Friction stir powder additive manufacturing of Al 6061 alloy: Enhancing microstructure and mechanical properties by reducing thermal gradient

anisotropy,


Introduction
Continuously increasing demand for weight reduction in space, aerospace, biomedical, defense, marine, automobile, and mechatronic applications combined with requirement of extremely good dimensional and geometrical accuracy, surface quality, surface integrity, and minimum damage to the components used in these applications, have compelled to explore use of Al-based and other lightweight alloys [1,2].Inability of the conventional subtractive and deformative manufacturing processes to meet all these requirements simultaneously for lightweight alloys has necessitated development of additive manufacturing (AM) processes which are not only energy and material efficient but produce better surface finish, and dimensional and geometrical tolerances with higher productivity [3].Use of fusion-based additive manufacturing (FBAM) processes to manufacture components from Al alloy has been explored over the last few years.But, AM processes are found to suffer from the following limitations: (i) they result in higher thermal gradient which makes challenging to manufacture components from the difficult to additively manufacture and weld Al alloys of 2xxx, 6xxx, and 7xxx series [4], (ii) presence of solidification related defects such as porosities and hot-cracking [5], (iii) columnar grains and cast microstructure along the deposition height yielding anisotropic mechanical properties in most of the cases [6], (iv) supply of excessive heat leads to spattering of powder particles deteriorating deposition quality whereas supply of insufficient heat results in inter-layer and intra-layer lack of fusion, and (v) frequent heating of the already deposited layers during deposition of new layer leads to formation of undesirable phases producing residual stresses [7].These limitations of FBAM processes for Al alloys have forced development of newer AM process having lower operating temperature that above-mentioned limitations can be overcome [8].
Solid-state additive manufacturing (SSAM) processes, which involve depositing material without melting it, hold the potential to surpass the limitations of FBAM processes [9].Their main advantages are: (i) consistently generate equiaxed grains possessing microstructure and mechanical properties similar to that of the wrought alloys [10,11], and (ii) yield lower residual stresses and thermal distortion as compared to FBAM processes [12].Major SSAM processes such as friction stir additive manufacturing (FSAM) [13,14], ultrasonic additive manufacturing (UAM) [15,16], cold spray additive manufacturing (CSAM) [17,18], and additive friction stir deposition (AFSD) [19,20] are reported in the literature.FSAM and UAM are hybrid sheet-lamination processes that necessitate the implementation of subtractive processes to shape the bonded layers [21].CSAM process suffers from the limitations such as smaller deposition rates and limited size and geometry of the product [22].AFSD process resembles very close to the layer-by-layer deposition concept and gives nearnet finished components via site-specific deposition [23].It involves continuous feeding of the feedstock material in solid rod form through the rotating spindle.The plasticization of the rod through frictional heat generation between the substrate/previous layer leads to metallurgical bonding, resulting in higher deposition rates and improved mechanical properties [6].Major concern of AFSD process is high compressive forces.They cause consolidation of feedstock material inside the rotating tool [5,21].Moreover, most of these SSAM processes have reported heterogeneous microstructure and anisotropy in mechanical properties along the build direction due to large temperature variation between initial and top deposition layers.Philips et al. [4] reported significant variation in microhardness values over a cross-section of Al 6061 alloy manufactured by the AFSD process.They found that mechanical properties deteriorate towards the previously deposited layers due to their frequent heating, and existence of thermal gradients along the build direction.Similarly, Zhao et al. [24] found reduced hardness from top sheet to bottom sheet in the FSAM of AleLi alloy.Higher value of the maximum temperature during the deposition coarsens the grain structure, dissolves hardening phases such as Mg 2 Si, decreases dislocation density, and reduces residual stresses and microhardness [25].On the other side, minimum temperature of the deposition (i.e., temperature during cooling) also plays a vital role in post-deposition microstructure and mechanical properties.By minimizing the ratio of temperature difference to the distance between any two points on the multi-layer deposition in the build direction (i.e., thermal gradient), there is an opportunity to improve the quality of the depositions [26].
Thus, it can be concluded from the past review work that (i) existing SSAM processes have their own limitations owing to their early development stage, (ii) development of newer SSAM process is needed to overcome limitations of the existing SSAM processes and for multi-layer deposition of lightweight alloys, and (iii) presence of thermal gradient along the build direction deteriorates microstructure and mechanical properties.Consequently, a novel SSAM process referred to as friction stir powder additive manufacturing (FSPAM) has been developed, using feedstock material in the form of powder.The present work aims for (i) solid state multi-layer manufacturing of aerospace grade Al 6061 alloy using FSPAM process, (ii) enhance its microstructure and mechanical properties through reduction of thermal gradient along the build direction by maintaining the substrate (i.e., AA6061-T4 Al alloy) close to its artificial aging temperature through its insitu control with the external heat source, and (iii) study of surface appearance, microstructure and porosity, element distribution and phase analysis, tensile strength and microhardness, and fretting wear performance of Al 6061 alloy multi-layer deposition.The study presented a new approach for manufacturing Al alloys using their feedstock in powder form and with improved microstructure and mechanical properties.

2.
Materials and methods

Materials
Aerospace grade Al 6061 alloy powder (Al:98.4%;Mg:1.1%;Si:0.5% by weight) having average particle size of 58 ± 20 mm was used for multi-layer deposition on commercially available substrate of AA6061-T4 Al alloy having avg.grain size of 27 mm.Fig. 1a shows morphology of the powder particles and Fig. 1b shows distribution of its major alloying elements obtained by the energy dispersive spectroscopy (EDS).Table 1 presents the chemical composition of the substrate plate having dimensions of 200 mm length; 100 mm width; and 6 mm thickness.It is finished using wire brush and cleaned by acetone to remove the oxide layers and contaminants before deposition.

Apparatus for FSPAM process
The working principle of FSPAM process for multi-layer deposition is schematically depicted in Fig. 2a.Guided by this principle, an indigenous apparatus was designed and developed for the FSPAM process, as showcased in Fig. 2b.It consists of the following components.
Powder feeder: The powder feeder, depicted in Fig. 2b, is positioned on top of the FSPAM apparatus and operates through a direct current motor.It ensures a consistent supply of feedstock material powder flow rate, ranging from 20 to 1590 mm 3 /min.To facilitate the continuous delivery of the feedstock material, a carrier gas of pressurized argon is employed at a flow rate of 0.2 l/min.Fixture: It is mounted on the worktable of FSPAM apparatus for mounting the substrate and backing plate.The substrate is kept over the backing plate which is held in the slot made for it in the fixture.Both substrate and backing plate are fixed in their position by means of the supporting plates as shown in Fig. 2b and c to restrict their movement during the deposition by the FSPAM process.Tool: It is made up of H13 tool steel having 46 Rockwell hardness on C-scale.It has 22 mm diameter and possesses radial and circumferential grooves on its bottom face (Fig. 2d).Such geometric features on the tool face helps in generation of more frictional heat due to better consolidation and mixing of feedstock material powder.It results in good metallurgical bonding among deposition layers.
External heat source with close-loop control: External heat source is kept in a through slot made in the backing plate (as shown in Fig. 2c) to protect it from the compressive load generated during the deposition process.It consists of copper coil which is placed inside the stainless-steel cover.
A temperature sensor is provided in 6 mm diameter blind hole made in the substrate for in-situ monitoring of substrate temperature and sending it to a temperature controller which changes power supply to the external heat source.It implies that the external heat source, substrate, temperature sensor, and temperature controller make close-loop control to maintain the required minimum temperature of the substrate.This loop minimizes adverse effects of thermal gradient in build (z) direction.
Pyrometer: It is fixed on the worktable.A laser light attached to it confirms desired location of temperature   measurement during deposition by the FSPAM process (Fig. 2b).In present case, temperature readings were obtained at a distance of 65 mm, which corresponds to the midpoint of the total deposition length.Pyrometer records the maximum temperature when FSPAM tool arrives at the measurement location during the deposition.It has measurement range from À30 to 1030 C with an accuracy of ±1.5 C. Its data are used to generate temperature versus time profile during the multi-layer deposition by FSPAM process.

Procedure for multi-layer deposition by FSPAM process
Twenty-seven full factorial experiments were conducted manufacturing single-layer single-track depositions of Al 6061 alloy [27].These experiments identified optimum values as 1200 rpm for tool rotation speed 'N'; 25 mm/min for substrate traverse speed 'v'; 6 for tool tilt angle 'Ø'; 22 mm for tool diameter 'D' having radial and circumferential grooves on its bottom; and 460 mm 3 /min for volumetric flow rate of feedstock material 'F' which were used in multi-layer deposition of Al 6061 alloy using the following procedure.
An external heat source is used to raise the substrate temperature near its artificial aging temperature (170 C for Al alloys) before starting the deposition process.The rotating tilted FSPAM tool then contacts the substrate for a fixed duration called the dwell period.The dwell period is set to ensure that the temperature exceeds the recrystallization temperature (ranging from 235 C to 290 C for Al alloys) required for powder deposition.The substrate is moved downward in the Z-direction by the stand-off distance (SOD), while feedstock material is supplied beneath the tilted FSPAM tool.Simultaneously, the substrate undergoes back and forth traverse motion in the X-direction, creating relative movement between the FSPAM tool and substrate.This movement initiates stirring between the tool and feedstock material, as well as between the feedstock material and substrate, causing severe plastic deformation of the feedstock material along with exerting high compressive forces.This thermomechanical processing shears the powder particles and triggers the deposition.After depositing a layer of the desired length (130 mm in this case), the substrate is moved downward by the SOD, and the supply of feedstock material is halted while the substrate travels back in the X-direction to the starting point for the next layer deposition.This process is repeated for subsequent deposition layers until the desired number of layers is achieved, as in the present case of 10 layers.The FSPAM deposition can be divided into three distinct regions: (i) advancing side (AS), where the rotating tool's local direction aligns with the deposition direction, (ii) center of the deposition, and (iii) retreating side (RS), where the rotating tool's local direction is opposite to the deposition direction.

Sample preparation from multi-layer deposition
Following samples were prepared from Al 6061 multi-layer deposition using the standard procedure to study its different metallurgical and mechanical characteristics.
(i) Sample for macrostructure, microstructure, porosity (from the microstructure), elemental analysis, and microhardness: 5 identical samples (of 11 mm thickness each) were prepared by cutting the multi-layer deposition along its build direction (X-direction) by wire spark erosion machining (WSEM) process on Ecocut machine (from Electronica Ltd.Pune, India).Each sample was polished up to 2000 grit SiC emery papers followed by fine polishing by the diamond paste of 3 mm and then 1 mm size.Two out of the five prepared samples were electropolished at 20 V DC voltage and 238 K temperature utilizing a mixture of 80 vol.% methanol and 20 vol.% perchloric acid to reveal grains and grain boundaries by electron back scattered diffraction (EBSD) technique.(ii) Sample for porosity (from Archimedes principle), phase analysis, fretting wear: Three cubic samples of 5 mm side each were cut from the multi-layer deposition on the WSEM machine.They were polished up to 600 grit SiC emery papers and cleaned using acetone.After identifying phases for top layer, each sample was further cut at the middle of the deposition height by the WSEM machine to identify the phases in the intermediate layer.(iii) Sample for tensile testing and fractography: Three samples of 2 mm thickness were prepared using the ASTM E8M standard by cutting and slicing the multi-layer deposition from its middle on the WSEM machine along the deposition direction (X-direction).The prepared sample is shown in Fig. 3.

Characterization of multi-layer deposition
The prepared samples were used to characterize multi-layer deposition as follows: (i) Study of appearance, microstructure, porosity: Maximum height 'Ry' and arithmetical average surface roughness 'Ra' (as per ISO 21920-2021 standard) values, macrostructure images, and photographs of Al 6061 multi-layer deposition were used to study its appearance.Macrostructure images were taken by a EZ4 HD stereomicroscope (from Leica, Germany), and they were fed to Image J software for analyzing height of individual layer of the multi-layer deposition.They were also visually analyzed for quality of bonding and interfaces between the layers.Values of 'Ry' and 'Ra' were measured on the top layer of the multi-layer deposition by tracing the probe of Handysurf surface roughness tester (from Zeiss, Germany) across the formed onion rings (i.e., along the deposition direction) using 0.8 mm as section length and 4 mm as the evaluation length.The surface roughness values were measured at three locations i.e., beginning of the deposition, at the center, and at the end of the deposition layer.The EBSD images i.e., inverse pole figure (IPF) and image quality (IQ) maps, of the electropolished samples were obtained by the NovaNano 450 scanning electron microscope (SEM) (from FEI, USA) to reveal their microstructure using 20 kV accelerating voltage, 1 tolerance angle, and 0.5 mm step size.Obtained EBSD images of the electropolished samples were processed using Image J software to calculate approximate grain size.Elemental composition was determined using a JSM-7610FPlus SEM (from JEOL, Japan) equipped with energy dispersive spectroscopy (EDS) facility.Phases present in the multi-layer deposition were identified by PANalytical x-ray diffractometer (XRD) (from Empyrean, Netherlands) using Cu ka radiation, 2q range from 20 to 90 with step size 0.01, and dwell time of 1 s.Initially phases were identified for the top layer of the multi-layer deposition by matching characteristic peaks with JCPDS database and then phases for the intermediate layer were identified in the further cut sample.Porosity of Al 6061 multi-layer deposition was determined using two methods: (i) calculating by the Image J software the ratio of porosity area and total area of the micrograph obtained by the SEM, and (ii) calculating bulk density and relative density using Archimedes principle of buoyancy.Mass of each prepared sample was measured on mass balance having least count of 0.01 mg and their volume was computed using their dimensions.Ratio of the measured mass and the computed volume gave bulk density 'r b ' as 2.69 g/cm 3 .Relative density 'r r ' of each sample was computed by Eq. ( 1) and its porosity was estimated using Eq. ( 2).

Relative density ðr
where, 'M a ' and 'M w ' are average masses of a sample measured in air and in water, and 'r w ' is density of water (1.0 g/cm 3 ).
(ii) Study of mechanical properties: Tensile testing was conducted on the prepared samples, applying a constant elongation rate of 0.5 mm/min.The fracture surfaces of the tested samples were then examined using JSM-7610FPlus SEM to determine the type of failure that occurred during the tensile test.Additionally, microhardness measurements were performed on the prepared samples using VMH-002 Vicker's microhardness tester (from Walter UHL, Germany) with a load of 200 g and a dwell time of 15 s, following the ASTM E384 standard [28].Total 25 indentations were made, of which 19 indentations were made on middle of each deposition layer and at the interface between the two consecutive layers thus giving distance between two consecutive indentations as 0.25 mm, and 6 indentations were made on the substrate.This method of indentation was repeated thrice to confirm the values of microhardness.Fretting wear testing was performed using a CM-9104 tribometer (from Ducom Instruments, India).A 6 mm diameter pin made of EN31 steel with a hardness of 62 HRC was used to reciprocate over the sample with a stroke length of 200 mm and a frequency of 25 Hz.The test was conducted for a duration of 30 min under three different normal force values: 5N, 10N, and 15N.The contact point of the pin was spherical in shape and was pressed against the stationary sample using a lever-arm and associated dead weights.The coefficient of friction was calculated by dividing the recorded friction force by the applied normal force.Wear samples were further analyzed using the obtained fretting wear tracks.Marsurf LD130 surface roughness measuring cum contour tracing equipment (from Mahr Metrology, Germany) was used to trace profile of the wear tracks by moving its probe twice in perpendicular direction of the fretting wear.Obtained avg.Maximum wear depth from the wear profile was used for further analysis.

3.
Results and discussion

Temperature profile of Al 6061 multi-layer deposition
In the metal additive manufacturing (AM) process, heat loss occurs through three primary pathways.Firstly, heat is lost radially through the surrounding powder bed or environment.Secondly, heat is lost vertically upwards from the top deposition layer through convection and/or radiation towards the ambient environment.Finally, it is lost vertically downwards through the substrate plate via conduction [29].Control over heat loss through former two pathways is difficult but heat loss through substrate can be modified as demonstrated in the present study.Fig. 4 depicts the temperature profile obtained during Al 6061 multi-layer deposition by FSPAM process using the external heat source which is in a close-loop with the substrate and temperature controller.It can be observed from Fig. 4 that the recorded value of the maximum temperature during the deposition varies in a range between 305 C and 358 C with an average maximum temperature of 321.8 C. Maximum temperature increased initially during the deposition of successive layers as heat available with already deposited layer assisted in more heat generation.A similar observation was reported by Seidi and Miller [30] while manufacturing Al 6061 by multi-layer friction surfacing process.After five deposition layers, maximum temperature reached its limit and did not increase further; instead, it was found to attain a nearly stable maximum temperature condition.This intriguing behavior could be influenced by factors such as the specific geometry of the deposition and the properties of the feedstock material used.The temperature obtained by a deposition layer (i.e., minimum temperature) during its cooling also plays a vital role in tailoring properties of final components.Minimum temperature affects the microstructure and mechanical properties of the substrate/ deposited layer since it controls thermal gradient during the deposition.It affects grain size, structures within the grains, and phase formation [25,31,32].It can be observed from Fig. 4 that average minimum temperature recorded by a pyrometer during the deposition of each layer is 159.6 C which is close to the set temperature (170 C) of the substrate.It is worth noting that temperature controller has no control over the temperature of the substrate when it remains above the minimum set temperature of the substrate.It effectively serves its purpose in regulating the substrate temperature only when it drops below the set limit.During the cooling of the first deposition layer, the temperature initially dropped below the set temperature of the substrate due to rapid cooling, as shown in Fig. 4.However, after a certain period, indicated by the magnified view enclosed in a red boundary, the temperature of the deposited layer started to increase again and returned to the pre-set temperature of the substrate.This is achieved using external heating of the substrate plate up to its artificial aging temperature through a temperature controller.This approach has led to minimizing thermal gradient (i.e., temperature difference ratio between any two points lying on the same or distinct deposition layers and distance between them) along the build direction.In the present case, the average difference between the maximum and minimum temperature of any deposition layer did not exceed 162.2 C. Smaller temperature difference helps in homogeneous microstructure and enhanced mechanical properties along the build direction [26], which has been explained in subsequent sections.

3.2.
Appearance of Al 6061 multi-layer deposition Fig. 5 depicts surface appearance of FSPAM manufactured multi-layer deposition of Al 6061 alloy showing its top and front view (Fig. 5a), enlarged view of the deposition end (Fig. 5b), multi-layers on AS (Fig. 5c), and multi-layers on RS (Fig. 5d).Upon close examination, it can be observed that the edges of the multi-layer deposition appear smoother and more straight on the AS (Fig. 5a, b and c) compared to the RS (Fig. 5d).It is attributed to the accumulation of more feedstock material on the AS, caused by an excess flow of material from RS [33].Consequently, the layers exhibit a sharper edge on the RS as relatively less material is available on this side.To address this issue, a practical solution involves reversing the direction of the tool rotation during the deposition of subsequent layers, either from clockwise to anticlockwise or vice  6a) revealing good bonding between them, and (ii) interfaces between different deposition layers i.e., interlayer interfaces (Fig. 6b) depicting nearly straight and parallel layers without significant bending and overlapping between them.Height of each layer is 0.5 ± 0.05 mm and bonding at the interface between two layers can be seen without major defects as shown in Fig. 6b.The FSPAM process, being SSAM process, avoids dilution unlike the FBAM processes.Dilution refers to mixing materials between different layers during AM, potentially affecting material properties.

Microstructure and porosity analysis
The cross-section of the multi-layer deposition of Al 6061 alloy prepared for the EBSD analysis is shown in Fig. 7a on which three locations are marked i.e., 'b' for top layer, 'c' for intermediate layer, and 'd' for first deposition layer whose IPF and image IQ maps are shown in Fig. 7b, c and 7d respectively.It can be observed in Fig. 7a that the absence of major defects between the substrate and first deposition layer and between  respectively for a multi-layer deposition of Al 6061 alloy which indicates large variation in grain size along the build direction when deposited without maintaining the temperature of the substrate as reported by Chaudhary et al. [34].However, still the difference in heat exposure time for different layers during the deposition may lead to variation in grain size, as reported by Mason et al. [35].Specifically, the first deposition layer experienced more heat exposure, followed by subsequent layers, leading to a slight coarsening of grains in the first layer compared to the other deposition layers.The absence of pin in FSPAM tool is another advantage as it results in homogeneous grains in width direction unlike its competitive FSAM process [9].Minimum grain size in the multi-layer deposition is found to be 1.09 ± 0.41 mm, which is a significant reduction (98.1%) as compared to the average particle size (58 ± 20 mm) of the asreceived Al alloy powder (Fig. 1a).Image quality maps shown in Fig. 7 showcases various deposition layers, revealing distinct grain boundaries.These boundaries can be categorized into two types: Large angle grain boundaries (LAGBs) with misorientations ranging from 2 to 15 , denoted by red and green lines, and high angle grain boundaries (HAGBs) featuring misorientations exceeding 15 , represented by blue lines.Percentage of HAGBs is more than 72% in top, intermediate and first deposition layers whereas LAGBs are only around 28%.The above observations indicate that continuous dynamic recrystallization (CDRX) has occurred extensively in multi-layer deposition of Al 6061 alloy during FSPAM, characterized by the severe refinement of grains [4].CDRX has led to the formation of new, refined and equiaxed grains, in the multi-layer deposition of Al 6061 alloy.
Computed porosity in the multi-layer deposition using Archimedes principle is 0.19%, which was further confirmed by 0.215% porosity calculated from SEM micrographs (Fig. 8), resulting in a percentage of error of 13.15%.This finding aligns with the porosity values reported by Wei et al. [36], showing similar porosity for Al 6061 alloy manufactured by the solidstate CSAM process.Porosity is influenced by the amount of heat supplied during the deposition.When the heat input is insufficient, it can result in improper consolidation of the powder feedstock, which leads to poor metallurgical bonding between the adjacent layers or tracks of material.As a result, voids or gaps can form between the layers or tracks, leading to porosity in the final components.Insufficient heat input can also hinder proper diffusion of alloying elements, affecting the homogeneity of the material and leading to porosity at grain boundaries.In case of FSPAM process, frictional heat and continuous external heating enhance heat generation during deposition thereby improving the flowability of the plasticized feedstock material.It leads to better consolidation of the feedstock material and intimate contact among the powder particles, thus minimizing porosity generation.On the other side, excessive heat input, such as in FBAM processes, can also contribute to porosity generation.If the heat input is too high, it may lead to localized overheating and rapid cooling, causing thermal stresses and cracking in the material.Rapid cooling can sometimes solidify the material before it has a chance to fully coalesce, creating discontinuities and porosity.Additionally, high heat input can cause excessive vaporization of volatile elements or entrainment of gases, which can get trapped within the material and lead to porosity upon solidification.Unlike FBAM processes, maximum temperature in FSPAM process is always below the melting point of the feedstock material, eliminating solidification and minimizing the probability of porosity formation.Pores within a deposition layer (Fig. 8b) play an important role in controlling toughness of the additively manufactured components during loading in the deposition direction since they induce delamination among the deposited layers.Their presence in interlayer interfaces critically affects tensile and fatigue strengths since they act as damage initiation sites under tensile and cyclic loading [37].Moreover, porosity also provides pathways for corrosive substances to penetrate the material, leading to localized corrosion and degradation.By achieving low porosity, the FSPAMmanufactured Al 6061 alloy components can exhibit improved resistance to corrosion and longer service life.

3.4.
Elemental and phase analysis Fig. 9 presents results of EDS for the multi-layer deposition of Al 6061 alloy showing major alloying elements (Fig. 9a), and distribution of main alloying constituents namely aluminum (Fig. 9b), magnesium (Fig. 9c), and silicon (Fig. 9d).It can be seen in Fig. 9a that the multi-layer deposition has high weight percentage of aluminum i.e., 96% along with 2.5% Magnesium (Mg) and 1.5% of Silicon (Si) by weight.It indicates similar distribution of its alloying elements with respect to its feedstock material powder.The uniform distribution of Mg and Si (denoted by blue-colored spots) in the multi-layer deposition, as shown in Fig. 9c and d is due to the better mixing of its feedstock material caused by its inherent movement from RS to AS and AS to RS during the deposition.Additionally, the circumferential and radial groove on the bottom face of the tool also contributes in enhancing the mixing of feedstock material.The presence of blue-colored spots of Mg and Si elements appear to form an irregular grain-like pattern, leaving distinct spaces between them, which are visually represented by the black coloration.Therefore, it can be concluded that majority percentage of Mg and Si are along the grain boundaries of the multi-layer deposition of Al 6061 alloy.These phases are likely to be MgeSi based precipitates which are responsible for enhancing the mechanical properties [38].But overgrowth of these phases due to higher value of maximum temperature may deteriorate the mechanical properties [35].
Fig. 10 presents results of XRD for multi-layer deposition of Al 6061 alloy, as-received Al 6061 alloy powder, and AA6061-T4 (substrate).The presence of Al as the major phase is confirmed in multi-layer deposition of Al 6061 alloy as shown in Fig. 10 (JCPDS: 89-2769).Moreover, XRD plot of multi-layer deposition has shown characteristic peak of Mg 2 Si at (0 2 2) crystallographic plane similar to the substrate AA6061-T4 alloy (JCPDS: 65e2988).It might have occurred due to reprecipitation.Generally, Mg 2 Si is present in AleMgeSi alloys as a hardening phase responsible for improved mechanical properties [4].No significant phase variation is observed on the top and intermediate layers of the deposition.Compared to as received Al 6061 alloy powder and substrate, slight shifting of peaks in XRD of the multi-layer deposition towards higher angle indicates the presence of compressive residual stresses [34,39].Lattice planes are compressed or brought closer together by the compressive stresses.This lattice compression results in a decrease in the interplanar spacing of certain crystallographic planes which causes a shift in the positions of the diffraction peaks to higher angles (smaller d-spacings) in XRD analysis.These stress are highly beneficial, increasing its fatigue life by reducing speed of crack propagation [40].AA6061-T4 (solutionized and natural aged substrate material) and AA6061-T6 (solutionized and artificial aged) alloy (Fig. 11a), and microhardness values taken along the middle of the cross-section consisting of substrate and ten deposition layers (Fig. 11b).Table 2 presents average values of their tensile properties (obtained from the three samples of each alloy) in the deposition direction.Average values of yield strength (YS), ultimate tensile strength (UTS), and % elongation of multi-layer deposition of Al 6061 alloy are 106.2± 5 MPa; 190.7 ± 11.9 MPa; and 16.7 ± 0.5% respectively.Values of YS and UTS are higher than the commercially available AA6061-O Al alloy by 104.3% and 54.1% respectively.It is due to considerable grain refinement caused by the severe plastic deformation and continuous dynamic recrystallization during FSPAM process, as evident from Fig. 7.These values were 109.4 ± 6 MPa; 197.3 ± 8 MPa; and 16.5 ± 0.4%, respectively, for Al 6061 alloy manufactured by FSPAM, without maintaining the substrate temperature as reported by Chaudhary et al. [34].The tensile properties obtained in both approaches (i.e., with and without maintaining the substrate temperature) are comparable since they were measured only in the deposition direction (i.e., X-direction).However, a comprehensive evaluation of the actual variation between these approaches can only be ascertained by measuring the tensile properties in the build direction.According to HallePetch relation (Eq.( 3)), refined grains play an essential role in increasing strength [4].
where 's y ' denotes the yield stress of a material (MPa); 's 0 ' denotes friction stress which is overall resistance of lattice of a material to dislocation movement; 'k y ' serves as the strengthening coefficient which quantifies the relative hardening of the grain boundaries within the material; and, 'd' is average grain size (mm).It states that as the grain size decreases, the yield strength of a material increases.In multilayer deposition of Al 6061 alloy, strengthening effect is attributed to the higher volume fraction of grain boundaries resulting from refined grains.These grain boundaries act as obstacles that impede the motion of dislocations, consequently enhancing the tensile properties through the mechanism known as grain boundary strengthening.Experimental data support the effectiveness of the HallePetch relation in materials with grain sizes spanning from 1 mm down to the range of hundreds or tens of nanometers.However, when the grain size becomes extremely small, such as around 10 nm or below, the yield strength tends to decrease with further reduction in grain size as per the inverse HallePetch relationship.In addition to the grain boundary strengthening, the strength enhancement in the multi-layer deposition of Al 6061 alloy is also attributed to the presence of Mg 2 Si precipitates, known as precipitation strengthening [41].Average values of YS and UTS of the multi-layer deposition of Al 6061 alloy are lower than commercially available AA6061-T6 alloy by 55.7% and 35.3% respectively.It might be due to the lesser number of strengthening precipitates Mg 2 Si as compared to AA6061-T6 alloy.The presence of interlayer porosities (Fig. 8b) also acts as stress concentration sites during the tensile loading causing faster propagation of cracks from them to nearby zones [37].However, very small percentage of porosity in multi-layer deposition of Al 6061 alloy might not be the dominant factor affecting its strength.Percentage elongation of the multi-layer deposition of Al 6061 alloy is 32.6% higher than AA6061-T6  Microhardness of the multi-layer deposition of Al 6061 alloy slightly increases from first deposition layer (83 HV) to top deposition layer (91 HV).Though initial deposition layers have a relatively smaller value of deposition temperature than subsequent layers (ref Fig. 4), their final microhardness value is smaller.It is due to their frequent heating caused by repeated thermal cycles (i.e., heat coming from subsequent deposition layers).The consequence of these repeated thermal cycles is the coarsening of the microstructure in the initial deposition layers.This coarsening process inevitably leads to a reduction in microhardness [35].Moreover, the frequent heating of the deposition layer might have led to the dissolution of hardening precipitates (Mg 2 Si), a major cause for decreasing the hardness.However, very small difference of 8.8% in microhardness (83 HV to 91 HV) is observed over the cross-section along the build direction.It is due to reduced thermal gradient achieved through in-situ control of the cooling temperature of each deposition layer by maintaining pre-set substrate temperature (refer Fig. 4).Whereas this difference for Al 6061 alloy was found to be 7.7% even for a deposition height of 2.5 mm when manufactured by FSPAM process, without maintaining the temperature of the substrate as reported by Chaudhary et al. [34].Average value (taken over ten deposition layers) of microhardness of Al 6061 multi-layer deposition is 87.7 ± 2.4 HV which is slightly higher than microhardness of commercially available wrought AA6061-T4 alloy (79 HV) and significantly higher than AA6061-O alloy (47 HV) though 18.8% less than its T6 wrought counterpart (108 HV).Attainment of better value of microhardness of the multi-layer deposition is due to refined grains and presence of Mg 2 Si hardening precipitates as evidenced from the XRD plots (ref.Fig. 10).It is also consistent with the previous studies on AA6061 alloy by Philips et al. [4].A very small reduction in microhardness of the substrate towards its interface with the first deposition layer might be due to heat affected zone caused by its frequent heating and cooling during multi-layer deposition.
Fig. 12 depicts SEM images showing fractography results at 500x and 1000x magnification for AA6061-O (Fig. 12a and b), multi-layer deposition of Al 6061 (Fig. 12c and d), AA6061-T4 (Fig. 12e and f), and AA6061-T6 (Fig. 12g and h) alloys.It can be observed from Fig. 12a and b that fracture surfaces of AA6061-O alloy have many small and large dimples indicating ductile behavior.It suggests that the AA6061-O alloy has good toughness and has undergone significant plastic deformation before failure as evident from elongation values reported in Table 2. Fracture surfaces of multi-layer deposition of Al 6061 alloy is found to have consisted of many dimples and few smooth regions, as shown in Fig. 12c and d.Dimples are more prevalent in some zones whereas they are absent in others as identified by smooth regions in Fig. 12d.It indicates ductile failure of the multi-layer deposition [24] along with some brittleness.The absence of dimples might be caused due to site-specific fine grains which are responsible for the distinct fracture morphology of a deposition made from powder form of feedstock material under severe plastic deformation.Crack propagation during the fracture might have inhibited in these zones.Consequently, there would have been a localized stress concentration causing the work-hardening rate to increase before the fracture [42], thus improving tensile properties.Similar fracture behavior has been observed in AA6061-T4 alloy also as shown in Fig. 12e and f.However, the smooth regions appear to be more in AA6061-T4 alloy than in multilayer deposition.AA6061-T6 alloy consists of large smooth regions, as shown in Fig. 12g and h, predominantly indicating brittle fracture.It is due to its solution heat-treated and artificially aged condition which has enhanced its hardness and strength and reduced its ductility, as evident from its values of tensile properties mentioned in Table 2. Dimples and ductile features may be present in AA6061-T6 but are generally limited compared to the other conditions.

Fretting wear behavior of Al 6061 multi-layer deposition and substrate
Fig. 13 depicts fretting wear results showing wear tracks (Fig. 13a), and coefficient of friction (COF) curves and wear depth profiles for the substrate and multi-layer deposition of Al 6061 alloy for the applied normal load of 5N (Figs.13b), 10N (Fig. 13c), and 15N (Fig. 13d).Values of COF spikes in the beginning of the fretting cycle for the multi-layer deposition and substrate for all the values of the applied normal loads which might be due to the presence of thin and soft oxide layers on them.X-ray photoelectron spectroscopy can be more useful in identifying the oxidation state [43].Once the soft oxide layer is broken, then the COF stabilizes.Multi-layer deposition of Al 6061 alloy has slightly smaller wear track than the substrate material AA6061-T4 alloy for all the values of the applied normal loads (Fig. 13a) due to relatively smaller value coefficient of friction (COF) which can be observed in Fig. 13b to d.The lower COF indicates reduced frictional forces and less resistance to sliding motion, resulted in relatively lesser surface damage than the substrate material.Lower COF of multi-layer deposition is also attributed to its relatively larger value of microhardness [44].Though applied normal load increased from 5N to 15N, values of COF for the multi-layer deposition and substrate were found to decrease consistently due to the formation of new oxide layers (i.e., Al 2 O 3 ) at the fretting zone, which acts as a solid lubricant [45].Presence of solid lubricant reduces friction and consequently decreases COF.However, maximum wear depth of the multi-layer deposition and substrate increases initially as applied normal load increases from 5N (Fig. 13b) to 10 N (Fig. 13c) but it decreases again as load increases from 10N to 15N (Fig. 13d) which might be due to stick-slip condition [46].The frictional force varies greatly as a function of sliding speed or distance during stick-slip rather than remaining constant.It increases to a critical value during the stick phase.Once a critical frictional force has been attained (to overcome the static friction), slip occurs at the fretting interface, and energy is released, reducing frictional force and consequently resulting in smaller maximum wear depth [47].

Conclusions
This paper presented solid state multi-layer manufacturing of Al 6061 alloy on AA6061-T4 substrate by the FSPAM process and enhancement of its microstructure and mechanical properties by reducing thermal gradient along the build direction.Following conclusions can be drawn from this work: The FSPAM process successfully produced defect-free and good-quality ten-layer single-track depositions of Al 6061 alloy without melting it.The layers are uniform, straight, and parallel, with minimal bending at their interfaces.Height of each layer is 0.5 ± 0.05 mm.The reduction in thermal gradient during deposition led to a homogeneous microstructure with fine and equiaxed grains (~1.12 mm), resulting from continuous dynamic recrystallization of the feedstock material.The process presented low porosity (0.19%) due to the intimate contact among powder particles caused by inherent compressive forces.Energy dispersive spectroscopy confirmed good mixing and consolidation of feedstock material, with no agglomeration of alloying elements beneath the FSPAM tool.Phase analysis revealed the presence of Al and hardening phase Mg 2 Si, indicating compressive residual stresses.The tensile properties (YS:106.2 ± 5; UTS:190.7 ± 11.9; Elongation: 16.7 ± 0.5%) and microhardness (87.7 ± 2.4 HV) of the multi-layer deposition are superior to AA6061-O alloy, with minor variations in microhardness (8.8%) along the build direction due to reduced thermal gradient.Fracture morphology showed many number of dimples indicating ductile nature of multi-layer deposition with 16.7% elongation.Dimples were absence at some zones indicating crack-impeding sites, likely due to site-specific fine grains.Fretting wear and microhardness results are in good agreement, showing a decrease in coefficient of friction with increased load due to oxide layer formation.The maximum wear depth initially increased and then decreased due to stick-slip conditions at higher applied loads.The approach of controlling substrate temperature to reduce thermal gradient is promising, but further optimization is needed.Initial results showed a reduction in temperature variation between layers, and ongoing research aims to enhance the efficiency and precision of the FSPAM process.

Fig. 1 e
Fig. 1 e Microstructure and EDS mapping of Al 6061 alloy powder: (a) morphology, and (b) distribution of major alloying elements Al, Mg and Si.

j o u r
n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 6 : 1 1 6 8 e1 1 8 4

Fig. 2 e
Fig. 2 e Apparatus and tool for FSPAM process: (a) schematic of working principle for manufacturing multi-layer deposition showing front and side view, (b) photograph of the developed apparatus and its different components, (c) location of the external heat source and temperature sensor in the apparatus, and (d) photograph of the developed circumferentially and radially grooved tool.

Fig. 3 e
Fig. 3 e Sample of Al 6061 multi-layer deposition prepared for the tensile testing and fractography: (a) photograph, and (b) dimensions as per the ASTM E8M standard.
versa.This rotation reversal effectively changes the AS and RS, thereby balancing the accumulation of feedstock material on both sides.Stand-off distance and feedstock material flow rate are found to play a vital role in controlling the flatness of the deposition layer.An essential observation indicates that if the already deposited layers have a good surface finish, it may facilitate strong metallurgical bonding for subsequent layers.Measured values of maximum height and arithmetical average roughness for the top deposition layer are 8.71 and 2.8 mm respectively.Measured values of the width and height of the multi-layer deposition are 22 ± 0.85 mm and 5 ± 0.1 mm respectively.The deviation in deposition width occurs because the plasticized material freely moves towards the AS and RS of the deposition.On the other hand, the deviation in deposition height is caused by constrained upward flow when the material exits the deposition zone on the trailing side of the tool.Fig.6depicts macrostructure of Al 6061 multi-layer deposition showing (i) interface between the substrate and the first deposition layer (Fig.

Fig. 5 e
Fig. 5 e Surface appearance of multi-layer deposition of Al 6061 alloy by FSPAM process: (a) photograph showing top and front view, (b) enlarged view of the deposition end, (c) multi-layers on the advancing side (AS) and, (d) multi-layers on the retreating side (RS).

Fig. 7 e
Fig. 7 e EBSD images of Al 6061 alloy multi-layer deposition by the FSPAM process showing (a) its cross-section, and inverse pole figure and image quality maps for its (b) top layer, (c) intermediate layer, and (d) first layer.

Fig. 8 e
Fig. 8 e SEM of the multi-layer deposition of Al 6061 alloy: (a) micrograph at 2000x, and (b) porosities in the enlarged view (at 3500x) of the red color area marked in Fig. 8a.

Fig. 11
Fig. 11 depicts results of tensile test and microhardness showing stress and strain curves of the multi-layer deposition of Al 6061 by the FSPAM process, AA6061-O (annealed),

Fig. 9 eFig. 10 e
Fig. 9 e Energy dispersive spectroscopy results for the multi-layer deposition of Al 6061 alloy showing: (a) elements weight percentage in it; and distribution elements such as (b) aluminum, (c) magnesium, and (d) silicon.

Fig. 13 e
Fig. 13 e Results of fretting wear for multi-layer deposition of Al 6061 alloy and substrate: (a) wear tracks for different values of the applied normal load, (b) coefficient of friction curves and wear depth profiles for the applied normal load of (b) 5N, (c) 10N, and (d) 15N.

Table 2 e
Tensile properties of Al AA6061-O, multi-layer deposition of Al 6061, AA6061-T4, and AA6061-T6 alloy along with standard deviation values.But it is smaller than AA6061-O Al alloy by 12.9% and very much close to the commercially available AA6061-T4 wrought alloy i.e., the substrate material.Thus, it can be concluded from the tensile testing that the multi-layer deposition of Al 6061 alloy by the FSPAM process possesses better tensile properties than Al 6061-O alloy, comparable to Al 6061-T4 alloy and lower than Al 6061-T6 alloy.