Thermal Stability of Friction Stir Processed Ultrafine Grained Alsingle BondMgsingle BondSc Alloy

8/10/2019 Thermal Stability of Friction Stir Processed Ultrafine Grained Alsingle BondMgsingle BondSc Alloy

1/10

Thermal stability of friction stir processed ultrafine grained

Al\Mg\Sc alloy

N. Kumar1, R.S. Mishra

Center for Friction Stir Processing and Department of Materials Science and Engineering, Missouri University of Science & Technology, Rolla,

MO 65409, USA

A R T I C L E D A T A A B S T R A C T

Article history:

Received 23 May 2012

Received in revised form

30 August 2012

Accepted 1 September 2012

Two different aspects of the thermal stability of grains were studied (i) thermal stability

of grain structure evolution during friction stir processing (FSP), and (ii) thermal stability of

grains after FSP. It was concluded that Al3(Sc,Zr) dispersoids are effective in limiting grain

growth to ultrafine grained regime (UFG) by stabilizing the microstructure during FSP. The

presence of these dispersoids before processing was more effective in refining the grain size

than the case in which these dispersoids precipitated during FSP. The mean grain sizes for

these cases were 0.40 0.17 m and 0.500.25 m, respectively. Isothermal grain growth

study revealed that UFG grain structure was maintained up to 723 K even after 16 h of

annealing whereas most of the UFG alloys reported in the literature show extensive grain

growth above 473 K. To understand the role of Al3(Sc,Zr) dispersoids in this alloy,

5086Al\H32 was friction stir processed and subjected to identical annealing conditions.

Annealing of FSP 5086Al\H32 at 623 K for 1 h resulted into abnormal grain growth (AGG).

The onset of AGG was observed for UFG sample only after annealing at 823 K for 1 h. This

was rationalized using Humphreys' model for AGG [F.J. Humphreys, Acta Mater., 45 (1997)

5031].

2012 Elsevier Inc. All rights reserved.

Keywords:

EBSD

Thermal stability

Friction stir processing

Grain refinement

Grain growth

1. Introduction

In the past two decades several grain refinement techniques

[15] have been developed to take advantage of HallPetch

strengthening[6,7]. Polycrystalline materials having grain size

below 1 m (ultrafine grained (UFG) and nanocrystalline) have

opened up entirely new domain of research due to their

difference in physical, chemical, and mechanical behaviors ascompared to coarse grained (CG) materials[8,9]. But, as grain

size decreases there is an increase in grain boundary area per

unit volume. Assuming a cube shaped grain it can be shown

that 50%reduction of grain leads to an increase in surface area

by ~16% and a reduction of 99% can cause surface area to

increase by ~3267%[10]. If grain boundary energy per unit

area of a material is assumed to be independent of its grain

size, there will be corresponding increase in total grain

boundary energy per unit volume. Hence, grain refinement

will always lead to an increase in the free energy of the system

thereby making the microstructure unstable. Such grains

when subjected to thermal cycle have tendency to grow to

minimize their energy by reducing the grain boundary areaper unit volume.

Thermal instability of the nanocrystalline[1115]and UFG

[1624]materials have been an issue and studied by many

researchers. Thermal stability of equal channel angular press-

ing (ECAP) processedUFG Al\3wt.%Mg was studied by Wang et

M A T E R I A L S C H A R A C T E R I Z A T I O N 7 4 ( 2 0 1 2 ) 1 1 0

Corresponding author at:Center for Friction Stir Processing and Department of Materials Science and Engineering, University of NorthTexas, Denton, TX 76203, USA. Tel.: +1 940 565 2316; fax: +1 940 565 4824.

E-mail addresses:[email protected](N. Kumar),[email protected](R.S. Mishra).1 Present address: Center for Friction Stir Processing and Department of Materials Science and Engineering, University of North Texas,

Denton, TX 76203, USA.

1044-5803/$ see front matter 2012 Elsevier Inc. All rights reserved.http://dx.doi.org/10.1016/j.matchar.2012.09.003

A v a i l a b l e o n l i n e a t w w w . s c i e n c e d i r e c t . c o m

w w w . e l s e v i e r . c o m / l o c a t e / m a t c h a r
http://-/?-http://-/?-http://dx.doi.org/10.1016/j.matchar.2012.09.003http://dx.doi.org/10.1016/j.matchar.2012.09.003http://dx.doi.org/10.1016/j.matchar.2012.09.003mailto:[email protected]:[email protected]://dx.doi.org/10.1016/j.matchar.2012.09.003http://dx.doi.org/10.1016/j.matchar.2012.09.003mailto:[email protected]:[email protected]://dx.doi.org/10.1016/j.matchar.2012.09.003http://-/?-http://-/?-http://-/?-http://-/?-


2/10

al.[25]. They studied the thermal stability of this alloy in the

range 443803 K. The UFG structure was found to be quite

unstable. The recrystallization and grain growth resulted in

elimination of UFG microstructure above 550 K. Stability of UFG

microstructure up to 700 K was reported by Furukawa et al.[26]

in a similar work on ECAP processed Al-5.5 Mg\2.2Li\0.12Zr

(wt%). Such a high resistance to grain growth was attributed to

very fine dispersion of Al3Zr dispersoids. In a similar work onECAP processed UFG Al\0.2 wt% Zr with grain size 0.7m, UFG

microstructure was retained only up to 573 K[16]. In another

study by Park et al. [18] on accumulative roll bonding (ARB)

processed UFG 6061Al, thermal stability was studied for 1 h in

the temperature range 373773 K. UFG microstructure was

maintained only up to 473 K. Roy et al. [21]have investigated

thethermal stability of a cryomilledUFG5083Al alloy. Similar to

Furukawa et al. [26] the stability of UFG microstructure was

reported up to 673 K. However, such a high level of stability of

grains was attributed to the dispersion of particles such as

oxides, carbides, andnitrides on thegrain boundaries. This brief

survey on the thermal stability of UFG alloys indicates that the

thermal stability is not very high and use of dispersoids pushes

the stability envelop towards the higher temperature range.

The rate of increase of grain boundary area per unit

volume is strongly influenced by the mode of deformation.

Gil Sevillano et al.[27]showed that rate of generation of grain

boundary area per unit volume was maximum for the sample

deformed under compression than those deformed either by

rolling, wire drawing or torsion for the same level of equivalent

strain. It indicates towards the fact that different processing

techniques introduce different types of microstructure. Hence,

UFG materials processed via different routes will have different

microstructures which would result into different response to

thermal treatment. Therefore, thermal stability of UFG alloys

processed via friction stir processing (FSP) is of interest in spite

of similar studies done on UFG materials processed by other

techniques[1624].

In the present worktwo different aspects of thermalstability

will be addressed. FSPhas potentialof refining grains toas small

as 25 nm [2830]. Butcommonly observed grains are larger than

1 m[31]. Zener pinning mechanism is known to inhibit grain

growth. This is realized by the use of precipitates or dispersoids

in the microstructure. Precipitates are generally not very

effective grain growth inhibitor due to either rapid dissolution

or excessive coarsening during FSP. In the present work the role

of dispersoids for grain stabilization to obtain UFG alloy will be

addressed. Another aspect is grain growth of UFG alloy after

processing.

2. Experimental Procedure

2.1. Processing of the Alloys

A twin-roll cast (TRC) Al-Mg-Sc sheet (nominal composition:

Al\4 Mg\0.8Sc\0.08Zr, wt%) of 3.75 mm thickness was

subjected to FSP in as-received (here afterwards referred to

as TRC-AR) and aged (here afterwards referred to as

TRC-Aged) conditions. The aging was carried out at 563 K for

22 h. It resulted into precipitation of Al3(Sc,Zr) dispersoids. To

evaluate the effectiveness of these dispersoids in inhibiting

grain growth during FSP, work hardenable 5086Al\H32 was

friction stir processed under same processing condition as

TRC alloy. A tool rotation rate of 325 revolutions per minute

(rpm) and a tool traverse speed of 8 ipm (203 mm/min) were

used to process these alloys. Tool tilt angle was 2.5 and

plunge depth was 0.097(2.5 mm). The tool material was tool

steel and tool had concave shoulder and stepped spiral pin

profile. The diameter of the shoulder was 12.0 mm. The pindiameter at root and the tip were 6.0 mm and 3.5 mm,

respectively. The height of the pin was 2.2 mm measured

from the root of the pin.

2.2. Annealing Heat Treatment for the Grain Growth Study

For thermal stability study, FSP TRC-Aged and 5086Al-H32

samples were subjected to isothermal annealing heat treat-

ment. FSP TRC-Aged alloy was annealed at 523, 623, 723, and

823 K. Time at each temperature varied from 116 h (at 823 K

only 1 h). FSP 5086Al-H32 was annealed at 523 K for 116 h

and at 623 K for 1 h.

2.3. Characterization of the Microstructure

Characterization of the microstructure was done by electron

backscattering diffraction (EBSD) microscopy using HKL EBSD

system fitted on FEI HeliosNanoLab 600 FIB/FESEM. Each sample

was mechanically polished using water based diamond suspen-

sion up to 1 m grit size and then polished using 0.02 m

colloidal silica suspension. EBSD was carried out in as-polished

condition. Centre of the FSP nugget on transverse cross-section

was chosen for EBSD analysis. A Tecnai G2 F20 S-Twin 200 keV

field-emission Scanning Transmission Electron Microscope

(S/TEM) was used for imaging nano-sized Al3(Sc,Zr) dispersoids.

TEM foil was prepared using twin-jet electropolishing technique.

The samples were mechanically polished down to 150m

thickness. A 2 mm disk was punched out and was further

polished down to 6080m. It was followed by twin-jet electro-

polishing at 30 C in CH3OH\20 vol.%HNO3electrolyte.

3. Results

3.1. Thermal Stability During FSP

The microstructure after FSP is shown in Fig. 1for two different

starting conditions. The accompanying inverse pole figure (IPF)

illustrates the various crystallographic axes associated with

different colors. The IPF has been plotted with respect to FSP

direction which is perpendicular to the plane of the paper.

Hence, these crystallographic directions basically represent

direction vector, in each grain, oriented parallel to FSP direction.

Fig. 1a corresponds to TRC alloy processed in AR condition at

325 rpm and 8 ipm (203 mm/min). EBSD micrographs corre-

sponding to TRC alloy processed in aged condition has been

shown inFig. 1b. It can be noticed that FSP of TRC-Aged alloy

resulted in a finer microstructure. Quantitativeanalysisof grain

size showed a mean grain size of 0.500.25 m and 0.40

0.16 m for FSP TRC-AR and FSP TRC-Aged, respectively. The

grain size distribution (GSD) histograms for these two condi-

tions have been shown inFig. 2. In both the cases GSD can be

2 M A T E R I A L S C H A R A C T E R I Z A T I O N 7 4 ( 2 0 1 2 ) 1 1 0


3/10

represented by theoretical Rayleigh or Louat's distribution[32]

and has been shown in the same figure. It is evident from the

GSD plot that some of the grains for FSP TRC-AR were larger

than1 m.But, almostall the grainswereless than1 mforFSP

TRC-Aged. The GSD histograms show that in both the cases

minimum grain size was same. It is an artifact of step size

selection during EBSD data acquisition. In both the cases

step-size was 0.10 m and minimum grain size was found to

be 0.12 m. A smaller step size may have some impact on the

distribution of grain size towards the smaller grain sizes. But it

is notexpectedto changethe average grain sizeand overallGSD

considerably.

TEM images for FSP TRC-Aged alloy have been shown in

Fig. 3. Fig. 3a shows Al3(Sc,Zr) particles distributed inside

matrix as well as at grain or subgrain boundaries. In Fig. 3b, a

high resolution TEM (HRTEM) image depicts a nano-sized

Al3(Sc,Zr) particle of approximately 56 nm radius.

FSP 5086-H32 alloys showed an average grain size of 1.1

0.81 m. The cumulative grain size distribution plot for this

alloy has been shown in Fig. 4 along with the cumulative

distribution of TRC Al-Mg-Sc alloy. From this figure, the effect

of dispersoid on the stability of grains during FSP can be

noted. FSP 5086Al-H32 alloy had ~ 40% grains larger than 1 m

whereas FSP TRC alloy had almost all the grains smaller than

1 m (for FSP TRC-AR a very small fraction of grains were

>1 m). Moreover, TRC-Aged alloy showed about 80% grains

below 0.50 m. It was about 60% for TRC-AR alloy.

3.2. Grain Growth Study

EBSD micrographs corresponding to annealing of FSP 5086Al\

H32 along with as-processed (AP) EBSD micrograph are shown

in Fig. 5. It can benoted thatannealing at523 K from 1 h (Fig.5b)

to 16 h (Fig. 5c) only resulted in recovery and annihilation of

substructure developed during FSP. InFig. 5a thin black lines

represent substructure developed in the grains during process-

ing.Dueto annealing at523 K for 1 h grain interiors appearto be

cleaner (Fig. 5b) than that of as-processed condition (Fig. 5a).

When sample was annealed at 623 K for 1 h, abnormal grain

growth (AGG) was observed. Very large grains (~100 m) can be

seen inFig. 5d along with very fine grains at the centre of the

image. The corresponding cumulative grain size distributions

for these four conditions are shown inFig. 6. In AP condition

~60% grains were smaller than 1 m. On annealing at 523 K for

1 h, larger grains grew at the expense of smaller grains as

evident from reduction of percentage of grains below 1 m

(38%). But fraction of grains below 1 m marginally changed

after annealing the sample at 523 K for 16 h. Cumulative

frequency curves related to these two conditions also indicate

that there was no significant change in the grain size

distribution on annealing for these two times. In the specimen

annealed at 623 K, the GSD distribution curve up to 34 m

follows GSD distribution curves related to samples annealed at

lower temperature very closely and afterwards become flat up

to 100 m. It is indicative of the presence of very large grains

due to AGG.

The EBSD images corresponding to annealing of FSP

TRC-Aged alloy are shown inFig. 7. It can be inferred that up

to 623 K for 16 h there was a marginal increase in the grain

size. The EBSD micrograph corresponding to 723 K shows

significant grain growth.Fig. 7d which represents microstruc-

ture annealed at 823 K for 1 h shows considerable grain

growth.

Fig. 8a shows the variation of average grain size as a

function of time and temperature. It can be noted that

Fig. 1 EBSD micrographs showing grain structure in FSP Al-Mg-Sc alloy, (a) AR+FSP and (b) Aged (563 K, 22 h)+FSP.

Fig. 2 Histogram showing the grain size distribution of FSP

TRC alloy in as-received (light colored bars) and aged (563 K,

22 h) conditions. Also shown are the theoretical Rayleigh or

Louat's distributions to represent the grain size distribution

of grain size cross-sectional area measurement.

3M A T E R I A L S C H A R A C T E R I Z A T I O N 7 4 ( 2 0 1 2 ) 1 1 0
http://localhost/var/www/apps/conversion/tmp/scratch_5/image%20of%20Fig.%E0%B2%80


4/10

annealing of samples at 523 K and 623 K up to 16 h did not

change the average grain size at all. Although, grain growth

was observed at 723 K when annealed between 1 to 16 h the

grain growth kinetics was not very fast. The sample annealed

at 823 K for 1 h showed a substantial grain growth and very

high grain growth kinetics as evident from the slope of the

curve. In Fig. 8b, thermal stability of FSP TRC-Aged alloy

annealed at different temperature for 1 h has been compared

with the thermal stability of UFG alloys processed by various

other routes. Comparison of the present data shows that most

of the UFG alloys have very poor thermal stability and above

~473 K UFG microstructure is lost. The FSP TRC-Aged alloy

maintained its UFG microstructure up to 723 K.

4. Discussion

As mentioned before FSP has the potential for producing

nanocrystalline grains. But, this advantage is lost due to

excessive grain growth of refined microstructure during FSP

because of high temperature involved during the process. The

thermal cycle a processed volume goes through during FSP is

shown schematically inFig. 9. The final grain size depends

largely on how much time a processed volume spends above

Tmin, minimum temperature to induce grain growth. One way

of cutting down time spent above Tminis to increase the rate

of cooling as shown by red (curve 1), blue (curve 2), and black

(curve 3) lines representing cooling rates in increasing order.

These different levels of cooling rates can be obtained by

using external cooling media such as liquid nitrogen, copper

backing plate, forced chilled air cooling, etc. Another method

of reducing the time above Tminwould be to reduce the peak

temperature (shown by curve 4). The peak temperature (Tpeak)

can be manipulated by controlling the tool rotation rate and/

or tool traverse speed. These two strategies have been widely

used to stabilize the processed microstructure during FSP.

Here, use of a third strategy has been shown to refine the

microstructure during FSP as outlined next.

In the present study, two different scenarios were inves-

tigated (a) precipitation of dispersoids during FSP, i.e., FSP

of as-received Al\Mg\Sc which contained Sc in solid solution

and (b) FSP in aged condition. The effectiveness of case (a) in

grain refinement was lower than that of case (b). Since, in case

(a) dispersoids have to precipitate out during FSP, their

effectiveness in acting as Zener pinning agent will depend

on the kinetics of stable precipitate nuclei formation and also

the volume fraction of the dispersoids. For coherent stable

particles or dispersoids containing alloys, limiting grain size is

given as

D 2r

3FV1

where D, r and FVare average grain size in FSP condition,

radius and volume fraction of precipitates or dispersoids,

respectively[10]. Ignoring the amount of Zr, the present alloy

system can be treated as a binary system (Al,Mg) \Sc.

Al\Sc literature shows that depending upon the rate of

cooling solid solubility of Sc in Al can vary anywhere from

0.35 wt.% (equilibrium cooling) to 0.6 wt.% (at a cooling rate of

100 K/s) [33]. In the twin-roll casting a cooling rate of 100 K/s is

easily obtained. Hence, 0.6 wt.% of the Sc can be easily

Fig. 3Transmission electron microscope (TEM) images of Al3(Sc,Zr) dispersoid in FSP TRC-Aged (a) a low magnification image

showing presence of Al3(Sc,Zr) dispersoids (b) a high resolution TEM image showing Al3(Sc,Zr) dispersoid.

Fig. 4 Comparison of grain size distribution of FSP TRC-AR

and TRC-Aged alloys with FSP 5086Al-H32 alloys.

http://localhost/var/www/apps/conversion/tmp/scratch_5/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_5/image%20of%20Fig.%E0%B3%80


5/10

assumed to be present in Al\Mg\Sc alloy. If all the Sc in solid

solution precipitates out in the form of Al3(Sc,Zr) dispersoids,

applying the lever rule it can be shown that it will correspond

to 0.0168 volume fraction of these dispersoids. Based on the

present TEM work (and also supported by the work of Lee et al.

[34]and Marquis et al.[35]on Al\Mg\Sc system), the average

size of Al3(Sc,Zr) dispersoids are expected to be ~5 nm in size.

Substituting these values (FV=0.0168 and r =5 nm) in Eq. (1),

limiting grain size D is ~200 nm. The experimentally observed

mean grain size in the case (b) was also 400 nm. The

discrepancy between the predicted and experimentally mea-

sured values is quite expected. Eq.(1)has been derived for the

case where Zener pinning force is maximum. We know, in

reality, not all the particles will exert maximum pinning force

on grain boundaries. Hence, the predicted values would be

smaller than the experimentally measured one, as is the case

here. The measurable coarsening of Al3(Sc,Zr) dispersoids

takes place only above 573 K, if heated for considerable

amount of time. In the case (a), a material volume in the

wake of the FSP tool is not expected to be above 573 K for a

considerable amount of time. Hence, the dispersoid size

should be about 23 nm. The works of Fuller et al. [36],

Marquis and Seidman[37], and Knipling et al.[38]support this

assumption. The mean grain size found in the case (a) was

0.5 m. Substituting these values (limiting grain size equal to

mean grain size and dispersoid size equal to 2 nm) in Eq. (1),

the volume fraction of the dispersoids was found to be 0.0026.

The estimated volume fraction in this case is lower than that

of case (b). This is an expected result. Given the very short

duration of FSP, complete precipitation of the dispersoids is

unlikely. So, the volume fraction of Al3(Sc,Zr) dispersoids

would be lower. Hence, in spite of having smaller dispersoid

size in the case (a), lower volume fraction of the dispersoids

resulted into a higher limiting grain size. Other noticeable

Fig. 5 EBSD micrographsshow thegraingrowth behaviorof FSP 5086Al-H32 alloy in different conditions;(a) as-processed, (b) 523 K,

1 h, (c) 523 K, 16 h, and (d) 623 K, 1 h.

Fig. 6 Cumulative grain size distribution of FSP

5086Al-H32 alloy in various thermo-mechanical conditions

(AP: as-processed).



6/10

difference due to the presence of Al3(Sc,Zr) dispersoids before

FSP was higher frequency of smaller grain sizes.

Grain growth study revealed that presence of stable

dispersoids like Al3(Sc,Zr) can stabilize very fine grain size at

temperatures as high as 723 K. Such a high level of thermal

stability of grains is hardly observed in any of the previously

reported UFG materials except for a report by Ma et al.[39]for

Al\4Mg\1Zr alloy. The highly thermal resistant Al3Zr disper-

soids were attributed for such a high thermal stability of this

alloy system. The UFG microstructure was maintained up to

698 K (mean grain size 0.98 m at this temperature) in this

case. Most of the UFG Al alloys lose their UFG structure above

473 K523 K.

The case of AGG for particle containing alloys have been

dealt with by many researchers, but only the model proposed

by Humphreys'[40]will be discussed here. The condition for

AGG is given as

RdR

dtR

dR

dt> 0 2

Fig. 7EBSD micrographs show the grain growth behavior of FSP TRC-aged alloy in different conditions; (a) 523 K, 16 h, (b) 623 K,

16 h, (c) 723 K, 16 h, and (d) 823 K, 1 h.

Fig. 8 (a) Grain size as a function of time and temperature for FSP TRC-aged alloy. (b) Thermal stability comparison plot for FSP

UFG TRC-aged alloy with literature data[1618,20,21].



7/10

where R and R are mean grain size and size of any grain,

respectively, in a polycrystalline material.

dRdt

M R

RZ

R

3

dR

dt

M

R 0:25Z 4

where for incoherent particles

Z incoh 3FVR=2r 5

and for coherent particles

Zcoh 3FVR=r 6

Here, , ,M, MandZ are average grain boundary energy,

grain boundary energy, grain boundary mobility, average

grain boundary mobility, and dimensionless particle pinning

parameter, respectively. For the present case substituting

Eqs.(3) and (4)into expression (2) and assuming= 1 and

M=M 1, expression (2) can be rewritten as

4Z1 X2 4 1Z 4> 0 7

forZ 0, i.e.,

RdR

dt RM

R

RZ

R

> 0 8

Again under the assumption , the Eq.(8)leads to thefollowing inequality,

X> 1= 1Z 9

where X R=R.

The variation of X with Z and mean grain size (R) withFv/r

has been shown in Fig. 10 for coherent and incoherent

particles. It depicts the regions of AGG and no grain growth.

When Z reaches unity, the requirement for minimum grain

size to initiate AGG becomes so large (as Z0, R from

Eq. (8)) that at the dispersion level higher than this grain

growth seizes completely. In Fig. 10a, it is represented by

curve Xminand inFig. 10b curve representing Z=1 (coh) and

Z=1 (incoh). InFig. 10b words coh represents coherent and

incoh incoherent particles.Fig. 10a has been plotted using

Eqs.(7) and (9)andFig. 10b using Eqs.(5) and (6). As per Jones

and Humphreys, coherency is lost at ~20 nm. Hence, in the

present analysis in plottingFig. 10a and b both aspects have

been considered. ParameterZ is a function of temperature. For

the present case it was calculated for all the annealing

temperatures. To calculate Z, volume fraction of dispersoids,

mean grain size (radius), and dispersoid radius were needed.

The volume fraction calculation was done at each tempera-

ture based on the solvus curve related to Al-Sc reported by

Ryset and Ryum [33]. The procedure of volume fraction

calculation at each temperature was same as outlined in the

discussion on thermal stability during FSP. The mean grain

size radius was calculated from the EBSD analysis. Since in

present work TEM analysis provides dispersoid size informa-

tion for FSP TRC-Aged in as-processed condition only, the

dispersoid radius was taken from the work of Jones and

Humphreys[41]on Al3Sc dispersoids coarsening as a function

of temperature. The calculated Z values are listed inTable 1.

The calculated Z values were superimposed on the grain

stability map shown in Fig. 10a. From the values ofZ, it is clear

that it decreases with increase in temperature. A similar

calculation by Charit and Mishra[42]for precipitates/disper-

soids in FSP UFG Al-Zn-Mg-Sc alloy showed a decrease in Z

with the increase in temperature.

TheZ values at 523 K and 623 K fall into no growth region

of the grain stability map. The grain size histograms corre-

sponding to annealing at 523 K and 623 K for 1 h are shown in

Fig. 10c. Clearly, both the histograms superimpose each other

completely. Also, mean grain size in these two conditions

were same as in as processed condition. The Z value for the

sample annealed at 723 K for 1 h also lies in the no growth

region but this time it is close to AGG region. Ideally, there

should not be any grain growth in this sample too since it is a

part of no growth region. The mean grain size in this case was

0.74 m. It can be attributed to lowering of Zener pinning

pressure due to coarsening of Al3(Sc,Zr) dispersoids. The grain

size distribution histogram is shown in Fig. 10d. An increasein

relative frequency of grains in the range 1.5 to 2.0 m can be

noted. But still grain size distribution is same as those

annealed at lower temperatures. If the sample is annealed at

723 K for longer period of time, it will cause further coarsening

ofAl3(Sc,Zr) dispersoids. That may result in reduction ofZvaluefurther (and reduction ofFv/r too) and consequential shift in

Fig. 9A schematic illustration of the thermal cycle experienced by a particular volume of material during FSP and its effect on

grain refinement process.

http://localhost/var/www/apps/conversion/tmp/scratch_5/image%20of%20Fig.%E0%B9%80


8/10

the position of datum point towards the AGG region and may

eventually become part of AGG region as shown by a blue

colored open circle inFig. 10a (and inFig. 10b also). In such a

scenario the sample should show AGG. The analysis of grain

size distribution histogram of the sample annealed at 723 K

for 16 h shows a bimodal grain size distribution as shown in

Fig. 10e. Apart from a peak in UFG regime, the presence of

another peak at~1 m can also be noted. The bimodal grain

size can be considered as an onset of AGG. The sample

annealed at 823 K for 1 h showed AGG. It is evident not only

from the EBSD micrographs shown inFig. 7d but also from the

grain size distribution histogram shown inFig. 10f. Apart from

the presence of multiple peaks, some grains larger than 10 m

can also be noted in Fig. 10f. It should be noted that the Z

Fig. 10 Grain stability maps (a) X Vs. Z and (b) mean grain size Vs. Fv/r; (cf) comparison of the predictions made by this map

with the experimentally observed grain growth behavior.

Table 1Values of different variables used in the calculation of particle pinning parameter, Z.

523 K 623 K 723 K 823 K

Dispersoid size (radius, nm) 5 5 9 47

Solubility of Sc in Al (wt%) Negligible Negligible 0.034 0.125

Dispersoid volume fraction 0.0168 0.0168 0.0159 0.0133

Mean grain size (radius, m) 0.26 0.26 0.26 0.26

Particle pinning parameter, Z 2.6 2.6 1.4 0.11

Fv/r, m1

3.36 3.36 1.8 0.28

http://localhost/var/www/apps/conversion/tmp/scratch_5/image%20of%20Fig.%E0%B1%B0


9/10

value corresponding to the sample annealed at 823 K falls in

AGG regime as shown in Fig. 10a. Hence, it satisfies the

condition for AGG which is evident from Fig. 10e. Similar

conclusions can be drawn based onFig. 10b too.

5. Conclusions

The following conclusions can be made based on foregoing

discussion:

I. Dispersoids like Al3(Sc,Zr) are effective means of

inhibiting the grain growth during FSP.

II. The presence of the dispersoids before FSP is much

more effective as Zener pinning agent than their pre-

cipitation during FSP.

III. Post-FSP annealing heat treatment showed a very high

thermal stability of UFG microstructure. UFG micro-

structure was maintained up to 723 K.

IV. Humphreys model for abnormal grain growth can be

utilized to explain the thermal stability of the present

alloy system.

Acknowledgment

Authors would like to acknowledge the financial assistance

from The Boeing Company, St. Louis and Intelligent System

Centre, Missouri S&T, Rolla to carry out this work. Authors are

also grateful to Dr. Saumya Nag at University of North Texas

for carrying out high resolution transmission electron micros-

copy for the present work.

R E F E R E N C E S

[1] Valiev RZ, Islamgaliev RK, Alexandrov IV. Bulknanostructured materials from severe plastic deformation.Prog Mater Sci 2000;45:103-89.

[2] Valiev RZ, Langdon TG. Principles of equal-channel angularpressing as a processing tool for grain refinement. Prog MaterSci 2006;51:881-981.

[3] Zhilyaev AP, Langdon TG. Using high-pressure torsion formetal processing: fundamentals and applications. Prog MaterSci 2008;53:893-979.

[4] Saito Y, Utsunomiya H, Tsuji N, Sakai T. Novel ultra-high

straining process for bulk materials development of theaccumulative roll-bonding (ARB) process. Acta Mater 1999;47:579-83.

[5] Tsuji N, Saito Y, Lee S-H, Minamino Y. ARB (accumulativeroll-bonding) and other new techniques to produce bulkultrafine grained materials. Adv Eng Mater 2003;5:338-44.

[6] Hall EO. The deformation and ageing of mild steel: IIIDiscussion of results. Proc Phys Soc B 1952;64:747-53.

[7] Petch NJ. The cleavage strength of polycrystals. J Iron SteelRes Inst 1953;174:25-8.

[8] Gleiter H. Nanocrystalline materials. Prog Mater Sci 1989;33:223-315.

[9] Gleiter H. Materials with ultrafine microstructures: retro-spectives and perspectives. Nanostruct Mater 1992;1:119.

[10] Humphreys FJ, Hatherly M. Recrystallization and relatedannealing phenomena. Oxford: Pergamon; 1995.

[11] Malow TR, Koch CC. Grain growth in nanocrystalline ironprepared by mechanical attrition. Acta Mater 1997;45:2177-86.

[12] Lu L, Wang LB, Ding BZ, Lu K. Comparison of the thermalstability between electro-deposited and cold-rollednanocrystalline copper samples. Mater Sci Eng A 2000;286:125-9.

[13] Andrievski RA. Stability of nanostructured materials. J MaterSci 2003;38:1367-75.

[14] Liu F, Yang G, Wang H, Chen Z, Zhou Y. Nano-scale graingrowth kinetics. Thermochim Acta 2006;443:212-6.

[15] Chauhan M, Mohamed FA. Investigation of low temperaturethermal stability in bulk nanocrystalline Ni. Mater Sci Eng A2006;427:715.

[16] Hasegawa H, Komura S, Utsunomiya A, Horita Z, FurukawaM, Nemoto M, et al. Thermal stability of ultrafine-grainedaluminum in the presence of Mg and Zr additions. Mater SciEng A 1999;265:188-96.

[17] Park KT, Kim YS, Lee JG, Shin DH. Thermal stability andmechanical properties of ultrafine grained low carbon steel.Mater Sci Eng A 2000;293:165-72.

[18] Park KT, Kwon HJ, Kim WJ, Kim YS. Microstructural andthermal stability of ultrafine grained 6061 Al alloy fabricatedby accumulative roll bonding process. Mater Sci Eng A

2001;316:145-52.[19] Zhilyaev A, Nurislamova G, Valiev R, Baro M, Langdon

T. Thermal stability and microstructural evolution inultrafine-grained nickel after equal-channel angular pressing(ECAP). Metall Mater Trans A 2002;33:1865-8.

[20] Geng H, Kang S, He S. Microstructural evolution and thermalstability of ultra-fine grained Al-4Mg alloy by equal channelangular pressing. J Mater Sci Technol 2004;20:315-8.

[21] Roy I, Chauhan M, Lavernia EJ, Mohamed FA. Thermalstability of bulk cryomilled ultrafine-grained 5083 Al alloy.Metall Mater Trans A 2006;37:721-30.

[22] Homola P, Slmov M, OenekV, UhlJ, Cieslar M. Annealingresponseof Al\0.22Sc\0.13Zr alloy processedby accumulativeroll bonding. Mater Sci Forum 2008;584586:899-904.

[23] Lugo N, Llorca N, Suol JJ, Cabrera JM. Thermal stability of

ultrafine grains size of pure copper obtained byequal-channel angular pressing. J Mater Sci 2010;45:2264-73.

[24] Nikitina MA, Islamgaliev RK, Kamalov AF. Thermal stabilityof the ultrafine-grained Al\Cu\Mg\Si aluminum alloy. RevAdv Mater Sci 2010;25:74-81.

[25] Wang J, Iwahashi Y, Horita Z, Furukawa M, Nemoto M, ValievRZ, et al. An investigation of microstructural stability in anAl\Mg alloy with submicrometer grain size. Acta Mater1996;44:2973-82.

[26] Furukawa M, Iwahashi Y, Horita Z, Nemoto M, Tsenev NK,Valiev RZ, et al. Structural evolution and the HallPetchrelationship in an Al\Mg\Li\Zr alloy with ultra-fine grainsize. Acta Mater 1997;45:4751-7.

[27] Gil Sevillano J, van Houtte P, Aernoudt E. Large strain workhardening and textures. Prog Mater Sci 1980;25:135-271.

[28] Su J-Q, Nelson TW, Sterling CJ. A new route to bulknanocrystalline materials. J Mater Res 2003;18:1757-60.

[29] Rhodes CG, Mahoney MW, Bingel WH, Calabrese M.Fine-grain evolution in friction-stir processed 7050 aluminum.Scr Mater 2003;48:1451-5.

[30] Mishra RS. Bulk nanomaterials from friction stir processing:Features and properties. In: Zehetbauer ZMJ, Zhu YT, editors.Bulk Nanostructured Materials. KGaA Weinheim: Wiley-VCHVerlag GmbH & Co; 2009. p. 255-72.

[31] Mishra RS, Ma ZY. Friction stir welding and processing. MaterSci Eng R 2005;50:178.

[32] Anderson MP, Grest GS, Srolovitz DJ. Computer simulation ofnormal grain growth in three dimensions. Philos Mag B1989;59:293-329.

[33] Royset NRJ. Scandium in aluminum alloy. Int Mater Rev2005;50:19-43.



10/10

[34] Lee S, Utsunomiya A, Akamatsu H, Neishi K, Furukawa M,Horita Z, et al. Influence of scandium and zirconium on grainstability and superplastic ductilities in ultrafine-grainedAl-Mg alloys. Acta Mater 2002;50:553-64.

[35] Marquis EA, Seidman DN, Dunand DC. Effect of Mg additionon the creep and yield behavior of an Al-Sc alloy. Acta Mater2003;51:4751-60.

[36] Fuller CB, Seidman DN, Dunand DC. Mechanical properties of

Al(Sc, Zr) alloys at ambient and elevated temperatures. ActaMater 2003;51:4803-14.

[37] Marquis EA, Seidman DN. Nanoscale structural evolution ofAl3Sc precipitates in Al(Sc) alloys. Acta Mater 2001;49:1909-19.

[38] Knipling KE, Karnesky RA, Lee CP, Dunand DC, Seidman DN.Precipitation evolution in Al-0.1Sc, Al-0.1Zr andAl-0.1Sc-0.1Zr (at%) alloys during isochronal aging. ActaMater 2010;58:5184-95.

[39] Ma ZY, Liu FC, Mishra RS. Superplastic deformationmechanism of an ultrafine-grained alloy produced by frictionstir processing. Acta Mater 2010;58:4693-704.

[40] Humphreys FJ. A unified theory of recovery, recrystallizationand grain growth, based on the stability and growth ofcellular microstructures - II. The effect of second-phaseparticles. Acta Mater 1997;45:5031-9.

[41] Jones MJ, Humphreys FJ. Interaction of recrystallization and

precipitation: The effect of Al3Sc on the recrystallizationbehavior of deformed aluminum. Acta Mater 2003;51:2149-59.

[42] Charit I, Mishra RS. Low temperature superplasticity in afriction-stir-processed ultrafine grained Al-Zn-Mg-Sc alloy.Acta Mater 2005;53:4211-23.


Thermal Stability of Friction Stir Processed Ultrafine Grained Alsingle BondMgsingle BondSc Alloy

Documents

Transcript of Thermal Stability of Friction Stir Processed Ultrafine Grained Alsingle BondMgsingle BondSc Alloy