The Influence of the Surface Roughness on the Microstructures and Mechanical Properties of 6061...

Surface & Coatings Technology 270 (2015) 272–283

1

The Influence of the Surface Roughness on the Microstructures and Mechanical

Properties of 6061 Aluminium Alloy Using Friction Stir Welding

Hasan I. Dawood1Kahtan S. Mohammed ,٭

1, Azmi Rahmat

1, Uday M. B.

2

1Universiti Malaysia Perlis,

School of Materials Engineering, Taman Muhibah-Jejawi-Arau 02600

Perlis, Malaysia

2UTM - Centre for Low Carbon Transport in cooperation with Imperial College London, Institute

for Vehicle Systems and Engineering, Universiti Teknologi Malaysia

81310 Skudai, Johor, Malaysia

٭ E-mail address: [email protected], Phone: +60174489057, Fax: 04-979 8178

Abstract

In the present study, four pairs of 6061 aluminium alloy workpieces with different surface

roughness were prepared for welding. The friction stir welding (FSW) technique was used for a

butt-joint configuration of a single pass. The influence of different surface roughness of the

workpieces coincided with a small welding tool shoulder diameter, and the tool pin was examined.

The results demonstrated that spherical nano-sized grains of the joints were produced. The

mechanical properties of the joints were significantly better at the least possible workpiece surface

roughness. The experimental results also indicated that the tensile strength of FSW 6061 aluminium

alloy was notably affected by joining at the different workpiece surface roughness selected.

However, an improvement of the Vickers microhardness in the heat affected zone (HAZ) was also

observed. The microhardness in the nugget zone (NZ) for the welded joint fabricated at the lowest

value of the workpiece surface roughness was higher than that of the base metal (BM). The

fractural surface of the cross-section of the tensile specimens has a gradient to change from brittle

fracture to ductile fracture.

Keywords: Surface roughness; Spherical nano-size grains; Friction stir welding; Microstructures;

Mechanical properties; 6061aluminium alloy

mailto:[email protected]


2

1. Introduction

FSW is a solid state joining method in which the relative motion between the welding tool

and the material of the work piece produces heat. This heat makes the material soft, thereby

enabling it to be joined via plastic deformation and a thermal cycle effect (i.e., dynamic

recrystallisation) caused by the rotational welding tool. Dynamic recrystallisation is associated with

materials in which the dynamic recovery is sufficient to cover slow passage of the sub grain

boundaries. Such materials undertake this softening by continuous division of their structures. This

division produces a fine grained microstructure without involving any nucleation or growth

mechanisms [1]. This microstructure, in turn, results in a significant grain refinement in the nugget

zone (NZ). This method relies on the direct conversion of the mechanical energy to thermal energy

forming the weld without any external source of heat[2]. In the FSW process, a non-consumable

rotating tool is forced down into the joint line under certain conditions, such as the controlled force,

welding speed, and rotational speed of the tool. All of these factors are important for a successful

FSW process because the welding occurs in a state under the melting point of the material [3].

Thus, all problems related to the solidification of a fused material are avoided. FSW produces an

asymmetric microstructure, represented by the retreating side (RS) and the advancing side (AS). It

is experimentally demonstrated that there are different strain levels in these two sides. Different

thermal cycles induce different properties and precipitate distributions in the welding zones [4].

Although Al-Mg-Si alloys are readily weldable, they suffer severe softening in the HAZ because of

the dissolution of magnesium-silicon (Mg2Si) precipitates during the welding thermal cycle. This

type of mechanical impairment poses a major problem in engineering design. It is more appropriate

to overcome or minimise the HAZ softening to improve the mechanical properties of weldments

[5].


3

The relationship between surface roughness and friction was considered in the present study

because it is relevant to the FSW process. One of the early attempts to explain friction was to relate

it to the surface roughness because the surface is not generally smooth; instead, the surface is rough

and wavy. Several investigators attempted to study the effect of surface texture of soft materials on

friction [6-8]. For example, the effect of surface roughness on frictional resistance using different

types of surfaces was studied by Rasp and Wichern [8]. In these researchers experiment, the

condition of the surfaces of the specimen varied greatly: from the as-received state, various states of

being abrasively scratched parallel to perpendicular to the simulated rolling direction, to being as

good as a mirror polished surface. These researchers determined that the arithmetic roughness value

and lubrication regime have a greater influence on the frictional resistance than the directionality.

However, the surface texture of deformed material cannot explain the true friction values during

sliding; thus, it is important to characterise the surface texture of the harder material.

The present study investigated the effect of surface roughness of harder materials on the

friction coefficient. In the literature, several researchers had focused on the effects of surface

texture of harder materials on the friction coefficient during sliding [9-14]. For example, the effects

of surface texture and surface roughness using steel discs of varying roughnesses and textures were

studied by Staph et al. [13]; they concluded that both the surface texture and the surface roughness

affect frictional behaviour. The relationship between the friction and surface roughness using

various lubricants was studied by Hu and Dean[14]. These researchers reported that a random

smoother surface could retain more lubrication and reduce friction. The influence of the

directionality of the surface grinding marks on the friction coefficient and the transfer layer

formation for Al–Mg alloy pin slid against EN8 steel flat under both dry and lubricated conditions

using inclined scratch test was studied by Menezes et al.[15]. The value of the roughness, as


4

represented by Ra, of the surfaces varied over a range via preparation using different grit emery

papers. The average friction coefficient and transfer layer formation were observed to rely primarily

on the directionality of the grinding marks but were independent of surface roughness on the harder

mating surface. Lundberg[16]also studied the influence of surface roughness on normal sliding

lubrication and reported that the Rmax was found to be the most significant surface roughness

parameter. The effect of the surface texture on the friction mechanism studied using the universal

testing machine was reported by Koura[17]. Steel specimens were prepared for various degrees of

roughnesses by grinding, lapping and polishing. The results indicated that the surface behaviour

and the friction during sliding depend on the degree of roughness. However, the correlation

coefficient between the friction coefficient and the roughness parameters were noted to depend on

the system. In addition, reports in the literature indicated that the surface roughness is the most

important parameter in describing the surface integrity of a material. The surface roughness also has

a great importance in specifying the function of the surface because a significant proportion of

material failure starts at the surface, due to either the discontinuity or deterioration of the quality of

the surface. The surface finish also plays an important role in corrosion resistance and may increase

the performance and reduce the costs of the life cycle of a part. Therefore, these consequent effects

raise an important issue that must be investigated. However, optimisation of the microstructure and

properties of the surface is an effective approach to enhance the global behaviour and service life

time of materials [18].

Plastic deformation induces grain refinement in the nanometre regime. This process has

been extensively investigated in various metals and alloys over the past decades, both in bulk forms

[19] and in the surface layers of bulk materials [20]. Optimising the surface properties by refining

the grains in the surface layers down to the nanometre scale (referred to as surface nano


5

crystallisation) promises to provide more practical industrial applicability. Surface mechanical

attrition treatment (SMAT) is a technique that can induce grain refinement into the nanometre

regime in the surface layer of bulk materials [21]. The SMAT transforms the original coarse grains

in the surface material into refined grains [22]. The grain refining mechanisms in SMAT are

implemented by the random and repetitive impact of milling balls to the sample’s surface. This

technique has been successfully applied in achieving surface nano crystallisation (SNC) in a

variety of materials, including pure metals, alloys and steels [23-26]. Friction sliding, another

technique for grain refinement down to the nano scale, is based on the principle of plastic

deformation-induced grain refinement [27]. However, these two processes suffer from either low

processing efficiencies and limited nano-structured layer thicknesses or structural inhomogeneity in

the surface layer; all of these negative characteristics have hindered the widespread application of

surface nano crystallisation technologies [28].

Regarding the FSW process, it is well-accepted that grain refinement that increases the

strength and ductility of aluminium alloys can lead to a wider use of FSW in a variety of industries,

including the defence industry. To obtain a microstructure with improved mechanical behaviour,

several methods based on plastic deformation have been developed for grain refinement in alloys,

such as rolling, Equal Channel Angular Processing (ECAP) and Equal Channel Angular Extrusion

(ECAE). Due to the importance of this topic, many relevant research studies have been reported.

Brian et al. [29] studied the grain refinement using nitrogen liquid as a cooler; they found that the

average nano-grain size is typically approximately 500nm. Grain refinement via FSW in alloys has

been reported by different researchers [30-36]. The grain structures were found to depend heavily

on the processing conditions, including processing parameters, tool geometry and cooling rate.


6

The present study involved development of a simple and efficient method of producing the

desired substructure in materials based on grain refinement into the nanometre regimes. The

investigation attempted to elucidate the effect of surface roughness of the workpiece before the

process of welding, and efforts were made to correlate these roughness parameters with the

coefficient of friction; the results revealed the effect of the small tool welding dimension on the

microstructural evolution, mechanical properties, nano-grain size refinement and HAZ

microhardness of 6061 aluminium alloy.

2. Materials and Experimental Procedures

The material used in this study was the 6061 aluminium alloy supplied by a local supplier

(Heap Sing Huat Metal and Machinery Sdn Bhd). The plates with a thickness of 4mm were sliced

into the required size (210 mm × 200 mm) using a cutting band saw machine (MODEL: UE-712A).

Three pairs of the workpieces were ground using coarser and finer abrasive papers of 240 to 1200

grits to remove material from the workpiece surfaces until the desired surface roughness was

achieved. The first pair of workpieces was ground with abrasive paper of 240 grit, and the other two

pairs were ground with abrasive papers of 240 to 600 and 240 to 1200 grit.

The average surface roughness (R) of the all workpieces were measured using Atomic Force

Microscopy (AFM) (Nanoscope Analysis; version 1.20; Veeco), and the surface roughnesses of the

four pairs of workpieces were found to be R1= 98 nm (as-received surface), R2= 60 nm, R3= 51nm

and R4= 37 nm; the surface profiles of the workpieces are presented in Fig. 1. The friction

coefficient, µ, was measured using a Pin-on-Disc geometry (CSEM tribometer) to calculate the heat

generated during FSW. The coefficient of friction was measured for each workpiece surface

roughness by taking the average temperatures determined by K-type grounded thermocouples with

http://en.wikipedia.org/wiki/Abrasive


7

a sheath diameter of 1mm attached to the workpiece surfaces at different points near the boundary

of the shoulder during the FSW process. For this purpose, an external oven was used to increase the

temperature of each workpiece surface to the desired value before using the CSEM tribometer. The

measurements of the friction coefficient were performed at a load of 15 N and at a velocity

of1mm/sec. After the test is completed, the average friction coefficient was considered for each

workpiece surface roughness. The test workpieces were rigidly clamped onto the backing plate in a

butt-joint configuration parallel to the rolling direction of the plates. The surface of each of the

plates was cleaned using acetone to remove dirt and grease. The chemical compositions of the BM

are presented in Table 1.

FSW was performed on a vertical milling machine, type KAMA (X6325; 3Hp; TRPER R8;

30 kN). Fig. 2a shows a schematic diagram of the FSW process. A non-consumable tool made of

medium carbon steel was used to fabricate the joints, which was heat-treated and quenched to

60HRC. The chemical compositions of the welding tool are listed in Table 2. For the welding

process, a tool with a shoulder diameter of 9 mm and a straight cylindrical pin with a diameter of 2

mm and a length of 3.7 mm were used, as shown in Fig. 2b. Under the action of a vertical axial

force imposed by the pin and shoulder of the tool, the pin is forced between the workpieces to

ensure the best possible contact between the shoulder and the top surface of the material. Frictional

heat is generated between the welding tool and the welded material, leading to softening of the

material around the welding pin. Accordingly, during the FSW process, the axial force (including

the friction pressure and the press depth) is not constant and gradually decreases after any step pass

of welding due to the temperature increase, which makes the material more flexible. Therefore, in

this study, the axial force was set to be in the range from the maximum value of 7.5 kN to the

minimum value of 3 kN, and the load cell with a digital screen was used.


8

The welding process was implemented using a rotational speed of 1750 rpm in the counter

clockwise direction, and the welding speed was 60 mm/min, which was normal to the rolling

direction. The single pass welding method was implemented to fabricate the joints. Five welds were

performed for each test parameter to ensure the accuracy and repeatability of the results. Afterward,

the best possible welded specimen was selected from each set of the of test parameters of the

specimens to study the microstructure and mechanical properties. Microstructural analysis was

performed on aluminium alloy joints perpendicular to the welding direction. The specimens for

metallographic examination were sectioned to the required sizes from the NZ, TMAZ, HAZ and BM

regions. Next, the samples were ground with 180 to 1200 grit silicon carbide papers. Final polishing

was performed using a diamond paste of particle size of 0.5µm. A solution of Keller’s reagent (3ml

of hydrochloric acid, 5ml of nitric acid, 2ml of hydrofluoric acid and 190 ml of distilled water) was

prepared to reveal the microstructures, according to the American Standard Testing of Materials

(ASTM: E407-07: ETD 2013). The microstructures of the specimens were observed using an optical

microscope (Olympus BX41RF) and a Field-Emission Scanning Electron Microscopy (FESEM)

apparatus (FEI Nova NanoSEM 450) equipped with an energy dispersive X-ray (EDX) analyser to

investigate the effect of plastic deformation on the grain size. The welding sample of the 6061

aluminium alloy was analysed via X-ray diffraction (XRD) using a copper target as the X-ray source

to identify the phases in the welded joints. The instrument used in this XRD work was a D8

Advance X-ray diffractometer (Bruker Analytical X-Ray Systems).

For the tensile tests, the FSW specimens were cut perpendicular to the welding direction

according to ASTM: E8/E8M-11, B557: ETD 2013. The specimens were loaded at the rate of

1kN/min using an INSTRON (Universal Testing Machine). Microhardness testing apparatus using

the conventional Vickers microhardness technique; the measurements were performed across the


9

joints and the BM at specific locations according to ASTM E384-11, ISO 6507-1: ETD 2013. For

the measurements of the microhardness, at least 19 Vickers type indentations on both sides of the

welded specimens were performed on polished cross-sections perpendicular to the welding

direction, using a load of 9.8 N (1 kg) and a dwell time of 15 sec using a Future-Tech (FV-700e)

microhardness tester (Future-Tech Corp., Japan). The microhardness was more closely measured,

with measurements being taken every 1 to 1.5 mm. The procedure of the microhardness test was

repeated until at least five impressions were made at widely spaced locations along the welding

line. The FESEM technique was also used to study the fracture surface of the tensile specimens.

3. Results and Discussion

3.1 Effects of surface roughness

Fig. 3a shows the specimen that represents the welded joint formed for the case that the

surface of the workpiece was not ground before the FSW process (i.e., the as-received surface). The

mechanism of formation of the onion rings is observed at the interface between the 6061 aluminium

alloy joints. The spacing of the layer in the onion ring pattern is equal to the forward motion of the

tool in one rotation, as was found in [37, 38]. Figs. 3(b, c and d) indicate that the characteristics of

the joint surfaces became smooth and homogeneous gradually after a slight grinding before

embarking on the FSW process. Hence, it can be inferred that the reduction in the surface

roughness of the workpieces plays an important role in controlling the quality of the weld surface,

considering that all the specimens were fabricated under the same conditions (i.e., rotational speed,

welding speed, axial pressure, tool dimensions and profile).

The flow of the material on the Advancing Side (AS) is different from that observed on the

Retreating Side (RS). The material on the RS never enters the rotational zone near the pin because


10

the material on the AS forms the fluidised bed near the pin and rotates around it. In the transition

zone, material transfer occurs mainly on the RS. This type of motion mechanism was also

supported by Ying et al. [39]. Accordingly, the long-term trend at the retreating side reflects more

flash, as seen in Figs. 3(a and b); in addition, the RS represents the location of the lowest values of

hardness at the same time. Figs. 3(c and d) show the absence of flash on the RS, probably due to the

lack of heat generation caused by the reduction of surface roughness, which makes the material less

flexible and difficult to extrude under the shoulder area.

3.2 Heat generation during the FSW process

The friction heat flow is a highly important factor for the FSW process because the

microstructure in the stir zone is strongly influenced by the friction heat flow [40]. The friction heat

input from the welding tool during the FSW process, qo, can be evaluated according to the report of

Frigaard et al. [41] as follows:

(1)

Where µ is the friction coefficient, F is the axial force, N is the rotational speed and D is the

shoulder diameter. If the contact area is Ac ((D/2)2×π), then the resulting heat flux at the stir zone is

(2)

During the transient heating period, when the temperature starts to build up under the tool

shoulder, Eq. 1 is believed to provide an adequate description of the heat generated during FSW.

This result is probably true as long as the individual contributions of the rotating movement of the

shoulder and the pin are responsible for generating the heat. The frictional heating contributes to the

32

23

4 DFNqo

C

o

A

qQ


11

formation of a plasticised layer of soft metal beneath the tool shoulder and around the pin. The

mechanism of plastic deformation and the variable frictional conditions at the tool interface are

accounted for by the use of a reasonable average value for the friction coefficient (µ). However, in

the thermal processing of aluminium alloys, it is generally accepted that local melting will occur if

the material is heated above the eutectic temperature at a rate that does not allow the eutectic phase

to fully dissolve. If the conditions for local melting are met, then the presence of a liquid film at the

tool interface will lubricate the tool shoulder, resulting in a lack of friction [41].

All of the previous studies considered that the friction coefficient was constant during FSW.

In this study, the variable friction coefficients were taken into consideration because the surface

roughness of the workpieces was reduced before the welding process. The friction coefficient was

measured at the average temperatures ranging from 211ºC to 301

ºC for different surface roughness

values of the workpieces; the results are represented in Fig. 4, which shows that the friction

experienced two stages. First, the friction coefficient rapidly increased to approximately 0.38.

Afterward, the friction coefficient increased slightly to reach a maximum of 0.41.

Four different friction coefficients were calculated using Eqs. (1 and 2) to study the effect of

the friction coefficients on heat generation. Fig. 5 shows the frictional heat flux per millimetre

square area for different workpiece surface roughness values or friction coefficients. Based on the

results, the high friction coefficient clearly leads to the highest heat flux. The frictional heat flux is

reduced by approximately 1.2 times when the friction coefficient is reduced from 0.41 to 0.34. This

result occurs because higher values of the friction coefficient correspond to a relatively higher value

of friction; consequently, higher energy will be produced.


12

3.3 Characterisation of the microstructure

Fig. 6 shows the microstructures of the welded joints of all specimens and taken at a

magnification of 100× using the OM. The microstructure, which is characterised by the central

welded joints, represents the NZ. The material was plastically deformed in this region, which

indicates a combined action of a high strain rate at elevated temperatures due to the FSW pin tool.

The NZ is surrounded by a thermomechanically affected zone (TMAZ). The data suggested that

this area is below the tool shoulder and that the material in this part is plastically deformed without

any dynamic recrystallisation. The heat affected zone (HAZ) has modified the microstructure and

the mechanical properties, which fall between the TMAZ and the unaffected base metal (BM).

Thus, there is no plastic deformation associated in this region, and it has only a thermal cycle. BM

is a remote material from the weld; from an experimental perspective, the BM has a thermal cycle

from the weld that is not affected by the heat in terms of the microstructure or mechanical

properties.

From Fig. 6, the BM is also observed to consist of large equiaxed grains and to contain

impurities. Generally aluminium alloys mainly include such impurities as iron, silicon, manganese,

copper and chromium, which are fairly coarse intermetallic compounds [42]. The strength and

hardness of some metal alloys may be enhanced by the formation of extremely fine uniformly

dispersed precipitates within the original phase matrix. It is obvious from Figs. 6(a, b, c and d) that

there are four distinct regions with different microstructural characteristics, divided by the white

dotted lines. These distinct regions are namely NZ, TMAZ, HAZ and BM. Some of the very fine

scale precipitates can be observed in the regions of TMAZ and HAZ. In the NZ, a clear refinement

of the microstructure is observed. In the HAZ, there are coarse equiaxed grains, while in the

TMAZ, there are elongated grains created between the NZ and the HAZ because of the shear


13

stresses. Interestingly, the width of the HAZ is found to decreases gradually with the decrease in the

workpiece surface roughness. Because of the decrease in the workpiece surface roughness causes

the decrease of the heat generation and the lower amount of heat conducted to the HAZ region,

resulting in reduction in the width of the HAZ.

Fig. 7 shows the microstructures in the NZ of the welded joints observed using the FESEM

technique. This technique offers the ability to gather information regarding finer grains compared to

the grains that can be characterised by OM. FESEM can readily distinguish between the clusters

and agglomerates of grains in a time-efficient manner. The FESEM technique was used to reveal

and measure the grain sizes in the weld nugget after etching. The grains in the NZ are observed to

be refined, with the mean size equal to~16.4 µm for a workpiece surface roughness of R1 (Fig. 7a).

As shown in Fig. 7b, the NZ also consists of spherical nano grains, with an average size of ~58.78

nm for a workpiece surface roughness of R2. The average grain sizes in the NZ for the workpiece

surface roughness values of R3 and R4 (Figs. 7 (c and d)) are ~56.69 and ~44.66 nm, respectively.

On one hand, the value of the minimum workpiece surface roughness (R4) indicates the quantitative

effects on grain refinement. On the other hand, the value of the maximum grain size was achieved

with the maximum value of workpiece surface roughness (R1), while the minimum grain size was

achieved with the minimum value of workpiece surface roughness (R4). Accordingly, it can be

concluded that the grain size in the NZ is decreased whenever the value of the workpiece surface

roughness is decreased. Hirata et al. [40] reported that the grain size in the nugget region was

decreased when the friction heat input was decreased. As a result, reduction in the workpiece

surface roughness value caused decreases in heat generation, resulting in further grain refinement.


14

An EDX analysis was performed for the NZ joints to investigate the phases that were

formed during the FSW process. The results illustrated in Figs. 8(a, b, c and d) show that the

chemical composition was varied, without referring to the formation of brittle intermetallic

compounds. This variation is attributed to the variation in dispersive precipitates by pinning the

dislocations up to certain values of the shear and thermal stresses. The results are found to confirm

the presence of a large percentage of carbon particles and a significant proportion of oxygen. The

existence of carbon in the NZ might be due to the action of carbon atom diffusion from the welding

tool during the FSW process. The tool pin profile was observed to tend to wear out after a few runs

of welding, leading to a conical form. This behaviour occurred particularly while working with

relatively higher tensile strength materials. In contrast, Kim et al.[43] reported that, at constant

rotational speed, the mechanical resistance of joints increases with increasing the travel speed,

which is attributed to the decreased heat input. Accordingly, the present study introduced lower

heat generation due to the effects of the lower workpiece surface roughness, the small size of the

welding tool dimensions and the welding speed being relatively high; therefore, the possibility of

welding tool erosion was taken into consideration.

Oxidisation of the Al–Al welded joints is expected. Frictional heat generation and an

atmospheric humidity of 75% enhance the oxidation rates of aluminium during the welding process.

The results of EDX analysis indicate that the formation of aggregation compounds (wt% C and O)

diminishes gradually whenever the value of the workpiece surface roughness is decreased. This

phenomenon explains the absence of any nano pores and void-related effects on the top surface of

the welded joint, when the micro-range surface roughness is modified.


15

3.4 X-ray diffraction (XRD) analysis

To determine the type of phase composition of the aluminium alloy used in the study of the

FSW process, the welded joints of the 6061 aluminium alloy were analysed via XRD patterns using a

copper target, and the results are shown in Fig. 9. The relevant parameters corresponding to the

diffraction peaks in the diagram were computed using X’ Pert High Score Plus software; the results

indicate that the aluminium alloy has a cubic structure. The relative intensities in the XRD peaks of

the 6061 aluminium alloy match with the pattern of the reference code 98-006-2706. The results

indicate the maximum and minimum diffraction peaks of the XRD patterns across the welded joints

of aluminium. As the data in the figure indicates, there is an absence of detrimental phases in the

product (i.e., intermetallic compounds). A high heat input was reported to result in a high growth

rate of intermetallic compounds [44]. However, this study focused on minimising the heat input to

the welded joints through the effects of surface roughness, the smaller tool pin, and the smaller

shoulder diameter.

After processing the XRD pattern and calculating the crystallite size, the results of the four

slightly different values of the crystallite sizes for all the specimens and the BM are presented in

Table 3. The scatter in the results of the crystallite size decreases with the decreased values of surface

roughness; however, in general, the results remain very close, except for the crystallite size at a high

rate of surface roughness values of R1 and R2. In addition, the relationship between the crystallite

strain and the values of surface roughness in the present 6061 aluminium alloy is also calculated. The

crystallite strain is observed to decrease slightly with a decrease in the roughness of the material

surface.


16

3.5 Measurement of the mechanical properties

3.5.1 Tensile strength of the joints

To determine the mechanical properties of the obtained welded joints, tensile engineering

stress-strain curves were performed, as shown in Fig. 10. Of the four joints, the joint fabricated with

a workpiece surface roughness of R4 exhibits a higher ultimate tensile strength (UTS) and

elongation at break compared to the other joints, which were fabricated at workpiece surface

roughnesses of R1, R2 and R3. As seen in this figure, the roughness of the material surface has a

perceptible influence on the tensile properties of the FSW joints. Similarly, the joint fabricated with

a workpiece surface roughness of R3 exhibits a moderate tensile strength when compared to that of

R4. However, the joint fabricated with a workpiece surface roughness of R1 exhibits an inferior UTS

and elongation at break compared to their counterparts.

There are two reasons for the value of UTS of the fabricated joints being less than the value

(355.6 MPa) of BM. One reason is the effect of the dissolved precipitates formed during FSW

process, which causes a reduction in the tensile strength of the joints. The other reason is the strong

tendency for intergranular cracking because of the combined effects of the weak grain boundaries

and the concentration of the grain boundary stress. Hence, cracks can expand rapidly along the

grain boundary. Therefore, the strength of the BM is higher than the strength of the welded joints.

3.5.2 Vickers microhardness tests

The average Vickers microhardness of the welded joints were measured transversely and

perpendicular to the direction of the welding line. The microhardness profiles for the specimens are

represented in Fig. 11. The highest microhardness gradient in the NZ is observed when the value of

the workpiece surface roughness is R4, while the lowest microhardness is obtained in the NZ when


17

the value of the workpiece surface roughness is the highest (R1). Hence, the main difference

between both cases of microhardness is attributed to the workpiece surface roughnesses. As

previously mentioned, the lowest value of workpiece surface roughness will generate less heat

during the friction process. Therefore, in the present study, the microhardness of the NZ was found

to increase with the decreasing heat generation. This result is in good agreement with the respective

findings of Hirata et al. [40].

As described in the literature, in alloys such as the 6061 aluminium alloy, the main

strengthening precipitate is magnesium-silicon (Mg5Si6), which is stable at temperatures lower than

200˚C [45, 46]. This precipitate exists in the BM; however, the precipitate is absent in the NZ and

in the HAZ region. In the FSW, the temperatures are expected to be higher than 200 to 250˚C, and

Mg5Si6 is easily dissolved [42]. This precipitate is the main reason for the hardening behaviour. The

nugget microhardness revival is due to dynamic recrystallisation of a very fine grain structure. An

increase of the microhardness in the NZ was found due to the development of a fine-grained

microstructure. According to the well-known Hall-Petch relationship [40], the microhardness

increases slightly with decreasing grain size;

(3)

where Hv is the hardness, H0 and KH are the appropriate constants associated with the hardness

measurements, and d is the grain size. The reduction in workpiece surface roughness caused

decreases in the grain size; thus it increased the apparent microhardness (Hv) of the NZ according to

the Hall-Petch formula (Eq. 3).

2/1

0

dKHH Hv


18

Fig. 11 shows that the microhardness gradient in the area of NZ is significantly higher than

TMAZ and HAZ; however, it is higher than the BM for the value of workpiece surface roughness

of R4. This variation in the microhardness between the welding regions and the BM is due to the

difference in microstructures [47]. On the AS, a higher value of microhardness than the RS through

the TMAZ and HAZ is observed. This result is due to the existence of more strain and plastic

deformation caused by the welding tool, resulting in remarkable refining effects.

Scialpi et al. [48] reported that the HAZ experience has a peak temperature; therefore, the

results indicate a decrease in hardness. Improvement of the properties of HAZ was proposed many

years ago. However, Fig. 11 indicates that the slope of the microhardness line through the HAZ

reaches the proficient level that is comparable to the literature findings [46]. In the present study,

the results exhibit a significantly lower heat generation, probably due to the effects of utilising

reduction of workpiece surface roughness and the small welding tool dimensions. The slight

increase of microhardness in the HAZ indicates that there is a great reduction in the transfer of heat.

This reduced heat transfer tends to reduce the extent of the HAZ, such that the temperature is not

high enough to dissolve all the precipitates during the FSW process. This result is in good

agreement with the findings of Uzun et al. [49]. Accordingly, it appears that most of the heat

generated by friction is located under the shoulder area (i.e., regions of NZ and TMAZ). Thomas et

al. [50] reported that if the shoulder diameter is large, then the heat generated by friction is high,

due to the large contact area, resulting in a wider TMAZ and HAZ. Therefore, according to this

topic, it can be concluded that, in the case of a small shoulder diameter and reduction in workpiece

surface roughness, the heat generated by friction and conducted to the HAZ is

correspondingly reduced. This process results in a narrow HAZ region and a desired level

of softening, which result in increased values of the microhardness.


19

Fig. 12 shows the data regarding the relationship between the Vickers microhardness and

the grain size in the NZ. The Vickers microhardness versus the reciprocal of the square root of

grain size for the NZ is plotted in this figure according to the Hall-Petch equation. The figure shows

that the microhardness is roughly proportional to d-1/2

. Thus, the value of the microhardness

increases with the decrease in the size of the grain.

3.6 Fractography

A special feature remains on the fractural cross-section of the materials, which is useful to

investigate the cause and mechanism of the fracture. To obtain a better understanding of the failure

micro-mechanism of the FSW zone, a fractographic examination using FESEM was performed on

the fracture surface of the tensile test specimens.

Figs. (13-16) demonstrate four types of fracture modes existing in the 6061 aluminium alloy

joints with surface roughness of R1-R4. The fracture surface of a cross-section of the specimen in

Fig. 13 exhibits the characteristics of brittle fracture: the cracks rapidly expand along those defects

and fractures under a relatively low stress. Part of the area of the fracture exhibits some features of

ductile fracture, as shown in Figs. 14-16. However, the local, well-connected weld may deform

plastically and form porous coalescing dimples. The figures (14-15) also show the existence of

microscopic voids of different sizes and forms, which confirms the presence of ductile fracture

[51]. The brittle fracture areas are limited for a workpiece surface roughness of R1 (Fig. 13); while

the other three cases of the fractural surface in cross-section of the workpieces of R2-R4 exhibited

partially ductile and brittle fracture modes. The fractural features of specimens R1-R3 indicate that

the failure area is at the TMAZ of the AS beside the NZ because of the higher temperature and

material flow at this side. Therefore, when the specimen breaks at the weld zone, it is a result of


20

ductile and brittle fracture [51]. Specimen of R4, was fractured from the HAZ near the BM region,

which proves that the weld is sound. Hence, it can be inferred that the position of fracture in the

weld reflects the location of minimum microhardness region. From the microscopic fractographs

(Figs. 14-16) of the fractural surface cross-section, some small equiaxed dimples can be found,

indicating that a ductile fracture mode has occurred in these regions. Yu et al. [52] reported that the

small dimples come from the homogeneous microstructure of the small grains at the failure region.

The fractural surface cross-section also shows many petals, such as rip edges and a cleavage plan.

With the decrease in the value of the workpiece surface roughness (R), the fracture mode is found

to have a tendency to change from brittle fracture to ductile fracture, which corresponds to the

change of the elongation.

4. Conclusions

In this study, 6061 aluminium alloy workpieces with different surface roughness were

welded successfully via the FSW technique. The results indicated that the reduction in surface

roughness of the workpieces plays an important role in controlling the quality of the weld. The RS

of the specimens exhibited disappearance of flash gradually due to the effect of the reduction in

workpiece surface roughness, which caused the reduction of heat generation and made the material

less flexible and difficult to extrude under the tool shoulder area. Characterisations of

microstructure revealed that the HAZ was narrow due to the effect of reduced workpiece surface

roughness, which induced a lower level of shear stress under the tool shoulder. This mechanism

reduces the heat generation and heat conducted through the workpiece, which leads to a narrow

HAZ region. The effect of surface roughness was also found to coincide with a small tool pin and a

small shoulder diameter, which had the responsibility for the main mechanism of grain refinement.

Accordingly, the spherical nano-grain size was achieved in an economical manner. Regarding the


21

tensile strength results, generally, the weakest part of the welded joints is TMAZ at the AS due to

the higher temperature and material flow than that of RS. The improvement of microhardness in the

NZ and HAZ was found to be due to further grain refinement and efficient heat recovery by

friction, respectively. Furthermore, fractographs of tensile specimens were in good agreement with

the partial brittle and ductile fracture, in contrast to the specimen with a workpiece surface

roughness of R1, for which a cleavage and brittle type fracture mode was observed. The high

ductility results are due to the existence of microscopic voids of different sizes. The presence of

microscopic voids confirms the existence of ductile fracture, where the specimen breaks outside the

NZ.

Acknowledgments

This work was supported under the grant No. 9001-00338 of the Universiti Malaysia Perlis

(UniMAP). The authors gratefully acknowledge the outstanding support provided by the staff in the

Materials Engineering School, Universiti Malaysia Perlis (UniMAP), UTM Centre for Low Carbon

Transport, in cooperation with Imperial College London, Transportation Research Alliance in

Universiti Teknologi Malaysia (UTM) and the School of Materials Engineering and Mineral

Resources in Universiti Sains Malaysia (USM).

References

[1] A. Rollett, F. Humphreys, G.S. Rohrer, M. Hatherly, Recrystallization and related annealing

phenomena, Elsevier, 2004.

[2] R. Nandan, T. DebRoy, H. Bhadeshia, Progress in Materials Science, 53 (2008) 980-1023.

[3] P. Staron, M. Kocak, S. Williams, A. Wescott, Physica B: Condensed Matter, 350 (2004) E491-

E493.


22

[4] G. Liu, L. Murr, C. Niou, J. McClure, F. Vega, Scripta materialia, 37 (1997) 355-361.

[5] J. Koo, M. Luton, N. Bangaru, R. Petkovic, D. Fairchild, C. Petersen, H. Asahi, T. Hara, Y.

Terada, M. Sugiyama, International Journal of Offshore and Polar Engineering, 14 (2004) 2-10.

[6] D. Bello, S. Walton, Tribology international, 20 (1987) 59-65.

[7] P.K. Saha, W.R. Wilson, R.S. Timsit, Wear, 197 (1996) 123-129.

[8] W. Rasp, C. Wichern, Journal of Materials Processing Technology, 125 (2002) 379-386.

[9] R. Lakshmipathy, R. Sagar, International Journal of Machine Tools and Manufacture, 32 (1992)

685-693.

[10] S. Malayappan, R. Narayanasamy, The International Journal of Advanced Manufacturing

Technology, 23 (2004) 636-643.

[11] Y. Xie, J. Williams, Wear, 196 (1996) 21-34.

[12] A. Määttä, P. Vuoristo, T. Mäntylä, Tribology International, 34 (2001) 779-786.

[13] H. Staph, P. Ku, H. Carper, Mechanism and machine theory, 8 (1973) 197-208.

[14] Z. Hu, T. Dean, International Journal of Machine Tools and Manufacture, 40 (2000) 1637-

1649.

[15] P.L. Menezes, S.V. Kailas, Journal of tribology, 128 (2006) 697-704.

[16] J. Lundberg, Tribology international, 28 (1995) 317-322.

[17] M. Koura, Wear, 63 (1980) 1-12.

[18] K. Lu, J. Lu, Materials Science and Engineering: A, 375 (2004) 38-45.


23

[19] R.Z. Valiev, R.K. Islamgaliev, I.V. Alexandrov, Progress in materials science, 45 (2000) 103-

189.

[20] L. Jian, J. Mater. Sci. Technol, 15 (1999) 193.

[21] H. Zhang, Z. Hei, G. Liu, J. Lu, K. Lu, Acta Materialia, 51 (2003) 1871-1881.

[22] B. Arifvianto, in: Instrumentation, Communications, Information Technology, and

Biomedical Engineering (ICICI-BME), 2009 International Conference on, IEEE, 2009, pp. 1-4.

[23] G. Liu, J. Lu, K. Lu, Materials Science and Engineering: A, 286 (2000) 91-95.

[24] G. Liu, S. Wang, X. Lou, J. Lu, K. Lu, Scripta Materialia, 44 (2001) 1791-1795.

[25] N. Tao, Z. Wang, W. Tong, M. Sui, J. Lu, K. Lu, Acta Materialia, 50 (2002) 4603-4616.

[26] X. Wu, N. Tao, Y. Hong, B. Xu, J. Lu, K. Lu, Acta Materialia, 50 (2002) 2075-2084.

[27] D. Hughes, N. Hansen, Physical Review Letters, 87 (2001) 135503.

[28] W. Li, N. Tao, K. Lu, Scripta Materialia, 59 (2008) 546-549.

[29] B. Thompson, K. Doherty, J. Su, R. Mishra, Friction Stir Welding and Processing VII, (2013)

9-19.

[30] D. Zhang, M. Suzuki, K. Maruyama, Scripta materialia, 52 (2005) 899-903.

[31] A. Feng, Z. Ma, Scripta materialia, 56 (2007) 397-400.

[32] W. Woo, H. Choo, D. Brown, P. Liaw, Z. Feng, Scripta materialia, 54 (2006) 1859-1864.

[33] C. Chang, X. Du, J. Huang, Scripta Materialia, 57 (2007) 209-212.


24

[34] G. Bhargava, W. Yuan, S. Webb, R.S. Mishra, Metallurgical and Materials Transactions A, 41

(2010) 13-17.

[35] C. Chang, X. Du, J. Huang, Scripta materialia, 59 (2008) 356-359.

[36] W. Yuan, S. Panigrahi, J.-Q. Su, R. Mishra, Scripta Materialia, 65 (2011) 994-997.

[37] M. Sutton, B. Yang, A. Reynolds, R. Taylor, Materials Science and Engineering: A, 323

(2002) 160-166.

[38] K. Krishnan, Materials science and engineering: A, 327 (2002) 246-251.

[39] Y. Li, L. Murr, J. McClure, Materials Science and Engineering: A, 271 (1999) 213-223.

[40] T. Hirata, T. Oguri, H. Hagino, T. Tanaka, S.W. Chung, Y. Takigawa, K. Higashi, Materials

Science and Engineering: A, 456 (2007) 344-349.

[41] Ø. Frigaard, Ø. Grong, O. Midling, Metallurgical and materials transactions A, 32 (2001)

1189-1200.

[42] L.-E. Svensson, L. Karlsson, H. Larsson, B. Karlsson, M. Fazzini, J. Karlsson, Science and

technology of welding and joining, 5 (2000) 285-296.

[43] Y. Kim, H. Fujii, T. Tsumura, T. Komazaki, K. Nakata, Materials Science and Engineering: A,

415 (2006) 250-254.

[44] Y.S. Sato, M. Urata, H. Kokawa, Metallurgical and Materials Transactions A, 33 (2002) 625-

635.

[45] Ø. Grong, D. Grong, Metallurgical modelling of welding, Institute of materials London, 1997.

[46] P. Moreira, T. Santos, S. Tavares, V. Richter-Trummer, P. Vilaça, P. De Castro, Materials &

design, 30 (2009) 180-187.


25

[47] G.R. Babu, K. Murti, G.R. Janardhana, ARPN Journal of engineering and applied sciences, 3

(2008) 68-74.

[48] A. Scialpi, M. De Giorgi, L. De Filippis, R. Nobile, F. Panella, Materials & Design, 29 (2008)

928-936.

[49] H. Uzun, C. Dalle Donne, A. Argagnotto, T. Ghidini, C. Gambaro, Materials & design, 26

(2005) 41-46.

[50] W. Thomas, E. Nicholas, Materials & design, 18 (1997) 269-273.

[51] K. Kumari, S.K. Pal, S.B. Singh, Journal of Materials Processing Technology, 215 (2015) 132-

141.

[52] L. Yu, K. Nakata, J. Liao, Journal of Alloys and Compounds, 480 (2009) 340-346.


26

List of Figures

Fig. 1: Typical 3D surface roughness maps of different surface roughness of the workpieces

measured using Atomic Force Microscopy (AFM); (a) R1 = 98 nm, (b) R2 =60 nm, (c) R3=51 nm

and (d) R4=37 nm.

Fig. 2: (a) Schematic diagram of the FSW process and (b) the welding tool dimensions.

Fig. 3: Typical images of 6061 aluminium alloy surface appearances of friction stir welded with the

following surface roughness values; (a) R1 = 98 nm, (b) R2 = 60 nm, (c) R3= 51 nm and (d) R4= 37

nm.

Fig. 4: Variation of the workpiece surface roughness (R1-R4) and the friction coefficient with the

average friction temperature for joining pieces of 6061 aluminium alloy via friction stir welding.

Fig. 5: Variation of the workpiece surface roughness and the friction coefficient (µ) on the

frictional heat flux for joining pieces of 6061 aluminium alloy.

Fig. 6: Optical images of 6061 aluminium alloy joints friction stir welded with different surface

roughness of the workpieces: (a) R1 = 98 nm, (b) R2 = 60 nm, (c) R3= 51 nm and (d) R4= 37 nm.

Fig. 7: FESEM micrographs showing grain microstructures in the weld nugget zone of 6061

aluminium alloy joints friction stir welded with different surface roughness of the workpieces: (a)

R1 = 98 nm, (b) R2 = 60 nm, (c) R3= 51 nm and (d) R4= 37 nm.

Fig. 8: EDX analysis in the weld nugget zone of the 6061 aluminium alloy joints friction stir

welded at different surface roughness of the workpieces: (a) R1 = 98 nm, (b) R2 = 60 nm, (c) R3= 51

nm and (d) R4= 37 nm.

Fig. 9: XRD pattern of 6061 aluminium alloy joints friction stir welded at different surface

roughness of the workpieces (R1-R4).

Fig. 10: Effects of using different surface topographies (R1-R4) on the tensile strain-stress curves.

Fig. 11: Effects of using different surface roughness of the workpieces (R1- R4) on the

microhardness distribution across the welded joints of 6061 aluminium alloy.

Fig. 12: Relationship between Vickers microhardness and average d-1/2

of the weld nugget zone

using different surface roughness of the workpieces (R1-R4).

Fig. 13: FESEM micrographs showing the fracture surface features of the cross-sectional area after

the tensile test of specimen R1: (a) 500× and (b) 1000×.


27








28

Fig. 1: Typical 3D surface roughness maps of different surface roughness of the workpieces

measured using Atomic Force Microscopy (AFM); (a) R1 = 98 nm, (b) R2 =60 nm, (c) R3=51 nm

and (d) R4=37 nm.

1

0.0 nm

5 µm

4

3

2

1

5 µm

1 2 3 4

2

3

4

5 µm

413.6 nm

1

21.7 nm

5 µm

4

3

2

1

5 µm

1 2 3 4

2

3

4

5 µm

248.4 nm

244.8 nm

23.3 nm

5 µm

4

3

2

1

5 µm

1 2 3 4

1

2

3

4

5 µm

1

25.6 nm

5 µm

4

3

2

1

5 µm

1 2 3 4

2

3

4

5 µm

252.8 nm

(a) (b)

(c) (d)


29

Fig. 2: (a) Schematic diagram of the FSW process and (b) the welding tool dimensions.

Tool shoulder

Axial force

Tool pin

Retreating side

Advancing side

Welding direction

Rotational direction

2mm

3.7mm

9mm

(a) (b)


30

Fig. 3: Typical images of 6061 aluminium alloy surface appearances of friction stir welded with the

following surface roughness values; (a) R1 = 98 nm, (b) R2 = 60 nm, (c) R3= 51 nm and (d) R4= 37

nm.

R1 R2

R3 R4

(a) (b)

(c) (d)

Retreating Side

Retreating Side Retreating Side

Retreating Side

Advancing Side

Advancing Side Advancing Side

Advancing Side


31

Fig. 4: Variation of the workpiece surface roughness (R1-R4) and the friction coefficient with the

average friction temperature for joining pieces of 6061 aluminium alloy via friction stir welding.

0.34

0.36

0.38

0.4

0.42

190 210 230 250 270 290 310

Fri

ctio

n c

oef

fici

ent

(µ)

Average temperature (˚C )

R4

R3

R2

R1


32

Fig. 5: Variation of the workpiece surface roughness and the friction coefficient (µ) on the

frictional heat flux for joining pieces of 6061 aluminium alloy.

26

28

30

32

34

36

R1 R2 R3 R4

Hea

t fl

ux

(w

att/

mm

2)

Workpiece surface roughness (nm)


33

Fig. 6: Optical images of 6061 aluminium alloy joints friction stir welded with different surface

roughness of the workpieces: (a) R1 = 98 nm, (b) R2 = 60 nm, (c) R3= 51 nm and (d) R4= 37 nm.

(a)

(d)

(b)

(c)

BM

HAZ

TMAZ

Interface

TMAZ

HAZ

BM

Interface

BM HAZ TMAZ

Interface

TMAZ HAZ

BM

NZ

NZ NZ

NZ

100µm 100µm

100µm 100µm


34

Fig. 7: FESEM micrographs showing grain microstructures in the weld nugget zone of 6061

aluminium alloy joints friction stir welded with different surface roughness of the workpieces: (a)

R1 = 98 nm, (b) R2 = 60 nm, (c) R3= 51 nm and (d) R4= 37 nm.

(a) (b) (a) Recrystallized nano grains

(b) Recrystallized nano grains

6061 aluminum alloy

6061 aluminum alloy

6061 aluminum alloy

(c)

(a)

(c)

(b)

Spherical nano grains

6061 aluminium alloy

6061 aluminum alloy

6061 aluminium alloy



(d) 6061 aluminium alloy


35

Fig. 8: EDX analysis in the weld nugget zone of the 6061 aluminium alloy joints friction stir

welded at different surface roughness of the workpieces: (a) R1 = 98 nm, (b) R2 = 60 nm, (c) R3= 51

nm and (d) R4= 37 nm.

(a)

(b)

(c)

Element C O Mg Al Si Fe Cu

Atomic % 31.58 2.32 0.68 64.24 0.57 0.21 0.4

Element C O Mg Al Si Cu

Atomic % 24.52 2.12 0.69 71.26 0.7 0.71

Element C O Mg Al Si Fe Cu

Atomic % 17.09 1.67 0.75 78.09 0.72 0.53 1.14

Element C O Mg Al Si Cu

Atomic % 12.01 1.13 0.77 83.26 0.77 2.06

(d)


36

Fig. 9: XRD pattern of 6061 aluminium alloy joints friction stir welded at different surface

roughness of the workpieces (R1-R4).

0

20000

40000

60000

80000

30 40 50 60 70 80 90

Inte

nsi

ty (

arb.u

nit

)

2-Theta

R1

R2

R3

R4

●(2

20

)

●(3

11

)

●(2

22

)

●(2

00

)

●(1

11

) ●Aluminum


37

Fig. 10: Effects of using different surface topographies (R1-R4) on the tensile strain-stress curves.

0

50

100

150

200

250

0 0.5 1 1.5 2 2.5 3 3.5

Ten

sile

str

ess

(Mpa)

Tensile strain (%)

R1

R2

R3

R4

100

32

6 10

30 30

All dimensions in mm


38

Fig. 11: Effects of using different surface roughness of the workpieces (R1- R4) on the

microhardness distribution across the welded joints of 6061 aluminium alloy.

0

20

40

60

80

100

120

-18 -16 -14 -12 -10 -8 -6 -4.5 -1 0 1 4.5 6 8 10 12 14 16 18

Vic

ker

s m

icro

har

dnes

s (H

V)

Distance from center of weld (mm)

R4 R3 R2 R1

Shoulder width

Pin width

AS RS


39

Fig. 12: Relationship between Vickers microhardness and average d-1/2

of the weld nugget zone

using different surface roughness of the workpieces (R1-R4).

60

80

100

120

140

0 0.04 0.08 0.12 0.16

Vic

ker

s m

icro

har

dn

ess

(HV)

Average grain size, d-1/2 (nm)

R4

R3

R2

R1


40



(b)

6061 Al Alloy 6061 Al Alloy

Brittle fracture

Rip edges

RS AS

(a)


41



(a) (b)


Ductile fracture

Brittle fracture

AS RS

Brittle fracture


42



(a) (b)


Brittle fracture

AS RS

Ductile fracture

Ductile fracture


43



(a)

6061 Al Alloy

(b)

6061 Al Alloy

RS AS

Ductile fracture Ductile fracture


44

List of Tables:

Table 1 Chemical compositions of the base metal

Table 2 Chemical compositions of the welding tool

Table 3 Relevant parameters of 6061 aluminium alloy from the XRD analysis


45

Table 1 Chemical compositions of the base metal

Element Al Si Fe Mg Cu Mn

wt % 97.57 0.525 0.339 1.062 0.120 0.080


46

Table 2 Chemical compositions of the welding tool

Element C Mn P Fe

wt % 0.44 0.79 0.012 Balance


47

Table 3 Relevant parameters of 6061 aluminium alloy from the XRD analysis

Surface roughness of the workpiece (nm) R1 R2 R3 R4 BM

Crystallite Size/(nm) 201.85 188.47 78.36 73.78 64.87

Crystallite strain (%) 0.089 0.027 0.005 0.003 0.0022

The Influence of the Surface Roughness on the Microstructures and Mechanical Properties of 6061...

Documents

Transcript of The Influence of the Surface Roughness on the Microstructures and Mechanical Properties of 6061...