Final Year Project(KME-753)

TOPIC- Analysis of Friction Stir Welding using Machine

Learning

JSS ACADEMY OF TECHICAL EDUCATION NOIDA-201301

• Devesh Kumar Pathak(2000910400047)

• Amit Ahirwar(200091040019)
Mentor- Kapil Panchal Sir • Anshul Sachan(2000910400027)
• Anuratna Dubey(2000910400031)
CONTENT TABLE

 Our Objective
 Material we will be using in FSW
 How ML will help
 Gaussian Regression Process
 What is Full Factorial

 Various Research Paper

 References
OUR OBJECTIVE

□ MATERIAL USED IN THIS EXPERIMENT

□ THE USE OF ML METHODS TO DIAGNOSE THE

TOOL
FAILURE AND TO DETECT THE TOOL DEFECTS DURING
FSW PROCESS.

□ ABOUT GAUSSIAN PROCESS REGRESSION

□ WHAT IS FULL FACTORIAL

What is FRICTION STIR WELDING(FSW)

FSW is a solid-state joining technology in

which the welded materials do not cross their
melting point temperature. This gives FSW
technology superior advantages to weld
difficult to weld materials as well as dissimilar
materials.
In addition, FSW technology has many
technical advantages such as low consumable
tooling costs, low processing time, severe
plastic deformation, low peak temperature, and
excellent-quality joints.
 Significance of FSW in Manufacturing

The significance of FSW in modern manufacturing is rooted in its ability to provide

strong, high-quality joints with reduced thermal impact, making it a valuable and
versatile process across various industries.
□ Works Well with Tricky Shapes:
□ FSW is like a contortionist. It can handle welding in tight spaces and around
curves, making it perfect for creating complex shapes without breaking a
sweat.
□ Speedy and Consistent:
□ FSW is like the Flash of welding. It's fast and consistent, which means you
can make a lot of products quickly and they'll all be top-notch quality.
□ Eco-Friendly Welding:
□ FSW is like a green superhero. It produces fewer fumes, uses less energy, and
doesn't need a bunch of extra materials. So, it's kinder to the
 FSW Joints Properties

The properties of Friction Stir Welding (FSW) joints are

critical factors that determine the overall performance and
suitability of the welded components.
• Tensile Strength
• Fatigue Strength
• Ductility
• Hardness
Why Machine Learning?

Machine learning (ML) offers several advantages in the

context of Friction Stir Welding (FSW), making it a powerful
tool for analysis and optimization.

□ Fault Detection and Quality Assurance

□ Reducing Trial and Error
□ Real-Time Monitoring and Control
□ Tool Wear Prediction
 How Machine Learning works in
FSW
❖ Data Collection:
❖ Gather data related to FSW process parameters (e.g., tool rotation speed, welding
speed, tool tilt angle) and joint properties (e.g., tensile strength, fatigue strength).
❖ Training Data:
❖ Divide the dataset into training and testing sets. The training set is used to train the
Gaussian Process Regression model.
❖ Feature Selection:
❖ Identify relevant features that significantly impact FSW joint properties. These could
include tool parameters, welding conditions, and material properties.
❖ Model Training:
❖ Train the Gaussian Process Regression model using the training data. This involves
learning the underlying relationships between process parameters and joint properties.
❖ Prediction:
❖ Use the trained model to predict joint properties based on new or unseen FSW process
parameters.
❖ Uncertainty Estimation:
❖ Gaussian Process Regression provides not only predictions but also estimates of
uncertainty. This can be valuable in FSW, where the process might have inherent
variability. Understanding the uncertainty in predictions can guide decision-making.
❖ Evaluation:
❖ Evaluate the model's performance on the testing set using appropriate metrics for regression
tasks. Common metrics include mean squared error, R-squared, or others depending on the
specific goals.
❖ Optimization:
❖ Use the trained model for process optimization. This could involve finding the set of
process parameters that maximize or minimize certain joint properties.
 Material Selection:
The 6xxx series of aluminum alloys is commonly used in
friction stir welding (FSW). Some of the most widely used alloys in the
6xxx series include:
6061 Alloy: This is one of the most commonly used general-purpose
aluminum alloys. It contains magnesium and silicon, which gives it a
good combination of strength, weldability, and corrosion resistance. It is
known for its versatility and is used in a wide range of applications,
including aerospace components, marine fittings.
❑ Elsevier BV(2005)

❑ Friction stir welding (FSW) is a relatively

solid-state
new joining process. This joining technique is
energy efficient, environment friendly, and versatile.
In particular, it can be used to join high-strength
aluminum alloys and other metallic alloys that are
hard to weld by conventional fusion welding. FSW is
considered to be the most significant development in
metal joining in a decade. Recently, friction stir
processing (FSP) was developed for microstructural
modification of metallic materials.
❑ Elsheikh, A. H. (2023)

❑ The use of ML to predict the mechanical and

metallurgical properties of FSW joints in which ML
methods are used to figure out the relationship
between the control factors of the welding
process and the process responses.
❑ The use of ML methods to real-time control
of
FSW process to enhance the quality of the FSW
joints.
❑ The use of ML methods to diagnose the tool
failure and to detect the tool defects during FSW
process.
 ❑ Mohammad Syahid
Mohd Isa(2021)

Al and Cu are employed in various industrial applications due to their high

plasticity, thermal conductivity electrical conductivity and characteristics. By
effectively joining dissimilar Al and Cu, the unique properties of composite
formed by these metals can be adequately addressed. Friction stir
welding(FSW), an energy-efficient solid-state welding process is capable
of joining dissimilar metals, has enormous potential in the future of various
industries. This present work comprehensively summarises all pertinent
topics related to Al and
Cu FSW, such as FSW process parameters, microstructural
characterization, mechanical properties, and electrical
characteristics of Al-Cu joints produced by FSW.
❑ Anandan, B., & Manikandan, M. (2021)
Friction stir welding (FSW) is a solid-state joining process that produces joints without melting
and recasting. This process gains attention for joining similar and dissimilar metals like the
Aluminium alloy 7050 and 2014A. FSW provides very low peak temperature and heat
distribution compared to the conventional joining techniques. In the FSW technique, the normal
working temperatures range from 200 to 550 °C based on its process parameters. The
aluminium alloys 7050 and 2014A are precipitant hardened alloys, and their precipitates
dissolve at temperatures above 350 °C. It leads to a degradation of the mechanical properties of
the weldment. Predicting peak temperature (PT) is the major phenomenon in getting good
quality weld joints with respect to the FSW process parameters. This study deals with the
prediction of peak temperature using machine learning (ML) approaches in various regression
analysis methods like linear regression (LR), polynomial regression (PR), support vector
regression (SVR), decision tree regression (DTR), and random forest regression (RFR). As a
result, the RFR analysis is strongly suitable to predict the peak temperature in the FSW
process. A tool rotation speed of 1000 rpm ensured the peak temperature of less than 300 °C
with good appearance, proper material mixing, and the absence of defects.
❑ Matitopanum, S., Pitakaso, R., Sethanan, K. (2023)

❑ This research aims to develop ensemble machine-

learning
methods for forecasting the ultimate tensile strength (UTS)
of friction stir welding (FSW). The substance utilized in the
experiment was a mixture of aluminum alloys
AA5083 and
AA5061. An ensemble machine learning model was created
to predict the UTS of the friction stir-welded seam, utilizing
11 FSW parameters as input factors and the UTS
as a
response variable. The proposed approach used the Gaussian
process regression (GPR) and the support vector
machine
 • Gite(2019, January 1)

Process parameters such as

“rotational speed, welding traverse speed, tilt
angle, axial force, probe
and shoulder profile”, etc.must be
optimized for successful
implementation of FSW.
Squillace, A. (2009, March
1) The effect of processing parameters on
mechanical and microstructural properties of
AA6082 joints produced by friction stir welding
was analysed in the present study. Different
welded specimens were produced by employing a
fixed rotating speed of 1600rpm and by varying
welding speeds from 40 to 460mm/min.
Saikumar, A. (2023, March 1)
 The majority of comparable research studies concluded that
important factors, such as welding speed, rotational speed, plunge
depth, spindle torque, shoulder design, base material, pin profile
and tool type have a significant impact on the weldability of the
aluminium joint when using the FSW method. FSW parameters are
selected as tool speed(1200rpm, 1400rpm and 1600rpm), welding
speed (200mm/min, 300mm/min and 400mm/min) and axial force
(4 kN, 6 kN and 8 kN). For the tensile strength analysis, the axial
force is highly influenced and the maximum tensile strength of 175
MPa is attained.
M. F. (2023, September 1)
 Process parameters have a great influence on the quality
of the weld joints. In particular, basic process variables
such as tool design, rotational speed, welding speed,
and axial force were reviewed in this study. By
optimizing these process parameters, weld strength can
be increased by minimizing welding defects. In addition,
the effects of the FSW method on mechanical properties
such as hardness, tensile, and fatigue behaviors were
discussed.
 GAUSSIAN PROCESS REGRESSION

 Gaussian process regression (GPR) is a non-parametric Bayesian

approach used for regression tasks. It models the relationship
between input variables and output variables without assuming a
specific functional form. GPR assumes that the underlying process
generating the data follows a Gaussian distribution.
 CONCEPTS ABOUT GPR

 GPR is employed to solve regression problems, where the objective is to

predict a continuous output variable based on input features.
 Given a set of input-output pairs (X,y), where X represents input data and y
represents corresponding outputs, GPR models the relationship between X
and y as a distribution over functions.
 GPR involves hyperparameters, such as the length scale in the kernel
function, which are typically optimized during model training to maximize
the likelihood of the observed data.
 Given a set of input points, the GP provides a predictive
distribution for the corresponding output values. This distribution
includes a mean prediction and uncertainty (variance) associated
with each prediction.

 GPR is often used in Bayesian optimization, where an acquisition

function (e.g., Expected Improvement) is used to guide the
selection of new input points for evaluation to optimize a target
function.
 Full Factorial
A full factorial is a statistical experimental design method used to study the effects of
multiple variables on a process or outcome. In a full factorial experiment, all
possible combinations of the variables and their levels are tested, allowing for the
identification of the main effects of each variable and their interactions

In a full factorial experiment, the "levels" refer to the different settings or values that
each factor can take. For example, if a factor is temperature, the levels could be
"high" and "low." If another factor is pressure, its levels could be "high" and "low" as
well. When using a full factorial design, all possible combinations of these levels are
tested. For instance, if there are two factors, each with two levels, a full factorial
experiment would test all four possible combinations of the factor levels. This allows
the investigator to study the effect of each factor on the response variable, as well as
the effects of interactions between factors on the response variable.
Advantages and Use Cases:
 Uncertainty Estimation:
 GPR provides not only predictions but also a measure of uncertainty. This is
particularly valuable in applications where understanding prediction
uncertainty is crucial.
 Small Data and Noisy Data:
 GPR is effective even with small datasets, and it can handle noisy
observations well.
 Non-linear Relationships:
 GPR is capable of capturing complex, non-linear relationships between inputs
and outputs without explicitly specifying a functional form.
 CONCLUSION :
 Friction stir welding (FSW) is a widely used joining process in various industries.The
full factorial design method has been used to study the effects of multiple parameters on
the FSW outcomes. The machine learning regression processes, such as Gaussian
process regression (GPR), support vector machine (SVM), and multi-linear regression
(MLR), can be used to predict the ultimate tensile strength (UTS) of FSW joints . The
studies have shown that the optimized parameters employed in these investigations can
lead to predictable microstructural and mechanical behavior expected for defect-free
FSW welds. The results of these studies demonstrate the potential of machine learning
regression processes and full factorial design methods to optimize FSW and improve
the quality of the welds.
 References:
❑ Mishra, R. S., & Ma, Z. (2005, August 1). Friction stir welding and processing. Materials Science and
Engineering: R: Reports; Elsevier BV. https://doi.org/10.1016/j.mser.2005.07.001

❑ Verma, S., Gupta, M., & Misra, J. P. (2018, January 1). Performance evaluation of friction stir welding
using machine learning approaches. MethodsX; Elsevier BV. https://doi.org/10.1016/j.mex.2018.09.002

❑ Elsheikh, A. H. (2023, May 1). Applications of machine learning in friction stir welding: Prediction of joint
properties, real-time control and tool failure diagnosis. Engineering Applications of Artificial Intelligence;
Elsevier BV. https://doi.org/10.1016/j.engappai.2023.105961

❑ Anandan, B., & Manikandan, M. (2022, October 1). Machine learning approach for predicting the peak
temperature of dissimilar AA7050-AA2014A friction stir welding butt joint using various regression
models. Materials Letters; Elsevier BV. https://doi.org/10.1016/j.matlet.2022.132879

❑ Matitopanum, S., Pitakaso, R., Sethanan, K., Srichok, T., & Chokanat, P. (2023, January 27). Prediction of
the Ultimate Tensile Strength (UTS) of Asymmetric Friction Stir Welding Using Ensemble Machine Learning
Methods. Processes; Multidisciplinary Digital Publishing Institute. https://doi.org/10.3390/pr11020391
 References:

❑ Anandan, B., & Manikandan, M. (2023, March 1). Machine learning approach with various regression models
for predicting the ultimate tensile strength of the friction stir welded AA 2050-T8 joints by the K-Fold
cross-validation method. Materials Today Communications; Elsevier
BV. https://doi.org/10.1016/j.mtcomm.2022.105286

❑ Ahmed, M. M. Z., Seleman, M. M. E. S., Fydrych, D., & Çam, G. (2023, April 8). Friction Stir Welding of
Aluminum in the Aerospace Industry: The Current Progress and State-of-the-Art Review .
Materials; Multidisciplinary Digital Publishing Institute. https://doi.org/10.3390/ma16082971

❑ B. London, M. Mahoney, B. Bingel, M. Calabrese, D. Waldron Proceedings of the Third International Symposium
on Friction Stir Welding, Kobe, Japan, 27–28 September (2001)
❑ Tashkandi, M. A. (2017, June 1). Lap Joints of 6061 Al Alloys by Friction Stir Welding. IOP Conference Series.
https://doi.org/10.1088/1757-899x/205/1/012005
 References:
 Cavaliere, P., De Santis, A., Panella, F., & Squillace, A. (2009, March 1). Effect of welding parameters on
mechanical and microstructural properties of dissimilar AA6082–AA2024 joints produced by friction stir welding.
Materials in Engineering. https://doi.org/10.1016/j.matdes.2008.05.044
 Gite, R. A., Loharkar, P. K., & Shimpi, R. (2019, January 1). Friction stir welding parameters and application: A
review. Materials Today: Proceedings. https://doi.org/10.1016/j.matpr.2019.07.613
 Saravanan, R., Malladi, A., Amuthan, T., Aneesh, V., Jerin, A., Anbuchezhiyan, G., & Saikumar, A. (2023, March 1).
Mechanical characterization of friction stir welded dissimilar aluminium alloy using Taguchi approach. Materials
Today: Proceedings. https://doi.org/10.1016/j.matpr.2023.03.278
 Kılıç, S., Öztürk, F., & Demi̇ rdöğen, M. F. (2023, September 1). A comprehensive literature review on friction stir
welding: process parameters, joint integrity, and mechanical properties. Journal of Engineering Research.
https://doi.org/10.1016/j.jer.2023.09.005
 Simoncini, M., Costa, A., Fichera, S., & Forcellese, A. (2020, December 31). Experimental Analysis and
Optimization to Maximize Ultimate Tensile Strength and Ultimate Elongation of Friction Stir Welded AA6082
Aluminum Alloy. Metals. https://doi.org/10.3390/met11010069