimens were mounted and polished using standard metallurgical procedures. All samples were etched using Keller’s reagent to document the macrostructure as recorded with a Nikon D1 camera. Surface topography was obtained in a scanning probe microscopy (SPM) using a diamond Berkovich probe mounted on the Hysitron TI 950™. Prior to SPM, the specimens were mechanically reground and repolished using 1.0- and 0.5-micron alumina on the pad followed by colloidal silica. Indentation experiments were conducted using the Hysitron TI 950™ instrument equipped with the nanoECR™ (electrical contact resistance) package and a conductive boron-doped diamond Berkovich probe with a tip radius of approximately 150 nm. The nanohardness of each transverse specimen was measured across the width approximately 1.3 mm below the crown surface. One hundred indents with a spacing of 250 μm were made using a 5-s loading to a peak of 10 mN, 5- s hold, and 5-s unloading segments, which corresponded to an average indentation contact depth of 485 nm. To measure the nanoconductivity, the nanoECR™ package was used, which enables simultaneous electrical measurements to be made during standard nanoindentation testing. During testing, a fixed voltage was applied to the sample via a conducting stage and the resultant current flow through the sample was measured through the conducting tip. Voltage was held constant at 2 V and the measured current was used to calculate the average current density based on the contact area of the indenter at peak loading. JANUARY 2013, VOL. 92 12-s WELDING RESEARCH Fig. 2 — Miniature tensile specimens fabricated from the FSW nuggets. A — Shown are the specimens from the FSW transverse microstructure with the specimen geometry superimposed; B — an end mill was used to machine the dogbone geometry; C — which was then sliced into individual specimens using wire EDM.
Welding Journal | January 2013
To see the actual publication please follow the link above