Analysis of the Influence Of Friction Stir Welding on the Microstructure and Mechanical Properties of Alloy UNS-C27200 (CU-ZN)

The copper alloy UNS-C27200 is a binary brass, composed of 38wt%Zn, beeing one of the most used in the industry. The union of Cu-Zn alloys by conventional welding by fusion in thin plates presents several challenges. The Friction Stir Welding (FSW) process performs the union of two, or more materials, through a rotating tool that travels throughout the welding region. One the great advantages of FSW is working below the melting point of the material, which reduces the change in the microstructure and properties of the material, when compared to conventional welding processes. Due to the lack of research on FSW process in lap brass joints with thin thicknesses, the present work aims to expand these studies in order to contribute to future research. Eight conditions were evaluated, varying rotation and welding speed. In the best results by visual inspection, analyzes of microstructure, temperature, microhardness and mechanical strength were performed. The better welding results were obtained with ω=1050 rpm and ν=20 mm/min. There was reduction of approximately 50% between Stir Zone and Base metal. The mechanical strength was affected by the pin hole resulting from the process. The microstructure revealed a homogeneous mixture in the stir zone and no defects were detected.


Introduction
The Cu without alloy element is very soft, ductile, easy to work cold, but has low mechanical strength.It has great application when require high corrosion resistance and high electrical conductivity.To improve some of its characteristics, cold working processes and/or alloy formation by solid solution are often used, where small amounts of alloying elements are often added 1,2 .
The most common and used copper alloys in the industry are binary brasses (Cu-Zn), zinc (Zn) being the main alloy element 1,2 .The copper alloy UNS-C27200 is a binary brass, composed of 38%Zn.For being a brass α+ß, exhibits better weldability 2 than the brasses-α and is more malleable than copper and zinc 2 .It has a relatively low melting point ranging from 800 to 1070°C, depending on the zinc value in its composition, and in general the higher the amount of zinc, the lower the melting point 3 .
In addition to the above-mentioned characteristics, the UNS-C27200 alloy offers very useful properties such as high strength, good conductivity, conformability, wear resistance and corrosion performance, deep stamping and intense gloss of strong golden color 3,4 .
The union of Cu-Zn alloys by conventional welding by fusion in thin plates presents several challenges 4 , because the temperatures obtained in these processes are high and cause partial evaporation of zinc 5 due to its low boiling point (907°C), bringing several problems in the weld region, such as surface irregularities, major distortions, cracks, color change due to oxidation, porosity and others arising from solidification 6,7 .
In order to minimize these problems, in 1991 TWI (The Welding Institute) developed a welding technique called Friction Stir Welding -FSW, which is a welding process that occurs in the solid state, where it is possible to join parts with working temperature below the melting temperature 8 .Figure 1 show the principal variables in FSW process.
Initially this technique was developed for welding Al alloys, but its great potential led several researchers 4,[10][11][12][13][14] to apply this technique in the welding of Mg, Cu, Ti, Al, lead, some steels, thermoplastics and different combinations of materials, particularly those with nearby melting temperatures and similar behavior, such as hot workability.
The FSW process in lap joints is not complex 5,9,[15][16][17] .A non-consumable rotating tool -formed by pin and shoulder -is dipped between the plates mixing the material and then the tool is displaced, forming the weld profile 18 .
The kinetic energy of the tool is transformed into thermal energy, generated by friction at the interface between the pin and the shoulder of the tool and the parts to be welded, causing the material to reach high temperature and is in a plasticized state 19 .As the tool moves, the plasticized material is stirred and forged by the shoulder face, where it consolidates and cools to form a solid phase bond between the two pieces 10 .Due to the intense plastic deformation that happens in the stir, the weld formation mechanism is described by some researchers 10,15,20 as a combination of extrusion and forging, where the working temperature is less than the melting point of the materials to be joined together.
The FSW process involves several variables 10,21,22 , which together contribute to obtain a quality welded joint, with rotation speed (ω) and welding speed (ν) being the most prominent parameters 23 .
Although there are several important results in the studies of FSW in butt joints 3,12,[22][23][24] , greater efforts are needed in research for FSW in lap brass joints with thin thicknesses, both with similar materials and in dissimilar materials, given the few publications found with this welding configuration 5,15,16,24 .

Materials and Methods
The welds were processed using rectangular, rolled and annealed plates of UNS-C27200 brass, hardness grade ½ hard, with 2 mm thicknesses.Tables 1 and 2 show the chemical compositions and mechanical properties of the material.
All samples were joined in the configuration of lap joints according to Figure 2a, with weld length of 50 mm.The samples dimensions were 100x20x2, length, width and thickness, respectively.A rotating tool with shoulder and tapered pin, in H13 tool steel, quenched e tempered with 50~55 HRC according to Figure 2b was used.A device is designed to provide the necessary process stiffness, plate fixing, and tool tilt angle relative to the Z axis (3°).
For temperature monitoring, a system with Arduino programming and CoolTerm freeware software was developed, 4 k-type thermocouples (T1, T2, T3 and T4) were used, with stainless steel tip and M6 x 12 thread.The thermocouples were mounted on the device to obtain the temperature at the bottom of the plates as shown in Figure 3.Although the measured values did not correspond to the actual temperatures, they served as the basis for observing the temperature distribution during the FSW process.
The welds were performed in a 3-axis Veker Machining Center, model MV-760-ECO, with computerized numerical control (CNC).The process parameters were based on literature data published by 2,5,7,16,18,22 and refined by preliminary tests to provide the best parameter capable of producing welding visually without defects.
Preliminary tests were produced according to the data shown in Table 3, maintaining constant the other parameters such as dwelt time = 5s, penetration speed = 10 mm/min, shoulder penetration = 0.2 mm and pin lentht = 3.7 mm.For each combination in Table 3, 4 replicates were performed, totaling 32 experiments.The microhardness tests were performed in a Buehler microdurometer on the Vickers HV scale with a load of 0.1kgf, using as reference the ASTM E384 standard 25 .
The selected samples were sectioned in the plane perpendicular to the welding direction, embedded in bakelite and were submitted to a sanding and polishing process according to standard procedures.They were chemically attacked by immersion of the samples in an ethyl alcohol-based reagent composed of 10.7% hydrochloric acid and 3.4% of ferric chloride for 25 seconds.
The weld strength was evaluated by shear tests in a universal test machine EMIC DL 10000, TRD 28 cell, traction load 100 kN with test speed of 5 mm/min.

Visual inspection
The welded joints were evaluated by visual inspection with the naked eye, without the aid of any instrument.Figure 4 reproduces the surface of welded joints with the combinations defined in Table 3.Not all parameters surveyed generated welds without apparent defects for UNS -C27200 brass samples.Figure 4 shows the tunnel defects in the stirred zone, which is commonly caused by insufficient heat 10 .It was observed on the advanced side and was linearly located along the weld line under welding conditions ω = 750, 850 and 950 rpm.The tunnel defect appeared in all samples welded with ν = 60 mm/min, being smaller with ν = 20 mm/min.From the visual inspections, the combination that showed the best results, such as filling the weld bead, no tunnels on the weld surface and minimal flash, were those obtained with ω = 1050 rpm and ν = 20 mm/min (combination C7).The other procedures and results will be based on the welds produced with these parameters.

Temperature
In order to better understand the mechanisms involved in the formation of microstructures and the behavior of zinc in alloy, temperature values were measured at the bottom of the plates 16 .Due to the characteristics of the FSW process, we can observe that the working temperature was below the melting temperature of the alloy under study (920°C) and the boiling temperature of zinc (907°C), as shown in Figure 5.The temperatures obtained are satisfactory, as it avoided the reduction of the percentage of zinc by evaporation and consequently the formation of pores in the welded region 3 .

Microhardness
Figure 6a shows the position where the microhardness was measured, where "BP" is the line between plates, "L1" and "L2" the lines of the upper plate and "L3" and "L4" the lines of the lower plate.Vertical distances are 0.5 mm. Figure 6b shows the hardness profiles found in the specimen.
The specimen showed a reduction in hardness in the stirred zone when compared to the base metal.Although large differences in hardness are observed on the advancing side, where there is a high concentration of heat due to the drag of material to the retreating side 26 , the stirred zone presents a considerable difference as observed in Figure 6b.The temperature peak is obtained precisely in the symmetry line of the joint, being a region of visual identification facilitated, with greater guarantee for the analysis 27 .There is also the contribution of the pin shape, because the tapered pin results in temperatures higher than the cylindrical pin 19 and the axial force is also higher, so that this increased heat generation in the hot condition, causing a significant reduction in the average hardness to 80.4 HV0.1 when compared to the base metal of 179 HV0.1.

Shear tensile tests
The strength of the joint was evaluated by shear tensile test, following the procedure adopted by 15,16 and ASTM Standard E8/E8M 28 , which the specimens and their dimensions are shown in Figure 2a.For this test four specimens were tested.The results of maximum stress obtained and two images, one from the top and other from base, of each specimen is shown in Figure 7.
The weld region remained intact in all specimens and the rupture of the base metal occurred on the output side of the tool, where the hole left by the pin caused a reduction of the resistant area, favoring the rupture.The joints showed some dispersion in the shear test, indicating a less stable result.It is possible that the temperature increase, that is, a higher heat intake influenced the contact condition, causing the tool to slip 21 .An 48% joint efficiency of Cu was reached in relation to base metal.Temperture doesn't affect the results on tensiles properties.

Microstructure
The typical macrograph of the cross-section of the lap welded UNS-C27200 alloy is shown in Figure 8.The macrostructure has a very similar appearance to studies of 18,23 where the shape of the stirred zone resembles a wine cup; this format may vary according to welding parameters, tool type and material characteristics.
Microstructural evolution in the perpendicular plane can be seen in Figure 9, where three different zones can be clearly identified, i.e., the stirred zone (SZ), the thermo-mechanically affected zone (TMAZ) and the base metal (BM) 23 .
The following microstructures show the structural changes obtained in the FSW process.In Figure 10a the SZ presents equiaxial and refined grains, resulting from the intense stir of materials and complete dynamic recrystallization 22,23 .In Figure 10b TMAZ, which due to severe plastic deformation and intense heat generation is the target of microstructural changes that, in turn, cause significant changes in mechanical properties.This region is not completely recrystallized and has structural changes from the region that encompasses the shoulder width to the pin near the root 20 .
Figure 10c and d shows the SZ surrounded by TMAZ with significantly different microstructure.It is possible to observe the progressive increase in the grain size of the TMAZ when moving away from the SZ 3 .The transition between SZ and TMAZ was more evident on retreating side than the advanced side.

Figure 1 .
Figure 1.Typical welding procedure by a rotating shoulder 9 .

Figure 3 .
Figure 3. (a) Adaptaded machine used in process, (b) welding device and (c) position of thermocouples type-k in welding device (dimensions in mm).

Figure 4 .
Figure 4. Surface appearances of the welds.

Figure 6 .
Figure 6.(a) Position of the lines for measuring microhardness; (b) Microhardness profiles in cross section with ω = 1050 rpm and ν = 20 mm/min.

Figure 7 .
Figure 7. Maximum stress obtained and images from the tested specimens.

Figure 9 .
Figure 9. Macrostructure of the welded joint and its zones.

Table 3 .
Combinations used for sample production.