MTN 300- Case Study

REPORT

Precipitation kinetics during friction stir

welding in 7xxx series alloys

By:
Akanksha Kushwaha (21118011)
Payal Odedara (21118058)

Under the guidance of

Prof. Aparna Tripathi, Professor MMED, IIT
Roorkee
TOPIC: Precipitation kinetics in friction stir
welded joints in 7XXX series alloys

Abstract
7XXX series alloys are widely used aviation industry due to their high
specific stiffness, high specific strength, high toughness and welding
performance. The cause for these excellent mechanical properties is
precipitation strengthening, which occurs during friction stir welding. In
this report, we have discussed how precipitates form, grow and coarsen
during the process, then microstructure development in friction stir weld,
and then what is its effect on precipitation evolution. Finally, we have
done a literature review especially to understand the kinetics of
precipitation during the welding in 7XXX alloys and different parameters
that can affect it.

Introduction
Friction stir welding is mainly used for those metals which are hard to
weld, for example, aluminium alloys which are precipitation hardenable.
It is a solid-state welding process which makes it suitable for those alloys
and metals which are hard to melt. It uses no filler material and gases
during the process. It is a very energy-efficient and environment-friendly
process. It is better than conventional processes as it introduces low
distortion and residual stresses to weld. There are no fumes, no arc
flash and no spatters.

Fig 1. The schematic of friction stir welding

FSW introduces deformation and heat to the metal during the process
which leads to microstructural changes in the weld which decided its
properties. To study these changes, we need to monitor the changes in
microstructure during the process itself.

In case of Aluminium alloys, mainly three types of microstructural zones

are formed
 1. Stir Zone (SZ)- It is the centre of the joint. It experiences
extensive plastic deformation, resulting in the formation of
dynamically recrystallized grains. This process leads to the
creation of grain sizes dependent on temperature. Additionally,
exposure to elevated temperatures occurs, causing the complete
dissolution of
strengthening precipitates in age-hardenable alloys.
2. Thermo-Mechanical Affected Zone (TMAZ)– this zone surrounds
the stir zone and is the combination of partly recrystallized and
partly recovered microstructure. This happens because this zone
is exposed to low temperatures and deformation in comparison to
the stir zone. The precipitates in this zone can be partly or fully
dissolved. The surviving precipitates in this zone may transform
and coarsen which depend on the time of exposure and
temperature.
3. Heat Affected Zone (HAZ)- This zone is only affected by the
temperature and precipitates here are coarsened and can
transform into equilibrium phases from metastable phase.

Fig 2.

A- Parent Metal
B- Heat Affected Zone (HAZ)
C- Thermo-Mechanical Affected Zone (TMAZ)
D- Stir Zone (SZ)
Now Due to formation of these complex microstructure regions,
Properties of weld is highly affected. In case of Aluminium alloys, they
can recover their strength at room temperature after being exposed to
high temperature. This recovery of strength happens maximum where the
amount of solute is maximum in matrix which is the case with stir zone.
Hence after months, strength at stir zone is very high compared to
TMAZ/HAZ due to natural aging.
Now to study these complex phenomena, There are mainly two
challenges. First that it is impossible to use conventional techniques to
follow up the transient microstructural changes during the FSW and
secondly microscopy should be performed at the scale which is needed
to study and observe precipitates on a small number of very local regions
of microstructure which can be done by Transmission Electron
Microscope.
The research conducted in the cited literature explores both of these
concerns and employs the FlexiStir system to investigate the
transformation of microstructures throughout the welding procedure.
FlexiStir is a precisely controllable welding system that can be installed
in a synchrotron X-ray beamline, enabling rapid acquisition of
microstructural data through small-angle X-ray scattering (SAXS).
In the literature FlexiStir system was applied to measure precipitate
formation and evolution for AA4779-TAF during friction stir welding.
(TAF – temper for age forming) material is underaged in this condition.

Experimental Setup:

1. FSW System (FlexiStir):

• Used a special machine called the FlexiStir system, designed for
in-situ synchrotron X-ray scattering experiments during welding.
• This system has six controlled actuators and axes for precise
movement during welding.
• The tool used in the FSW process is fixed in place with respect to
the X-ray beam, while the workpiece (the aluminum alloy) is moved
during welding.
Synchrotron X-ray Scattering:
• X-ray scattering is a technique used to study the structure of
materials at the atomic and molecular level.
• High-energy X-rays (70 keV photons) were used because they can
penetrate the thickness of the aluminum alloy and a copper
backing sheet.
• The X-ray beam passed next to the tool shoulder during welding.
Fig 3. Experimental set up of FlexiStir system (a) Side view, showing
configuration of
tool, plate, and backing bar. (b) Isometric view showing region analysed to
study dynamic
precipitate evolution behind the tool and the lateral scan to study
precipitate dissolution in the heat affected zone (HAZ).

Literature review
Discussion on the microstructure evolution during friction stir welding
(FSW) of an aerospace aluminium alloy using in-situ measurement
combined with simulation was done. The study
focuses on understanding the changes in precipitates, which are crucial
for determining the weld properties in aerospace aluminium alloys. The
study introduces the FlexiStir system, a portable FSW unit designed for
use in a high-energy synchrotron beamline, enabling real- time studies
of microstructural changes during welding. The experiments were
performed using the FlexiStir system, and the measurements were
interpreted with the aid of a
microstructure model, showing good qualitative agreement between the
model predictions and SAXS measurements. The study also discusses
the complex microstructural changes and property variations across the
weld, as well as the implications for the aerospace industry.
The study reveals that the FlexiStir system, combined with SAXS
measurements and microstructural modelling, provides valuable insights
into the dynamic precipitate evolution during FSW of aerospace
aluminium alloys. The results demonstrate the complex
microstructural changes that occur during welding, such as partial
reprecipitation and dissolution of precipitates in different zones of the
weld, and the retention of supersaturated solute for post-weld natural
aging. The findings also highlight the sensitivity of the
microstructural changes to factors such as welding speed and distance
from the weld centreline. The study emphasizes the importance of
understanding microstructural evolution during welding, especially in
aerospace applications, where the properties of the weld are critical for
performance.

Model
The fundamental model used in the above approach has two important
parts.
1. A process model which can describe the welding process
2. Second microstructure model that can predict the effect of
the thermal and deformation cycle on grain evaluation and
precipitate.
The initial model relies on a Computational Fluid Dynamics framework,
where the fundamental approach involves representing deforming solid
material as a high-viscosity fluid. The characteristics of this fluid are
influenced by both strain rate and temperature. The model exhibits
strong predictive capabilities for determining the temperature zone
based on process parameters, provided that the selection of thermal
coefficients and material properties is accurate.
Fig 4. Thermal predictions of process model compared with thermocouple
measurements
 Second model takes temperature, strain, strain rate as input from the
process model which are its output parameters. For AA7449, this model
especially tracks three precipitate populations which are grain boundary
η, grain interior equilibrium η and grain interior metastable η’.
The current framework relies on classical kinetic theory, serving as an
expansion of the numerical approach proposed by Kampmann and
Wagner (KWN). In the Kampmann and Wagner methodology, distribution
of particle size for individual precipitates is discretized into numerous size
classes. The progression of these size classes over time is segmented
into small intervals. During each time interval, the application of classical
nucleation theory computes new particles count for every type.
Subsequently, these new particles are incorporated into size distribution
at a dimension a little exceeding than critical radius.

Microstructural Model
A novel numerical approach, drawing from the Kampmann and Wagner
methodology, has been formulated to forecast the progression of
precipitate dispersion during friction stir welding (FSW). Specifically
tailored for FSW of AA7449, a prominent aerospace aluminum alloy, this
model encompasses both transient and stable precipitate phases,
accounting for their interconversion. Furthermore, it anticipates the
development of precipitates along grain boundaries and their impact on
the size of precipitate-free zones. Rigorous calibration against outcomes
from isothermal and quench experiments has validated the model's
accuracy. Upon application to weld scenarios, the model demonstrates
commendable qualitative consistency with empirical observations.
Additionally, it has facilitated the exploration of diverse FSW parameters
and their influence on projected precipitate dynamics.

Homogeneous Nucleation in Matrix

Where,
s - incubation time for nucleation,
t – time,
c - interfacial energy of the matrix/nucleus interface,
N0 - number of nucleation sites per unit volume,
T – temperature,
Z - Zeldovich non-equilibrium factor,
b* - rate of atomic attachment to a growing embryo,
k - Boltzmann constant,
Rc – critical radius for nucleus formation,
which depends on the volume free energy change for nucleation delta Gv
as follows:

considering a pseudo-binary alloy system

Limiting condition:

One approach to understanding heterogeneous nucleation involves

modifying the principles of homogeneous nucleation theory to
accommodate the presence of a finite number of heterogeneous
nucleation sites and the lower energy barrier required for nucleus
formation at these locations.

Transition from Metastable to equilibrium phase

• It is a continuous process.
• It is dependent on the critical radius reached by the metastable
phase, which is temperature-dependent.
• As temperature increases, the critical radius decreases

Precipitation growth

1)Within the grains

Assuming the spherical morphology of the particles and considering one

diffusing element E then the growth rate:

Where,
DE - diffusion rate of E in the matrix,
R - particle radius,
cmE - concentration of E in the matrix,
Interfacial Composition

2. At the Grain Boundaries

Carolan and Faulkner formulated the growth rate of grain boundary

precipitates by establishing a correlation between the solute input to the
boundary and the necessary flux for the precipitate to reach a specific
dimension.

Following a methodology similar to the one described earlier, we

investigate the equilibrium concentrations at the interface between the
curved particle and matrix. Then, by choosing interfacial compositions
that maintain a flux balance similar to precipitates within the grain interior,
we determine the growth rate of the particles.

Results and Discussion

SAXS Data Analysis:
Dataset 1: Dynamic Evolution of Precipitates Trailing the Tool

The data collection was conducted under the conditions of a reference

welding speed of 5 mm/s.
Findings:
Analysis of the maps indicates a progressive increase in the volume
fraction of precipitates trailing behind the tool during the cooling
process of the weld.
There is a noticeable rise in the volume fraction of reprecipitated
material as one moves from the central line towards the periphery of
the shoulder.
While the average particle radius diminishes from the center of the
weld, it remains relatively stable as one moves further away from the
tool.
Analysis:
These findings validate the occurrence of reprecipitation behind the
tool, demonstrating significant variation with positional changes.
The equilibrium between reprecipitated and retained solutes will have
implications for potential natural aging post-welding.
Challenges:
Constraints include the limited range of particle sizes captured by the
Small Angle X-ray Scattering (SAXS) setup (0–20 nm).
Another challenge lies in the inability to directly sample beneath the
tool, where microstructural transformations occur rapidly.
Integration with Modeling:

Fig 5. Maps showing the measured evolution of

(a) precipitate volume fraction and
(b) precipitate radius for the region behind the tool (welding speed
5 mm s−1).

Dataset 2: Lateral Scan in the HAZ

Purpose:
These measurements give insight into the microstructural state of the
HAZ near its maximum temperature as the tool center passes.
They provide unique data on the volume fraction of precipitates
surviving from the initial condition.
Experimental Setup:
Lateral scans were performed perpendicular to the tool direction.
(welding speed 2 to 8 mm/s)
Observations:
The lowest volume fraction is closest to the shoulder for all speeds,
ranging from 15% to 35% of the initial precipitate volume fraction
(which is 4.7%).
As distance from the shoulder increases, differences in welding speed
become more apparent.
At the fastest speed (8 mm/s), the retained volume fraction increases
strongly with distance.
For the slowest speed, the volume fraction remains relatively constant
at 15–35% of the initial value across the distance studied.
Interpretation:
The observations align with the effect of welding speed on the
thermal field.
Slower welding speeds allow more time for heat to travel, leading to
greater precipitate dissolution further from the centerline.
Faster speeds result in a rapid fall-off of peak temperature, reducing
the amount of precipitate dissolution.

Fig 6. shows the measured volume fraction as a function of lateral

distance for the welding speeds studied. It can be seen that for all
speeds, the lowest volume fraction is closest to the shoulder, and in all
cases is between 15 and 35% of the initial precipitate volume fraction
(which is 4.7%). As the distance from the shoulder increases, the
difference in welding speed becomes more apparent. At the fastest
speed (8 mm s−1) the retained volume fraction increases strongly with
distance, whereas for the slowest welding speed, the volume fraction
remains at 15–35% of the initial value across the full distance studied
Scanning Transmission Electron Microscope

Fig 7. Dark field scanning transmission electron microscopy (STEM)

images of the microstructure showing the precipitates present in
(a) base material (b) stir zone (c) thermomechanically affected zone
(welding speed 5mm/s).

In this investigation, low-magnification images capturing the overall

microstructure of friction stir welds in a specific alloy exhibited features in
line with previous observations. However, the necessity for meticulous
microscopy arose to authenticate the findings derived from small-angle X-
ray scattering (SAXS) measurements. Specialized high-resolution images
were captured to precisely scrutinize the distribution of precipitate size in
different weld zones, especially the base material, stir zone, and thermo-
mechanically affected zone.

The utilization of HAADF scanning transmission electron microscopy

facilitated the identification of unique precipitate distributions within
each weld zone. The base material showcased numerous uniformly
distributed small precipitates, primarily recognized as n in the
underaged temper state. Grain boundaries were embellished with
slightly larger particles, expected to be n-phase.
The TMAZ displayed larger plate-shaped precipitates with a
diminished number density, projected to be n-phase. The metastable
n in the base material was anticipated to either dissolve or undergo
transformation into h, contingent on particle size.
The stir zone exhibited a more complex bimodal distribution of particles,
comprising large particles (approximately up to 50 nm) concurrently with
smaller nanoscale particles. This pattern resembled findings from
previous friction stir welds of the identical alloy, where the large particles
were recognized as n phase precipitated on refined high-angle grain
boundaries during the initial phases of post-weld cooling. Small particles,
identified as n or n’
precipitates emerged within grains or on boundaries of sub-grains later
in the cooling process.
The alignment of numerous smaller particles suggested precipitation on
sub-grain boundaries or individual dislocations. Despite the absence of
larger particles on grain boundaries, they were located within grains,
potentially originating from nucleation at grain boundaries during
cooling. The considerable growth of grains in the region immediately
behind the pin could account for their displacement away from the grain
boundaries.

Microstructural model

Fig 8. Anticipated changes in the precipitate volume fraction concerning

the distance behind or ahead of the centre of the tool.
(a) weld centreline (0mm lateral position) (b) 2mm from centreline, (c)
4mm from centreline,
(d) 7mm from centreline (HAZ edge). Welding speed 5mms-1.

In this section, the authors discuss a microstructural model for the

evolution of precipitate phases during friction stir welding (FSW) of
AA7449 aluminium alloy. The model is compared to small-angle X-ray
scattering (SAXS) data, and its calibration is detailed elsewhere. The
significance of dynamic effects in the stir zone (SZ) is highlighted, with
refined grain size influencing nucleation sites for grain boundary h phase
formation during cooling. The evolution of precipitate volume fractions is
predicted at different distances from the weld centreline.
The microstructure of the base metal consists of h0 precipitates within
grains and a small fraction of grain boundary h phase. The model
predicts a complex sequence of changes, particularly in the SZ, where
grain boundary h precipitates first, suppressing quench-induced
precipitates in grain interiors. The evolution of volume fractions is
detailed for various positions along the welding direction.
Zooming in on the SZ, the model predicts the transformation of
metastable h0 phase to h, followed by dissolution of h precipitates and
subsequent grain boundary h precipitation
during cooling. Grain refinement leads to heterogeneous nucleation,
influencing the final microstructure. The model suggests a bimodal
population of precipitates on refined grain boundaries and smaller h0
precipitates within grains.
As the distance from the weld center increases, thermal and
deformation differences alter the balance of kinetic processes. Reduced
grain boundary precipitation and increased h0 reprecipitation within
grains are predicted. Experimental comparisons with SAXS and
scanning transmission electron microscopy (STEM) observations show
qualitative
consistency, with some differences attributed to model simplifications.
The model successfully predicts mean precipitate sizes within a
reasonable range of agreement.
The complexity of precipitate evolution, including bimodal or trimodal
distributions, poses challenges in interpreting SAXS results. Despite
these complexities, the model provides valuable insights into the
microstructural changes during FSW of AA7449 alloy.

Effect of welding speed

Fig 9. Predicted evolution of precipitate volume fraction as a function of

distance ahead or behind the tool center (positive positions are behind the
tool center)
(a) weld centerline (0mm lateral position) (b) 2mm from centerline, (c)
4mm from centerline, (b) 7mm from centerline (heat affected zone edge).
Welding speed 5mm/s.

The model predicts that as weld speed changes, the amount of

precipitate in the Heat Affected Zone (HAZ) varies. At slower speeds,
more precipitates dissolve, especially farther from the weld centre. Even
at the slowest speed, some precipitates remain due to peak temperature
control. Closest to the tool, both speeds dissolve a large fraction of
precipitates, with slow speed leading to re-precipitation. The model
slightly overestimates dissolution at slow speeds and underestimates at
fast speeds. Further from the centreline, slow speed causes more
dissolution, while fast speed shows little change. The findings suggest
that slower speeds lead to a wider zone of reduced hardness due to
expanded thermal fields.
These results, based on specific welding conditions, highlight the
complexity of microstructural changes in the HAZ.
In conclusion, the study presents a comprehensive analysis of the
microstructural evolution during FSW of aerospace aluminium alloys,
demonstrating the potential of the FlexiStir
system, in-situ measurements, and microstructural modelling to provide
valuable insights
into the complex phenomena occurring during welding. The findings have
significant implications for understanding and optimizing the weld
properties in aerospace applications, providing a valuable contribution to
the field of materials science and engineering. The study also highlights
the challenges and advancements in in-situ measurement techniques and
microstructural modelling, offering a promising approach for studying
dynamic
microstructural changes during welding processes.

Conclusions:
• FlexiStir system with synchrotron X-ray scattering tracks dynamic
precipitate evolution in AA7449-TAF during FSW.
• Interpretation relies on the microstructural model, clarifying
contributions from various precipitate phases.
• Full precipitate dissolution seen in Stir Zone (SZ), with partial
reprecipitation on grain boundaries due to refined grain size, aiding
heterogenous nucleation.
• In the Thermomechanically Affected Zone (TMAZ), partial precipitate
dissolution occurs, followed by post-weld reprecipitation mainly as
fine particles within grains due to lack of grain refinement.
• Heat Affected Zone (HAZ) shows dissolved precipitates depending
on distance from weld centerline and welding speed. Near shoulder,
significant original volume fraction loss observed for all speeds.
Faster welding speeds result in less dissolution away from
centerline. Additionally, even with relatively unchanged volume
fraction, model predicts significant coarsening and η' to η
transformation within HAZ.

References
[B. Zhou, "The Advancement of 7XXX Series Aluminum Alloys for
1Aircraft Structures: A Review,"
]metals, p. 1, 2021.
[R. Mishra, "Friction stir welding and processing," Science direct.
2
]
1