Understanding the track integrity in wire arc additive manufacturing using mechanistic model and machine learning - Master of Science thesis

Aug 5, 2026

This thesis investigates track integrity in wire arc additive manufacturing by combining mechanistic modeling with machine learning to improve prediction of defects and process quality. A part-scale finite element framework is used to model thermal behavior under different deposition patterns, and a reduced order model is developed to capture key temperature-field characteristics efficiently. Data-driven machine learning approaches—including support vector machines, logistic regression, and multilayer perceptrons—are evaluated for humping prediction and performance analysis.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages120

Reading Time5 h 2 min

AccessFree to download

List of Tables

List of Figures

Acknowledgements

Abstract

Chapter 1. Introduction

Background and Motivation

Wire Arc Additive Manufacturing

Current Status

Deposited Track Integrity Issues

Modeling and Machine Learning

Goals and Objectives

Thesis Structure

Chapter 2. Literature Study

Wire Arc Additive Manufacturing

Numerical Methods

Reduced Order Model

Humping

Machine Learning

Chapter 3. Modelling Approach

Part Scale Finite Element Model

Modelling of Temperature Field using Different Deposition Patterns

Reduced Order model

Chapter 4. Machine Learning Approach

Support Vector Machine (SVM)

Data-driven Machine Learning Framework

Logistic Regression and Multilayer Perceptron in WEKA

Chapter 5. Results and Discussion

Part-scale Finite Element Model

Modelling of Temperature Field using different deposition pattern

Reduced Order Modeling and Machine Learning for Humping Prediction

Chapter 6. Summary, Conclusions, and Future Work

References

Understanding the track integrity in wire arc additive manufacturing using mechanistic model and machine learning by Rakshith Reddy Sanvelly A thesis submitted to the graduate faculty in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Major: Mechanical Engineering Program of Study Committee: Tuhin Mukherjee, Major Professor Aishwarya Rajiv Pawar Jakob D. Hamilton The student author, whose presentation of the scholarship herein was approved by the program of study committee, is solely responsible for the content of this thesis. The Graduate College will ensure this thesis is globally accessible and will not permit alterations after a degree is conferred. Iowa State University Ames, Iowa 2025 Copyright © Rakshith Reddy Sanvelly, 2025. All rights reserved. LIST OF CONTENTS LIST OF TABLES ......................................................................................................................... iv LIST OF FIGURES ........................................................................................................................ v ACKNOWLEDGEMENTS ........................................................................................................... xi ABSTRACT.................................................................................................................................. xii CHAPTER 1. INTRODUCTION ................................................................................................... 1 1.1 Background and Motivation.................................................................................................. 1 1.2 Wire Arc Additive Manufacturing ........................................................................................ 3 1.3 Current Status ........................................................................................................................ 4 1.4 Deposited Track Integrity Issues ........................................................................................... 6 1.5 Modeling and Machine Learning .......................................................................................... 9 1.6 Goals and Objectives........................................................................................................... 12 1.7 Thesis Structure ................................................................................................................... 12 CHAPTER 2. LITERATURE STUDY ........................................................................................ 14 2.1 Wire Arc Additive Manufacturing ...................................................................................... 14 2.1.1 Process Principles and Variants.................................................................................... 14 2.1.2 Microstructure, Thermal Behavior, and Distortion Control ......................................... 15 2.1.3 Surface Finish, Mechanical Performance, and Advances ............................................ 20 2.1.4 Digital Tools and Industrial Applications .................................................................... 21 2.2 Numerical Methods ............................................................................................................. 22 2.3 Reduced Order Model ......................................................................................................... 29 2.4 Humping .............................................................................................................................. 35 2.5 Machine Learning ............................................................................................................... 40 CHAPTER 3. MODELLING APPROACH ................................................................................. 45 3.1 Introduction .........................................................................................................................

FAQs

What problem does the thesis address in wire arc additive manufacturing?

It focuses on deposited track integrity issues, particularly defects such as humping that affect part quality and reliability.

How is the thermal behavior modeled in this work?

A part-scale finite element model is used to simulate the temperature field, including variations caused by different deposition patterns.

Which machine learning methods are evaluated for prediction?

Support vector machines, logistic regression, and multilayer perceptrons in WEKA are assessed within a data-driven learning framework for humping prediction.

Aug 5, 2026

Upload by Seraphina