Lecture Notes in Mechanical Engineering

Ramesh K. Agarwal Editor

Recent Advances
in Manufacturing
Engineering and
Processes
Proceedings of ICMEP 2022
Lecture Notes in Mechanical Engineering

Series Editors
Fakher Chaari, National School of Engineers, University of Sfax, Sfax, Tunisia
Francesco Gherardini , Dipartimento di Ingegneria “Enzo Ferrari”, Università di
Modena e Reggio Emilia, Modena, Italy
Vitalii Ivanov, Department of Manufacturing Engineering, Machines and Tools,
Sumy State University, Sumy, Ukraine

Editorial Board
Francisco Cavas-Martínez , Departamento de Estructuras, Construcción y
Expresión Gráfica Universidad Politécnica de Cartagena, Cartagena, Murcia, Spain
Francesca di Mare, Institute of Energy Technology, Ruhr-Universität Bochum,
Bochum, Nordrhein-Westfalen, Germany
Mohamed Haddar, National School of Engineers of Sfax (ENIS), Sfax, Tunisia
Young W. Kwon, Department of Manufacturing Engineering and Aerospace
Engineering, Graduate School of Engineering and Applied Science, Monterey, CA,
USA
Justyna Trojanowska, Poznan University of Technology, Poznan, Poland
Lecture Notes in Mechanical Engineering (LNME) publishes the latest develop-
ments in Mechanical Engineering—quickly, informally and with high quality.
Original research reported in proceedings and post-proceedings represents the core
of LNME. Volumes published in LNME embrace all aspects, subfields and new
challenges of mechanical engineering.
To submit a proposal or request further information, please contact the Springer
Editor of your location: Europe, USA, Africa: Leontina Di Cecco at Leontina.
[email protected] China: Ella Zhang at [email protected] India: Priya
Vyas at [email protected] Rest of Asia, Australia, New Zealand: Swati
Meherishi at [email protected] Topics in the series include:
. Engineering Design
. Mechanical Structures and Stress Analysis
. Engine Technology
. Nanotechnology and Microengineering
. Control, Robotics, Mechatronics
. MEMS
. Theoretical and Applied Mechanics
. Fluid Mechanics
. Engineering Thermodynamics, Heat and Mass Transfer
. Manufacturing
. Precision Engineering, Instrumentation, Measurement
. Tribology and Surface Technology

Indexed by SCOPUS and EI Compendex. All books published in the series are
submitted for consideration in Web of Science.
To submit a proposal for a monograph, please check our Springer Tracts in
Mechanical Engineering at https://link.springer.com/bookseries/11693
Ramesh K. Agarwal
Editor

Recent Advances
in Manufacturing
Engineering and Processes
Proceedings of ICMEP 2022
Editor
Ramesh K. Agarwal
Department of Mechanical Engineering
and Material Science
Washington University in St. Louis
St. Louis, MO, USA

ISSN 2195-4356 ISSN 2195-4364 (electronic)

Lecture Notes in Mechanical Engineering
ISBN 978-981-19-6840-2 ISBN 978-981-19-6841-9 (eBook)
https://doi.org/10.1007/978-981-19-6841-9

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Singapore Pte Ltd. 2023
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Organization

Conference Chairs

Prof. Ramesh K. Agarwal, Washington University in St. Louis, USA

Prof. Omar S. Es-Said, Loyola Marymount University, USA

Program Chairs

Assoc. Prof. Ken Mao, Warwick University, UK

Prof. Rong Liu, Carleton University, Canada

Student Program Chair

Prof. Yong Suk Yang, Pusan National University, Korea

International Technical Committees

Francisco Velasco, Universidad Carlos III de Madrid, España

Jingdong Wang, Changchun University of Science and Technology, China
Kazem Abhary, University of South Australia, Australia
Waail Mahmod Lafta, Martinus Rail Pty Ltd, Australia
Cherng-Yuan Lin, National Taiwan Ocean University, Taiwan
Eka Lutfi Septiani, Universitas Internasional Semen Indonesia, Indonesia
Chitphong Ketthanom, Vongchavalitkul University, Thailand
Jing YU, Dalian Maritime University, China

v
vi Organization

Rahul Chhibber, Indian Institute of Technology, India

Saiful Amri Mazlan, Universiti Teknologi Malaysia, Malaysia
Chih-Yuan Chen, National Taipei University of Technology, Taiwan
Zaini Ahmad, Universiti Teknologi Malaysia, Malaysia
Mohd Sukri Mustapa, Universiti Tun Hussein Onn Malaysia, Malaysia
S. Eiamsa-ard, Mahanakorn University of Technology, Thailand
Ivitskiy Igor, National Technical University of Ukraine, Ukraine
Mohd Rasidi Bin Ibrahim, Universiti Tun Hussein Onn Malaysia, Malaysia
Shiau Wei Chan, Universiti Tun Hussein Onn Malaysia, Malaysia
Kowit Piyamongkala, King Mongkut’s University of Technology North Bangkok,
Thailand
Tien-Dat Hoang, Thai Nguyen University of Technology, Vietnam
Jayavel S., Indian Institute of Information Technology, Design and Manufacturing,
Kanchipuram, Melakottaiyur, India
Dhaneshwar Mishra, Manipal University Jaipur, India
Gobinda Gopal Khan, Tripura University (A Central University), India
Deepa Prabhu, Manipal Institute of Technology, India
Nuchnapa Tangboriboon, Kasetsart University, Thailand
Minh Tuan Pham, Ho Chi Minh City University of Technology, Vietnam
Thitipong Wutisart, Dhonburi Rajabhat University, Thailand
Rafail Apakashev, Ural State Mining University, Russia
Mark Khazin, Ural State Mining University, Russia
Marina Menshykova, University of Aberdeen, UK
Zartasha Mustansar, NICHE-NUST Interdisciplinary Cluster for Higher Education,
Pakistan
Mohammad Junaid Iqbal Khan, NICHE-NUST Interdisciplinary Cluster for Higher
Education, Pakistan
Nuchanaporn Pijarn, Ubon Ratchathani University, Thailand
Rifky Ismail, Universitas Diponegoro, Indonesia
Shoroog Alraddadi, Umm Al-Qura University, Saudi Arabia
Menandro Marquez, Mapua University, Philippines
Krit Jiamjiroch, Thammasat University, Thailand
Hassan A. El-Hofy, Egypt Japan University of Science and Technology, Egypt
Mohanad Alabdullah, Deakin University, Australia
Pham Son Minh, Ho Chi Minh City University of Technology and Education,
Vietnam
Wahyu Dwi Lestari, University of Pembangunan Nasional Veteran Jawa Timur,
Indonesia
Mehmet Karakose, Firat University, Turkey
Silviu-Mihai Petrisor, Land Forces Academy from Sibiu, Romania
Constantin Dulucheanu, University “Stefan cel Mare” of Suceava, Romania
Radu Godina, NOVA University Lisbon, Portugal
Panarat Rattanaphanee, Suranaree University of Technology, Thailand
Markus Brillinger, Graz University of Technology, Austria
Samet AKAR, Çankaya University, Turkey
Organization vii

İlhan Asiltürk, Selcuk University, Turkey

Izirwan Bin Izhab, Universiti Malaysia Pahang, Lebuhraya
Zafer Bingül, Kocaeli University, Turkey
Vesna Spasojevic Brkic, University of Belgrade, Serbia
Mehmet Savsar, Kuwait University, Kuwait
Pirat Khunkitti, Khon Kaen University, Thailand
Elisha D. Markus, Central University of Technology, South Africa
Nguyen Dinh Ngoc, Thai Nguyen University of Technology, Vietnam
Abbas Amini, Western Sydney University, Australia
Baris Burak Kanbur, Nanyang Technological University, Singapore

International Reviewer Committees

Anh Van Nguyen, Cranfield University, UK

Ajeng Yulianti Dwi Lestari, Universitas Islam Indonesia, Indonesia
Sukangkana Talangkun, Khon Kaen University Thailand, Thailand
Naruebodee Srisang, Prince of Chumphon Campus, Thailand
Lu-Lin Li, National United University, Taiwan
Mohammed H. Rady, Universiti Tun Hussein Onn Malaysia, Malaysia
Mohd Azhar Harimon, Universiti Tun Hussein Onn Malaysia
Manjunath Shettar, Manipal Institute of Technology, India
Achmad Chafidz Mas Sahid, Universitas Islam Indonesia, Indonesia
Shazarel Shamsudin, Universiti Tun Hussein Onn Malaysia, Malaysia
Wahyu Dwi Lestari, Universitas Pembangunan Nasional Veteran Jawa Timur,
Indonesia
Thanh-Hai Nguyen, Ho Chi Minh City University of Technology (HCMUT),
Vietnam
Ali Mamedov, American University of the Middle East, Kuwait
Pham Quang Trung, Ho Chi Minh City University of Technology, Vietnam
Rhobert Endozo Alvarez, Batangas State University, Philippines
Inoka Manthilake, University of Moratuwa, Sri Lanka
Preface

During June 6–8, 2022, the 12th International Conference on Manufacturing

Engineering and Processes (ICMEP 2022) was successfully held via web platform.
Since 2021, ICMEP has been successfully held for 11 years. It always aims at
providing international and excellent platform for scholars and researchers from
all over the world to exchange ideas in the field of manufacturing engineering and
process and so on.
All delegates from all over the world communicated thoughts, shared experience,
etc., again with this form. It is great memories. All of them were active and positive.
After a fair and rigorous review by technical committees, 21 papers from Morocco,
Egypt, Indonesia, Vietnam, Thailand and so on were selected as a volume and
presented during the conference. We sincerely hope that all scholars of the related
research areas will benefit from it.
We are so appreciated to acknowledge all those who supported ICMEP2022. And,
especially thanks to technical committees and reviewers for taking time to help us
review papers rigorously. And, thanks for all participants’ support. We will try our
best to make ICMEP a more influential and professional platform. And, we look
forward to meeting you all physically next year!

St. Louis, USA Ramesh K. Agarwal

ix
Contents

Electromechanical System and Engineering Design

Finite Element Analysis of Ankle Foot Design Based on Indonesian
Anthropometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
W. D. Lestari, N. Adyono, L. Z. Faiza, P. Bagas, and G. D. Widagda
Automated Disinfection System for Polyethylene Terephthalate
Bottles for Bacteria, Fungi, and Viruses Using UVC LED Camera . . . . . 11
Gerson Orihuela, Esleiter Reyes, and Deyby Huamanchahua
Design of the CNC Router Structure for Machining Wood
Materials Using Reliability-Based Design Optimization Method . . . . . . . 25
Huu Loc Nguyen and Van Thuy Tran
Development of Intelligent Drone for Cassava Farming . . . . . . . . . . . . . . . 37
Orapadee Joochim, Kridtat Satharanond, and Wirachat Kumkun
Calculation and Optimization of the Two Stage Worm-Gear
Reducers Speed Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Le Thuy Anh and Nguyen Huu Loc
Project-Based Learning in Product Design Engineering: A Case
Study of External Gear Pump Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Muhammad Izzuddin Sabaruddin,
Mohammad Habib Fadillah Mohammad Mazni Latiff,
Nur Saidah Awang Abdullah, Wahyu Caesarendra, and Juliana Zaini

Machining and Manufacturing

Tool Wear Analysis on Time-Domain and Frequency-Domain
Data of Machining S45C Using Signal Processing Technique . . . . . . . . . . . 69
P. Jirapipattanaporn, A. Chanpariyavatevong, W. Lawanont,
and W. Boongsood

xi
xii Contents

Optimization of the Weight and Size Characteristics of the Power

Bracket for Additive Manufacturing Based on Topological
Optimization Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
A. V. Zhdanov, S. A. Baranov, and L. V. Belyaev
Identification and Active Control of the Chatter Phenomenon
in the Milling Process Using a Pneumatic Actuator . . . . . . . . . . . . . . . . . . . 87
Peyman Enteshari Najafabadi, Reza Nosouhi, Milad Soleimani,
and Mahmood Mirkhanzadeh
SHS Extrusion of a Composite Material Based
on Aluminum-Magnesium Spinel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Andrey Chizhikov, Alexander Konstantinov, Pavel Bazhin,
and Mikhail Antipov

Fluid Mechanical Engineering and Engineering Thermodynamics

Visual Observation of the Vortex Flow from the Same Diameter
Circular Cylinders Which Sandwiched a Partition Disk . . . . . . . . . . . . . . . 109
Yoshifumi Yokoi
Experimental Study of the Porous Emitter Installed in a Gas
Porous Burner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Panuwat Chanmak, Wapee Sripunya, and Bundit Krittacom
Experimental and Numerical Investigations of Effects of Wall
Friction and Vibrational Amplitude on Convection Phenomena
in Granular Materials Under Vertical Vibrations . . . . . . . . . . . . . . . . . . . . . 127
Tanapon Yachai, Itthichai Preechawuttipong, and Pawarut Jongchansitto
Comparative Study of Radiative Flux from Mushroom BLEVE
Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Anas Mbarki, Aziz Ettahir, and Abderraouf Guelzim

Applied Mechanics and Engineering Mechanics

Belt Layer Effects on Non-pneumatic Tire Performance by Finite
Element Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Juthanee Phromjan and Chakrit Suvanjumrat
Analysis of Hybrid Viscoelastic Sheets Adhered to the Hull
of a GFRP Vessel to Reduce Impact Damage . . . . . . . . . . . . . . . . . . . . . . . . . 159
Patrick Townsend, Juan C. Suárez Bermejo, and Alvaro Rodríguez-Ortíz
On the Contact Between Viscoelastic Tube with Inner Nonuniform
Coating and Several Rigid Inserts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Kirill E. Kazakov and Svetlana P. Kurdina
Failure Modes of Skull Bone of a Dog Head Under Impact Loading . . . . 177
Reda M. Othman, M. A. Hassan, and Wael M. Khierldeen
Contents xiii

Seismic Retrofit of RC Building with Elastic Stage

of Buckling-Restrain Braces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Thanawat Thipprapan, Amphon Jarasjarungkiat, Panumas Saingam,
and Arthit Petchsasithon
Response and Damage Identification by Long-Gauge Strain Sensor
of Beam-Like Bridges Under the Loads of Vehicle: A Comparison
of the Analytical Model and in Situ Measurement . . . . . . . . . . . . . . . . . . . . 197
Pham-Truong Son, Tien Hoang, Denis Duhamel,
Gilles Foret, Franziska Schmidt, François-Baptiste Cartiaux,
and Véronique Le Corvec

Metallic Materials and Corrosion Protection

Relation Between Binging Energy, Isothermal Bulk Modulus,
and Surface Tension of Liquid Metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Yu. N. Starodubtsev and V. S. Tsepelev
Corrosion Protection Method with Sacrificial Anode for Steel
Members in Boundary with Ground Using Water Swelling Rubber . . . . 219
J. Xu, S. Kainuma, K. Takaki, M. Yang, and T. Tomoda
Electromechanical System
and Engineering Design
Finite Element Analysis of Ankle Foot
Design Based on Indonesian
Anthropometry

W. D. Lestari, N. Adyono, L. Z. Faiza, P. Bagas, and G. D. Widagda

1 Introduction

The challenges faced by amputees today are in carrying out daily activities and trans-
portation. Prosthetics or orthotic devices have evolved to provide solutions to these
problems in line with the needs of prosthetic users. An overview of currently devel-
oping prosthetic limbs can be found in various references (Ghaith 2012; Jimenez-
Fabian et al. 2015; Kerkum et al. 2015; Corazza et al. 2014; Omasta et al. 2012;
Rigney et al. 2015; Veneva et al. 2014). There are two types of ankle prosthesis that
are currently being developed based on the type of movement, namely passive and
active prostheses. Passive prostheses are more affordable for amputation rehabili-
tation, where active prostheses are growing rapidly but require power to generate
energy. However, both types of ankle foot prosthetics have the same procedure,
namely, how amputees can imitate normal walking functions. Passive devices may
include springs to ensure storage and energy and dampers to suppress vibration. In
the past, solid ankle-based cushioned heels were considered in designing ankle foot
prosthetics (Casillas et al. 1995; Rao et al. 1998; Torburn et al. 1990). Several criteria
that must be considered in designing passive prostheses are portability capacity in the
form of size and level of lightness. Although active prostheses have several advan-
tages, they require an actuator and a motor, so they are still limited in portability.
The next criterion is how to keep the prosthesis simple and affordable.
Most of the research on ankle foot prosthetics that was developed previously and
currently uses robots and exoskeletons. Although this device can be controlled, it
is still quite complex and heavy, making it difficult for portability. Researchers are
currently facing the challenge of creating lightweight prosthetic designs. This study
developed the ankle foot design to be a lighter and more flexible structure. This study

W. D. Lestari (B) · N. Adyono · L. Z. Faiza · P. Bagas · G. D. Widagda

Department of Mechanical Engineering, Faculty of Engineering, University of Pembangunan
Nasional “Veteran” Jawa Timur, Jawa Timur 60294, Indonesia
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 3
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_1
4 W. D. Lestari et al.

proposes an ankle foot design with a compliant mechanism (CM) model commonly
known as a flexible mechanism. This mechanism is a special type of mechanism in
mechanical engineering which has a friction-free system, no backlash and monolithic
fabrication. Therefore, this compliant mechanism model has a lightweight. Several
researchers have used this model to create various designs (Howell et al. 2013).
Compliant mechanisms can be found in a variety of applications such as aerospace,
biomechanics and soft robotics (Dhote et al. 2015, 2018). The research group of
(Dao and Huang 2015, 2017a) uses the compliant model mechanism to make design,
modeling and optimization in several applications such as a one degree-of-freedom
(DOF) mechanism (Dao 2016), a two DOF mechanism (Huang and Dao 2016a), a
displacement sensor (Dao and Huang 2017b), a two DOF flexure-based mechanism
(Huang and Dao 2016b). Most of the previous studies still have a complex structure.
In particular, there are still shortcomings in creating an ankle joint that fits the natural
human foot. In this study, we propose a new prosthetic ankle foot design to support
smoother movement. So, this study develops a new ankle foot topology using a
compliant mechanism model, chooses a material that is easy to manufacture and
performs simulations to analyze the strength of the material.
Computational-based material strength simulation is one of the most effective
steps taken to reduce trial error in manufacturing prosthetics from the design results.
In addition, having a simulation beforehand will save costs, time and minimize the
energy expended when the strength testing of the prosthetic model. Computational-
based strength simulation is able to analyze a model to measure the value of defor-
mation, stress, and strain that occurs in detail. This research is developing a biome-
chanical design and simulation for ankle foot prosthetics. The designs are made by
mimicking actual human feet. The selected prosthetic materials are silicone for the
skin and AISI 304 for the support core. Ankle foot prosthetic models are designed to
be able to withstand the average human weight on the walking cycle mechanism. In
addition, the model must also have a safe factor of safety before the manufacturing
process is carried out. The results of this study are expected to improve the quality
of life of people with disabilities, help people with disabilities in normal activities
and can be produced on a mass scale.

2 Material and Methods

2.1 Biomechanical Design

The initial stage in this research process is a literature study to look for similar
research references and forms of prosthetic ankle foot models that have been studied.
The next step is to determine the ankle foot design criteria to be made. The design
criteria in this study can be seen in Table 1. Based on these criteria, the design process
will be continued and the results will be analyzed. The model analysis aims to analyze
how strong and feasible the design to produce and apply to patients with disabilities.
Finite Element Analysis of Ankle Foot Design Based on Indonesian … 5

Table 1 Design criteria

Criteria Description
Strength (Chuan et al. 2010) Able to withstand weight up to 100 kg
Light (Tayyari 2017) Has a lighter mass than the real leg which is less than 2.2686 kg
Ankle foot type Model compliant mechanism by imitating the human foot in
general
Convenience Easy to use in walking position

Fig. 1 Ankle foot design

If the analysis is not strong enough to withstand the load, then return to the model
design stage.
Design and analysis were made by Solidworks software. In making the ankle foot
design in this study, the lowest mass of Indonesian people was taken, which was
39.8 kg. The weight of the leg below the knee according to Tayyari (2017) is 5.7%
of the human weight. So if the calculation is carried out, the weight of the prosthetic
foot must be below 2.2686 kg so as not to burden the user. The ankle foot prosthetic
designed using a compliant mechanism model and using AISI 304 material. The
material was chosen because it has properties that has good formability or shaping
power, making it easier to make the core support on the ankle foot design according
to the desired shape. The mass of the designed artificial foot is 0,498,285 kg. The
design results are shown in Fig. 1.

2.2 Simulation

The finite element method was used to analyze the ankle foot design in this study.
The analysis was carried by Solidworks software to analyze the structural behavior of
the model and determine the overall deformation, equivalent stress and strain energy.
The characteristics of the AISI 304 material that are input into the simulation process
are Young’s modulus of 190,000 MPa and Poisson’s ratio of 0.29. Testing based on
ISO 10328 standard which applies angle and support to the ankle foot. Ankle foot
6 W. D. Lestari et al.

Fig. 2 Simulation illustration of the midstance phase

prosthetic with a compliant mechanism model limits it to three kinds of fix support in
the gait cycle. The three positions are the heel strike position, the midstance position
and the toe-off position. In this study, the test was carried out on one of the conditions
of the walking cycle, namely midstance. In this condition, the ankle foot has a fixed
support located on the soles of the forefoot and back at an angle of 0° .
The magnitude of the load and the load direction given is based on gait analysis
data from the percentage of body weight or vertical force and ankle angle in each
phase of the gait cycle (Bovi et al. 2011). The body weight or the amount of load that
is applied in accordance with the predetermined design criteria is 100 kg, so that the
given load is 980 N. The illustration of the simulation carried out with the boundary
condition of the midstance position of the human gait cycle is shown in Fig. 2.

3 Results and Discussion

The design testing of the prosthetic ankle foot was carried out in a walking condition
in the midstance phase. In this condition, fixed geometry is placed on the entire base
of the foot. The test is assumed that the support at the foot is fix. While the load is
directed from the top to the bottom of the foot (Fig. 2). There are three types of data
taken, namely in the form of Von Misses stress, displacement and strain.

3.1 Stress Analysis Von Misses Stress

The Von Misses stress analysis parameter is used to determine the results of the stress
that occurs in compliant ankle foot prosthetics during the walking mechanism in the
midstance phase. The analysis results show that the maximum stress occurs in the
midstance position of 91.18 MPa. In the midstance position, because it is supported
by two supports at the front and rear, so that the stress distribution is more even.
The area of maximum stress that occurs in the ankle foot design with the midstance
Finite Element Analysis of Ankle Foot Design Based on Indonesian … 7

Fig. 3 Von Mises stress ankle foot midstance phase

position is shown in Fig. 3. The safety factor of ankle foot prosthetics was evaluated
using the following formula:

σ yield
n= (1)
σ actual
where n is the safety factor, σ yield is the yield strength of the material, σ actual and is
the maximum stress that occurs. Based on these equations, the safety factor results of
2268. These results indicate that the stress that occurs is still below the yield strength
limit of AISI 304 material with a safety factor above 1.5 which is quite safe. In other
words, the ankle foot prosthetic design by adopting a compliant mechanism is still
categorized as safe because it is still in the elastic area of the material.

3.2 Static Analysis Displacement

The total deformation is used as an analysis parameter to determine the changes in

the shape of the model that occur when it is loaded. The result of the total deforma-
tion shows that the midstance position experiences a displacement of 004,887 mm.
The shape of the deformation that occurs at that position is illustrated in Fig. 4. The
material stiffness value in the ankle foot design in the midstance phase can be deter-
mined by using the formula K = F/d, where K is stiffness, (N/mm) is force and (N)
is displacement (mm). Based on these equations, it can be seen that the amount of
stiffness that occurs in the midstance position is 20053,2024 N/mm.
8 W. D. Lestari et al.

Fig. 4 Total deformation results of the ankle foot design in midstance position

3.3 Static Analysis Strain

The next analysis parameter is to measure the stored strain energy. Based on the
given load according to the magnitude of the gait cycle mechanism, the results show
that at the midstance position, the maximum strain that occurs is 0.0003626. The
area of maximum strain energy is located at the curvature between the toe and the
sole of the foot with the maximum stress point as shown in Fig. 5.

4 Conclusion

Ankle foot prosthetics are mobility aids for people with disabilities to assist with
daily activities. The purpose of the design and analysis of the compliant ankle foot
prosthetic model is to determine the strength and feasibility before this model is
fabricated. Based on the above discussion, it can be concluded that the ankle foot
compliant prosthetic design is capable of performing motion with a gait cycle mech-
anism in the midstance phase with a load of 100 kg. This model has the advantage
of being flexible because the elastic curve adopts the concept of a compliant mech-
anism and utilizes strain energy. In addition, the design made has a safety factor
in accordance with the requirements. For further research, it is necessary to make
artificial ankle foot according to the material used in the simulation so that their real
capabilities are known.
Finite Element Analysis of Ankle Foot Design Based on Indonesian … 9

Fig. 5 Strain results of the ankle foot design in the midstance position

References

Bovi, G., Rabuffetti, M., Mazzoleni, P., Ferrarin, M.: A multiple-task gait analysis approach: kine-
matic, kinetic and EMG reference data for healthy young and adult subjects. Gait Posture 33(1),
6–13 (2011). https://doi.org/10.1016/j.gaitpost.2010.08.009
Casillas, J.M., Dulieu, V., Gras, P., Cohen, M., Marcer, I., Didier, J.P.: Bioenergetic comparison of a
new energy-storing foot and sach foot in traumatic below-knee amputees or in vascular amputees.
Ann. Readapt. Med. Phys. 38(2), 57–63 (1995). https://doi.org/10.1016/0168-6054(96)89302-0
Chuan, T.K., Hartono, M., Kumar, N.: Anthropometry of the Singaporean and Indonesian
populations. Int. J. Ind. Ergon. 40(6), 757–766 (2010). https://doi.org/10.1016/j.ergon.2010.
05.001
Corazza, F., Parenti-Castelli, V., Stagni, R., Cappello, A., O’connor, J.J., Leardini, A.: Biomechanics
of the intact and replaced human ankle joint. J Mech Med Biol 06(01), 39–46 (2014). https://doi.
org/10.1142/s0219519406001819
Dao, T.P.: Multiresponse optimization of a compliant guiding mechanism using hybrid taguchi-grey
based fuzzy logic approach. Math. Problems Eng., 2016 (2016). https://doi.org/10.1155/2016/
5386893
Dao, T.P., Huang, S.C.: Design, fabrication, and predictive model of a 1-dof translational flexible
bearing for high precision mechanism. Trans. Can. Soc. Mech. Eng. 39(3), 419–429 (2015).
https://doi.org/10.1139/tcsme-2015-0031
Dao, T.P., Huang, S.C.: Compliant thin-walled joint based on zygoptera nonlinear geometry. J.
Mech. Sci. Technol. 31(3), 1293–1303 (2017a). https://doi.org/10.1007/s12206-017-0228-8
Dao, T.P., Huang, S.C.: Design and analysis of a compliant micro-positioning platform with
embedded strain gauges and viscoelastic damper. Microsyst. Technol. 23(2), 441–456 (2017b).
https://doi.org/10.1007/s00542-016-3048-3
Dhote, S., Yang, Z., Zu, J.: Modeling and experimental parametric study of a tri-leg compliant
orthoplanar spring based multi-mode piezoelectric energy harvester. Mech. Syst. Signal Process.
98, 268–280 (2018). https://doi.org/10.1016/j.ymssp.2017.04.031
10 W. D. Lestari et al.

Dhote, S., Zu, J., and Zhu, Y.: A nonlinear multi-mode wideband piezoelectric vibration-based
energy harvester using compliant orthoplanar Spring, Appl. Phys. Lett., 106(16) (2015). https://
doi.org/10.1063/1.4919000
Ghaith, F.A.: Nonlinear finite element modeling of nano-indentation.: Proc Int Conf Adv Robo Mech
Eng Des, pp. 1–4 (2012)
Howell, L.L., Magleby, S.P., Olsen, B.M.: Handbook of Compliant Mechanisms. Wiley (2013)
Huang, S.C., Dao, T.P.: Design and computational optimization of a flexure-based XY positioning
platform using FEA-based response surface methodology. Int. J. Precis. Eng. Manuf. 17(8),
1035–1048 (2016a). https://doi.org/10.1007/s12541-016-0126-5
Huang, S.C., Dao, T.P.: Multi-objective optimal design of a 2-DOF flexure-based mechanism
using hybrid approach of grey-taguchi coupled response surface methodology and entropy
measurement. Arab. J. Sci. Eng. 41(12), 5215–5231 (2016b). https://doi.org/10.1007/s13369-
016-2242-z
Jimenez-Fabian, R., Flynn, L., Geeroms, J., Vitiello, N., Vanderborght, B., Lefeber, D.: Sliding-bar
MACCEPA for a powered ankle prosthesis. J. Mech. Robot. 7(4), 1–11 (2015). https://doi.org/
10.1115/1.4029439
Kerkum, Y.L., Buizer, A.I., Van den Noort, J.C., Becher, J.G., Harlaar, J., Brehm, M.A.: The effects
of varying ankle foot orthosis stiffness on gait in children with spastic cerebral palsy who walk
with excessive knee flexion, PLoS ONE, 10(11) (2015). https://doi.org/10.1371/journal.pone.014
2878
Omasta, M., Paloušek, D., Návrat, T., Rosický, J.: Finite element analysis for the evaluation of
the structural behaviour, of a prosthesis for trans-tibial amputees. Med. Eng. Phys. 34(1), 38–45
(2012). https://doi.org/10.1016/j.medengphy.2011.06.014
Rao, S.S., Boyd, L.A., Mulroy, S.J., Bontrager, E.L., Gronley, J.K., Perry, J.: Segment velocities in
normal and transtibial amputees: prosthetic design implications. IEEE Trans. Rehabil. Eng. 6(2),
219–226 (1998). https://doi.org/10.1109/86.681188
Rigney, S.M., Simmons, A., Kark, L.: Concurrent multibody and finite element analysis of the
lower-limb during amputee running. Proc Eng Med Biol Soc, pp. 2434–2437 (2015)
Tayyari, S.: Occupational ergonomics. Occup. Ergon. 3, 6221 (2017). https://doi.org/10.4324/978
0203484913
Torburn, L., Perry, J., Ayyappa, E., Shanfield, S.L.: Below-knee amputee gait with dynamic elastic
response prosthetic feet: a pilot study. J. Rehabil. Res. Dev. 27(4), 369–384 (1990). https://doi.
org/10.1682/jrrd.1990.10.0369
Veneva, I., L., Vanderborght, B., Lefeber, D., Cherelle, P.: Propulsion system with pneumatic artifi-
cial muscles for powering ankle-foot orthosis. J. Theor. Appl. Mech., 43(4), 3–16 (2014). https://
doi.org/10.2478/jtam-2013-0029
Automated Disinfection System
for Polyethylene Terephthalate Bottles
for Bacteria, Fungi, and Viruses Using
UVC LED Camera

Gerson Orihuela , Esleiter Reyes , and Deyby Huamanchahua

1 Introduction

Faced with the global crisis, which is being experienced due to SARS-CoV-2. Health
problems are a topic to be addressed worldwide, such as infectious diseases caused
by bacteria, viruses, parasites, and fungi, practically by pathogenic microorganisms,
which are contagious and are transmitted by inhalation of particles indirectly or
directly (Bloom and Cadarette 2019). Likewise, the risks of contracting the virus
have been increasing exponentially due to new sources of contagion such as markets,
supermarkets, and crowded places, with the most frequent transmission of the virus
through microdroplets or droplets that come off the nose (May 2021). These can
be deposited on the surfaces of nearby materials, which are mostly handled and
discarded, likewise, they can survive a couple of hours or up to several days depending
on the material or residue (Quitral Q 2020).
Faced with this, all materials must have optimal handling from their production,
consumption, and final disposal, so there must be a control and hygiene treatment
for the comfort of people working in the recycling sector, to guarantee their health
and well-being (Soliz Torres et al. 2020). The spread of the virus occurs most often
in enclosed and poorly ventilated places, such as recycled waste treatment plants
where recycled materials, many of them contaminated, accumulate to be treated for
reintegration (Huayanay 2020). Considering that the main transmission mechanism
is by contact and inhalation (Li et al. 2020), category C (UVC) ultraviolet light
offers an efficient alternative in cleaning air and surfaces, which by eliminating said
microbial particles reduces the probability of human contagion by the virus that
circulates in an environment or space where there is contaminated recycled material
accumulated with viruses and bacteria (Cervino et al. 2021; Manuel et al. 2020).

G. Orihuela · E. Reyes (B) · D. Huamanchahua

Department of Mechatronics Engineering, Universidad Continental, Huancayo, Perú
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 11
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_2
12 G. Orihuela et al.

According to studies and the ideal treatment with ultraviolet rays, it reduces the
microbial load in contaminated material where UVC rays of the short wavelength of
200–280 nm are applied which generates a germicidal effect (Rattanakul and Oguma
2018; Kim et al. 2017; Hsu et al. 2021). UV rays possess a germicidal action that kills
or neutralizes the action of viruses, bacteria, and other primitive microorganisms, so it
is mainly used in the sterilization of surfaces and local medical, pharmaceutical, and
food industry environments (Hsu et al. 2021; Raeiszadeh and Adeli 2020). The germi-
cide lamp depends on the power and geometric shape considering the appropriate
time of dose of UVC light emission, eliminating up to 99.9% of microorganisms
(Nnadi et al. 2021; Dwivedi et al. 2021; Heilingloh et al. 2020).
UV irradiation has proven to be an effective physical means for the control of
microorganisms, including viruses, with the UVC wavelength being most effective
in practical application, followed by that of UVBs and finally that of UVA (Ou
and Petersen 2008). The germicidal capacity against viruses lies in their ability to
penetrate the viral structure, especially UVC and UVB (Nishisaka-Nonaka et al.
2018), which directly affect the genome preventing its replication and therefore
leading to inactivation (Bono et al. 2021).
UVC treatment does not generate organoleptic variations, nor does it generate
chemical residues and is effective in disinfecting the surfaces of products and mate-
rials (Raeiszadeh and Adeli 2020). Currently, UV light is used for the disinfection
and sterilization of surgical products, cleaning of cell phones, purification of environ-
ments (Manuel et al. 2020; Rattanakul and Oguma 2018; Casini et al. 2019) process
by which it performs to eliminate all forms of microbial life and bacteria (Kim et al.
2017; Dwivedi et al. 2021). It is also used in the disinfection of algae in ballast water
based on a new treatment system with multiple ultraviolet waves in an approximate
time interval of 20–30 s with a UV photocatalysis system (Lu et al. 2021).
Therefore, the objective of this article is to design a disinfection machine for
recycled PET-type bottles using UVC radiation emitted by an LED camera, through
an engineering development with the benefits of UV type C light, based on the
mechatronic methodology of the Association of German Engineers (VDI 2206),
composed of a mechanical, electrical-electronic and control system (Graessler and
Hentze 2020), which makes it possible to sterilize recycled PET materials that accu-
mulate in recycling plants for the process of insertion in society, since according to
studies shows that 99.99% of viruses and bacteria are removed from the surface of
materials and waste that are disposed of in urban areas and markets in general.

2 Methodology

The methodology used in the article is based on the VDI 2206 model. The method-
ology that represents the decomposition of the system is based on a V model for
mechatronic systems that represents a logical sequence of activities composed of the
left side the decomposition of the system and the right side the subsystems, which
are constantly verified and validated (Graessler and Hentze 2020).
Automated Disinfection System for Polyethylene Terephthalate Bottles … 13

Fig. 1 Activities to be carried out in the VDI2206 structure

In Fig. 1, the approach of combining activities is shown, and the guideline suggests
three main phases followed by a fourth that accompanies the other three: First, the
System design that has an idea of the result between the influence that specifies
the essential particularities logical and physical design motive, second the Domain
Specific Design which establishes the need for further illustration through more
detailed interpretations and calculations separately for each technical system, and
the third phase is the System Integration which consists of evaluating the overall
performance of the system when individual domains are integrated and the fourth
is a cross-sectional Modeling and Simulation phase that creates and examines the
qualities of the system with the support of structural models, tools, and computer-
aided simulation instruments. The cycles of this project are the design of the LED
camera and the functional simulation prototype. The prototype process includes 3D
CAD software, Ladder programming, and the Factory IO integration system.

3 Materials

The mechanical system of the machine: Composed of the main motor (0.75 HP) that
drives the conveyor belt (Siegling Transilon MT black E8/2 U0/V5H with k1% =
8 N/mm) generating the transfer of recycled PET bottles toward the LED disinfection
chamber of approximately three-square meters of area, after that the exit conveyor
will be activated.
Electrical System–electronic machine: Composed of the main sensor and three
secondary sensors (Diffuse Sensors) (Gutiérrez and Iturralde 2017). The objective
of the main sensor is to detect the recycled PET-type material when it is inside the
14 G. Orihuela et al.

chamber, generating the stoppage of the conveyor belt and the start of the disinfection
process. The ideal time to sanitize recycled bottles is 26 s (s) (Woo et al. 2019;
Minamikawa 2008; Muramoto et al. 2008).
Machine control system: Employing a programmable logic controller (modular
PLC) composed of programming lines in charge of giving instructions to the machine
and storing them in its memory, represented by a control panel by pushbuttons, a
liquid crystal display (LCD) for observing the count of disinfected materials, and
light-emitting diodes (LEDs).
The prototype design of the UVC LED camera was developed in the Inventor
program, to develop the virtual simulation of the Factory IO program with the use of
PLC programming in the TIA Portal engineering system (Factory 2021; Download
Inventor 2021; Software 2019). For the design, we must consider that when working
with UV radiation with a range of 200–280 nm, the rays must be isolated, since
produces a modification in DNA and this causes cancer (Muramoto et al. 2008;
Buonanno et al. 2020). For this reason, strips of anti-ultraviolet ABS plastic sheet
have been used to block the projection wave of the irradiation of ultraviolet rays
at the beginning and the end of the disinfection chamber stroke (Rodríguez-Tobías
et al. 2013).
Figure 2 shows the structural scheme of the prototype of the PET bottle disinfec-
tion system, whose structure is based on the process control system with the input
and output signal. The vision system captures the input signal of the bottles to the
conveyor belt with diffuse sensor inspection. The electronic system that turns on the
feeding actuators of the PET bottles, the disinfection chamber that projects category
C ultraviolet rays, and the removal of the disinfected material, generating a feedback
signal or process concluded. The electromechanical system generates the mobiliza-
tion of the conveyor belt, the illumination of the UVC LEDs, and the transport of the
decontaminated PET bottles to their storage. The energy generates the impulse, the
heat, the movement, the vibrations, and the light of the prototype.

4 Results

4.1 Disinfection Process Time

The maximum disinfection efficiency is 254 nm (nm), which depends on the time of
26 s (s) for the sterilization of the recycled PET bottles to be disinfected.
Figure 3 shows the disinfection chamber made up of four fundamental compo-
nents; the first component is the chamber liner made of AISI 304 stainless steel from
the American Iron and Steel Institute (Mundial 2021). The second component is the
protection by strips of anti-ultraviolet ABS plastic sheet that blocks the projection
wave of ultraviolet radiation (Rodríguez-Tobías et al. 2013). The third component
is the LED type lights that project category C ultraviolet rays, and finally, the fourth
component is the main stainless steel structure that ensures resistance, generates
Automated Disinfection System for Polyethylene Terephthalate Bottles … 15

Fig. 2 Structural diagram of the disinfection system

Fig. 3 LED camera that projects ultraviolet rays type C

the shape, and designs stability of the UVC LED camera, which has dimensions of
2000 cm long, 1200 cm high, and 1500 cm wide.

4.2 Conveyor Belt

It is based on the weight of the mass of 200 kg of PET bottles that will be transported
approximately by the conveyor belt, and then the resistance of the conveyor belt to be
used will be determined, consulting Eqs. 1 and 2. The maximum belt traction force
(F 1 ) is calculated, which is the result of multiplying the tangential force (Fu) of 200
16 G. Orihuela et al.

pounds equivalent to 1962 N by the smooth dry steel factor C 1 of 1.5 at an angle of
180° (Siegling 2016).

F1 = Fu(C1 ) (1)

With the result of 2943 N, maximum traction of the belt is divided with the width
of the belt (b0 ) of one meter, it is compared with the maximum range factor of the
type of tractor element of 2 that must have a conveyor belt multiplied by the tolerance
k 1 % of 8 N millimeters (N/mm) so that it works properly with the values provided
by the technical manual (Siegling 2016).

F1 /b0 ≤ 2(k1 %) (2)

Therefore, with the result obtained of 2943 N/mm less than and equal to 16 N/mm,
it is stated that the conveyor belt will perfectly resist the force exerted by the motor
since the value is within the adequate working range as shown in Fig. 4.

Fig. 4 Automated machine conveyor belt

Automated Disinfection System for Polyethylene Terephthalate Bottles … 17

Fig. 5 Voltage regulation

4.3 Electronic Circuit and Programmable Logic Control

Diagram

Figure 5 shows the electronic circuit that regulates the voltage from alternating current
(VAC) of 220 V to a direct current voltage (VDC) of 24 V for the correct operation
of the programmable logic controller.
Figure 6a shows the circuit diagram of the operation of the input and output PLC
devices, and the outputs are seen as contactors KM1, KM2, and KM3, as well as the
start and stop LEDs of the electronic receivers.
Figure 6b shows the power circuit, a part that generates the highest current
consumption of the controller, this circuit oversees exciting the relays to activate
the output devices, these work through semiconductor devices to control or modify
the voltage or current of the machine.
Figure 6c shows the connection diagram of the input devices of the PLC that
consists of a start, stop safety, emergency stop buttons, also inductive sensors, and
control elements of the electronic diagram.

4.4 Ladder Programming Blocks

Figure 7a shows the segment with Ladder programming for the connection of Factory
IO engineering programs with the TIA portal software.
Figure 7b shows the segment of the safety system and control drive of the
automated machine.
Figure 8a shows the enumerator segment of the system, revealing the total amount
of materials disinfected in a full working day.
Figure 8b shows the segment of the drive of the first conveyor belt that feeds the
disinfection system, starting the count of material entering the C category UV LED
chamber.
Figure 9a shows the segment of the drive of the ultraviolet led camera category
C, with the estimated waiting time, while the material of the recycled PET bottles is
18 G. Orihuela et al.

Fig. 6 a Programmable logic control, b power circuit and c PLC input connection diagram

Fig. 7 a Connection factory IO–TIA portal and b security system

disinfected for 26 s, after which it generates the drive of the output conveyor belt for
the warehouse of the recycling plant.
Figure 9b shows the segment of the outputs of the programmable logic controller
that activates each drive component of the machine.
Automated Disinfection System for Polyethylene Terephthalate Bottles … 19

Fig. 8 a System counter and b drive of the conveyor belt and counter

Fig. 9 a Disinfection and exit of the material and b component activation signal outputs
20 G. Orihuela et al.

Fig. 10 Automated disinfection in factory IO of polyethylene terephthalate bottles for bacteria,

fungi, and viruses; by UVC LED camera

4.5 Automated PET Bottle Machine

Figure 10 shows the final prototype of the automatic polyethylene terephthalate bottle
disinfection machine that uses a 254 nm short-wave category C UV LED camera for
the disinfection of bacteria, fungi, and viruses.

5 Discussion

SARS-CoV-2 is a highly variant and contagious virus due to its rapid transmission
through droplets that settle on the surfaces of surrounding materials or by materials
handled and discarded. Therefore, the disinfection system focuses on PET bottle
recycling treatment plants with poor ventilation and closed environments for waste
storage processes. These droplets remain impregnated on the surface of any material
and at the same time in the process of transferring or displacing the contaminated
material to reinsert the recycled polyethylene terephthalate material into society, the
risk of contagion increases since the virus can survive for hours or days depending
on the material where it resides.
That is why automated disinfection is important and necessary, with an LED
camera that projects type C ultraviolet rays, sterilizes PET bottles against the threat
of SARS-CoV-2 and highly contagious variant particles, thus efficiently eliminating
impregnated viruses, generating security for the population of the recycling sector
that plays a very important role in the collection and transformation of recycled
material.
In this process, the recycled PET bottles enter through the conveyor belt to the
UVC LED camera which has a memory program that applies 26 s to sterilize 99.99%
of fungi, batteries, and viruses. With a short wavelength of 254 nm and an irradiation
Automated Disinfection System for Polyethylene Terephthalate Bottles … 21

dose of 220 (J/m2 ) Joule per square meter, the disinfected bottles are removed by the
conveyor belt for their reinsertion process. This proposal is considered an emerging
technology that agrees with the following studies (Rattanakul and Oguma 2018;
Kim et al. 2017; Hsu et al. 2021; Raeiszadeh and Adeli 2020; Nnadi et al. 2021;
Dwivedi et al. 2021) and (Heilingloh et al. 2020) which consists of control surfaces.
The recycled PET bottles, at the illumination of the wave range of 200–280 nm, it is
confirmed that there is greater sterilization with the emissions of 254 nm.

6 Conclusion

In conclusion, it is highlighted that the machine works with ideal automated treat-
ment, with category C short-wave UV radiation, and this reduces the microbial
load found on the surface of polyethylene terephthalate materials in the recycling
plant. This UVC disinfection strategy is highly effective in eradicating up to 99.99%
of microorganisms and viruses; this method being an alternative to help disinfect
surfaces in the current pandemic, so this research shows how to safely disinfect the
surface to be treated.
Implement the proposal both in different recycling and disinfection sectors, which
will allow its effectiveness to be verified, and also to adapt and improve both the model
and the proposed methodology and tools.
Finally, this research will serve for the development and improvement of similar
machines; since its purpose is that viruses, fungi, and bacteria do not remain on the
surfaces of the materials that are processed for their reintegration into our society.

References

Bloom, D.E., Cadarette, D.: Infectious Disease threats in the twenty-first century: strengthening the
global response. Front. Immunol. 10, 549 (2019). https://doi.org/10.3389/fimmu.2019.00549
Bono, N., Ponti, F., Punta, C., Candiani, G.: Effect of UV irradiation and TiO2 -photocatalysis on
airborne bacteria and viruses: an overview. Mater. (basel) 14(5), 1075 (2021)
Buonanno, M., Welch, D., Shuryak, I., Brenner, D.J.: Far-UVC light efficiently and safely inactivate
airborne human coronaviruses. Res. Square (2020)
Casini, B., et al.: Evaluation of an ultraviolet C (UVC) light-emitting device for disinfection of high
touch surfaces in hospital critical areas. Int. J. Environ. Res. Public Health 16(19), 3572 (2019)
Cervino, C.O., Almandoz, J.C., Mignone, M., Irusta, A., Leiton, G.: The use of UV-C radiation
for disinfection in the face of the COVID-19 pandemic: new portable UV-C disinfection system,
UNIMORON-Desinfector© (2021)
Download Inventor 2022: Autodesk.com, [Online]. Available in: https://latinoamerica.autodesk.
com/products/inventor/free-trial. Last accessed 06 Apr 2021
Dwivedi, V. et al.: Rapid and efficient inactivation of SARS-CoV-2 from surfaces using UVC
light-emitting diode device (2021)
Factory I/O—Next-Gen PLC Training. Factoryio.com. [Online]. Available in: https://factoryio.com/
download-archive/. Last accessed 15 Apr 2021
22 G. Orihuela et al.

Graessler, I., Hentze, J.: The new V-Model of VDI 2206 and its validation. Autom. 68(5), 312–324
(2020)
Gutiérrez, M., Iturralde, S.: Basic fundamentals of instrumentation and control, First Edition,
Editorial EUPSE, Ecuador (2017)
Heilingloh, C.S., et al.: Susceptibility of SARS-CoV-2 to UV irradiation. Am. J. Infect. Control,
48(10), 1273–1275 (2020)
Hsu, T.C., et al.: Perspectives on UVC LED: Its progress and application. Photonics 8(6), 196 (2021)
Huayanay, L.: Airborne transmission in closed spaces of SARS-Cov-2. An. Fac. Med. 81(3), 342–
347 (2020)
Kim, D.K., Kim, S.J., Kang, D.H.: Bactericidal effect of 266 to 279 nm wavelength UVC-LEDs
for inactivation of Gram-positive and Gram-negative foodborne pathogenic bacteria and yeasts.
Food Res. Int. 97, 280–287 (2017)
Li, Q., et al.: Early transmission dynamics in Wuhan, China, of novel Coronavirus-infected
pneumonia. N. Engl. J. Med. 382(13), 1199–1207 (2020)
Lu, Z., Wang, Y., Zhang, S., Zhang, K., Shi, Y., Meng, C.: Multi-wave UV-photocatalysis system
(UVA+UVC+VUV/Cu-N-TiO2) for efficient inactivation of microorganisms in ballast water.
Mater. Express 11(9), 1608–1614 (2021)
M. Manuel et al., “Development of a UV-C air sterilizer for the control of airborne transmission of
COVID-19”, Innovación y Desarrollo Tecnológico y Social, vol. 2, num. 2, pp. 167–203, 2020.
Material Mundial: AISI 304 Stainless Steel Technical Sheet, Properties, Density, Hardness.
Materialmundial.com, [Online]. Available: https://www.materialmundial.com/acero-inoxidable-
ss-astm-sae-aisi-304-ficha-tecnica/. Last accessed 15 Sept 2021
May, U.: The risks—know them—avoid them. Scpns.com. [Online]. Available in: http://www.scpns.
com/wp-content/uploads/2020/06/The-Risks-Know-Them-Avoid-Them.pdf. Last accessed 20
Oct 2021
Minamikawa, T. et al.: Inactivation of SARS-CoV-2 by deep ultraviolet light-emitting diode: a
review. Jpn. J. Appl. Phys., 2008 (2021)
Muramoto, Y., Kimura, M., Kondo, A.: Verification of inactivation effect of deep ultraviolet LEDs
on bacteria and viruses, and consideration of effective irradiation methods. Jpn. J. Appl. Phys.
2008 (2021)
Nishisaka-Nonaka, R., et al.: Irradiation by ultraviolet light-emitting diodes inactivate influenza
a virus by inhibiting replication and transcription of viral RNA in host cells. J. Photochem.
Photobiol. B 189, 193–200 (2018)
Nnadi, D.B.N., Araoye, T.O., Egoigwe, S.V., Vincent, D.A.: Application of robotics UV-light device
in averting the spread of Coronavirus. Res Square (2021)
Ou, Y., Petersen, P.M.: Application of ultraviolet light sources for in vivo disinfection. Jpn. J. Appl.
Phys. (2008), 60(10), 100501, (2021)
Raeiszadeh, M., Adeli, B.: A critical review on ultraviolet disinfection systems against COVID-19
outbreak: applicability, validation, and safety considerations. ACS Photonics 7(11), 2941–2951
(2020)
Rattanakul, S., Oguma, K.: Inactivation kinetics and efficiencies of UV-LEDs against Pseudomonas
aeruginosa, Legionella pneumophila, and surrogate microorganisms. Water Res. 130, 31–37
(2018)
Rodríguez-Tobías, H., Morales, G., Rodríguez-Fernández, O., Acuña, P.: Mechanical and
UV-shielding properties ofin situsynthesized poly(acrylonitrile-butadiene-styrene)/zinc oxide
nanocomposites. J. Appl. Polym. Sci. 127(6), 4708–4718 (2013)
Siegling, F.: Conveyor belts, and processing, conveyor belt calculation. 2016. Retrieved from: https://
docplayer.es/11410149-Calculo-de-la-banda-transportadora.html. Last accessed 20 Oct 2021
“Software”: Siemens.com, [Online]. Available in: https://new.siemens.com/mx/es/productos/aut
omatizacion/industry-software/automation-software/tia-portal/software.html. Last accessed 27
Aug 2019
Soliz Torres, M.F., Durango Cordero, J.S., Yépez Fuentes, M.A., Solano Peláez, J.L.: The right
to health in the recycling trade: community actions against COVID-19. Quito, EC: Universidad
Automated Disinfection System for Polyethylene Terephthalate Bottles … 23

Andina Simón Bolívar, Ecuador Headquarters/VLIR-UOS/Zero Waste Campaign, Ecological

Action/Global Alliance for Alternatives to Waste Incineration (2020)
Woo, H., et al.: Efficacy of inactivation of human enteroviruses by dual-wavelength germicidal
ultraviolet (UV-C) light-emitting diodes (LEDs). Water (basel) 11(6), 1–1131 (2019)
Quitral Q, Y.A.: Libraries in the face of the COVID-19 pandemic: fundamentals and actions in Latin
America. Bibl. Univ., (2020). https://doi.org/10.22201/dgb.0187750xp.0.0.992
Design of the CNC Router Structure
for Machining Wood Materials Using
Reliability-Based Design Optimization
Method

Huu Loc Nguyen and Van Thuy Tran

1 Introduction

The uncertainties of the structural system (i.e., loads, materials, dimensions, models,
etc.) in deterministic design optimization are taken into account in an indirect and
subjective method, based on partial safety factors determined in design ranges. There-
fore, the level of reliability can be reduced when using deterministic optimal solutions
(Beck and Santana Gomes 2012; Thuy and Nguyen 2018). The objective of a RBDO
method is to optimize structures, and make sure that a minimum reliability level is
upholded by the optimum structure.
The optimization algorithms of the RBDO method use the reliability methods
to evaluate the probability constraints, and the objective functions are utilized to
prescribe the reliability (Chiralaksanakul and Mahadevan 2005; Royset et al. 2001).
In RBDO methods, the commonly used design parameters are the mean of random
system parameters and solve a mathematical nonlinear programming problem to
determine the optimized cost according to the prescribed probabilistic constraints.
Therefore, the solution from RBDO supplies not only improves the design quality but
also increases the reliability of the design, reduces the production cost and improves
the model more beautiful (Youn and Choi 2004a; Tu et al. 2001).
There are two different methods used to evaluate the RBDO incorporating prob-
abilistic constraints, the performance measure approach (PMA) and the reliability
index approach (RIA) (Youn et al. 2003; Tu et al. 1999; Cheng et al. 2018). According

H. L. Nguyen (B)
Faculty of Mechanical Engineering, Ho Chi Minh City University of Technology (HCMUT), 268
Ly Thuong Kiet Street, District 10, Ho Chi Minh City, Vietnam
e-mail: [email protected]
Vietnam National University Ho Chi Minh City, Linh Trung Ward, Thu Duc City, Ho Chi Minh
City, Vietnam
V. T. Tran
Faculty of Engineering - Technology, Pham Van Dong University, Quaij ng Ngãi, Vietnam

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 25
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_3
26 H. L. Nguyen and V. T. Tran

to the RIA method, the probability constraint is defined as reliability. Nevertheless,

RIA does not converge for some problems or converges slowly. To avoid this short-
coming, PMA is recommended by solving an inverse problem for the first-order
reliability method (FORM) (Youn and Choi 2004b; Enevoldsen and Sørensen 1994;
Padmanabhan et al. 2006).
For reliability analysis, FORM has proven to be effective and widely used. There-
fore, to evaluate the objective functions and the number of probabilistic constraints,
the previous RBDO methods used FORM. Over the years, the structural and engi-
neering disciplines have adopted. Several first-order RBDO methods have been
shown to be more effective than previous RBDO methods (Loc et al. 2019).

2 Theoretical Background

RBDO problem in system parameter design is introduced as follows (Nguyen 2015;

Wang and Ma 2017):
⎧ ⎫
⎪
⎪
X1⎪
⎪
⎪
⎪ ⎪
⎪
⎪ X2⎪⎪
⎪
⎪
⎪ ⎪
⎨. ⎪ ⎬
Find X = (1)
⎪. ⎪
⎪ ⎪
⎪
⎪ ⎪
⎪
⎪. ⎪
⎪ ⎪
⎪
⎪ ⎪
⎩ ⎪ ⎭
Xn

To minimize f (X),
Subject to:
P(g j (X, p) ≥ 0)) ≥ R ∗j ; j = 1, 2, . . . n g
h k (X, m p ) ≤ 0; k = 1, 2, . . . n h
X il ≤ X i ≤ X in ; i = 1, 2, . . . n.
where X i is the design variable (deterministic or random), X is the design variable
vectors (including deterministic and random), p are random parameter vectors (mean
value is mp ), gj (X, p) is the limit state function, f (X) is the objective function, n are
the numbers of design variables, ng are the numbers of probability constraints, Rj *
are the desired reliability and X i l , X i n are the upper and upper limits of the design
variable.
The inverse reliability analysis method is used to transform the RBDO problem
into a deterministic optimization problem. To solve this problem, it is first necessary
to determine the limit state function value gF correspond with the desired reliability
Rj * . The steps are as follows (Fig. 2).
Find the value of limit state function g F correspond with a failure probability F:
Design of the CNC Router Structure for Machining Wood Materials … 27
( )
F = P g(X ) < g F (2)

The new limit state function is obtained after estimation by an approximation

method:

g , (X ) = g(X ) − g F (3)
( )
(Assume the) most probable point (MPP) of a function P g , (X ) < 0 =
P g(X ) < g F is u*, according to the FORM, if the probability of failure F is
known in advance, the index of reliability β is determined as follows:
| |
β = |Φ−1 (F)| (4)

In Fig. 1, the most probable point u* is the point of contact of a circle of radius
gF and the function g(X ) − g F = 0. In other words, it is the point of contact that
determines the shortest distance from the limit state function surface g(X ) − g F = 0
to the origin O in U space.
At the MPP u*:
|| ||
u ∗F = −||u ∗F ||a ∗ = −βa ∗ (5)

In the kth loop:

u k+1 = −βa k {u} (6)

where
( )
∇g u k
a k = || ( k )|| (7)
||∇g u ||

Fig. 1 Find MPP by inverse

reliability analysis
28 H. L. Nguyen and V. T. Tran

Fig. 2 Algorithm to find

MPP according to the inverse
reliability analysis method

Convergence conditions are used to finish the loop:

|| k+1 || || ( ) ( )||
||u − u k || ≤ ε1 v à ||g u k+1 − g u k || ≤ ε2 (8)

After determining the MPP u*, the value of the function g F is determined as
follows:
( )
g F = g u∗ (9)

In summary, the algorithm to find the MPP according to the inverse reliability
analysis method is presented as Fig. 2.
After determining gF , the RBDO problem turns into a deterministic optimization
problem and the first constraint of the RBDO problem (1) become gj (X, p) ≥ gF .

3 Numerical Example

RBDO approach is used to optimize the design of the CNC router structure as shown
in Fig. 3. Supposing that the CNC router structure is composed of four elements
linked together such as the head (1), crossbeam (2), column (3) and the base (4). To
simplify the calculations, it is accepted that the head, crossbeam and column are the
Design of the CNC Router Structure for Machining Wood Materials … 29

Fig. 3 Model of a CNC router structure

cantilever beam correspond with the length L 1 , L 2 , L 3 and the base L 4 which is fixed
by 2 supports.
The applied forces F X1 , F Y 1 , F Z1 and M Z1 are put into the computational model.
During machining, force components F X1 , F Y 1 , F Z1 and torque MZ1 from the cutting
tool are applied to the spindle of the machine body structure (Nguyen and Tran 2022).
If we consider the planar problem, the element (1) is subjected to bending and
compression. The compressive stress in the Z direction is usually small compared
to the bending stress, so it can be ignored. The planar problem with element (1) is
then only subjected to bending and torsion. Elements (2), (3) and (4) are similarly
considered. Therefore, the structural design problem of the machine body is to solve
the problems for elements (1), (2), (3) and (4), respectively, which are considered
as bars and beams. Then, combining the solutions obtained together, we have the
preliminary dimensions of the body. We separate elements (1), (2) and (3) from the
model, convert the model to s simple form and set the force conditions (cutting force
is constant) as shown in Fig. 4.
Preliminary selection of body structure geometry dimensions: The basis for
preliminary selection of the geometric dimensions of the body structure is as follows:
The CNC router on the market were analyzed, and it is found that the working area of
the medium CNC router according to X is less than 1000 mm, Y less than 2000 mm
and Z less than 100 mm. In addition, wood workpiece in sheet form is common, so
the thickness in the Z direction is usually less than 100 mm. On that basis, the dimen-
sions of the machine structure are preliminary selected as follows: L 1 = 250 mm, L 2
= 1000 mm and L 3 = 350 mm.
For the preliminary design of the CNC router body structure, the desired reliability
R* = 0.999 is selected and the RBDO method is used. Two design variables that
represent the height V i and width H i are determined and given in Table 1.
30 H. L. Nguyen and V. T. Tran

Fig. 4 Simple model of elements (1), (2) and (3)

Table 1 Dimensions of the upper and lower limits of the geometry parameters
Elements Parameters Basic values, mm Lower limits Upper limits
Vimin , Himin , mm Vimax, Himax , mm
Head 1 V1 200 150 250
H1 25 20 30
Crossbeam 2 V2 275 225 325
H2 25 20 30
Column 3 V3 350 300 400
H3 35 30 40

Material for all elements is steel C45, yield strength σ y = 360 MPa and stan-
dard deviation S y = 6 MPa. Coefficient of variation of forces and moment is 0.1.
Mechanical properties of steel materials C45 are in Table 2.
The processing material is Goldden oak wood, whose scientific name is Homa-
lium Caryophyllaceum Benth with the density of 1000 kg/m3 , durability at 70 MPa,
hardness at 7.7 HB and humidity of 15%. The components of the applied force are
defined rely on a cutting force in a woodworking process. Components of the applied
force are presented in Table 3 (Wang and Ma 2017).
According to the strength criteria, the limit state function is written as follows:

Table 2 Mechanical properties of steel material C45

C Si Mn S P Cr
0.42–0.5 0.15–0.35 0.50–0.80 ≤ 0.025 ≤ 0.025 0.2–0.4
Design of the CNC Router Structure for Machining Wood Materials … 31

Table 3 Applied forces

Elements Force Force Force Moment Moment Torque Length Li ,
FZi , N FXi , N FYi , N MXi , Nmm MYi , Nmm MZi , mm
Nmm
Head 1 215 390 390 0 0 1170 250
Crossbeam 215 390 390 97,500 97,500 1170 1000
2
Column 3 185 120 196 533 48,843 16,607 350

|( ( ))2 ( )2
( ) FZi 1 1 M Zi
g σiy , σieq = σ y − + 6FXi L i + + 4
Hi Vi Hi Vi2 Vi Hi2 3.07Vi Hi2
(10)

According to the stiffness criteria, the limit state function is written as follows:
|
( )2 ( )2
FXi L i3 FYi L i3
g(u i lim , u i max ) = u i lim − u i max = u i lim − + (11)
3EIYi 3EIXi

The RDBO problem for the elements is presented as follows:

Design variable: Vi , Hi .
Objective function: m i = ρ L i Vi Hi → min.
Subject to:
⎧
⎪
⎪ Vi min ≤ Vi ≤ Vi max
⎨
( ( ) Hi min) ≤ H((
i ≤ Hi max ) ) (12)
⎪
⎪ P g σ , σ > 0 = P σi y − σieq > 0 ≥ Ri∗
⎩ i y ieq
P(g(u i lim , u i max ) > 0) = P((u lim − u i max ) > 0) ≥ Ri∗

where limit displacement uilim = 0.035 mm and desired reliability Ri * = 0.999.

The inverse reliability analysis method is utilized to shift the RBDO problem into
the deterministic optimization problem. After four loops, the function gF at the MPP
has the following value in Table 4.
Thus, the RBDO problem is transformed into the deterministic optimization
problem as follows:
Design variable: Vi , Hi .

Table 4 Calculation results

The limit state Head 1 Crossbeam 2 Column 3
function gi F
giFσ = gi σ (u ∗ ) 0.0327 0.0362 0.0298
giFS = gi S (u ∗ ) −7.65 × −8.21 × −6.82 ×
10−4 10−4 10−4
32 H. L. Nguyen and V. T. Tran

Objective function: m i = ρ L i Vi Hi → min.

Subject to:
⎧
⎪
⎪ Vi min ≤ Vi ≤ Vi max
⎨
( Hi min)≤ Hi ≤ Hi max F (13)
⎪
⎪ g σiy , σieq = σi y − σieq ≥ giσ
⎩
g(u i lim , u i max ) = u i lim − u i max ≥ giFS

The genetic algorithm method in Tool Optimization of MATLAB is utilized to

solve the design optimization problem. The results achieved after optimizing of the
machine head, horizontal beam and vertical column of the CNC router are presented
in Table 5.
Equivalent stresses σeq1 , σeq2 , σeq3 and displacement u1max , u1max , u1max , respec-
tively, of machine head, crossbeam and vertical column of CNC router are all smaller
than the limit strength value σy and displacement ulim .
The MCS method is used to analyze the reliability of the machine structure after
optimal design. The reliability analysis results according to the MCS method are
compared with the desired reliability of the body structure. Simulation results with
sample number N = 5.106 are presented in Figs. 5 and 6.

Table 5 Optimal design results

Elements Mass mi , kg Height Width Stress Maximum displacement uimax , mm
Hi , mm Vi , mm σeqi , MPa
Head 1 13.326 29 230 3.56 0.0199
Crossbeam 2 34.67 28 300 4.22 0.0184
Column 3 33.66 35 353 1.40 0.0195

Fig. 5 Monte Carlo simulation according to the strength criteria: a yield strength σy and stress
σmax and b limit state function g(X)
Design of the CNC Router Structure for Machining Wood Materials … 33

Fig. 6 Monte Carlo simulation according to the stiffness criteria: a displacement umax , ulim and b
limit state function g(X)

The results of the reliability analysis of the MCS method: according to the strength
criteria RMCSσ = 0.99977 and stiffness criteria RMCSu = 0.99973. Because RMCSσ
and RMCSu are more than R* = 0.999, the body structure after optimizing the design
according to RDBO method is satisfactory.
After the complete design, CAD model has the form as shown in Fig. 7a, an
analysis and a test are conducted according to strength and stiffness on the CAE
system as shown in Fig. 7b, c. The results of the test in accordance with the stiffness
and strength criteria is satisfactory.
After the complete design, CAD model has the form as shown in Fig. 7a, an
analysis and a test are conducted in accordance with the stiffness and strength on the
CAE system as shown in Fig. 7b, c. The results of the test in accordance with the
stiffness and strength criteria are satisfactory.

4 Conclusions

This paper applies the RBDO approach to analyze and design CNC router struc-
ture. With desired reliability R*, the inverse reliability analysis method is utilized to
transform the RBDO problem into a deterministic optimization problem.
Genetic Algorithm method in Tool Optimization of MATLAB is utilized to solve
the design optimization problem. The results obtained after design optimization
include equivalent stresses σeq1 , σeq2 , σeq3 and displacement u1max , u1max , u1max ,
respectively, of machine head, crossbeam and vertical column of CNC router are all
smaller than limit strength value σy and displacement ulim .
MCS is used to analyze the machine structure reliability after an optimal design
according to strength and stiffness criteria. The reliability analysis results according
to the MCS method are compared with the desired reliability R* of the body structure.
34 H. L. Nguyen and V. T. Tran

a) b)

c)

Fig. 7 CAD and CAE model

The obtained findings highlighted that body structure after optimizing the design
according to the RDBO method is satisfactory.

Acknowledgements We acknowledge Ho Chi Minh City University of Technology (HCMUT),

VNU-HCM for supporting this study.

References

Beck, A.T., de Santana Gomes, W.J.: A comparison of deterministic, reliability-based and risk-based
structural optimization under uncertainty. Probabilistic Eng. Mech. 28, 18–29 (2012)
Van Thuy, T., Nguyen, H.L.: Investigation on influence of cutting parameters on spindle vibration
of CNC wood milling machine. MATEC Web Conf. 213, 01007 (2018)
Chiralaksanakul, A., Mahadevan, S.: First-order approximation methods in reliability-based design
optimization (2005)
Royset, J., Kiureghian, A.D., Polak, E.: Reliability-based optimal design of series structural systems.
J. Eng. Mech. 127, 607–614 (2001)
Youn, B.D., Choi, K.K.: An investigation of nonlinearity of reliability-based design optimization
approaches. J. Mech. Des. 126, 403–411 (2004a)
Tu, J., Choi, K.K., Park, Y.H.: Design potential method for robust system parameter design. AIAA
J. 39, 667–677 (2001)
Youn, B.D., Choi, K.K., Park, Y.H.: Hybrid analysis method for reliability-based design optimiza-
tion. J. Mech. Des. 125, 221–232 (2003)
Design of the CNC Router Structure for Machining Wood Materials … 35

Tu, J., Choi, K.K., Park, Y.H.: A new study on reliability-based design optimization (1999)
Cheng, Q., Zhao, H., Zhao, Y., Sun, B., Gu, P.: Machining accuracy reliability analysis of multi-axis
machine tool based on Monte Carlo simulation. J. Intell. Manuf. 29, 191–209 (2018)
Youn, B.D., Choi, K.K.: A new response surface methodology for reliability-based design
optimization. Comput. Struct. 82, 241–256 (2004b)
Enevoldsen, I., Sørensen, J.D.: Reliability-based optimization in structural engineering. Struct. Saf.
15, 169–196 (1994)
Padmanabhan, D., Agarwal, H., Renaud, J.E., Batill, S.M.: A study using Monte Carlo simulation for
failure probability calculation in reliability-based optimization. Optim. Eng. 7, 297–316 (2006)
Loc, N.H., Van Thuy, T., Trung, P.Q.: Reliability-based analysis of machine structures using second-
order reliability method. J. Adv. Mech. Des. Syst. Manuf., 13, JAMDSM0063-JAMDSM0063
(2019)
Nguyen, H.L.: Reliability based Design and Analysis of Mechanical Systems, National University
publisher HCMC, Viet Nam (2015)
Wang, G., Ma, Z.: Hybrid particle swarm optimization for first-order reliability method. Comput.
Geotech. 81, 49–58 (2017)
Nguyen, H.L., Tran, V.T.: Applying FCCCD response surface method in studying the cutting power
of wood materials. Solid State Phenom. 329, 25–31 (2022)
Development of Intelligent Drone
for Cassava Farming

Orapadee Joochim, Kridtat Satharanond, and Wirachat Kumkun

1 Introduction

At present, cassava is cultivated in Thailand as it is the main economic crop of

the country. The industrial sector needs the quality cassava to be processed into
various products. Therefore, it is necessary to maintain the cassava to meet the quality
requirements. However, the quality control is difficult because most of the cassava
plantation areas are large. Moreover, at the moment, farmers are still facing the
problem of Cassava Mosaic Disease (CMD), which is spreading in large numbers. It
can spread out in the immediate vicinity quickly. When there is a problem or damage,
the caregiver is unable to figure out the problem and the amount of severity it caused.
Therefore, the actual production quantity is not known whether it is sufficient for
production or not. This paper thus proposes an idea to develop a drone to detect
anomalies by creating a system to detect, find and analyse the abnormalities in cassava
fields and collecting credentials to show the location of the disease on a large diagram.
Presently, the number of cassava plantation business is increasing. To meet both
domestic and international markets, which are more than 9,000,000 rai (1 rai is
equal to 1600 km2 ) at present. In the past, they have faced the crisis of CMD. As a
result, cassava is damaged at a level of 80–100%, wasting both time and productivity.
Hence, the desired profit proportion is not obtained. The problem arises from the use
of diseased cassava cuttings for further planting. Additionally, the whitefly pests are
vector, causing widespread outbreaks quickly. Walking to survey the farmer’s plot is
slow and not thorough. It takes a lot of walking time to explore. However, there are
no studies related to artificial intelligence (AI) for pest inspection of cassava plants
with drone.

O. Joochim (B) · K. Satharanond · W. Kumkun

Institute of Field Robotics, King Mongkut’s University of Technology Thonburi, Bangkok,
Thailand
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 37
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_4
38 O. Joochim et al.

This study aims to help cassava farmers for detecting the abnormalities in plots by
using the developed drone for survey and preliminary analysis of the abnormalities
so that people walk in to check what the malfunction is. The main emphasis is on the
speed of the survey. This will be useful for farmers and general users to access the
location of the abnormal area to shorten the time spent exploring. In this research,
the structure of the drone is designed, developed and equipped with a camera to be
used for surveying. The intelligent detection system to find the abnormal cassava
leaves is studied and implemented using machine learning models.

2 Developed Drone Structure

The drone used for cassava farm surveying is developed in this research. The survey
drone has to bear the burden while surveying the multispectral camera weighing 500 g
and flying for a minimum of 15 min. The 6-bladed multi-rotor drone is designed in
Hex-V format to provide the stability during the flight. There is a system to determine
the coordinates with GPS, and two positions for accuracy in indicating the coordinates
are sent back to the base by Telemetry. The structure is designed to be lightweight
with composite materials. Figure 1 shows the structure of developed drone. The
drone is tested by lifting a load of 5 kg, and it is found that the structure is able to
withstand the forces that occurred without damage. Safety factor is 1.6 (see Fig. 2).

3 Cassava Abnormality Detection

The detection of cassava abnormalities uses artificial intelligence techniques

combined with colour image analysis and enhancement through image processing
where the human vision is important. It is also one of the complex image acquisition
mechanisms. It provides information needed for simple tasks such as object recog-
nition, and for complex tasks such as planning, decision making and researching.
The interest of vision or visual information is image processing which is the process
of dealing with the digital image data with the aim of improving and developing
the visual information in order to be able to interpret to better describe the image.
Image processing process can be divided into three steps. The initial process is image
data acquisition and image enhancement, namely image sharpness correction, image
conversion and image signal recovery. The intermediate process is the isolation of
the elements of an image and the extraction of important features or identities and
the advanced process is the interpretation and image recognition (Chowalit 2016).
For artificial intelligence (AI) techniques, the detection of abnormalities of cassava
leaves from drone photographs using deep learning techniques is applied, which is an
adaptation of the original models with advanced object detection capabilities. In this
Development of Intelligent Drone for Cassava Farming 39

Fig. 1 Structure of developed drone

study, teaching AI to learn is divided into two steps: preparing the data for creating
AI to detect cassava plants and creating AI to detect the location of abnormal cassava
plants.

3.1 Data Set Preparation

Data sets play a key role in teaching AI to learn the unusual locations of cassava
plants. However, because the drone photos are too large and the details of the cassava
40 O. Joochim et al.

Fig. 2 Simulation of drone: a simulation of drone while carrying the weight and b load simulation
to determine safety factor

plants are not clearly visible. Therefore, it is necessary to cut the image into 36 sub-
images before creating a data set. The split images have a filename that represents
the order of the segments from the original image. The format of the file name is
“DJI_XXXX_YY”, where XXXX refers to the numbered sequence of the image
taken by the drone, while YY refers to the part number from the original image (see
Fig. 3). The cropped image is 912 × 608 pixels.

3.2 Image Identification and Amount of Data

To pinpoint for teaching AI to learn what unusual cassava leaves are. Visual Geometry
Group Image Annotator (VIA) is utilized, which is a web platform for generating
the location data within the image data for AI. The programme can create a location
file from the bounding box for using as the sample data for developing AI object
Development of Intelligent Drone for Cassava Farming 41

Fig. 3 Example of segmenting an image into 36 sub-images

Fig. 4 Example of creating the dataset

detection. The amount of data used to teach AI will be adjusted for each version of
the data to study the trend of AI’s ability to detect abnormal locations of cassava.
As a result, the total of 4 versions of the data set is implemented. They are used to
teach AI. Agronomists who specialize in CMD assist in validation of the data set
(see Fig. 4). Cassava abnormality data set is selected (2500 images).

3.3 Creating AI to Detect Cassava (Object Detection)

To create an AI to detect cassava plants, the research project selected 4 AI algorithm

techniques, i.e. Mask R-CNN, EfficientDet, RetinaNet and YOLOv5, for the object
42 O. Joochim et al.

detection and localization of cassava plants. There are advantages in common. The
developers of each technique have taught AI to learn the basic of the object detection
from Big Data; therefore, it can take specific data that would like to detect. Figure 5
shows the example of prediction results of AI algorithm techniques. Figures 6, 7, 8
and 9 illustrate the structures of AI algorithm techniques.

Fig. 5 Example of prediction results: a Mask R-CNN, b RetinaNet and c EfficientDet, d YOLOv5

Fig. 6 Sample of Mask R-CNN structure (Kaiming et al. 2017)

Development of Intelligent Drone for Cassava Farming 43

Fig. 7 Sample of RetinaNet structure (Toan et al. 2019)

Fig. 8 Sample of EfficientDet structure (Mingxing and Adams 2020)

Mask R-CNN is an instance segmentation model that helps us locate pixels

smarter. “Instance Segmentation” means grouping individual objects within a scene
regardless of whether they are of the same type. In addition to specify the class
and bounding box location, it also can display coloured pixels in the corresponding
region (Kaiming et al. 2017).
RetinaNet is able to deal with unequal class problems. For example, there are two
classes, Disease A and Disease B. The labelled figures of Disease A are more than of
Disease B. RetinaNet can solve this problem well and able to use all the information
in learning effectively. Although the amount of data used to learn is small, it still has
good performance (Tsung-Y et al. 2018).
EfficientDet is different from other models. Other models are normally developed
based on the idea that the model has the depth because it allows the model to remember
information well, but EfficientDet is the model developed by the idea that the model
has both depth to remember information well and the ability to extract features of
the data as much as possible, making EfficientDet more efficient than other models.
It also uses less computational resources compared to other models (Mingxing and
Quoc 2019).
YOLOv5 is developed with a focus on real-time object detection regardless of the
version. The higher version is more accurate, making YOLO popular for applications
44 O. Joochim et al.

Fig. 9 Sample of YOLOv5 structure (Renjie et al. 2021)

that require real-time object detection, such as detecting objects from a video camera
and detecting fast moving objects (Raphaël et al. 2021).
By comparing the best results of all models, RetinaNet-50 is the most efficient
model with 0.675 mAP, the highest, and precision and recall at 0.75 and 0.77, respec-
tively. This means that the model can detect irregular cassava leaves with the high
accuracy and can cover a large number of labelled spots. Compared to Mask R-
CNN, which is a benchmark model to compare its performance. The research project
increased the mAP from 0.034 to 0.675, an increase of 0.641 from the original, so it
is chosen as the model to be put into practise.

4 Conclusion

This research develops the drone to help cassava farmers for detecting the area in
plots that may be caused by CMD. The drone used for the cassava farm surveying is
developed. The 6-bladed multi-rotor drone is designed in Hex-V format for providing
the stability during the flight. It can be concluded that the developed drone can
be used to inspect and find the abnormal area that may be caused by CMD. The
intelligent detection system for finding the abnormal cassava leaves is investigated
Development of Intelligent Drone for Cassava Farming 45

and implemented using machine learning models. The results show that RetinaNet-
50 is the best performance with 0.675 mAP, 0.75 precision, 0.774 recall and takes
around 0.06 s to process per image. Comparing with Mask R-CNN benchmark, mAP
is improved up to 0.641 with RetinaNet-50 trained on round 4 data set. The images
are collected, which are 2500 labelled training images successfully.

Acknowledgements This work would not have been possible without the financial support of
National Science and Technology Development Agency (NSTDA), Thailand.

References

Chowalit, K.: Thailand. Plant Leaf Disease Diagnosis System From Color Imagery Using Co-
Occurrence Matrix and Artificial Intelligence Approach. (Master Thesis, Suranaree University
of Technology, Thailand (2016)
Kaiming, H., Georgia, G., Piotr, D., Ross, B.G.: Mask R-CNN, IEEE International Conference on
Computer Vision (ICCV) (2017)
Mingxing, T., Adams, Y.: EfficientDet: Towards Scalable and Efficient Object Detection (Google
Research) (2020)
Mingxing, T., Quoc, V.L.: EfficientNet: Rethinking Model Scaling for Convolutional Neural
Networks. In Proceedings of the 36th International Conference on Machine Learning, Long
Beach, California, PMLR 97 (2019)
Raphaël, C., Hassan, N.N., Ola, S., Abderrahmane, S.: A Deep Learning Object Detection Method
for an Efficient Clusters Initialization (2021). https://arxiv.org/pdf/2104.13634.pdf
Renjie, X., Haifeng, L., Kangjie, L., Lin, C., Yunfei, L.: A Forest Fire Detection System Based on
Ensemble Learning (MDPI Forest) (2021)
Toan, M.H., Phong, H.N., Noi, Q.T., Young, W.L., Kang, R.P.: Deep RetinaNet-Based Detection and
Classification of Road Markings by Visible Light Camera Sensors. MDPI Special Issue “Deep
Learning-Based Image Sensors” (2019)
Tsung-Y, L., Priya, G., Ross, G., Kaiming, H., Piotr, D.: Focal Loss for Dense Object Detection.
Facebook AI Research (FAIR). https://arxiv.org/pdf/1708.02002.pdf (2018)
Calculation and Optimization of the Two
Stage Worm-Gear Reducers Speed Ratio

Le Thuy Anh and Nguyen Huu Loc

1 Introduction

Worm gear drives are used in transmission systems because of their ability to achieve
high reduction ratio with compact size (Rai and Barman 2019). Currently, there
have been a number of studies in the world on worm gear optimization (Mogal and
Wakchaure 2013; Padmanabhan et al. 2013; Godwin Raja et al. 2017; Alexandru
2010). The growing demand for compact, efficient, and reliable reducers forces the
designer to use the optimal design approach.
Rai and Barman (2019) presented the power loss reduction of worm gear drive
(objective function) using simulated annealing (SA), together with design variables
(gear tooth number z2 , helix angle γ , and coefficient of friction) and constraints on
bending stress, defection of worm and linear pressure. The obtained results showed
that the performance of SA is better than that of other approaches used previously
to minimize power loss. The reduction percentage in power loss using SA ranged
from 78.02 to 22.98%. Mogal and Wakchaure (2013) used genetic algorithm (GA) to
optimize the worm using design variables, including gear ratio, face width, and pitch
circle diameters. The obtained results were 17.91% reduction in volume of worm and
worm wheel, 20.34% reduction in center distance between worm and worm wheel,
and 51.05% reduction in deflection of worm. Padmanabhan et al. (Padmanabhan
et al. 2013) built an optimization model for worm gear drives in terms of design
variables m, z2 , and p and design constraints on stress, using ant colony optimiza-
tion (ACO) method and objective functions on capacity, mass, efficiency, and shaft
distance, obtaining problem values with modulus between 4 and 8 mm. Godwin Raja

L. T. Anh · N. H. Loc (B)

Faculty of Mechanical Engineering, Ho Chi Minh City University of Technology (HCMUT), 268
Ly Thuong Kiet Street, District 10, Ho Chi Minh City, Vietnam
e-mail: [email protected]
Vietnam National University Ho Chi Minh City, Linh Trung Ward, Thu Duc City, Ho Chi Minh
City, Vietnam

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 47
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_5
48 L. T. Anh and N. H. Loc

Ebenezer et al. (2017) used optimization algorithms including SA, FA, CS, and GA
by MATLAB solver with design variables as gear ratio, face width, and pitch circle
diameters. The result was a 42.06% reduction in worm drive mass. The design param-
eter values reduced by 13.04% in gear ratio, 35.53% in face width, and increased by
32.97% in pitch circle diameter of worm.
Some other studies also gave similar results on optimizing geometric parameter
values (Alexandru 2010; Dudas 2005; Park et al. 2013; Su and Peng 2008; Chong
et al. 2002), increasing efficiency values (Padmanabhan et al. 2013; Krol and Sokolov
2021), reducing power loss (Rai and Barman 2019; Patil et al. 2019), volume reduc-
tion (Mogal and Wakchaure 2013; Godwin Raja Ebenezer et al. 2017; Patil et al. 2019;
Golabi et al. 2014), low noise (Chul Kim et al. 2020), yet with old and incomplete
formulas. Optimizing the design of a worm drive is complex and time consuming
because of the variety of objectives, design variables, constraints, empirical formulas,
various graphs, and tables (Padmanabhan et al. 2013; Chandrasekaran et al. 2015;
Elkholy and Falah 2015; Marjanovic et al. 2012).
At present, the speed ratio distribution in the two stage worm-gear reducer is
mainly done according to the graph (Miltenović et al. 2017). Based on the pre-
selected parameters and the total speed ratio uh of the reducers, researchers look up
the speed ratio u1 of the fast drive in the graph, then calculate the speed ratio u2 of
the low drive according to the formula uh = u1 u2 .
This method not only has low accuracy due to the need to look up information in
the graph but is also inconvenient because of the necessity to carry lookup documents,
and it is very difficult to program automatic calculations. In addition, studies only
focus on the worm type in the below position, while the worm in the above position
and two stage worm-gear reducers with bevel gear have not had enough research.
In this paper, the speed ratio distribution in a two stage worm-gear reducer with
the worm in the above position, spur gears (straight and helical) and bevel gears is
presented. The calculation is based on the constant condition of the torque on the
worm shaft and the uniform contact strength condition of the drives in the reducer,
satisfying the oil-immersed lubrication conditions and the minimum volume.

2 Fundamental of Gear and Worm Drives Calculations

For well-lubricated gear and worm drives, design calculations are required according
to contact strength. To save manufacturing materials, when designing, engineers
must ensure equal contact strength of worm drive and gear pairs. The problem of
distribution of gear speed ratios of two stage worm-gear reducers to ensure that
uniform contact strength and lubrication conditions with minimization volume is
considered as an optimization problem and is stated as follows
Calculation and Optimization of the Two Stage Worm-Gear Reducers … 49
⎧ ⎫
⎪
⎪ u 12 ⎪⎪
⎨ ⎬
u 34
X= (1)
⎪
⎪ ... ⎪⎪
⎩ ⎭
u n−1,n

Along with constraints:

• Equal contact strength between the drives

σ H 12 ≈ σ H 34 ≈ ... ≈ σ H (n−1),n (2)

• Gear oil immersion lubrication conditions 1 ≥ dd2 ≥ 23 .

4
• The parameters have standard values: gear ratio, gear width, gear width factor,
and diameter factor…
• Constraint volume: minimum volume d 2 = d 4 .
In addition, depending on the goal of the problem, there may be many different
constraints.

2.1 Speed Ratio Distribution in Two Stage Worm-Spur Gear

Reducer

The scheme of the two stage worm-spur gear reducer is shown in Fig. 1: 1–2 is worm
drive, and 3–4 is spur gear pair.

Fig. 1 Kinematic scheme of

two stage worm-spur gear
reducer
50 L. T. Anh and N. H. Loc

For Worm Drive 1–2. According to Nguyen (2020), the center distance of the
worm drive is determined by the formula

( )|( )
q 3 5400 2 T2 K H
aw = 1 + (3)
z2 [σ H ] (q/z 2 )

where q is the diameter factor; z2 is the number of worm gear teeth; aω is the center
distance of worm drive; T 2 is the torque on the worm gear shaft [Nm]; K H is the
contact calculated factor; [σ H ] is the allowable contact stress of the worm gear wheel.
From (3), it is possible to determine the allowable torque on the worm gear shaft
( )3
1 a12 (q/z 2 )[σ H 2 ]2
[T2 ] = · (4)
5400 1 + q/z 2
2 K H2

Substituting a12 = 21 d2 (1 + q/z 2 ) into formula (4), we have

d23 [σ H 2 ]2
[T2 ] = (5)
8 · 54002 · u 12 · tan γ K H 2

where d 2 = mz2 is the pitch circle diameter and the worm wheel; u12 is the speed
ratio of worm drive, and γ is lead angle

z1 z2 q 1
tan γ = = or = (6)
q u 12 · q z2 u 12 · tan γ

For Second Stage of Spur Gear Pair 3–4. For well-lubricated gear pairs, we
calculate by contact strength (Nguyen 2020; ISO 6336-2:2006), and we have
|
2T3 · 103 · K H (u 34 ± 1)
σH 3 = Z M Z H Zε ≤ [σ H 3 ] (7)
u 34 bω34 d12

where Z M is the elasticity factor that takes into account gear material properties
(modulus of elasticity and Poisson’s ratio); Z H is the factor taking into account the
shape of the contact surface; Z ε is the factor taking into account the effect of total
contact length (contact ratio factor); T 3 is the torque on the driving shaft; bω34 is the
width of driving gear 3; d 3 is the pitch diameter of the driving gear 3; and u34 is the
speed ratio of worm drive.
From (7), the allowable torque on the driving gear can be determined by

u 34 bω34 d32 [σ H 3 ]
[T3 ] = (8)
2× 103 · (Z M Z H Z ε )2 K H 3 (u 34 ± 1)

where bw34 = ψ bd34 d 3 is width of spur gear; ψ bd34 , ψ ba34 is a width ratio; and
ψ ba34 is taken from the sequence of standard numbers: 0.1, 0.125, 0.16, 0.2, 0.25,
Calculation and Optimization of the Two Stage Worm-Gear Reducers … 51

0.315, 0.4, 0.5, and 0.63 depending on the wheel position relative to the bearings
ψbd34 = 21 ψba34 (u 34 ± 1).
From (8) and replacing d4 = u 34 · d3 , we have

ψba34 d43 [σ H 3 ]2
[T3 ] = (9)
2K d3 K H 3 u 234
/
where K d = 3 2 × 103 · (Z M Z H Z ε )2 , since the gear material is steel, then calculated
value K d = 756 for straight and K d = 680 for helical spur gears.
To ensure the conditions of equal contact stress, from (5), (9), with u 12 = uu34h , uh
is the speed ratio of the two stage worm-gear reducer, we infer

d43 ψba34 [σ H 3 ]2 K H 2
u 334 = 4 · 54002 · tan γ · uh (10)
d23 K d3 [σ H 2 ]2 K H 3

The obtained formula depends on the following conditions:

1. Worm gear—gear material;
2. The width ratio of the gear drive, type of the spur gear: straight or helical;
3. The relationship between d 2 and d 4 : d2 = d4 or 1 ≥ dd24 ≥ 23
To ensure oil-immersed lubrication conditions, we have the formula
( )3
d 3 ψba34 [σ H 3 ]2 K H 2 3
1≤ u 334 = 4 · 5400 · tan γ 43
2
· uh ≤ (11)
d2 K d3 [σ H 2 ]2 K H 3 2

2.2 Speed Ratio Distribution in Two Stage Worm-Bevel Gear

Reducer

The scheme of the two stage worm-bevel gear reducer is shown in Fig. 2 (1–2 is worm
drive, and 3–4 is bevel gear drive). The contact stress for bevel gears is determined
|
T3 · 103 · K Hβ3 .u 234
σH 3 = 2 · Z M Z H Zε ≤ [σ H 3 ] (12)
0.85 · (1 − 0.5 · ψbe )2 · ψbe · d43

From here, we deduce the formula to determine the allowable torque value

d43 0.85 · (1 − 0.5ψbe )2 · ψbe · [σ H 3 ]2

[T3 ] = (13)
(2 · Z M Z H Z ε ) · 103
2
K Hβ3 · u 234

Or shorten to
52 L. T. Anh and N. H. Loc

Fig. 2 Scheme of two stage worm-bevel gear reduce

d43 0.85 · (1 − 0.5ψbe )2 · ψbe · [σ H 3 ]2

[T3 ] = · (14)
2 · K d3 K Hβ3 · u 234
/
with K d = 3 2 · 103 · (Z M Z H Z ε )2 .
The conditions of equal contact stress [T3 ] = [T2 ]; from (5) and (14), we deduce

d43 0.85 · (1 − 0.5 · ψbe )2 · ψbe · [σ H 3 ]2 d23 [σ H 2 ]2

· = ·
2 · K d3 K Hβ3 · u 234 8 · 54002 · u 12 · tan γ K H2
(15)

Replacing u 12 = uh
u 34
and after simplifying, we get
( )3
d4 K H 2 [σ H 3 ]2
u 334 = · 4 · 54002 · tan γ · 0.85(1 − 0.5 · ψbe )2 ψbe uh
d2 · K d K Hβ3 [σ H 2 ]2
(16)

To ensure oil-immersed lubrication conditions, we choose 1 ≥ d2

d4
≥ 2
3
, and
from there deduce the formula
( )3
d4
1≤ u 334 = · 4 · 54002 · tan γ · 0.85
d2
( )3
K H 2 (1 − 0.5ψbe )2 · ψbe [σ H 3 ]2 3
· uh ≤ (17)
K Hβ3 K d3 [σ H 2 ]2 2

To ensure the minimum volume, d2 = d4 , we should choose the smallest value

u34 in the range of the above gear ratio values.
Calculation and Optimization of the Two Stage Worm-Gear Reducers … 53

3 Calculation Results and Applications

3.1 Speed Ratio Distribution in Two Stage Worm-Spur Gear

Reducer

For worm drive, we choose the lead angle γ = 10◦ (< 30◦ ) and the material for
making worm gear: tinless bronze and brass are used with intermediate slip velocities
(<5 m/s). This bronze has high mechanical strength, but low score resistance, so it
is used together with quenched (>45 HRC) ground and polished worms (Grote and
Antonsson 2009).

[σ H 2 ] = (276 ÷ 300) − 25vs ≈ 200 MPa

Calculated factors K H 2 = K v · K β = 1.1 (Nguyen 2020), where K v is internal

gearing dynamics caused by manufacturing errors (accuracy grade 7, slip velocities
vs = 3 − 7.5 m/s), and K β ≈ 1 is the factor that takes into account the unevenness
of the load distribution in the contact area due to distortion of the worm and wheel
shaft.
For spur gear pairs, material selection for gear train with driving gear being made
of C45 steel, normalization or refining with the hardness of 300HB, σ0H lim = 2 ·
H B + 70= 670 MPa, we select [σ H 3 ] = 603 MPa.
For design, we select K H 3 = K Hβ = 1.01 depending on factor ψbd and hardness
(Grote and Antonsson 2009). The width ratio taken from the sequence of standard
numbers depending on the wheel position relative to the bearings: for symmetrical
arrangement, we select ψba34 = 0.315; ψba34 = 0.4; and ψba34 = 0.5.
The calculation results and the combination of selecting the gear ratio u34 of the
second stage are according to 2 standard ranges (series 1 is the priority series in
bold), and the smallest value should be selected to ensure the smallest volume and
weight as shown in Table 1.
Analysis of the results on Minitab software obtained the dependence of the
minimum gear ratio of the gear pair u34 and the gear pair ratio of the two speed
reducer uh in the form of the quadratic regression equations with testing for lack of
fit (R-square = 99.7%) presented in Table 2.
We draw a graph showing the relationship between speed ratio uh of two stage
worm-spur gear reducer and speed ratio u34 of the second stage to ensure equal
strength, lubrication and minimum volume with respective width ratio in Fig. 3.
54 L. T. Anh and N. H. Loc

Table 1 Calculation results of the distribution of minimum gear speed ratio u34
uh Spur gear Bevel gear
ψba34 = 0.315 ψba34 = 0.4 ψba34 = 0.5
Straight Helical Straight Helical Straight Helical
22.4 1.49 1.66 1.62 1.80 1.74 1.94 1.23
25 1.55 1.72 1.68 1.86 1.81 2.01 1.28
28 1.61 1.79 1.74 1.94 1.88 2.09 1.33
31.5 1.67 1.86 1.81 2.01 1.95 2.17 1.38
35.5 1.74 1.93 1.88 2.10 2.03 2.26 1.44
40 1.81 2.01 1.96 2.18 2.11 2.35 1.50
45 1.88 2.09 2.04 2.27 2.20 2.44 1.56
50 1.95 2.17 2.11 2.35 2.28 2.53 1.61
56 2.03 2.25 2.19 2.44 2.36 2.63 1.68
63 2.11 2.34 2.28 2.54 2.46 2.73 1.74
71 2.19 2.44 2.37 2.64 2.56 2.84 1.81
80 2.28 2.54 2.47 2.75 2.66 2.96 1.89
90 2.37 2.64 2.57 2.86 2.77 3.08 1.96
100 2.46 2.73 2.66 2.96 2.87 3.19 2.03
110 2.54 2.82 2.75 3.05 2.96 3.29 2.10
125 2.65 2.94 2.87 3.19 3.09 3.43 2.19
140 2.75 3.06 2.98 3.31 3.21 3.57 2.27
160 2.87 3.20 3.11 3.46 3.35 3.73 2.38
180 2.99 3.32 3.24 3.60 3.49 3.88 2.47
200 3.10 3.44 3.35 3.73 3.61 4.02 2.56

Table 2 Regression equations determines the minimum gear ratio

Two stage reducer Factor ψ ba Regression equations R-square (%)
Worm-straight spur gear 0.315 u34 = 1.205 + 0.01598u − 99.7
0.000034u2
0.4 u34 = 1.304 + 0.01731u − 99.7
0.000036u2
0.5 u34 = 1.406 + 0.01866u − 99.7
0.000039u2
Worm-helical spur gear 0.315 u34 = 1.337 + 0.01781u − 99.7
0.000038u2
0.4 u34 = 1.452 + 0.01923u − 99.7
0.000040u2
0.5 u34 = 1.567 + 0.02062u − 99.7
0.000043u2
Worm-bevel gear 0.285 u34 = 0.9954 + 0.01323u − 99.7
0.000028u2
Calculation and Optimization of the Two Stage Worm-Gear Reducers … 55

Fig. 3 Speed ratio distribution of two stage worm-spur gear reducer

3.2 Speed Ratio Distribution in Two Stage Worm-Bevel Gear

Reducer

We select worm drive the same as two stage worm-spur gear reducer. For bevel gear,
material selection for gear train with driving gear being made of C45 steel, normal-
ization or refining with the hardness of 300 HB: σ0H lim = 2H B + 70 =670 MPa,
we select [σ H 3 ] = 603 MPa. Preliminary selection of factor K H3 = K Hβ = 1.07.
The calculation results and the combination of selecting the bevel gear ratio u34
of the second stage are according to standard ranges (series 1 is the priority series),
and the smallest value should be selected to ensure the minimum volume and weight
as shown in the last column of Table 1.
The results of the quadratic regression equations showing the dependence of
minimum gear ratio u34 on uh are presented in Table 2.
We construct a graph showing the relationship between speed ratio uh of two stage
worm-bevel gear reducer and speed gear ratio u34 of the second stage to ensure equal
strength, lubrication and minimum volume in Fig. 4

Fig. 4 Speed ratio

distribution of two stage
worm-bevel gear reducer
56 L. T. Anh and N. H. Loc

Fig. 5 Kinematic scheme and front view drawing of two stage worm-spur gear

In this section, the above research results are applied to select the gear ratio and to
design the transmission systems for loads chain system with the kinematic scheme
shown in Fig. 5.
The loads chain system includes: 1—motor; 2—V belt drive; 3—double stage
worm-helical gear reducer; 4—coupling; 5—loads chain sprocket; and 6—loads
chain. Input data for design: force on chain F = 23,000 [N]; chain velocity v = 0.2
[m/s]; pith diameter of loads chain sprocket d = 315 [mm].
According to the power 4.825 [kW], with the common speed ratio of the two stage
reducer u r = 50, we choose the gear ratio of the gear pair according to the priority
set of value u 34 = 2.5. To optimize the product volume of the reducer, we should
choose d2 ≈ d4 , illustrated in drawing Fig. 5.

4 Conclusion

This paper has established an analytical formula for the distribution of speed ratios
in a two stage worm-gear reducer with either spur gear or bevel gear. The research
addresses the combination of three goals, which are to ensure the conditions of equal
contact stress and conditions of oil-immersed lubrication as well as the minimum
volume. Formulas in Table 2 not only allow quick and accurate determination of
speed ratios u12 , u34 according to the total speed ratio uh , but also facilitate program-
ming to design automatically. The paper has constructed a series of speed ratios
and graphs in the two stage worm-gear reducer (Figs. 3 and 4) to look up the gear
ratios corresponding to each gear width ratio value. In addition, we can also extend
this calculation method to apply to another scheme of two stage and triple stage
worm-gear reducers.
Calculation and Optimization of the Two Stage Worm-Gear Reducers … 57

This paper results can be used to design and calculate the distribution of speed
ratios in two stage worm-gear reducers, as a scientific basis for the distribution of
speed ratios for other types of reducers and used as reference when researching, as
well as teaching and learning.

Acknowledgements We acknowledge Ho Chi Minh City University of Technology (HCMUT),

VNU-HCM for supporting this study.

References

Alexandru, A.T.: Worm gears with optimized main geometrical parameters and their efficiency.
Mechanika, Nr.1.81 (2010)
Chandrasekaran, M., Padmanabhan, S., Srinivasa Raman, V.: Swarm intelligence optimization of
worm and worm wheel design. ARPN J. Eng. Appl. Sci. 10(13) (2015). ISSN 1819-6608
Chong, T.H., Bae, I., Park, G.-J: A new and generalized methodology to design multi-stage gear
drives by integrating the dimensional and the configuration design process. Mech. Mach. Theory
37(3), 295–310 (2002)
Chul Kim, S., Gon Moon, S., Hyeon Sohn, J., Jun Park, Y., Ho Choi, C., Ho Lee, G.: Macro
geometry optimization of a helical gear pair for mass efficiency and transmission error. Mech.
Mach. Theory 144 (2020)
Dudas, I.: The Theory and Practice of Worm Gear Drives. British Library Cataloguing in Publication
Data (2005)
Elkholy, A.H., Falah, A.H.: Worm gearing design improvement by considering varying mesh
stiffness. Int. J. Mech. Mechatron. Eng. 9(9)
Godwin Raja Ebenezer, N., Saravanan, R., Ramabalan, S., Navaneethasanthakumar, S.: Worm gear
drive optimization considering industry constraints based on nature inspired algorithms. World
Sci. News (WSN) 87, 205–221 (2017). EISSN 2392-2192
Golabi, S.I., Fesharaki, J.J., Yazdipoor, M.: Gear train optimization based on minimum
volume/weight design. Mech. Mach. Theory 73, 197–217 (2014)
Grote, K.-H., Antonsson, E.: Springer Handbook of Mechanical Engineering (2009)
ISO 6336-2:2006: Calculation of load capacity of spur and helical gears—Part 2: Calculation of
surface durability (pitting) (2006)
Krol, O., Sokolov, V.: Selection of worm gearing optimal structure for machine rotary table.
Diagnostyka 22(1) (2021)
Marjanovic, N., Isailovic, B., Marjanovic, V., Milojevic, Z., Blagojevic, M., Bojic, M.: A practical
approach to the optimization of gear trains with spur gears. Mech. Mach. Theory 53, 1–16 (2012)
Miltenović, A., Banić, M., Miltenović, D.: Load capacity of worm gear transmission from aspect
of maximal use of available resources. MATEC Web Conf. 121 (2017)
Mogal, Y.K., Wakchaure, V.D.: A multiobjective optimization approach for design of worm and
worm wheel based on genetic algorithm. Bonfring Int. J. Man Mach. Interface 3(1) (2013)
Nguyen, H.L.: Fundamentals of Machine Design, 1st edn. Publishing House, Vietnam National
University of Ho Chi Minh City (2020)
Padmanabhan, S., Chandrasekaran, M., Srinivasa Raman, V.: Design optimization of worm gear
drive. Int. J. Min. Metall. Mech. Eng. (IJMMME) 1(1) (2013). ISSN 2320-4060
Park, N., Sohn, J., Baek, G., Oh, C.: Optimal design of worm gear system using in CVVL for
automobiles. In: Proceedings of ASME 2013 International Mechanical Engineering Congress &
Exposition (IMECE 2013), November 15–21, San Diego, CA, USA (2013)
Patil, M.B., Ramkumar, P., Shankar, K.: Multi-objective optimization of two-stage helical gearbox
with tribological constraints. Mech. Mach. Theory 138, 38–57 (2019)
58 L. T. Anh and N. H. Loc

Rai, P., Barman, A.G.: Design optimization of worm gear drive with reduced power loss. IOP Conf.
Ser. Mater. Sci. Eng. 635 (2019)
Su, D., Peng, W.: Optimum design of worm gears with multiple computer aided techniques. Proc.
ICCES 08, 575–658 (2008)
Project-Based Learning in Product
Design Engineering: A Case Study
of External Gear Pump Design

Muhammad Izzuddin Sabaruddin,

Mohammad Habib Fadillah Mohammad Mazni Latiff,
Nur Saidah Awang Abdullah, Wahyu Caesarendra, and Juliana Zaini

1 Introduction

A gear pump is a pump that uses the meshing of two gears to seal an amount of
liquid and transmit it when the gears are moving. There are two types of gear pumps,
namely the external gear pump and the internal gear pump. However, the scope of
this project is to focus on the external gear pump. As part of the Product Design
Engineering (PDE) module.
This project requires research and a survey of the types of gear pump characteris-
tics and functions, such as size, type of gear, etc., that are desirable from a customer
standpoint. From there, the design problem can be understood, and multiple concept
designs and alternative concept designs were generated. Concepts were weighted
using a Pugh matrix, with evaluation criteria such as manufacturability and dura-
bility being one of the important ones. A parametric design for the gear pump was
performed using an Excel spreadsheet to determine an optimal spur gear configura-
tion. A risk assessment was performed on the pump to reduce the overall potential
risk of the pump.
External gear pumps effectively manage solids-free and lubricating fluids and
perform well in sectors such as fuel and food industry as lubricant oil pumps and
high viscosity fluid pumps, respectively. A gear pump has three cycles: filling, transfer
and delivery as presented in Carletti et al. (2016). In the filling stage, the unmeshed
gears expand the pump’s inlet volume, creating a vacuum that allows the fluid to be
pushed in by external pressure. In the transfer stage, the trapped fluid between the
gear teeth and the cavity of the wall casing is transferred from the pump inlet to the
outlet as the gears rotate. Fluid is forced out under pressure in the delivery stage as the

M. I. Sabaruddin · M. H. F. Mohammad Mazni Latiff · N. S. Awang Abdullah ·

W. Caesarendra (B) · J. Zaini
Manufacturing Systems Engineering, Faculty of Integrated Technologies, Universiti Brunei
Darussalam, Jalan Tungku Link, Gadong 1410, Brunei
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 59
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_6
60 M. I. Sabaruddin et al.

Table 1 Component types with their functions

No Components Type Function
1 Seal Standard parts To prevent water leak
2 Cover Special purpose parts To prevent damage to the internal components
3 Drive Shaft Special purpose parts To drive and turn the idler gear
4 Driving Gear Special purpose parts To alter the speed and/or direction generated by the
motor
5 Inlet – For fluid intake
6 Driven Gear Special purpose parts To transfer motion from the driving gear
7 Bushing Standard parts To reduce vibration and friction between moving
parts
8 Casing Special purpose parts To prevent damage to the internal components

gears interlock on the outlet side. A helical, spur and herringbone gear is used in an
external gear pump. Spur gears can handle high differential pressures, whereas helical
gears are silent and reduce trapped fluid. Herringbone gears combine the benefits of
helical and spur gears but are more expensive to manufacture and more prone to fluid
trapping. According to Carletti et al. (2016), the components of external gear pumps
are classified into three categories: standard components, standard assemblies and
special-purpose components. The function of each component is presented in Table
1.

2 Methodology

2.1 Design Problem Formulation

An Engineering Design Specification (EDS) was used to define the design require-
ments of the external gear pump, that is, the requirements of the customer and
the team. In order to know the customer requirements, a survey was created and
distributed to know if the available external gear pump on the market meets the
customer needs and expectations. According to the survey, most of the respondents
are students, engineers and practitioners, general assistants and self-employed, and
their preferred external gear pumps are the following: medium size, lightweight,
medium water displacement, complex build, low noise and low vibration. These
functions are quantified in Table 2 and are directly related to the team method to
meet these requirements using the House of Quality (HoQ) in Fig. 1. The HoQ
shows that the type of gear plays the most crucial factor in achieving the customer’s
requirements, followed by the cost of production, lifespan and casing thickness.
In this study, five alternative design concepts were proposed in Table 3. These
alternative design concepts were then evaluated based on a few criteria such as
Project-Based Learning in Product Design Engineering: A Case Study … 61

Table 2 Product function and engineering characteristic of the external gear pump
Function Engineering Units Limits
characteristics
Medium size Dimension of overall Cubic centimetres The total height, width
components and depth is at least 9 cm
by 13 cm by 24 cm
Lightweight Weight of all Gram 100–300 g
components
Medium water Displacement Cubic centimetres/rev 20 cc/rev
displacement
Complex Components’ design
assembly
Regular services Casing’s design
Low noise Number of teeth Have more than 9 teeth

manufacturability, durability, rigidity, functionality, efficiency, lightweight and low

cost. A weight is given to quantitatively measure the alternative design concepts as
presented in Table 4. The weight is based on the customer survey results and potential
user surveys. The alternative design concept with the highest weight is selected as
the final concept design.

3 Results and Discussion

3.1 Gear Parametric Design

After the design concept is made, the gear design needs to be considered for the
performance. The team decided on the design using a gear calculator in Excel to
calculate the gear parameters such as the number of teeth, pitch diameter and gear
face width. After careful deliberation, the pitch diameter was set to 5.5 cm, the
pressure angle was set to 20° to reduce the undercut, and the number of teeth was set
to 18 to pair with the pressure angle. The gear parameters set were set into an Excel
file where it calculated the theoretical displacement value of 17.12 cm3 /rev.
The gear parameter value was tested with different slip coefficient values and
different pressure values which in turn affect the volumetric efficiency, actual flow
and the time it takes to fill a 5-L tank. From the simulated values, with increasing
pressure and a higher value of slip flow coefficient, the lower the flow value, and thus
the time spent to fill a 5-L tank increases.
62 M. I. Sabaruddin et al.

Fig. 1 HoQ for the team’s external gear pump

Project-Based Learning in Product Design Engineering: A Case Study … 63

Table 3 Alternative design

Design Standard Concepts
concepts
1 Physical principle Spur gear
Material PLA
Geometry Box shaped with rounded edges
2 Physical principle Spur gear
Material PETG
Geometry Box shaped
3 Physical principle Spur gear
Material PLA with transparent acrylic cover
Geometry Box shaped
4 Physical principle Herringbone gear
Material PLA
Geometry Box shaped
5 Physical principle Herringbone gear
Material PETG
Geometry Pill shaped

3.2 Final Concept Design

Figure 2 illustrates the final concept design based on the evaluation of Concept #3
and the parametric design selected. The design was created using SOLIDWORKS
2020 and Gear DXF.

3.3 Failure Mode Analysis

Failure cannot be permanently eliminated in the manufacturing industry. However,

it can be reduced using an analysis called failure mode and effects analysis, FMEA,
by calculating the Risk Priority Number (RPN). This RPN is based on three factors:
severity (S), occurrence (O), and detection (D) (Bhattacharjee et al. 2020). Each
factor has a rating of 10, 1 is the least and 10 is the maximum. Equation (1) is the
formula to find the RPN value. Hence, the range value for the RPN value is between
1 and 1000.

RPN = S × O × D. (1)

A thorough discussion was conducted among the team members to identify the
possibility of failure modes that can happen to the external gear pump. The RPN
rating of the external gear pump is presented in Table 5. A descriptive of the RPN
rating for the following failure mode: (1) External gear pump leaked, (2) External
gear pump severe pressure fluctuation and (3) External gear pump power loss.
To reduce the failure mode, the following is the recommended action for each
failure mode:
 64

Table 4 Final evaluation of alternative design concepts

Evaluation criteria Importance Concepts
Weight (%) Concept #1 Concept #2 Concept #3 Concept #4 Concept #5
Rating Weighted rating Rating Weighted rating Rating Weighted Rating Weighted Rating Weighted
rating rating rating
Manufacturability 15 3 0.45 3 0.45 2 0.3 2 0.3 2 0.3
Durability 10 2 0.2 3 0.3 3 0.3 2 0.2 3 0.3
Rigidity 15 2 0.3 3 0.45 4 0.6 2 0.3 3 0.45
Functionality 15 3 0.45 3 0.45 3 0.45 3 0.45 3 0.45
Efficiency 20 3 0.6 3 0.6 4 0.8 4 0.8 4 0.8
Lightweight 10 4 0.4 2 0.2 2 0.2 3 0.3 2 0.2
Low cost 15 3 0.45 2 0.3 2 0.3 2 0.3 2 0.3
Total percentage 100 2.85 2.75 2.95 2.65 2.8
M. I. Sabaruddin et al.
Project-Based Learning in Product Design Engineering: A Case Study … 65

Fig. 2 Final concept design

Table 5 RPN rating

Failure mode Severity (S) Occurrence (O) Detection (D) RPN
Effects S rating Causes O rating Control tests D rating
External gear Improper 3 Improper 6 Very easy 1 18
pump leaked fluid transfer seal inside
and spills the external
gear pump
Cracks
External gear Fluid 5 Cavitation 7 Likely 4 140
pump severe transfers
pressure fluctuate
fluctuation
External gear Improper 3 Motor 1 None 5 15
pump power shaft RPM speed
loss fluctuation

1. External gear pump leaked

a. Proper installation of the gasket
2. External gear pump severe pressure fluctuation
a. Install pressure gauge
3. External gear pump loss of power
a. Automatic tapping gearbox
b. RPM meter on the shaft.
66 M. I. Sabaruddin et al.

4 Conclusions

To conclude the project, proper and strategical planning from the lab practical helps
team members solve the problems encountered during planning and studying, espe-
cially with assistance from the instructor. From the class lecture, combined with the
multiple labs practical, the team member was able to identify the possible problem
that can occur in the manufacturing process. The team also developed another skill,
creating the possible survey question using Google Forms services elicited from
developing an external gear pump. With the second wave of the pandemic affecting
Brunei, various limitations arise, one of which is that the campus is not accessible to
students, making the fabrication of the project difficult. Moreover, the limited supply
of material, such as PLA puts a constraint on team members.

Acknowledgements The first, second and third Authors would like to thank to Dr. Wahyu
Caesarendra as a lecturer for Product Design Engineering (PDE) module and Dr. Juliana Zaini
as a lecturer for Computer Aided Design (CAD) module.

References

Bhattacharjee, P., Dey, V., Mandal, U.K.: Saf. Sci. 132, 104967 (2020)
Carletti, E., Miccoli, G., Pedrielli, F., Parise, G.: Arch. Acoust. 41, 285 (2016)
Information on https://www.michael-smith-engineers.co.uk/resources/useful-info/external-gear-
pumps. Accessed on 9 Jan 2022
Information on https://dienerprecisionpumps.com/external-gear-pumps/. Accessed on 8 March
2022
Information on https://www.engineersedge.com/pumps/helix_spur_gear_pump.htm. Accessed on
8 March 2022
Machining and Manufacturing
Tool Wear Analysis on Time-Domain
and Frequency-Domain Data
of Machining S45C Using Signal
Processing Technique

P. Jirapipattanaporn, A. Chanpariyavatevong, W. Lawanont,

and W. Boongsood

1 Introduction

Nowadays, the machining process is important in many industries and engineering

applications. Because the machining process is the first for preparing materials for the
next steps. Milling is one machining process is widely used in industrial (Al-Zubaidi
et al. 2011). Due to high material removal rate, it ensures accurate cuts. And milling
is perfect for producing individual pieces in small or large batches (Liu et al. 2019).
However, the milling process is necessary to use a cutting tool to cut material. Simul-
taneously, tool wear is usually occurs in the machining process. And tool wear is one
of the main reasons leading to quality of the workpiece, machine power consump-
tion, and production cost (Siddhpura and Paurobally 2013). Commonly, tool wear
investigation can be classified into two categories. The direct method is observa-
tion by optical microscope or scanning electron microscope (Čerče et al. 2015). The
advantage of the direct method is that it can record geometric changes or degrees of
cutting tool. However, the equipment for this method is expensive and fragile. More-
over, to track the process, it is also necessary to stop the machine, in which result in
delays (Snr 2000). And on the word of Ambhore et al. (Ambhore et al. 2015) “Direct

P. Jirapipattanaporn (B) · A. Chanpariyavatevong · W. Lawanont (B) · W. Boongsood

School of Manufacturing Engineering, Suranaree University of Technology, Nakhon Ratchasima,
Thailand
e-mail: [email protected]
W. Lawanont
e-mail: [email protected]
A. Chanpariyavatevong
e-mail: [email protected]
W. Boongsood
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 69
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_7
70 P. Jirapipattanaporn et al.

methods are rather less beneficial on largely cutting tools because the visual inspec-
tion equipment has limitations in area of inspection.” Indirect method observes a rela-
tion between related parameters, such as cutting force during machining process by
a load cell, vibration measured by dynamometers, and machine power consumption.
García-Ordás et al. (García-Ordás et al. 2017) comment on direct method methods is
constrained to use expensive equipment or high-performance computer to analyze.
However, the methods mentioned above are complex and difficult, complicated,
and expensive equipment. For example, the measuring tool is difficult in real situa-
tions by taking pictures. This is difficult because setup systems are time-consuming.
In the world of work, a professional technician can able to identify changes in tools
during machining (Chen et al. 2009).
Therefore, the sound signal is one of alternative methods to investigate the defor-
mation degree of tool wear during the machining process. Weller et al. (Li et al. 2019)
observed changing of cutting tool in machining the 6061 aluminum. It was found that
a degree of tool wear is related to time–frequency-domain data. The study mentioned
that different frequency ranges are affected by the degree of tool wear. Moreover,
they investigated degree of tool wear in machining AISI 1045 steel by using sonic
signals (Weller and Schrier 1969). After the 2000s, there are many studies published
online that are related to sound and signal processing in the machining process
(Vaseghi 2008). Most of the studies used a microphone connected to a computer or
data acquisition device to receive audio signals and proceed with data analysis. In
Kothuru et al. (2018), the study investigated the sound of end mill during machining
6061 aluminum. The study used a microphone, which was connected to computer,
and converted the sound signal to frequency spectrums of pre-processed data samples.
Likewise, Salgado and Alonso (Salgado and Alonso 2007) observed the sound of
cutting AISI 1040 steel. They carried out the study by using a microphone to record
sound and used singular a spectrum process for the analysis. Juan Emilio Noriega-
Linares (Noriega-Linares and Navarro Ruiz 2016) developed a prototype for a low-
cost acoustic sensor based on the Raspberry Pi platform to analyze and monitor envi-
ronmental sound. Another use case of Raspberry Pi was from Mitrović et al. (2017)
that proposes a solution for water flow sound classification with the Raspberry Pi
platform. From the aforementioned studies, Raspberry Pi met our requirements in
many aspects whether to its ease of use to programming with Python and its ability
to control the sampling rate and time of the recording sound to be the same in each
experiment.
In this study, we proposed a milling process monitoring method using acoustic
signal during machining S45C steel plate due to widely used in many industries
(İpekoğlu et al. 2019) related to mechanical parts with high-speed steel end mills
under no cooling. We adopted signal processing (Rioul and Vetterli 1991) technique
to analyze milling sound. Sound of the milling process was recorded using a micro-
phone array which was controlled by Raspberry Pi. The organization of this study
starts with the introduction, methodology, experiment and results, and conclusion
and discussion, respectively.
Tool Wear Analysis on Time-Domain and Frequency-Domain Data … 71

2 Methodology

Overall. This study conducted an experimented on several scenarios. The purpose

of the experiment was for the data collection process. This section discussed the
material used in the experiment as well as the experiment parameters. Later in the
section, data collection devices and data visualization techniques are explained.
Tool and Materials. This study performed the experimented by using the three-
axis computer numerical control (CNC) in face milling process. The cutting tool was
an uncoating high-speed steel end mills. The details of the cutting tool are shown
in Table 1. The experiment was conducted with fixed cutting parameters under no
cooling. Table 2 shows the parameters used in the experiment. S45C steel plate with
a thickness of 15 mm and the dimension of 70 (W) × 300 (H) mm was used in
the experiment. After that, we observed the tool wear for flank wear following ISO
8688-2:1989 using a light microscope with ×5 magnification and a scanning electron
microscope.
Data Collection. This section discussed instruments and methods for collecting
data of milling sound. The experiment was performed according to cutting parameters
shown in Table 2. Face milling processing was carried out by varying the process
time starting from 1 to 6 min, respectively. During this experiment, new end mill was
replaced in every experiment. This resulted in a total six experiments with each of
them differing in process time.
For the sound recording device, a Re-speaker 2-Mics Pi Hat microphone array
was used to record the milling process sound produced from the CNC machine. The
microphone array was programmed to control the recording time and sampling rate
of the recorded sound. In every experiment, the sound was recorded at the sampling
of 22,050 Hz. The microphone array was set up 40 cm away from the front of the
workpiece while maintaining the same level of the workpiece as shown in Fig. 1.
The device starts recording the ambient sound for 10 s. Then, the CNC machine’s
spindle starts, during this period, it takes about 15 s before the tool start milling
the workpiece. During the milling process, the microphone continues to record the
sound until the milling process is finished, which will take about 1 min. After that,
the device stops the recording 10 s after the milling process is completed. When the

Table 1 Details of HSS end

Material high-speed steel
mill
Dimeter 10 mm
Cutting edge 2 Flute
Cut length 25 mm

Table 2 Cutting parameter

Cutting speed (rpm) Feed rate (mm/min) Depth of cut (mm)
800 90 0.5
72 P. Jirapipattanaporn et al.

Fig. 1 Work area and setting

of a data collection device

recording is done sound data will be stored in WAV file type, and this file was used
later for the data visualization.
Data Visualization. In this study, Python programming was implemented to
generate the graphs for the data visualization. Graph that we considered in this
work is time-domain and frequency-domain (Stein and Shakarchi 2011) graph. The
main library that we are using to import sound data and generate the digital signal
is Librosa (McFee et al. 2015). First, we generated the time-domain graph of every
milling process sound we recorded to see what the sound wave looks like in each
condition of welding. This can be considered as the raw data of the recorded sound.
To emphasize, time-domain data can be misleading as this type of data only shows
the magnitude of the recorded data and not the whole frequency distribution of the
data.
To extract more data from the time-domain data, a frequency-domain graph was
generated to see the difference between the range of frequency that occurred in
each milling condition. To generate the frequency domain, the imported sound data
need to be passed through Fourier transform (Vaseghi 2008) algorithm. This was
processed locally on the data collection device with The Python programming tech-
nique. Fourier transform (Vaseghi 2008) is the method to convert data from the
time-domain function to the frequency-domain function.

3 Result and Discussion

From the experiment, milling S45C with no cooling, the study found that the end
mill showed signs of wear in flank wear as shown in Fig. 2a. This relates to the
assumption that heat generated and hard particle in S45C steel is the main effect
of rapid tool wear and shortened the tool life span (Kataoka and Shamoto 2019).
Figure 2b shows that the cutting tool used in the experiment causes tool wear over
time, when the machining time increased tool wear increased.
Tool Wear Analysis on Time-Domain and Frequency-Domain Data … 73

700

600

Tool Wear (µ m)
500

400

300

200
0 1 2 3 4 5 6 7
Time (min)

(a) Tool wear in milling s45c (b) Relationship between tool wear
and time

Fig. 2 a Tool wear in milling s45c and b relationship between tool wear and time

After visualizing all the milling process sound data with the time-domain graph,
we found that each time-domain graph in every experiment had the same pattern as
shown in Fig. 3. The time-domain graph can indicate the occurred milling process
procedure in each period based on the magnitude of the graph. However, based on
the same Fig. 3, it was not possible to identify the relation between the waveform
and the tool wear in the experiment because there is too many data points in the
time-domain graph.
To identify the relationship between the sound of the milling process and tool
wear. We further investigate the sound in the milling period that has occurred by
generating the frequency-domain graph using the original time-domain data. After
generating the frequency-domain graph, it was found that the tool wear is related
to some frequency range shown in Fig. 4, which is the frequency domain of one of

Fig. 3 Time-domain graph of the milling process

74 P. Jirapipattanaporn et al.

Fig. 4 Frequency domain of the milling process

our experiments. Every frequency-domain graph shows the same pattern in every
experiment. As seen in Fig. 4, we found that the frequency in the range between 0
and 3000 Hz has a similar magnitude and pattern occurred.
However, after investigating all the frequency-domain graphs, there are some
frequency ranges that the magnitude was changed according to the tool wear. To
emphasize that results, we further analyzed the frequency range between 3000 and
4000 Hz by enlarging the graph. As a result, Fig. 5 shows the frequency domain of one
of our experiments that had a significant difference between the frequency magnitude
that occurred in each tool wear condition. The experiment that we selected to show
is sound after milling with the cutting tool for 1 min, 4 min, and 6 min as shows in
Fig. 5a–c, respectively. Each figure shows the image of tool wear compared with the
frequency domain in a focused frequency range. After comparing all the frequency-
domain graphs, it was found that there are two frequency bands that its magnitude
changed significantly according to the tool wear. These two frequency bands are
3350 and 3755 Hz. The magnitude of these two frequency bands increased after the
tool went through the milling process for a prolonged period. And the magnitude
increased significantly compared to the tool wear change.

4 Conclusion

In this study, signal processing techniques have been applied for the milling process
to monitor the wear of a cutting tool by milling sound. All milling sounds are recorded
with a microphone array, and the data was visualized using signal processing tech-
nique to transform the time-domain data into frequency-domain data. It was found
in this study that each frequency band can indicate tool wear increased and used
Tool Wear Analysis on Time-Domain and Frequency-Domain Data … 75

(a) after 1 minutes of milling

(b) after 4 minutes of milling

(c) after 6 minutes of milling

Fig. 5 Frequency-domain graph and tool wear

for monitoring the milling process. The frequency band from 0 to 3000 Hz of every
milling condition has a similar magnitude and pattern on the graph. Therefore, we
can conclude that these frequency bands are the normal frequency band of the milling
process. Furthermore, two frequency bands have a relationship with the tool wear
over time. From the graph, the magnitude of the frequency band 3350 and 3755 Hz
increased as the tool wear level increased. Thus, this can be summarized that we are
can track 3350 and 3755 Hz frequencies to monitor the wear of the cutting tool.
76 P. Jirapipattanaporn et al.

Finally, by identifying the frequency band relate to the tool wear level, the results
can be further used as a foundation to develop a smarter monitoring system with
the implementation of machine learning or artificial intelligence technology. Imple-
menting such technologies on monitoring the milling process in this suggested
frequency band for tool wear level would greatly reduce delays in manufacturing,
resulting in cost reduction of the process.

References

Al-Zubaidi, S., Ghani, J.A., Che Haron, C.H.: Application of ANN in milling process: a review.
Model. Simul. Eng. 2011 (2011)
Ambhore, N., et al.: Tool condition monitoring system: a review. Mater. Today: Proc. 2(4–5),
3419–3428 (2015)
Čerče, L., Pušavec, F., Kopač, J.: A new approach to spatial tool wear analysis and monitoring.
Strojniški Vestnik-J. Mech. Eng. 61(9), 9 (2015)
Chen, T.-H., et al.: Study of sound signal for tool wear monitoring system in micro-milling processes.
In: International Manufacturing Science and Engineering Conference (2009)
García-Ordás, M.T., et al.: A computer vision approach to analyze and classify tool wear level in
milling processes using shape descriptors and machine learning techniques. Int. J. Adv. Manuf.
Technol. 90(5), 1947–1961 (2017)
İpekoğlu, G., et al.: Investigation of microstructure and mechanical properties of friction stir welded
dissimilar St37/St52 joints. Mater. Res. Express 6, 046537 (2019)
Kataoka, R., Shamoto, E.: Influence of vibration in cutting on tool flank wear: fundamental study
by conducting a cutting experiment with forced vibration in the depth-of-cut direction. Precis.
Eng. 55, 322–329 (2019). LNCS Homepage, http://www.springer.com/lncs. Last accessed 21
Nov 2016
Kothuru, A., Nooka, S.P., Liu, R.: Application of audible sound signals for tool wear monitoring
using machine learning techniques in end milling. Int. J. Adv. Manuf. Technol. 95(9), 3797–3808
(2018)
Li, Z., Liu, R., Wu, D.: Data-driven smart manufacturing: tool wear monitoring with audio signals
and machine learning. J. Manuf. Process. 48, 66–76 (2019)
Liu, M.-K., Tseng, Y.-H., Tran, M.-Q.: Tool wear monitoring and prediction based on sound signal.
Int. J. Adv. Manuf. Technol. 103 (2019)
McFee, B., et al.: Librosa: audio and music signal analysis in python. In: Proceedings of the 14th
Python in Science Conference (2015)
Mitrović, A., Djukanović, S., Radonjić, M.: Implementation of signal classification using frequency
spectrum features on the raspberry Pi platform. In: 2017 25th Telecommunication Forum
(TELFOR). IEEE (2017)
Noriega-Linares, J.E., Navarro Ruiz, J.M.: On the application of the raspberry Pi as an advanced
acoustic sensor network for noise monitoring. Electronics 5(4), 74 (2016)
Rioul, O., Vetterli, M.: Wavelets and signal processing. IEEE Signal Process. Mag. 8(4), 14–38
(1991)
Salgado, D., Alonso, F.: An approach based on current and sound signals for in-process tool wear
monitoring. Int. J. Mach. Tools Manuf. 47(14), 2140–2152 (2007)
Siddhpura, A., Paurobally, R.: A review of flank wear prediction methods for tool condition
monitoring in a turning process. Int. J. Adv. Manuf. Technol. 65(1–4), 371–393 (2013)
Snr, D.E.D.: Sensor signals for tool-wear monitoring in metal cutting operations—a review of
methods. Int. J. Mach. Tools Manuf. 40(8), 1073–1098 (2000)
Tool Wear Analysis on Time-Domain and Frequency-Domain Data … 77

Stein, E.M., Shakarchi, R.: Fourier Analysis: An Introduction, vol. 1. Princeton University Press
(2011)
Vaseghi, S.V.: Advanced Digital Signal Processing and Noise Reduction. Wiley (2008)
Weller, E., Schrier, H.M., Weichbrodt, B.: What sound can be expected from a worn tool?
Optimization of the Weight and Size
Characteristics of the Power Bracket
for Additive Manufacturing Based
on Topological Optimization Algorithms

A. V. Zhdanov, S. A. Baranov, and L. V. Belyaev

1 Introduction

Currently, one of the intensively developing technological methods of manufacturing

products is the method of selective laser melting. This method has a number of
technological advantages over traditional methods of material processing, allowing
you to obtain products of almost any geometric complexity from a large range of
materials (Sadeghilaridjani 2021; King et al. 2015). However, it is not economically
feasible to manufacture products whose geometry is not adapted to the peculiarities
of additive manufacturing (Bajaj et al. 2020).
One of the methods of such adaptation is the use of topology optimization.
Topology optimization is a mathematical tool that allows you to optimize the distribu-
tion of material in a part at the design stage. There are various topology optimization
algorithms developed over the past three decades. For instance, set level approach,
homogenization method, SIMP method and density approach are among the main
topology optimization techniques (Rozvany 2009; Sigmund and Maute 2013).
The purpose of this article is to investigate the possibility of optimizing the weight
and size characteristics for additive manufacturing based on the selected algorithm
of topological optimization.

2 Materials and Methods

The object of the study was a bracket used as a swivel mount and experiencing
heavy loads, with overall dimensions of 176 × 170 × 95 mm (Fig. 1), made of
EN-AW 2024 material and weighing 1.75 kgf. EN-AW 2024 alloy is a typical 2xxx

A. V. Zhdanov (B) · S. A. Baranov · L. V. Belyaev

Vladimir State University, 87 Gorky Street, Vladimir 600000, Russia
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 79
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_8
80 A. V. Zhdanov et al.

series (Al–Cu–Mg) wrought Al alloy, and it has the advantages of low density, high
specific strength, excellent fatigue resistance and good machinability (Chen et al.
2019; Zhang et al. 2019; Bonfils-Lahovary et al. 2024). The EN-AW 2024 is resilient
material with increased strength values, which is mainly used in the aerospace and
automotive industries. The properties of the EN-AW 2024 material and its chemical
composition are provided in Tables 1 and 2.
The use of aluminum is justified by the fact that the design of the bracket should be
lightweight and at the same time withstand the loads applied to it. At the same time,

Fig. 1 3D-model of power bracket

Table 1 Physical and mechanical properties of the material EN-AW 2024

Parameter Value
Modulus of elasticity, MPa 70,000
Poisson’s ratio 0.3897
Shear modulus, MPa 27,000
Density, kg/m3 2780
Tensile strength, MPa 440
Yield strength, MPa 290

Table 2 Chemical composition of the material EN-AW 2024 [% weight]

Si Fe Cu Mn Mg Cr Zn Ti Other Al
≤0.5 ≤0.5 3.8–4.9 0.3–0.9 1.2–1.8 ≤0.1 ≤0.25 ≤0.15 ≤0.15 90.7–94.7
Optimization of the Weight and Size Characteristics of the Power … 81

the use of this material imposes a number of significant restrictions on the physical
characteristics of the product. These include lack of load capacity, high probability
of fatigue failure and temperature resistance. Nevertheless, it is worth pointing out
that in the process of optimization, fine-tuning of the design is possible, taking into
account all complicating factors.
The original design of the bracket is an integral structure of complex geometric
shape, consisting of several holes and stiffeners. Fixing the bracket to the surface
of the fit is carried out through the lower plane and the corresponding hole in it.
The fastening of the mating part of the moving structure is carried out through holes
having a cylindrical shape.
The mechanical and physical properties of the material used are listed below.
Topological optimization was carried out in the SOLID WORKS 2020 software
product (Dassault Systems, France). This software product uses the SIMP method
for topology optimization (SOLID WORKS 2020).
The optimization conditions were as follows:
• the optimized geometry must fit within the original part envelope; minimum
material feature size (wall thickness or central rib): 18.4 mm;
• the part was fixed through holes in the base;
• the load was 10 kN and applied at an angle of 90° to the base plane (horizontal
plane);
• the aim of the study is to reduce the weight by 60%.
To determine the effectiveness of the topological optimization process, the stress–
strain state of the initial and obtained geometries was evaluated under the same
loading conditions.

3 Results and Discussion

As a result of the topological optimization of the initial geometry of the product

based on the finite element method, a new geometry of the bracket in question was
obtained (Fig. 2).
In the process of topology optimization, the product was transformed and geomet-
rically underwent some complications. Thus, the resulting geometry had a less mono-
lithic base, but an increased number of support points. Geometric cutouts were formed
at the base of the product to reduce the weight of the entire structure. The stiffeners
were also changed in the direction of complicating the geometry, but reducing the
total mass and area.
The analysis of the stress–strain state of the part allowed us to obtain pictures
of the von Mises stress distribution, deformations and the safety factor, which are
shown in Figs. 3 and 4.
Stress fields have shown that the greatest load falls on the central stiffener of the
product at the base. Nevertheless, these stresses are moderate values that are not
critical for the applied loads and the material used.
82 A. V. Zhdanov et al.

Fig. 2 Topological transformation of geometry

The von Misses stresses values were reduced and now they are about 133 MPa
on average, over the entire length of the edge. The graph of rib loading along the
length shows a sharp decrease in stresses in the area from 50 to 80 mm. This is due
to the distribution of the load over the newly formed geometry, which allows us to
delegate the load along the entire length of the edge. In addition, the graph shows
that the average values of the emerging stresses do not exceed 140 MPa. This is a
significant improvement.
The maximum values of these parameters are given in Tables 3 and 4.
A comparison of the results obtained from the tables above shows the possibility
of reducing the mass of the part without losing its mechanical properties.
The reduction of weight saves huge amount of material and processing energy
thus huge amount of money. It also shows that the capability of topology optimization
can be fully utilized with additive manufacturing techniques, as the manufacturing
constraints in the conventional methods are no longer available. As a result of the
work, the weight of the product decreased by 31.5%, and the average values of stresses
Optimization of the Weight and Size Characteristics of the Power … 83

(a) (b)

MPa
500
400
300
200

100
0
0 20 40 60 80 mm
(c) (d)

Fig. 3 Diagrams of stresses (a), deformations (b), safety factor (c), and stress distribution graph
along the edge length (d) for the initial geometry of the product

MPa
140
130
120
110
100
90
80
0 20 40 60 80 mm

(a) (b)

Fig. 4 Stress plots (a) and stress distribution graph along the edge length (b)
84 A. V. Zhdanov et al.

Table 3 Results of static

Parameter Value
analysis of the model without
topological changes Von Mises stresses, MPa 117.858
Deformation, ESTRN 0.002
Safety margin 2.461
Weight, kgf 1.180

Table 4 Results of static

Von Mises stress, MPa Deformation, ESTRN Safety margin
analysis of the model with
topological changes 113.416 0.001 2.174

arising during the load decreased by almost a half. The results require practical
confirmation in working conditions, which is the next stage of the study.

4 Conclusion

From the case study, it can be concluded that topology optimization is a powerful
design concept to reduce the weight of structural products. The reduction of weight
saves huge amount of material and processing energy thus huge amount of money.
It also shows that the capability of topology optimization can be fully utilized with
additive manufacturing techniques, as the manufacturing constraints in the conven-
tional methods are no longer available. From the case study result, which is 31.5%
weight and 11.5 von Mises stress reduction, respectively, it can also be concluded
that topology optimized design for additive manufacturing can reduce huge portion
of the mass thus result in lightweight design.

Acknowledgements This work was supported by the Russian Foundation for Basic Research
(Grant 20-08-00310) and was carried out using the equipment of the interregional multispecialty
and interdisciplinary center for the collective usage of promising and competitive technologies in the
areas of development and application in industry/mechanical engineering of domestic achievements
in the field of nanotechnology (CCP VlSU).

References

Bajaj, P., Hariharan, A., Kini, A., Kürnsteiner, P., Raabe, D., Jägle, E.A.: Steels in additive manu-
facturing: a review of their microstructure and properties. Mater. Sci. Eng., A 772, 138633
(2020)
Chen, G., Liu, J., Shu, X., Gu, H., Zhang, B., Feng, J.: Beam scanning effect on properties opti-
mization of thick-plate 2A12 aluminum alloy electron-beam welding joints Mater. Sci. Eng. A
744, 583–592 (2019)
Optimization of the Weight and Size Characteristics of the Power … 85

de Bonfils-Lahovary, M., Josse, C., Laffont, L., Blanc, C.: Influence of hydrogen on the propagation
of intergranular corrosion defects in 2024 aluminium alloy. Corrosion Sci. 148:198–205
King, W., Anderson, A., Ferencz, R., Hodge, N., Kamath, C., Khairallah, S., Rubenchik, A.:
Laser powder bed fusion additive manufacturing of metals; physics, computational, and materials
challenges. Appl. Phys. Rev. 2, 041304 (2015)
Rozvany, G.: Struct. Multidiscipl. Optim. 37(3), 217–237 (2009)
Sadeghilaridjani, M.: Review of powder bed fusion additive manufacturing for metals. Metals 11,
1391 (2021)
Sigmund, O., Maute, K.: Struct. Multidiscipl. Optim. 48(6), 1031–1055 (2013)
SOLID WORKS 2020. https://help.solidworks.com/2019/English/SolidWorks/cworks/c_simp_
method_topology.htm. Last accessed 21 Nov 2021
Zhang, S., Zhang, T., He, Y., Liu, D., Wang, J., Du, X., Ma, B.: Long-term atmospheric corrosion
of aluminum alloy 2024–T4 in coastal environment: surface and sectional corrosion behaviour.
J. Alloys Compd. 789, 460–471 (2019)
Identification and Active Control
of the Chatter Phenomenon in the Milling
Process Using a Pneumatic Actuator

Peyman Enteshari Najafabadi, Reza Nosouhi, Milad Soleimani,

and Mahmood Mirkhanzadeh

1 Introduction

Given the advances in metal cutting in recent years, the issue of metal machining
with milling and turning machines is still in the spotlight, and these machines are
progressing every day. In today’s advanced industries, especially the aerospace, mili-
tary and medical industries, components of high dimension precision and roughness,
are needed. Reduction of production time is also one of the most important issues of
concern to the industry. However, one of the problems making these demands fail is
the chatter vibration phenomenon. This vibration causes poor surface quality, reduced
dimensional accuracy, increased production time and damage to the machines and
tools. Many studies have been conducted on this subject, and they all try to suppress
or reduce the negative effects of this vibration using different methods.
Generally speaking, any phenomenon that alters the cutting forces in machining
process periodically, can be candidate of the chatter source. There are two main
different types of chatter vibrations. The first type is caused by the cutting process
phenomena including: friction between the tool and the workpiece, effects of the
elevated temperature on the cutting tool and the workpiece which increase the tool
length and consequently, alters the tool length contact length with the workpiece,
and the mode coupling effect; which occurs when relative vibration between the
tool and the workpiece is present in at least two directions on the cutting plane. The
second type, on the other hand, is due to the successive waves created on the surface

P. E. Najafabadi · R. Nosouhi (B)

Modern Manufacturing Technologies Research Center, Najafabad Branch, Islamic Azad
University, Najafabad, Iran
e-mail: [email protected]
Department of Mechanical Engineering, Najafabad Branch, Islamic Azad University, Najafabad,
Iran
M. Soleimani · M. Mirkhanzadeh
R&D Department, Naein Abzar Co. Ltd, Isfahan, Iran

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 87
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_9
88 P. E. Najafabadi et al.

of the workpiece which alternatively changes the cutting forces. The latter is called
regenerative chatter and is the main cause of the chatter phenomenon.
In the literature, several methods have been employed to address the chatter
phenomenon. Some researchers have worked on determination of the machining
parameters (i.e., mainly the spindle speed and depth of cut) so that the chatter would
not occur. In these strategies (Tyler et al. 2015; Dai et al. 2018; Grossi et al. 2017), the
main objective is the stability lobes diagram, by which the stable chatter-free zones
in machining are distinguished from the unstable zones. These methods, however,
do not alter the dynamic behavior and the stable zones of the machine tool; it enables
the machine tool user to carry out the machining process with the highest achievable
material removal rate while avoiding the destructive chatter phenomenon.
Some other researchers have approached differently toward the chatter problem.
They have devised methods by which the dynamic behavior of the machine could
be altered. These methods are divided into two main categories: namely active and
passive methods. The distinction criteria between these two methods are that in the
active chatter control methods (Shi et al. 2018; Fallah et al. 2019), the vibration
parameters of the machine are detected and measured precisely, and an excitation in
accordance with the vibration parameters is applied to the machine system. While in
the passive methods (Wang et al. 2019; Fujimaki et al. 2020), there is no feedback
from the machine, the dynamic behavior of the machine is altered via some methods
that are known to be helpful in all machining chatter conditions. Some examples of
these methods are the redesign of the dynamically weak components of the machine,
and the variable helix and pitch angles end mills.
The advantage of the active method is that in this case, the dynamic state of
the machine is controlled at any moment and when the system malfunctions, the
controller acts and changes the system to a more stable state. Such systems typically
use a controller that is specifically designed for the vibration characteristics of the
system, an accelerometer or strain gage as the vibration sensor, and an actuator,
usually for damping the vibrations (Quintana and Ciurana 2011).
The study of the machine dynamic behavior would eventually lead to a diagram
which is called stability lobes diagram (SLD). The horizontal axis of the diagram is
the spindle speed, while the vertical axis is the depth of cut. The lobes of the machine
tool stability are depicted in this diagram, below which the machining is stable
or chatter-free, and over which the machining is unstable and chatter phenomenon
occurs. The calculations of the SLD are, in fact, based on the dynamics of the machine
and the tool giving the operator a graph for each particular material. This chart shows
how much depth of cut at each selected spindle speed is accessible before occurring
the chatter and reaching unstable state.
According to Fig. 1, it can be said that if, by any means, the diagram could
be shifted upward, higher cutting depths can be selected for the cutting operation,
which results in higher material removal rates and lower production time, which
is economically more advantageous. Furthermore, due to the periodic nature of the
diagram, shifting the diagram into right or left directions might, in some cases,
stabilize an unstable point. The shifting of the SLD upward primarily, and to the
Identification and Active Control of the Chatter Phenomenon … 89

Fig. 1 Milling machine

used with accelerometer

right and left directions secondarily, is the main objective of the research works in
chatter suppression/control literature.
Recent research efforts rely on the use of active and passive methods of chatter
phenomenon control. Active methods are based on the measurement of the vibra-
tion parameters which enable the controller to detect the chatter occurrence and its
parameters, and then excitation of the machine in accordance to the chatter parameters
using an actuator. The most commonly used actuators include piezoelectric stacks and
magnetostrictive actuators. Passive methods are designed based on changes in system
structure and by changing the system natural frequencies by adding and deducting
masses, changing machining parameters or redesigning machine tool accessories
(Munoa et al. 2016).
Ahrens et al. (2016) used an electromagnetic actuator, tried to suppress the chatter
in the grinding process using the active method. The idea of an active chatter damper
is to estimate the shape of the waves on the tool and create a force against the forces
on the workpiece that is the cause of the surface waves. Bort et al. (2016) attempted
to control the chatter using a model-predictive-control (MPC). By using an MPC, the
process time and roughness improved by 46% and 22%, respectively. Burtscher and
Fleischer (2017) presented an adaptive tuned mass damper (ATMD) that its design
and results are based on the genetic algorithm. When the machine’s frequency values
change through the working position, the ATMD reduces the chance of the chatter
occurring by adjusting and dominating these frequency values. When the damper
was activated, the vibrations were reduced by 36%.
As a conclusion of the review of previous studies, it is clear that none of the
abovementioned studies, has used pneumatic actuators to control the chatter. These
actuators, in addition to their many capabilities, are inexpensive and affordable and,
as far as their operating frequency allows, can be used as a chatter control actuator.
The compressed air for the actuator is available in almost all machining workshops,
which enables the machine shop owners to use this system with minimum cost.
90 P. E. Najafabadi et al.

2 Materials and Methods

We need to obtain SLD for the milling machine to know chatter frequency occurs
at what speed range. To obtain SLD, the basis for predicting chatter in theory, the
machine’s natural frequencies are required. For this purpose, an experimental modal
analysis is carried out in order to obtain the dynamic behavior parameters of the
machine.

2.1 Experimental Modal Analysis

Modal analysis is the process of determining the dynamic properties of a system

in terms of natural frequencies, damping coefficients and mode shapes (He and
Fu 2001). We apply them to create a mathematical model for the system dynamic
behavior. The basis of this technique is based on the relationship between vibrational
response at one point of the structure with the excitation at the same point or another.
The equipment used in the experiments is an impact hammer (Bruel & Kjaer brand),
which analyzes the input data using PULSE LabShop software Version 6.1. The
accelerometer is made in DJB Instruments with part number of A/120/V, and the
hammer used was type of RION PH-51. The milling machine is a DECKEL FP4M
type, made in Tabriz Machinery Manufacturing Company. Since the structural chatter
phenomenon occurs in collet and collet holder when the tool is involved with the
workpiece, the accelerometer is directly installed beneath the collet holder. The
hammer impact point is on the collet holder nut. The vibration bandwidth in structural
chatter is between 0 and 200 Hz (Quintana and Ciurana 2011). Figure 1 shows the
milling machine with the accelerometer attached to it.
At the beginning of the modal experiments, the force range applied by the oper-
ator to the object with the hammer is measured. The coherence diagram illustrates
the relationship between the final response and the input impulse to stimulate the
system. If the vertical value of the diagram is less than 0.8, the response obtained
is considered as a noise signal. This indicates that the stimulus generated in the
system has been caused by an external factor. The way to obtain natural frequen-
cies from the obtained graphs is by simultaneously examining the coherence graph
and the frequency response graph. When the number shown on the Y-axis of coher-
ence graph is between 0.8 and 1. The corresponding peak in the frequency response
diagram is the natural frequency. Finally, for the first and second modes, the values
of 16 and 48 Hz were obtained.
In order to obtain the SLD, the cutting force coefficient is also needed. The force
coefficients are mostly influenced by the workpiece material and the cutting tool
parameters (i.e., geometry, size and material). The materials used in this study are
Inconel 617 superalloy and austenitic stainless steel 316. The cutting forces are
measured using a Kistler 9443B dynamometer. The end mill is a WC tool with
Identification and Active Control of the Chatter Phenomenon … 91

the diameter of 12 mm manufactured in YG-1 company under the code number

G9453912.
After measurement of the machining forces, the cutting coefficients are derived
from the following Eqs. 1 and 2:

4F yc
K tc = (1)
ac

−4F xc
Kr c = (2)
ac

where F is the average of the force measured, a is the depth of cut and c is the
feedrate per tooth.
Finally, the stability lobes are calculated as follows:
Selection of a chatter frequency from transfer functions around a dominant mode.
Solution of the eigenvalue, by use of transfer function gotten in modal test, from

a0 Λ2 + a1 Λ + 1 = 0 (3)

Calculation of the critical depth of cut from

2π Λ R ( )
alim = − 1 + κ2 (4)
N Kt

where Λ R is the real part of the eigenvalue real part, κ is obtained by dividing the
real and imaginary part of eigenvalue and N is the tool tooth flute. Calculation of the
spindle speed from for each stability lobe k = 0, 1, 2, …, from

60
n= (5)
NT
where T is tooth-passing period. Repeat the procedure by scanning the chatter
frequencies around all dominant modes of the structure evident on the transfer
functions.
Pneumatic Actuator
To damp the vibration of the machine actively, an actuator is required for excitation
of the system in accordance to the chatter parameters.
In this study, after investigations, it was decided to use a pneumatic actuator. One
advantage of this actuator is its cost-effectiveness compared to other actuators for
its use in the first modes of the system. The compact cylinder, has force of 1000 N,
made by SNS Company in model of SDA 63X40-S.
An air solenoid valve is required to command the cylinder to generate the signal
sent to the valve. The VUVG-L14-M52-MT-G18-1P3 model solenoid valve made
by FESTO Company is used for this purpose. The valve is able to withstand pressure
92 P. E. Najafabadi et al.

up to 800 kPa and has an operating frequency of 16.5 Hz. To produce desired signal
an AFG-2125, GW INSTEK arbitrary function generator is employed.

2.2 Equipment

Since the purpose of this study is to control the vibrations and motion of the machine
tool, a sensor is required to measure the vibration amplitude during machining. In this
research, a MicroSense 5810 non-contact capacitance displacement sensor precision
made by MicroSense Inc. is used. This sensor, which is shown in Fig. 2, measures
the spindle vibration amplitude by voltage variation and has the ability to measure
vibrations in frequencies up to 100 kHz.
Due to the fact that this sensor shows the values in an analog way, it is necessary
to connect it to a data acquisition card (DAC). The 12-bit POS-760 data card made
by Advantech Company is used for this purpose.
The active controller constantly receives and processes the input signals online.
Therefore, a host computer is needed to establish a connection between the input and
output data. LabView 2018 software is employed to design the controller. Finally,
the control cycle is depicted in Fig. 3. The spindle vibration is measured during
machining using the displacement sensor. Whenever the chatter phenomenon occurs,
the controller detects the chatter as the vibration amplitude increases, and it excites
the pneumatic actuator in the chatter frequency.

3 Results and Discussion

The capacitance displacement sensor and the DAC are used to detect the chatter
by checking the displacement versus time diagram of the device in the LabView
software. The experiments are carried out in different spindle speeds with and without
the controller to study the advantage of the active control. When the chatter occurs,
the displacement vs. time diagram shows a peak, as it can be seen in Fig. 4.

Fig. 2 Capacitance
displacement sensor
Identification and Active Control of the Chatter Phenomenon … 93

spindle
pneumatic capasitance
cylinder sensor

solenoid
DAC
valve

signal host
generator computer

Fig. 3 Active control cycle

Displacement (mm)

Time (0.01s)

Fig. 4 Displacement versus time diagram

This is the method of detecting the chatter by a displacement–time diagram.

The horizontal axis of the graph represents the time with a resolution of 0.01 s,
and the vertical axis represents the spindle housing displacement by monitoring
this diagram online during the machining, operator can stop the operation when the
chatter phenomenon is started.
According to the experiments performed when the controller was switched off,
the average voltage range was 0.3 V. As a result, this value was defined as a condition
for the controller to send a signal via the signal generator to the solenoid valve if the
amount of change has been exceeded during operation. The signal sent through the
generator to the solenoid valve was 16 Hz square wave, with 75% duty cycle and a
voltage of 5. The force applied to the spindle by the actuator was 154 N. Figure 5
illustrates the displacement sensor and cylinder installed on machine tool.
Workpieces require machining on each to determine the depth of cut at different
speeds. For this purpose, the workpieces are designed with an inclined surface. Thus,
depth of cut will be increasing during the machining. Since the feedrate did not affect
the chatter, it was considered constant in all experiments with value of 12 mm/min.
94 P. E. Najafabadi et al.

Fig. 5 Installed capacitance

sensor and cylinder

The results of the experiments on the austenitic stainless steel 316.

The machining experiments are carried out without the controller at the spindle
speeds of 315, 400, 500, 1250, 1600 and 2000 rpm with a feedrate of 12 mm/min to
reach critical depth of cut before occurring chatter. The above-mentioned experiments
are performed once again with the presence of the controller.
Figure 6 is the final workpieces obtained from tests. It can be seen that when the
actuator is exciting the system at the spindle speeds in the first mode in the SLD,
the critical depth of cut increases. At the covering speeds of second mode, excitation
of the machine in the first mode frequency has a negative effect and decreases the
critical depth of cut.
Table 1 contains the final values obtained for the critical depth of cut at different
speeds. The last column shows the percentage increase or decrease of the cutting
depth in the presence of the operator, with a positive and a negative sign, respectively.
At spindle speeds 315, 400 and 500 rpm, which lie in the first mode of the machine
in the SLD, the amount of the critical of depth, in the presence of the actuator is
increased up to 47%. The higher spindle speeds, which lie in the second mode of
the machine, show a sharp decrease in critical depth of cut. The reason for such
a decrease is that the excitation with the frequencies far from the mode frequency

Fig. 6 a Final austenitic stainless steel 316 and b Inconel 617 workpieces
Identification and Active Control of the Chatter Phenomenon … 95

Table 1 Obtained values for austenitic stainless steel 316

Spindle speed (rpm) Depth of cut (mm) Depth of cut (mm) Increase or decrease
without controller with controller depth-of-cut
percentage
315 4.85 6.15 + 21
400 6.9 8.45 + 18
500 6.65 12.7 + 47
1250 10.25 6.6 − 55
1600 12.75 4.45 − 186
2000 8.95 6.8 − 31

would only add extra energy to the system, which eventually makes the system more
prone to the chatter.

3.1 The Results of the Experiments on the Inconel 617 Super

Alloy

Inconel 617 tests conducted at speeds of 315, 400 and 500 rpm with and without the
presence of the controller. When the actuator is on, at the covering speeds of the first
mode in the SLD, the critical depth of cut is increased.
In Table 2, the final values obtained for the Inconel 617 super alloy are shown
when the controller is on and off. The three tests for Inconel were all in the first
mode. In the first mode, critical depth of cut increases.
Note that the chatter phenomenon is mainly occurring in the roughing operations,
where the depth-of-cut values are high enough. The roughing operations, specifi-
cally, in hard-to-cut materials such as austenitic stainless steels and super alloys, are
usually carried out in low cutting speeds. Therefore, although the pneumatic actu-
ator is unable to suppress the chatter phenomenon in higher spindle speeds due to the
limitation of the actuator bandwidth, the excitation of the machine in the first mode
is adequate, and the excitation of the machine in the second mode sounds redundant.

Table 2 Obtained values for Inconel 617 super alloy

Spindle speed (rpm) Depth of cut (mm) Depth of cut (mm) Increase or decrease
without controller with controller depth-of-cut
percentage
315 6.95 7.5 +7
400 5.65 6.3 + 10
500 4.75 5.3 + 10
96 P. E. Najafabadi et al.

4 Conclusions

The main purpose of this study was to identify the chatter phenomenon and control it
using the active method. As explained in previous chapters, an active control system
requires data online to identify the chatter. An actuator is also used to actively control
the vibrations. So, it can receive the signal and apply it to the desired area using the
host computer and the controller defined for it.
After reviewing the final results, it is found that the pneumatic actuator is capable
of increasing the critical depth of cut and mitigating the chatter in the first mode.
But at speeds related to the second mode, the presence of the actuator reduces the
cutting depth compared to the absence of it. The reason for this is due to the operating
frequency of the solenoid valve. The frequency of the solenoid valve is 16 Hz, which
corresponds to the frequency of the device first mode (16 Hz). The frequency value
of the device second mode is 48 Hz. As a result, at the speeds of the second mode,
the presence of the actuator stimulates the vibration more and helps the chatter to
occur faster. Using this actuator in the first mode of the device increases the average
cutting depth for austenitic stainless steel 316 and Inconel 617 super alloy by 28 and
10 percent, respectively. Therefore, due to the low price of this type of actuator, it
can be said that the pneumatic actuator can be more economical than other actuators
in order to damp the first modes of the device.
The employed pneumatic actuator bandwidth is limited to the first mode frequency
due to the limitation in operating frequency of the valve. Although this is one
of the disadvantages of the pneumatic system, since the chatter occurs mostly in
roughing operations, and the roughing operations of the hard-to-cut materials are
mainly carried out in the low cutting speeds, the first mode speeds are high enough
in economic machining of these materials.

References

Ahrens, M., Dagen, M., Denkena, B., et al.: An active damping method for chatter vibration in
plunge grinding using electromagnetic actuators. Procedia CIRP 46, 197–200 (2016)
Bort, C.M., Leonesio, M., Bosetti, P.: A model-based adaptive controller for chatter mitigation and
productivity enhancement in CNC milling machines. Rob. Comput. Integr. Manuf. 40, 34–43
(2016)
Burtscher, J, Fleischer, J.: Adaptive tuned mass damper with variable mass for chatter avoidance.
CIRP Ann 66(1), 397–400 (2017)
Dai, Y., Li, H., Xing, X., et al.: Prediction of chatter stability for milling process using precise
integration method. Precis. Eng. 52, 152–157 (2018)
Fallah, M., Moetakef-Imani, B., Hosseini, A., et al.. Boring Bar chatter control using feedback
filtered-x normalized least mean square algorithm. IFAC-Papers OnLine; 52(10), 358–363 (2019)
Fujimaki, S., Shibayama, T., Hayasaka, T., et al.: Proposal of “curved-profile wiper turning” for
efficient, stable, and smooth finishing. Precis. Eng. 61, 152–159 (2020)
Grossi, N., Sallese, L., Scippa, A., et al.: Improved experimental-analytical approach to compute
speed-varying tool-tip FRF. Precis. Eng. 48, 114–122 (2017)
He, J., Fu, Z.F.: Modal Analysis. Butterworth-Heinemann Publications (2001)
Identification and Active Control of the Chatter Phenomenon … 97

Munoa, J., Beudaert, X., Dombovari, Z., et al.: Chatter suppression techniques in metal cutting.
CIRP Ann. 65(2), 785–808 (2016)
Quintana, G., Ciurana, J.: Chatter in machining processes: a review. Int. J. Mach. Tools Manuf.
51(5), 363–376 (2011)
Shi, F., Cao, H., Li, D., Chen, X., et al.: Active chatter control in high speed milling processes based
on H∞ almost disturbance decoupling problem. Procedia CIRP 78, 37–42 (2018)
Tyler, C.T., Troutman, J., Schmitz, T.L.: Radial depth of cut stability lobe diagrams with process
damping effects. Precis. Eng. 40, 318–324 (2015)
Wang, C., Zhang, X., Yan, R., et al.: Multi-harmonic spindle speed variation for milling chatter
suppression and parameters optimization. Precis. Eng. 55, 268–274 (2019)
SHS Extrusion of a Composite Material
Based on Aluminum-Magnesium Spinel

Andrey Chizhikov, Alexander Konstantinov, Pavel Bazhin,

and Mikhail Antipov

1 Introduction

An urgent task for the modern scientific and technical complex is the development
of new promising materials and products based on them. These materials include
composites based on oxide ceramics. Oxide ceramic materials are widely used in
various industries as cutting ceramics (Guo et al. 2018; Pronin et al. 2019; Stolin
et al. 2018), refractory materials (Ren et al. 2021), protective coatings (Tang et al.
2021; Kozerozhets et al. 2021), as materials for fuel cells (Wang et al. 2022), etc.
Among a wide range of oxide ceramic materials, materials based on magnesium
aluminum spinel MgAl2 O4 can be singled out. These materials have a high melting
point (2135 °C), high hardness, high chemical and corrosion resistance to acids and
molten metals (Ganesh 2013; Shi et al. 2020). The described properties make mate-
rials based on aluminum-magnesium spinel promising for use in the chemical and
metallurgical industries for the manufacture of protective covers for thermocouples,
various crucibles and molds, and refractory linings (Meng et al. 2018; Ma et al.
2016).
The synthesis of materials based on MgAl2 O4 is complicated due to the fact that
the formation of this compound occurs as a result of an endothermic reaction of
magnesium and aluminum oxides and requires the use of high temperatures. There
are also low-temperature methods for the synthesis of spinels, such as the sol–gel
method (Wu et al. 2022) and hydrothermal synthesis (Panasyuk et al. 2019). However,
these methods are complicated by the need to process acids, solutions, gels and the
complexity of the equipment. The high melting point of MgAl2 O4 makes it difficult
to obtain finished products from this material.

A. Chizhikov (B) · A. Konstantinov · P. Bazhin · M. Antipov

Merzhanov Institute of Structural Macrokinetics and Materials Science, Russian Academy of
Sciences, Chernogolovka, Moscow, Russia
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 99
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_10
100 A. Chizhikov et al.

A promising approach to obtaining products from composite materials based on

aluminum-magnesium spinel is a combination of self-propagating high-temperature
synthesis (SHS) and shear plastic deformation (Bazhin et al. 2019). This combination
is implemented in the SHS-extrusion method (Bazhin et al. 2015). The flow of SHS
occurs due to the release of internal heat of the system during the chemical reaction
between the components and is characterized by high process temperatures (> 1800
°C). This makes it possible to carry out the synthesis of refractory compounds without
the use of external heating sources. At the same time, since the synthesis products
are in a heated state, this makes it possible to form finished products by shear plastic
deformation.
The purpose of this work was the SHS extrusion of a ceramic composite material
based on aluminum-magnesium spinel reinforced with titanium diboride particles,
as well as the study of the effect of deformation rate on the structure of the resulting
products.

2 Materials and Methods

Powders of boron oxide (≥ 99%, < 1 µm), aluminum (≥ 99.5%, < 5 µm), magnesium
(≥ 99%, 250–450 µm) and titanium (98%, 280 µm) were used as initial reagents
in this work. These powders were pre-dried at a temperature of 50 °C for 12 h,
after which they were mixed in a ratio of 2B2 O3 f-2Al-3Mg-2Ti in a drum mill for
12 h. Next, cylindrical blanks weighing 25 g and a relative density of 0.65 were
prepared from the obtained powder mixture by cold uniaxial pressing. The synthesis
of materials was carried out in the SHS mode (Vidyuk et al. 2021; Shaikh et al.
2019; Bazhin et al. 2021). The SHS process occurred due to the reduction of boron
oxide with aluminum and magnesium with the formation of the corresponding metal
oxides. The reduced boron reacted with titanium to form titanium diboride. Since
the formation of spinel occurs with the absorption of heat, the reactions listed above,
which have a high thermal effect, create favorable conditions for the synthesis of
spinel. To obtain products based on the synthesized material, the SHS-extrusion
method was used, which combines the SHS process and high-temperature shear
deformation.
Thermodynamic calculations were carried out using the ISMAN-Thermo soft-
ware package developed to estimate the probable phase composition of the reac-
tion products, their phase state, and the adiabatic combustion temperature of the
system. An experimental study of the combustion temperature was carried out using
tungsten-rhenium thermocouples in a setup that simulates the real conditions of SHS
extrusion.
The phase composition was studied by X-ray phase analysis (XRD) on a DRON-
3M diffractometer. The microstructure of the obtained materials was studied by
scanning electron microscopy (SEM) using a Carl Zeiss LEO 1450 VP electron
microscope. The results of energy dispersive analysis are given in mass percent.
SHS Extrusion of a Composite Material Based … 101

3 Results and Discussion

According to the results of thermodynamic calculations, the adiabatic combustion

temperature of the system under study was 2827 °C. According to the results of an
experimental study of the combustion characteristics of the system under study, the
actual combustion temperature was 2027 °C. The strong discrepancy between the
calculated and experimental results of determining the combustion temperature is
explained by the high heat removal that occurs between the sample and the walls of
the metal tooling despite the presence of asbestos insulation on the sample.
Since the combustion of the material during SHS occurs in a non-equilibrium
mode, and also due to the error of the thermocouple measurement method, it can
be assumed that the actual combustion temperature of the material exceeded the
melting temperature of aluminum oxide (2072 °C). Also, the combustion temperature
significantly exceeded the melting point of the initial components (the most refractory
of them is titanium—1668 °C). Thus, in the course of synthesis, a liquid phase
appears in the material, as a result of which a suspension is formed from liquid
initial components, partially liquid synthesis products and solid particles of synthesis
products. The formation of a suspension leads to the fact that the material acquires
viscoelastic properties and the ability to plastically deform (Chizhikov et al. 2019).
Due to the ability to high-temperature plastic deformation, this object of study is
promising for the production of products by the SHS-extrusion method. To control the
process of SHS extrusion, there are a number of technological parameters: delay time
before applying pressure, strain rate, pressing pressure as well as design parameters
of the forming matrix.
In the course of establishing the influence of technological and structural param-
eters on the SHS-extrusion process, ceramic hollow rods based on aluminum-
magnesium spinel and titanium diboride up to 100 mm long and with an outer
diameter of 5, 6 and 7 mm were obtained. Figure 1 shows a general view of the
fracture of the obtained hollow rods.
One of the most important parameters affecting the SHS-extrusion process is the
delay time. At low values of this parameter, chemical transformations and structure
formation processes are not completed in the material, there is an excess amount
of the liquid phase, since in this case, the synthesis proceeds with the formation of
a sufficiently large amount of the liquid phase. At high values of this parameter,
the material undergoes a complete crystallization process, as a result of which the
material loses its ability to plastic deformation and extrusion. Thus, there is an optimal
delay time interval in which the material has the ability to plastically deform, and also
has a sufficient level of viscoelastic properties necessary for the formation of hollow
rods. For this material, the optimal delay time was 4–5.5 s. Another key technological
parameter affecting the SHS extrusion of ceramic refractory hollow rods is the strain
rate. The “jet swelling” effect (Barus effect) (Lombardi and Tammaro 2021; Wang
et al. 2018), which is the basis for the formation of ceramic hollow rods in this work,
manifests itself at sufficiently high extrusion rates. In the course of the work, it was
found that there is a threshold value of the strain rate, below which the formation
102 A. Chizhikov et al.

Fig. 1 SEM results of the

fracture of the resulting
hollow rod

of ceramic hollow rods does not occur. Although the extrusion still occurs. For the
studied model composition, such a threshold value was 65 mm/s. At a lower value of
the strain rate, there is no accumulation of a sufficient amount of elastic deformation
necessary for the occurrence of normal stresses at the exit from the forming hole,
which stretch the material jet. In this case, extrusion occurs in a stationary mode.
As a result, based on the studied composition, rods were obtained, which are a
compact core, mainly consisting of titanium diboride, which is covered with a shell
of aluminum-magnesium spinel with a thickness of less than 1 mm, Fig. 2.
In this case, there is also a limiting value of the rate of deformation of the press,
above which extrusion does not occur. For the studied model composition, this value
was 90 mm/s. Above this value, the viscosity of the material, which depends on
the strain rate, increases so much that it completely prevents extrusion. Another
important technical parameter is the molding pressure. It was found in the work
that high molding pressure is not required to obtain ceramic hollow rods. There is
a threshold value above which increasing the molding pressure has no effect on the
length and quality of the resulting products. For the studied model composition, this
value was 30 MPa. This is due to the fact that the viscosity of the studied materials
depends more on the extrusion rate than on pressure. Structural parameters also have
SHS Extrusion of a Composite Material Based … 103

Fig. 2 SEM results of the

section of the rod obtained at
a strain rate below 65 mm/s

a great influence on the process of SHS extrusion. So, with a decrease in the angle
of the conical part of the matrix, the contact area of the synthesized material with
the matrix increases, while the magnitude of heat removal increases. As a result,
the near-matrix regions rapidly lose their ability to undergo plastic deformation, and
blockage of the exit hole of the matrix occurs. The optimal value of the angle of the
conical part for the studied model composition is 160º.
It was shown that by changing the diameter of the exit section of the matrix, it is
possible to obtain hollow rods with different inner diameters. In this case, the outlet
of the matrix must necessarily be less than the diameter of the quartz caliber into
which the extrusion is carried out. An important parameter is also the length of the
forming channel of the matrix. With insufficient length during extrusion, there is no
sufficient accumulation of elastic deformation, as a result of which tensile normal
stresses do not arise, and hollow rods do not form. When the optimal value of the
length of the forming belt is exceeded, the elastic deformation is relaxed directly in
the forming channel, which also prevents the formation of ceramic hollow rods. It
has been established that an increase in the outer diameter of the obtained ceramic
hollow rods due to an increase in the diameter of the quartz caliber, the speed of
the press plunger and the mass of the workpiece leads to a decrease in the size of
104 A. Chizhikov et al.

Fig. 3 XRD results: a hollow rod and b rod obtained at a strain rate below 65 mm/s

the structural components. This is facilitated by an increase in heat removal that

occurs when the material contacts the walls of the quartz caliber. According to the
XRD results, the obtained ceramic hollow rods had a uniform distribution of the phase
composition along the entire length. It is also shown that under the conditions of SHS
and shear deformation of the studied model composition, the formation of aluminum
and magnesium intermetallic compounds occurs, as well as the deformation of the
crystal lattice of the aluminum-magnesium spinel. Thus, by adjusting the parameters
of SHS extrusion, it is possible to influence the phase composition of the resulting
products.
According to the SEM results, the resulting hollow rods had a composite structure.
The basis of the rods was octahedral grains characteristic of spinel, 10–40 µm in
size. Cubic grains of titanium diboride were distributed over the entire volume of
the obtained hollow rods and had a size of less than 1 µm. The XRD results showed
that the resulting rods mainly consist of two phases: MgAl2 O4 and TiB2 . At the
same time, the products also contain traces of unreacted titanium, magnesium oxide,
which did not react with Al2 O3 , and Al12 Mg17 intermetallic compound. The XRD
results also showed that the resulting hollow, Fig. 3a, and compact rods, Fig. 3b, had
a similar phase composition. At the same time, the intensity of the TiB2 peaks in the
compact rod is higher due to its larger amount.

4 Summary

Ceramic rods based on the MgAl2 O4 -TiB2 composite material with a length of more
than 100 mm and an outer diameter of 5–7 mm were obtained by SHS extrusion.
SHS Extrusion of a Composite Material Based … 105

Thermodynamic calculations of the adiabatic combustion temperature of the speci-

fied material were carried out, which amounted to 2827 °C. As a result of an experi-
mental study of the combustion characteristics of this material, it was shown that the
actual process temperature due to heat removal conditions is much lower than the
adiabatic one and is 2027 °C. The influence of technological and design parameters
on the extrusion process has been studied. It is shown that the strain rate has a signif-
icant effect on the process of extrusion of the specified material. This parameter has
a threshold value, for the studied material it was 65 mm/s, above which hollow rods
are formed during extrusion. When extruding at a speed below 65 mm/s, a compact
rod is formed, which is a core based on TiB2 in a shell of MgAl2 O4 .

Acknowledgements This work was supported by the Russian Science Foundation, project no.
20-73-00235.

References

Bazhin, P.M., Stolin, A.M., Alymov, M.I., Chizhikov, A.P.: Inorg. Mater. Appl. Res. 6, 187 (2015)
Bazhin, P.M., Kostitsyna, E.V., Stolin, A.M., Chizhikov, A.M., Bychkova, MYa., Pazniak, A.:
Ceram. Int. 45, 9297 (2019)
Bazhin, P.M., Kostitsyna, E.V., Chizhikov, A.P., Konstantinov, A.S., Neganov, L.E., Stolin, A.M.:
J. Alloys Compd. 856, 157576 (2021)
Chizhikov, A.P., Bazhin, P.M., Stolin, A.M., Alymov, M.I.: Dokl. Chem. 484, 79 (2019)
Ganesh, I.: Int. Mater. Rev. 58, 63 (2013)
Guo, X.L., Zhu, Z.L., Ekevad, M., Bao, X., Cao, P.X.: Adv. Appl. Ceram. 117, 16 (2018)
Kozerozhets, I.V., Panasyuk, G.P., Azarova, L.A., Belan, V.N., Semenov, E.A., Voroshilov, I.L.,
Danchevskaya, M.N.: Theor. Found. Chem. Eng. 55, 1126 (2021)
Lombardi, L., Tammaro, D.: Phys. Fluids 33, 033104 (2021)
Ma, B., Yin, Y., Zhu, Q., Zhai, Y., Li, Y., Li, G., Yu, J.: Refract. Ind. Ceram. 56, 494 (2016)
Meng, J.X., Chen, W.Q., Zhao, J.Z., Liu, L.: High Temp. Mater. Process 37, 581 (2018)
Panasyuk, G.P., Kozerozhets, I.V., Danchevskaya, M.N., Ivakin, Y.D., Murav’eva, G.P., Izotov,
A.D.: Dokl. Chem. 487, 218 (2019)
Pronin, A.I., Myl’nikov, V.V., Rybalkin, A.A.: Glass Ceram. 76, 99 (2019)
Ren, X.M., Ma, B.Y., Wang, L.L., Liu, G.Q., Yu, J.K.: Ceram. Int. 47, 31130 (2021)
Shaikh, N., Patel, K., Pandian, S., Shah, M., Sircar, A.: Arab. J. Geosci. 12, 11 (2019)
Shi, Z.Q., Zhao, Q.L., Guo, B., Ji, T.Y., Wang, H.: Mater. Des. 193, 108858 (2020)
Stolin, A.M., Bazhin, P.M., Konstantinov, A.S., Chizhikov, A.P., Kostitsyna, E.V., Bychkova, M.Y.:
Ceram. Int. 44, 13815 (2018)
Tang, S.W., Liu, P.F., Su, Z., Lei, Y., Liu, Q., Liu, D.S.: High Temp. Mater. Process 40, 77 (2021)
Vidyuk, T.M., Korchagin, M.A., Dudina, D.V., Bokhonov, B.B.: Combust. Explos. Shock Waves
57, 385 (2021)
Wang, J.N., Wang, T., Xu, J., Yu, J.C., Zhang, Y.M., Wang, H.P.: J. Appl. Polym. Sci. 135, 8 (2018)
Wang, H., Lei, Z., Zhang, H., Li, Y., Jing, J., Jin, Y., Yang, Z., Peng, S.: J. Electrochem. Energy
Convers. Storage 19, 011007 (2022)
Wu, Q., Feng, G., Jiang, F., Miao, L., Jiang, W., Liang, J., Liu, J., Tan, J., Zhang, X., Lao, X.: Ceram.
Int. 48, 3351 (2022)
 Fluid Mechanical Engineering
and Engineering Thermodynamics
Visual Observation of the Vortex Flow
from the Same Diameter Circular
Cylinders Which Sandwiched a Partition
Disk

Yoshifumi Yokoi

1 Introduction

In the experiment of the flow from two dimensional bluff cylinders, such as a circular
cylinder, it is carried out by providing an end plate in order to maintain two dimen-
sional conditions at the cylinder edge. In the end plate, in a leading edge, the flow
is divided and the separated flow is unified in a trailing edge. It is imagined that the
unified flow becomes complicated flow structure. Nevertheless, influence condition
of the structure of the complicated flow is not made into the object of interest. So,
there is no consideration about the size and shape of the end plate. The knowledge
about the influence of the size of the end plate to the size of the bluff cylinder is
hardly found. Although the size of the end plate is enough enlarged compared with
the projection width of the bluff cylinder, the concrete reason is not shown. Szepessy
and Bearman (1992) and Norberg (1994) have reported influence of as opposed to
the flow and the vortex shedding in the space and the end plate itself which the end
plate makes. The study on the influence of the size of the end plate is seen by a
Hara and Suzuki (1989, 1991), Yamada (1992). Those study investigates about the
influence of the size of the end plate exerted on the cylinder base pressure and vortex
shedding in the test section inserted into the end plate. Even if it provides an end
plate and prepares the flow of the inspection section, it does not have the meaning of
rectification, if the turbulence element which generated interference by the wake is
fed back. It is required to solve the mechanism which the turbulence element feeds
back, and an aspect in engineering. Yokoi (2019) divided into the circular cylinder
span central part, prepared the plate, varied the size (the length of a partition plate)
of the partition section, and investigated influence on the aspect of a flow by visual-
ization experiment. And the following results are reported. The Karman vortex street
from which a phase differs on the front side and the back side of the partition plate

Y. Yokoi (B)
National Defense Academy of Japan, 1-10-20 Hashirimizu, Yokosuka, Kanagawa, Japan
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 109
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_11
110 Y. Yokoi

may be formed. Two flows which pass over the partition plate and join are forming
a complicated flow field. Transfer of the information on the turbulence is dependent
on the size of the partition plate. In order to verify the universality of the obtained
matter, it is important to carry out by changing experiment environment (e.g., an
apparatus, a method, and an observation viewpoint). Therefore, in this study, the size
of a partition disk was varied and visualization observation was carried out for the
flow around a circular cylinder with a partition disk using the closed circuit water
channel apparatus.

2 Experimental Apparatus and Method

2.1 Outline of Experimental Apparatus

Experimental apparatus consists of the closed circuit water channel apparatus, a

circular cylinder with partition disk, a visualization apparatus, and a video recording
apparatus. The overall experimental apparatus is shown in Fig. 1. The experiment
situation in the test section of the closed circuit water channel apparatus is shown
in Fig. 2. The closed circuit water channel apparatus is a small closed circuit water
channel with a water volume of 0.4 m3 , a test section is a product made from trans-
parent acrylics, the cross-sectional size is 0.3 m × 0.3 m, and the length is 1.2 m.
The upper part of the test section is opened wide, and the plate for circular cylinder
attachment is prepared in the part. The mirror leaned under the test section bottom
is prepared, a camera is placed by the side, and the aspect of the flow seen from
the circular cylinder axis is photographed. The range of the flow velocity in the test
section is 0.07–0.98 m/s, and adjustment of the flow velocity is carried out by oper-
ating an operation controller (digital operator). The water level of the test section at
the time of the experiment was adjusted to 0.2 m.
The circular cylinder with a partition disk consists of two circular cylinders of
the same diameter, and one disk. The circular cylinder is metal, and three kinds of
diameters of 10, 16, and 20 mm were prepared. The circular cylinder with a diameter
of 10 mm is a product made from brass, and the circular cylinder with a diameter of
16 mm and a diameter of 20 mm is a product made from aluminum. The thickness of
the disk is a product made from transparent acrylics in 2 mm. The disk does not have
cutting off the corners, and the attachment hole (ϕ4 mm) is provided in the center of
the disk. Twelve disks with which diameters differ are prepared. As for the structure
of the circular cylinder with a partition disk, the disk with a thickness of 2 mm is put
between the circular cylinder with a length of 199 mm and the circular cylinder with
a length of 99 mm so that full length may be set to 300 mm. The circular cylinder
with a partition disk is set in the plate for circular cylinder attachment prepared in
the test section upper part of the small closed circuit water channel. When it does so,
the partition disk will be located in a place with a depth of 100 mm at the time of the
test.
Visual Observation of the Vortex Flow from the Same Diameter Circular … 111

4
5
2
7
1

9 6 3

Fig. 1 Photograph of the overall experimental apparatus: 1 closed circuit water channel apparatus,
2 test section, 3 mirrors, 4 tracer tank, 5 the amount adjustment clamp of tracers, 6 halogen lamp,
7 personal computer, 8 CCD video camera, and 9 tripod mounts

Fig. 2 Experiment situation in the test section: the flow is flowing into the right from the left. Two
tracer introduction small tubes are aslant inserted from the upstream side at the flow

The visualization apparatus consists of a tracer ink introduction apparatus and the
halogen light (one piece) of 500 W. As the tracer ink, the poster color (the product
made by TURNER; red, blue) by which specific gravity adjustment was carried out
was used. Introduction to the flow field of tracer ink was carried out for the purpose
of the circular cylinder upstream side stagnation point using the stainless steel tube
of 1.5 mm of outer diameters. The introductory depth of water of the tracer was made
into the position (the depth of water of 150 mm and position of 50 mm) distant about
112 Y. Yokoi

50 mm, respectively, from the partition plate. The video recording apparatus is a
digital camcorder (WAT-250D2, product made by Watec), and a personal computer
(LaVie PC-LL990FD by NEC Corp.). The digital camcorder (it is hereafter called
a video camera) is equipped with the zoom lens. Between the video camera and
the personal computer, the signal converter (USB conversion unit AX-200, made
in Carrots Systems) is prepared, and the picture signal from a video camera can be
taken now into a direct personal computer. Here, since photography of an aspect
that it saw from the circular cylinder axis through the mirror set in the lower part
of the test section is carried out, in this side, with the image, the bottom and depths
side becomes the water surface side. The video camera is being fixed with the tripod
mount so that the photography position may not change for every photography time.

2.2 Experimental Procedure

Water is filled so that it may become a closed circuit water channel with the depth of
water of 0.2 m in predetermined amount of water and a predetermined test section.
Preparation operation is carried out at this time and fine-tuning of depth of water is
also carried out. Secondly, the diameter of the circular cylinder and the size of the
partition disk are selected, and the circular cylinder with a partition disk is assembled.
The circular cylinder with a partition disk is set in the attachment plate. In the
section of the circular cylinder which sandwiched the partition disk, the tracer ink
introduction small tube is set in the upstream side of the stagnation point of the
circular cylinder. The test flow velocity operates and sets up the inverter for the flow
velocity setup. Two kinds of tracer ink in which colors differ is poured in into the
flow after adjustment of the test flow velocity, and flow visualization is carried out.
The aspect of the visualized flow takes a photograph with the video camera and takes
the image into the personal computer. If the photography is completed, the aspect
of the flow will be similarly observed about another partition disk and another flow
velocity. And also in order to make it legible in the case of observation, it illuminated
using the halogen light.

2.3 Experimental Parameters

The main experiment parameters are the circular cylinder diameter d (10, 16, and
20 mm), the diameter D of the partition disk, and the main flow velocity U. The
diameters of the partition disk are 24, 32, 40, 48, 56, 64, 72, 80, 88, 96, 160, and
240 mm. If diameter ratio D/d formed into the-less dimension for the cylinder diam-
eter is used, the applicability of diameter ratio D/d will be set to 1.5–240. Moreover,
in order to compare, the state where the partition disk was not used was also carried
out. In this case, diameter ratio D/d is set to 1.0. The main flow velocity wrote the
inverter input frequency of the digital operator for operation with 2, 3, and 4 Hz, and
Visual Observation of the Vortex Flow from the Same Diameter Circular … 113

became 0.065, 0.098, and 0.131 m/s. The range of Reynolds number Re (= Ud/ν, ν
is the kinematic viscosity of water) corresponding to each flow velocity was Re =
660 to Re = 2630.

3 Experimental Results and Discussion

The experiment was performed from confirming the repeatability of the experimental
result described by previous report (Yokoi 2019). Figure 3 shows the aspect of the
flow carried out in the state without the partition disk (Fig. 3a), and the state using
the partition disk (diameter of 240 mm) of the maximum size (Fig. 3b), respectively.
The main flow velocity at this time was 0.098 m/s. It is shown by Fig. 3a that the red
tracer streak line and the blue tracer streak line have overlapped. Roll up of those
vortexes can be seen synchronize. On the other hand, in Fig. 3b, the tracer streak
line of two colors can be recognized clearly. There, it is shown that the timing of the
vortex shedding from the circular cylinder set to each section differs. It seems that
the 90-degree phase gap has occurred in the vortex shedding. There is nothing that
is interrupted over the spanwise of the circular cylinder, when there is no partition.
Therefore, the information on roll up of the vortex and the information on the timing of
separation propagate to the spanwise. As a result, the vortex shedding constitutes the
same phase. If the partition is provided, the information transmitted to the spanwise
will be interrupted there. Therefore, the vortex is discharged independently in each
section. If it thinks such, it will be easy to understand that a phase gap occurs. Figure 4
shows the aspect of the flow in each section when changing the main flow velocity.
Here, the diameter of a partition disk is 240 mm and the largest disk is used. The
main flow velocity is U = 0.065 m/s in Fig. 4a and is U = 0.131 m/s in Fig. 4b. It
is shown that 90-degree phase difference has occurred in both of the figures in the
vortex shedding which sandwiched the partition disk. The phase gap is because the
communication of information of the spanwise is intercepted physically, and it was
shown that it is not dependent on the flow velocity.

(a) (b)

Fig. 3 Comparison of the flow pattern by the existence of a partition disk, U = 0.098 m/s, a without
disk, b with disk, D/d = 15
114 Y. Yokoi

(a) (b)

Fig. 4 Comparison of the flow pattern according to the variation of main flow velocity at the
diameter ratio D/d = 15, a U = 0.065 m/s, b U = 0.131 m/s

It is one of the most interesting matters of this study to investigate the influence
which the size of the partition disk has on roll up of the vortex or the phase difference
at the time of the vortex shedding. Figure 5 shows the result of having observed the
existence of the phase difference in the both sides when varying the size of the
partition disk. An abscissa is diameter ratio D/d and an ordinate is Reynolds number
(Re = Ud/ν). Here, the diameter ratio D/d is defined by the ratio of the partition disk
diameter and the circular cylinder diameter. In the figure, when ● mark does not have
phase difference and ⊚ mark has phase difference, ◎ mark shows the state where
these two states were changed intermittently. And ϕ10, ϕ16, and ϕ20 which are
written in the figure are the circular cylinder diameter, and they show the result when
varying the flow velocity (Reynolds number). In the figure, when diameter ratio D/d
is 7 or more, phase difference occurs, and when diameter ratio D/d is three or less, it
is shown that there is no phase difference. And it can grasp that the existence of phase
difference is the same distribution, without being dependent on Reynolds number.
Thus, it is found that depend for the partition effect by the disk on the size, and it is
not dependent on the flow velocity. Yokoi and Kamemoto (1993) have described that
the characterization of the vortex formation is carried out when have been rolled by
the separating shear layer, and turbulence is supplied by the recirculating flow near
the separation position. When the outer end of a partition disk is in the upstream
side rather than the position rolled by the first vortex, relation of the spanwise of the
recirculating flow cannot be intercepted. Since the information on the turbulence is
transmitted near the separation position, the vortex shedding in the both sides of the
partition disk is considered to become the same phase.
It is important to grasp the position in which the separating shear layer rolls
around and the vortex is formed at the beginning. Taneda (1972) has described that
the maximal domain of the twins vortex formation produced behind the circular
cylinder is about 3d from behind the circular cylinder in the case where the circular
cylinder leaves suddenly, and is about 2.5d from behind the circular cylinder in
the case where the circular cylinder carries out uniform acceleration. If they are
converted into diameter ratio D/d used by this study, 3d behind circular cylinder will
be diameter ratio D/d = 7, and 2.5d behind circular cylinder will be diameter ratio
D/d = 6. Nishioka (1992) described that the position where the isolated vortex which
Visual Observation of the Vortex Flow from the Same Diameter Circular … 115

Fig. 5 Distribution of phase 3500

difference generating Without phase diference
3000 Intermittent synchronization
With phase diference
20
2500

Re
16
2000 20
16
1500
20 10
1000 16 10
10
500
0 5 10 15 20 25
D/d

constitutes the vortex street is produced is 3d in position from the circular cylinder
center. It is D/d = 6 when expressed with the diameter ratio. In the visualization
experiment which Yokoi and Hirao (2008) performed, the observation points of
vortex shedding frequency were the position of 2.5d circular cylinder behind, and
the position of 5.5d circular cylinder behind. They are D/d = 6 and D/d = 12 when
those positions are expressed with the diameter ratio. It became clear that are the
vortex shedding without phase difference from the result in these references and the
experimental result at the time of the diameter ratio D/d < 3, and the partition disk
is not working.
When the diameter ratio D/d exceeds 3, probably because the partition disk affects
the exchange of the information on turbulence, phase difference occurs intermittently
in roll up of the vortex. When the circular cylinder is seen from an axis, each Karman
vortex seen in two sections divided with the partition disk will overlap or shift.
Figure 6 shows the intermittent existence of the phase difference as an example.
Here, the diameter ratio is D/d = 4.5 and Reynolds number is Re = 1570. Figure 6a
shows the case of the same phase. Red tracer ink overlaps with blue tracer ink, and
seems to be one Karman vortex. On the other hand, Fig. 6b shows the case where
there is phase difference. Red tracer ink and blue tracer ink show each Karman vortex.
Probably, it will be difficult to recognize it as the Karman vortex, if a color exception
is not carried out. Although it is very interesting to get to know the characteristic of
both phase difference, it is very difficult to carry out in the visualization experiment.
So, it is expected that details will be clarified by carrying out comparison examination
of the waveform obtained by the point measurement using a hot wire anemometer,
for example with wind tunnel equipment. When there is a position where the first
vortex is formed inside the partition disk, the exchange of each other information is
intercepted on the both sides of the partition disk. Therefore, vortex street formation
which became independent each other will be carried out. However, since the flow
which passed over the partition disk joining and inducing turbulence not a little is
foreseen, the size in which the partition disk had a margin is needed. In this study,
since the size is not clarified, the elucidation is expected from future study. It seems
116 Y. Yokoi

(a) (b)

Fig. 6 Appearance of intermittent phase difference, U = 0.098 m/s, D/d = 4.5, a same phase, b
different phase

that it is hard to receive the influence of unification of the flow produced after passing
the partition disk when the partition disk is large enough. Therefore, the Karman
vortex from which the phase differs on the both sides of the partition disk as shown
in Fig. 1b will be formed. Thus, in order not to receive the mutual interference but
to obtain the independent Karman vortex, 7 or more diameter ratio D/d is needed.

4 Conclusions

In the closed circuit water channel experiment, the visualization experiment which
provides the partition disk and controls transfer of turbulence of the circular cylinder
spanwise was performed. The following conclusions were obtained.
(1) It was confirmed also with another apparatus that the Karman vortex street from
which the phase differs on the both sides of the partition disk may be formed.
(2) When the diameter ratio D/d is 3 or less, there is no effect as the partition disk.
(3) As for the size of a partition disk, it is desirable that it is larger than the position
in which the first vortex is formed, and it needs more than diameter ratio D/d
= 7 to obtain the Karman vortex which did not receive the mutual interference
but became independent.
(4) The matters of the conclusion (2) and (3) were able to be seen not depend on
Reynolds number in this experiment range.

References

Hara, S., Suzuki, T.: Effects of end-plate dimensions on hydrodynamic forces on circular cylinders
oscillating sinusoidally in water. Trans. Jpn. Soc. Mech. Eng. 55(513), 1368–1371 (1989)
Hara, S., Suzuki, T.: Effects of end-plate dimensions on hydrodynamic forces on square cylinder
oscillating sinusoidally in water. Trans. Jpn. Soc. Mech. Eng. 57(533), 14–17 (1991)
Visual Observation of the Vortex Flow from the Same Diameter Circular … 117

Nishioka, M.: On the formation process of Karman vortex street. In: Special Publication of National
Aerospace Laboratory, SP vol. 18, pp. 41–45 (1992)
Norberg, C.: An experimental investigation of the flow around a circular cylinder: influence of
aspect ratio. J. Fluid Mech. 258, 287–316 (1994)
Szepessy, S., Bearman, P.W.: Aspect ratio and end plate effects on vortex shedding from a circular
cylinder. J. Fluid Mech. 234, 191–217 (1992)
Taneda, S.: Observation on unsteady separated vortices. Trans. Jpn. Soc. Aeronaut. Space Sci.
20(226), 600–610 (1972)
Yamada, H.: Effect of end plates on spanwise characteristics of flow around a circular cylinder.
Trans. Jpn. Soc. Mech. Eng. 58(552), 2368–2373 (1992)
Yokoi, Y.: Visual observation of the vortex flow structure from a circular cylinder with a partition
plate. Fedconf19, Prepr. of Jpn. Soc. Mech. Eng., pp. OS3–03–1-OS3–03–4. Toyohashi (2019)
Yokoi, Y., Hirao, K.: Vortex flow around an in-line forced oscillating circular cylinder. Trans. Jpn.
Soc. Mech. Eng. 74(746), 2099–2108 (2008)
Yokoi, Y., Kamemoto, K.: Initial stage of a three-dimensional vortex structure existing in a two-
dimensional boundary layer separation flow (visual observation of laminar boundary layer sepa-
ration over a circular cylinder from the side of a separated region). JSME Int. J., Ser. B 36(2),
201–206 (1993)
Experimental Study of the Porous
Emitter Installed in a Gas Porous Burner

Panuwat Chanmak, Wapee Sripunya, and Bundit Krittacom

Nomenclature

d PA Diameter of the porous absorber, cm

d PE Diameter of the porous emitter, cm
HE Heat exchanger
H PE Thickness of porous emitter, cm
LPG Liquefied petroleum gas
PA Porous absorber
PE Porous emitter
Qw Water flow rate fed into heat exchanger, L/min
T Temperature profile of porous burner, °C
V mix Volumetric premixed-gas velocity, m3 /h
X Burner length, cm
ε Effectiveness of the heat exchanger
φ Porosity
Φ Equivalent ratio

P. Chanmak · W. Sripunya · B. Krittacom (B)

Development in Technology of Porous Materials Research Laboratory (DiTo-Lab), Department of
Mechanical Engineering, Faculty of Engineering and Architecture, Rajamangala University of
Technology Isan, Nakhonratchasima, Thailand
e-mail: [email protected]
P. Chanmak
Department of Mechanical Engineering, Faculty of Industrial Education, Rajamangala University
of Technology Suvarnabhumi, Nonthaburi, Thailand

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 119
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_12
120 P. Chanmak et al.

1 Introduction

The benefit of the porous emitters (PE) is to absorb energy and generate higher
heat radiation. This is the reason why many engineers and researchers have been
encouraged to conduct numerical studies and experimental investigations of porous
emitters (PE) for the development of thermal efficiency (Vafai 2000; Jugjai et al.
2002; Jugjai and Polmart 2003; Thongtem and Jugjai 2008; Chuenchit and Jugjai
2012; Lekpradit and Krittacom 2013; Krittacom et al. 2011; Pranpanya and Krittacom
2016; Kaewchart and Krittacom 2017). If the two porous materials are separated from
each other (see in Fig. 1) (Krittacom 2009), higher heat radiation between the porous
absorber (PA) and the porous emitter (PE) is obtained. From the advantages of the
porous emitter (PE), many applications of heating work are widely used (Jugjai et al.
2003; Jugjai and Phothiya 2006, 2007; Krittacom and Amatachaya 2011).
According to where the flame was located, the gas porous burner was often divided
into two categories (Krittacom and Kamiuto 2009). The first type was the stabilized
flame inside the porous matrix or embedded flame with high porosity (φ) of 0.5–0.99.
The second one was the surface flame (spread flame) where the porosity (φ) of porous
burner became below 0.5. In the last five years, the application of a low porosity case
of the porous burner played a role as PA without PE was experimentally studied by
our previous works (Sripunya et al. 2016; Amatachaya and Krittacom 2017). The heat
exchanger of staggered tube banks was installed below the porous burners (PB) or the
porous absorber (PA). Water was used as the working fluid and liquefied petroleum
gas (LPG) was adopted as fuel. The porous media packed as a porous matrix was
made of an alumina-cordierite ball (Al-Co). Noticeably, the results of our experiment
confirmed that the temperature (T ) at the exhaust gas outlet zone was existed high
(over 420 °C). From this behavior, the high temperature at the exit zone should be
not ignored. Thus, to improve the combustion mechanism and thermal efficiency of
the heat exchanger, heat recirculation from such temperature was needed. The PE
was one of the choices to solve this point.

Fig. 1 Property of porous absorber (PA) and porous emitter (PE) (Krittacom 2009)
Experimental Study of the Porous Emitter Installed in a Gas Porous Burner 121

From the advantage of the porous emitter (PE) on the gas porous burner, the PE was
incorporated with the PA in the present work. The PE was installed at the outlet zone
of exhaust gas below the tube-bank heat exchanger. Similar to earlier experiments
(Sripunya et al. 2016; Amatachaya and Krittacom 2017), the operating fluid in cold-
line and fuel continued to be water and liquefied petroleum gas (LPG). To propose
the performance of the present application and the combustion mechanism from the
influence of PE, the temperature distribution along the porous burner (T ) and the
heat exchanger-effectiveness (ε) were explained.

2 Experimental Apparatus

Figure 2 presents the schematic diagram of the porous emitter installed in a gas porous
burner. The experimental equipment consisted of eight sections, i.e., the LPG and air
mixture, the injection chamber, the porous absorber (PA), the combustion zone and
the heat exchanger of staggered tube-bank (HE), the porous emitter (PE), the exhaust
gas outlet, the working-fluid system, and the measurement equipment. The premixed
gas (air and LPG) was a mixture and flow into the injection chamber. The porous
absorber (PA) and the porous emitter (PE) were made of an alumina-cordierite ball
(Al-Co) with a mean diameter of 0.5 cm and a porosity (φ) of 0.413. The 19 mm
outside diameter and 1 mm thickness of the SUS 304-stainless tube were modified
to form 15 tubes for a staggered tube-bank heat exchanger. Pitch measurements for
the transverse (ST), longitudinal (SL), and diagonal (SD) directions were 35 mm,
35 mm, and 39 mm, respectively. The working fluid of the heat exchanger was water
at a continual temperature of 29 ± 1 °C. The PE was installed at the outlet zone of
exhaust gas. In the present study, alumina-cordierite (Al-Co) was also employed as
the examined PE with three different thicknesses (H PE ): 2.5, 5, and 10 cm.
To understand the present work, the experimental procedure was described as
follows in this paragraph. In the beginning, the premixed-gas mixture (air and LPG)
was supplied into the injection chamber. The mixture flowed through the porous
absorber as a preheating process and, then, burned at the exit of the porous surface.
The absorption of energy from high temperatures occurred in the PA and the PE. The
heat transfer process between the working fluid (water) and the high temperature of
flame was obtained at the combustion zone and HE. Twenty N-type thermocouples
were attached along the length of the testing apparatus (PA, HE, and PE) to monitor
the profile of temperature (T ). The inlet and outlet temperatures of the water (T c, in
and T c, out ) were also monitored by the N-type thermocouples.
122 P. Chanmak et al.

Fig. 2 Schematic diagram of the porous emitter installed in a gas porous burner

3 Results and Discussion

3.1 Influence of Thickness of Porous Emitter (HPE )

The influence of the thickness of porous emitter (H PE ) on temperature distribution

(T ) in the length of the experimental burner is presented in Fig. 3. The testing condi-
tions included a volumetric premixed-gas velocity (V mix ) of 30 m3 /h, a diameter of
the porous emitter (d PE ) of 0.5 cm having appeared porosity (φ) of 0.413, an equiv-
alent ratio (Φ) of 0.8, and a water flow rate fed into the HE (Qw ) of 30 L/min. As
the HPE rose, it was discovered that the status of T increased because a higher thick-
ness of porous emitter can raise the energy absorption (Krittacom and Amatachaya
2011; Krittacom and Kamiuto 2009; Peamsuwan and Krittacom 2014; Krittacom
and Tongbai 2018).
Figure 4 demonstrates the influence of PE thickness (H PE ) on the effectiveness (ε)
with the same experimental conditions as shown in Fig. 3. According to the findings,
which can be seen in Fig. 4, the value of ε marginally rose as greater HPE was attained
as a result of the little ability in energy absorption between the working fluid (water)
of a heat exchanger and the high temperature of the flame (Krittacom and Kamiuto
2009).

3.2 Influence of Volumetric Premixed-Gas Flow Rate (Vmix )

Figure 5 depicts the impact of the volumetric premixed-gas velocity (V mix ) on the
temperature profile (T ) along the experimental length. The experiment was conducted
Experimental Study of the Porous Emitter Installed in a Gas Porous Burner 123

1400
dPE = 0.5 cm Non PE
= 0.8 HPE = 2.5 cm
1200 HPE = 5 cm
Vmix = 30 m3/hr
Qw = 30 L/min HPE = 10 cm
1000
HPE = 2.5 cm
800
T ( o C)

HPE = 5 cm
HPE = 10 cm
600

400 PA

200
HE PE

0
0 20 40 60
X (cm)

Fig. 3 Influence of H PE on T

100
dPE = 0.5 cm

= 0.8
Vmix = 30 m3/hr
80
Qw = 30 L/min

60
(%)

40

20

0
0 2 4 6 8 10 12
HPE (cm)

Fig. 4 Influence of H PE on ε

in this situation: a PE thickness (H PE ) of 10 cm, a diameter of the porous emitter

(d PE ) of 0.5 cm having appeared porosity (φ) of 0.413, an equivalent ratio (Φ) of 0.8,
and a water flow rate fed into the HE (Qw ) of 30 L/min. The results showed that the
trend of T was explicitly enhanced with increasing V mix , as shown in Fig. 5. When
the fuel was initially introduced to the system, more potent combustion should have
emerged, followed by an increase in T (Glassman 1996).
124 P. Chanmak et al.

1400
3
dPE = 0.5 cm Vmix = 10 m /hr
3
Vmix = 15 m /hr
HPE = 10 cm
1200 3
Vmix = 20 m /hr
= 0.8 3
Vmix = 25 m /hr
Qw = 30 L/min 3
Vmix = 30 m /hr
1000 3
Vmix = 35 m /hr

800
T (ο C)

600

400 PA

200
HE PE
0
0 20 40 60
X (cm)

Fig. 5 Influence of V mix on T

In relation to Figs. 5, 6 illustrates how volumetric premixed-gas velocity (V mix )

affects the effectiveness (ε). The experiment’s setting was comparable to Fig. 5. The
experiment revealed the following: the value of ε was decreased as increasing V mix
because the system had an increased amount of hot gas. This was further explained by
an effectiveness equation of the difference on the staggered tube-bank heat exchanger
(Sripunya and Krittacom 2021). Moreover, the value of ε explicitly increased with
installed the PE as compared in Fig. 6. This finding was the fact phenomena resulted
from the benefit of PE (Jugjai et al. 2002; Lekpradit and Krittacom 2013; Jugjai and
Phothiya 2007; Krittacom and Amatachaya 2011).

4 Conclusion

The research investigation of the porous emitter (PE) installed in a gas porous burner
with LPG as fuel and the heat exchanger of staggered tube-bank with the water
as working fluid was performed. The combustion phenomena, i.e., the profile of
temperature on a burner length (T ) and the HE effectiveness (ε) were investigated. In
the present study, alumina-cordierite (Al-Co) was employed as the examined porous
emitter with different thicknesses (H PE ). The T tended to increase with the H PE and
the V mix . The quantity of ε was decreased as increasing V mix but increased slightly
as increasing H PE . Moreover, the value of ε explicitly increased with the installation
of the PE.
Experimental Study of the Porous Emitter Installed in a Gas Porous Burner 125

100
HPE = 10 cm
Non PE
= 0.8 dPE = 0.5 cm
Qw = 30 L/min
80

60
(%)

40

20

0
10 20 30 40
Vmix(m3/hr)

Fig. 6 Influence of V mix on ε

References

Amatachaya, P., Krittacom, B.: Combustion mechanism of gas porous burner installed an in-line
tube-bank heat exchanger. Energy Procedia 138, 50–55 (2017)
Chuenchit, C., Jugjai, S.: Effect of tube length on combustion characteristics of a self-aspirating
radiant tube burner (SRTB). J. Res. Appl. Mech. Eng. 1(2), 19–23 (2012)
Glassman, I.: Combustion, 3rd edn. Academic Press (1996)
Jugjai, S., Polmart, N.: Enhancement of evaporation and combustion of liquid fuels through porous
media. Exp. Thermal Fluid Sci. 27(8), 901–909 (2003)
Jugjai, S., Wongpanit, N., Laoketkan, T., Nokkaew, S.: The combustion of liquid fuels using a
porous medium. Exp. Thermal Fluid Sci. 26(1), 15–23 (2002)
Jugjai, S., Phothiya, C.: Cyclic flow reversal combustion (CFRC) of liquid fuel within porous
medium. In: The 20th Conference of the Mechanical Engineering Network Thailand (18–20
October 2006) (in Thai)
Jugjai, S., Phothiya, C.: Liquid fuels-fire porous combustor-heater. Fuel 86(7–8), 1062–1068 (2007)
Jugjai, S., Chuwisit, K., Chalermpornchai, K., Iampai, N.: Potential for porous-meidum combustion
technology as applied to liquid fuel without spray atomization. In: The 17th Conference of the
Mechanical Engineering Network Thailand (15–17 October 2003) (in Thai)
Kaewchart, K., Krittacom, B.: Comparison of combustion behavior between solid porous burners
installed the porous emitter and non-porous emitter. Energy Procedia 138, 2–7 (2017)
Krittacom, B.: Studies on thermal characteristics of open-cellular porous burners. Doctor Engi-
neering Dissertation, Oita University (2009)
Krittacom, B., Kamiuto, K.: Radiation emission characteristics of an open-cellular porous burner.
J. Therm. Sci. Technol. 4(1), 13–24 (2009)
Krittacom, B., Tongbai, P.: The effective thermal conductivity (ETC) of spherical packed-bed porous
media for modification of the color surface as black color. Mater. Sci. Forum 911, 100–104 (2018)
Krittacom, B., Amatachaya, P., Srimuang, W., Inla, K.: The pack-bed sphere liquid porous burner.
Int. J. Mech. Mech. Eng. 5(9), 1917–1921 (2011)
126 P. Chanmak et al.

Krittacom, B., Amatachaya, P.: The effect of porous emitter on the combustion of diesel oil porous
burner. In: Proceeding of the 2nd TSME International Conference on Mechanical Engineering,
pp. 19–20. TSME-ICOME (19–21 October 2011)
Lekpradit, T., Krittacom, B.: Enhancement of combustion of liquid fuel on porous burner using
porous emitter. Eng. J. Res. Dev. 24(2), 36–44 (2013). (in Thai)
Peamsuwan, R., Krittacom, B.: Estimating correlation of effective thermal conductivity on the
spherical packed bed porous media. Appl. Mech. Mater. 575, 282–286 (2014)
Pranpanya, S., Krittacom, B.: Effect of porous–emitter porosity on the combustion behavior of a
bi-layer liquid porous burner. In: The 30th Conference of the Mechanical Engineering Network
Thailand (5–8 July 2016) (in Thai)
Sripunya, W., Krittacom, B.: Improvement of thermal effectiveness of the tube-bank heat exchanger
in a porous burner with installed the porous emitter. In: The 3rd International Conference on
Energy and Power (18–20 November 2021).
Sripunya, W., Amatachaya, P., Krittacom, B.: The effect of factors on effectiveness of the tube-
bank heat exchanger installed in a porous burner using LPG as fuel. In: 30th Conference of the
Mechanical Engineering Network Thailand (5–8 July 2016) (in Thai)
Thongtem, P., Jugjai, S.: Nonpremixed porous burners for gaseous fuels. In: The 22nd Conference
of the Mechanical Engineering Network Thailand (15–17 October 2008) (in Thai)
Vafai, K.: Hand Book of Porous Media. CRC Press, New York, USA (2000)
Experimental and Numerical
Investigations of Effects of Wall Friction
and Vibrational Amplitude
on Convection Phenomena in Granular
Materials Under Vertical Vibrations

Tanapon Yachai, Itthichai Preechawuttipong ,

and Pawarut Jongchansitto

1 Introductions

In the engineering context, granular materials can be defined as a cluster of solid parti-
cles, whose macroscopic behaviors are controlled by the interaction forces between
particles, for example, rock, soil, sand, plastic pellets, pharmaceuticals, etc. Indeed,
particle mixing and segregation often occur in many industrial processes involved the
transportation and handling of granular mixtures under vibrated or shaken systems.
Convection is one of key driving mechanism for such phenomena (Windows-Yule
et al. 2013, 2014; Windows-Yule 2016). The strength of the granular convection
depends on many parameters such as number of particles (Windows-Yule et al.
2013, 2014; Windows-Yule 2016), aspect ratio of container dimensions (Windows-
Yule and Parker 2014; Hsiau et al. 2011), material properties (Windows-Yule et al.
2013, 2014), dimensionless vibrational acceleration (Γ) (Zhang et al. 2014), and wall
friction (Song and Zhang 2020). Among these parameters, the wall friction domi-
nates compared to the others (Windows-Yule and Parker 2014; Then et al. 2020).
Only, a few previous studies considered the strength of convection due to the friction
coefficient between particle and wall container. Grossman (Grossman 1997) numeri-
cally examined the effects of boundary conditions on convection in vibrated granular
systems. The results showed that the net shear force exerted by the walls on the parti-
cles determines the strength and direction of the convection rolls. Hsiau et al. (2002)
and Zeilstra et al. (2008) performed experiments and simulations to investigate the
effect of sidewalls roughness on the convection under vertically vibrated granular
bed. The results showed that the strength of the convection increases with the side-
walls roughness. Nevertheless, the effect of the particle–wall friction on convection

T. Yachai (B) · I. Preechawuttipong · P. Jongchansitto

Department of Mechanical Engineering, Faculty of Engineering, Chiang Mai University, 239
Huay Kaew Rd., Muang 50200, Chiang Mai, Thailand
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 127
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_13
128 T. Yachai et al.

phenomena has been still unclear. Therefore, the present study is aimed to investigate
the influence of wall friction and vibrational amplitude on convection phenomena
in a vertically vibrated granular system by means of experiments and reproduced
numerical simulations.

2 Methodology

The experiment and simulation setup as well as the analysis of convection through
the dimensionless angular velocity is presented in this section.

2.1 Experimental Setup

Figure 1A shows a preparation of our experiments. A particle was made from the
polymethyl methacrylate (PMMA). The particle has a mean diameter (Dm ) of 10 mm
and a thickness of 5 mm. The particle density obtained from the supplier datasheet is
1190 kg/m3 . Each particle was then randomly placed inside the container covering
the front and back sides by a clear PMMA sheet, as shown in Fig. 1b. The aspect
ratio between bed height and bed width was equal to 1. A granular sample consisted
of 4000 particles with a uniform distribution of five different diameter sizes D ∈
[9.0, 9.5, 10.0, 10.5, 11.0] mm. A fluorescent sticker with the diameter of 5 mm was
also attached to each particle. This sticker is glowed in the dark under UV light, as
presented in Fig. 1c. Note that the center of the sticker was coincident to the centroid
of the particle. This is very novel and useful technique for extracting the location of
the particles from an image processing. To vary the coefficient of friction between
particle and wall container, seven types of materials consisting of a polytetrafluo-
roethylene (PTFE) tape, a polyethylene (PE) tape, a masking tape, a sandpaper No.
10, 000, a sandpaper No. 5000, and a polyurethane (PU) tape were attached to the
sidewalls of the container. These materials provide the coefficient of the particle–wall
friction μw ∈ [0.15, 0.26, 0.36, 0.46, 0.61, 0.81, 0.95], respectively. The coefficient
of interparticle friction (μp ) is equal to 0.5. Note that μw and μp were obtained from
the measurement based on the ASTM D4918-97 standard (ASTMD4918-97 2007).
The granular sample under a gravity was subjected to a vertical vibration with a
sinusoidal wave. The vertical vibration was provided by a vertical testing machine
through a scotch-yoke mechanism, as shown in Fig. 1a. The level of vibration was
controlled by dimensionless acceleration (Γ). The dimensionless acceleration can be
defined as Γ = Aω2 /g, where g is the gravity, A is the vibrational amplitude, and ω
= 2π ƒ is the vibrational frequency. In this study, the dimensionless acceleration was
fixed at 5 for a chaotic state of granular system (Chen 2020), while the vibrational
frequency was adjusted suitable for the vibrational amplitude of 0.5Dm and 1.0Dm ,
respectively. During the experiments, the center of each particle was recorded by a
high-speed camera with 240 fps. Each considered frame of the video was used to
Experimental and Numerical Investigations of Effects of Wall Friction … 129

Fig. 1 a Experimental preparation and sample of tested particles. b A snapshot of granular sample
under visible light. c A snapshot of granular sample under UV light. d Result obtained from image
processing. Note that the plus sign represents the center of the particle

extract the location of the particle center by an image processing technique. Figure 1d
presents the obtained image after performing the image processing.

2.2 Numerical Setup

The simulations were carried out by means of the non-smooth contact dynamics
(NSCD) method, which was originally developed by Moreau and Jean (Jean 1999).
This method is suitable for simulating large assemblies of rigid massive particles.
In this method, the equations of motion (i.e., Newton’s second laws) are integrated
over a small-time step and combined with the kinematic constraints resulting from
contact interactions. An open-source software named “LMGC90” (http://mimetics-
engineering.fr/index.php/en/lmgc90-2/), which was developed in Montpellier, was
used for the simulations in this study. The 2D simulations were compared with
the experiments under the same conditions. The experiments and simulations were
performed up to 300 s. The time step of simulation used for updating the particle
position was set to 1/27 fs. For a comparison between experiments and simulations,
an average velocity field was plotted to reveal the convection. The average velocity
field can represent the motion of each particle corresponding to the flow behavior
of granular media. To determine the average velocity field, the center locations of
130 T. Yachai et al.

Fig. 2 Schematic of a phase of vibration corresponding to the sinusoidal wave. Note that the dotted
line is the reference line of the container movement

each particle at each phase of 4π were exported every 4 s. It is clearly seen that the
selected phase of 4π is the highest compaction after a rearrangement of the particles,
as illustrated in Fig. 2. These locations were then employed to compute the velocity
of each particle for every 4 s. Eventually, the average velocity field was determined
by averaging all the velocity data for 300 s.

2.3 Dimensionless Angular Velocity (W)

In the previous studies, the granular temperature (T ) (Windows-Yule and Parker

2014) and the dimensionless convection flow rate (J) (Windows-Yule 2016) are
commonly applied as the indicators to describe the quantity of convection. The
granular temperature is used to quantify the kinetic energy of the flow, while the
dimensionless convection flow rate is used to quantify the mass flow rates at the
horizontal plane crossing the center of convection. However, the major disadvantage
of both indicators is that they cannot explain the direction of convection and the
convection at low velocity. Therefore, the dimensionless angular velocity (W ) was
introduced in the present study to describe the quantity of the strength convection.
The dimensionless angular velocity is defined as the ratio of the angular velocity of
the convection to the angular frequency of the sinusoidal wave of vibration, which
can be expressed as:

WL + WR
W = (1)
2ω
where W L and W R are the average angular velocity of the left and right convection
rolls, respectively. The center of rotation for each convection roll can be obtained
under the concept that a point having the highest number of velocity vectors rotating
around that point, as represented by the blue circle in Fig. 3a. The W L and W R can
be determined by:
Experimental and Numerical Investigations of Effects of Wall Friction … 131

Fig. 3 a An illustration of a velocity field for determination of the center of the convection. b The
number of vectors rotating around point (N r ). c An illustration of the normal convection. d An
illustration of the reverse convection

Σ NL ( |vi | sin θ ) Σ NR ( |vi | sin θ )

i=1 |ri | i=1 |ri |
WL = and WR = (2)
NL NR

where N L and N R are number of vectors on left and right convection roll, respectively.
θ is the smallest angle between the magnitude of a radial vector |ri | the magnitude
of a velocity vector |vi | of ith vector. The term of sinθ is positive when vi rotates
counterclockwise about the considered point. On the other hand, sinθ is negative
when vi rotates clockwise about the considered point. Figure 3b shows a surface
plot of the number of vectors rotating about each considered point. Two blue crosses
represent the center of the convection having the highest number of rotating vectors.
Hence, W varies from −1 to 1. The dimensionless angular velocity is equal to 1,
corresponding to the largest strength of the normal convection, as shown in Fig. 3c.
On the contrary, the largest strength of the reverse convection (see in Fig. 3c) exhibits
when W = −1. When the dimensionless angular velocity is close to 0, it means that
the convection does not occur.

3 Results

In this section, experimental and simulation results of the convection due to the
effects of the wall friction and the vibrational amplitude are discussed.
132 T. Yachai et al.

3.1 Effect of the Wall Friction

Figure 4A shows the evolution of the dimensionless angular velocity (W ) as function

of the particle–wall friction obtained from the experiment and simulation results for
both the vibrational amplitude of 0.5Dm and 1.0Dm . The trend of both experiment and
simulation results is well correlated. It can be said that the dimensionless angular
velocity starts to increase and then stabilizes at a constant value after μw > 0.6.
Figure 4b presents the average velocity fields of points a, b, and c at different levels
of μw . For μw < 0.26, it was observed that the convection does not appear, i.e.,
W ≈ 0. The strength of the convection then tends to increase with μw until reached
its maximum value. Finally, the strength of the convection is steady after μw > 0.6.
It was also found that the magnitude of the dimensionless angular velocity for the
experiments is slightly greater than that for the simulations. This can be explained
by the front and rear covers of the container used in the experiments affect to the
flow behavior in granular system, which is consistent to the previous study (Fortini
and Huang 2015).
Let us consider the strength of the convection when μw > μp , it is clearly seen
that only the effect of the particle–wall friction cannot cause an enhancement of
the dimensionless angular velocity. Therefore, the particle at the container wall is
required the interparticle friction from an adjacent particle to migrate downward into

Fig. 4 a Dimensionless angular velocity as a function of the particle–wall friction obtained from
experiments and simulations for the vibrational amplitude of 0.5Dm and 1.0Dm . b Examples of the
average velocity field of points a–c corresponding to Fig. 4a
Experimental and Numerical Investigations of Effects of Wall Friction … 133

Fig. 5 Free-body diagram

of particles at the container
wall while the container is
moving down

the bottom of the container (Song and Zhang 2020). This can be explained by a
free-body diagram, as illustrated in Fig. 5, where F n, wall is the normal force acting
by the container wall, f w is the particle–wall friction force, F n, AB is the normal force
between particle A and particle B, and f p is the friction force between particle A and
particle B.

3.2 Effect of the Vibrational Amplitude

It was observed that when the vibrational amplitude was increased, the strength of
the convection is increased for any μw . This is due to the fact that larger vibrational
amplitude provides the larger displacement of the container, thus leading to more
spaces between the particles for the movement (Zhang et al. 2014; ASTMD4918-
97 2007). Time for the particle movement is also increased when the vibrational
amplitude was increased, thus causing the higher velocity. Comparison with the
simulations, the strength of the convection due to the vibrational amplitude obtained
from the experiment’s overestimates for any μw .

4 Conclusion

This work pointed out the first step for a better understanding of the effect of the
wall friction and the vibrational amplitude on convection phenomena in 2D granular
system under vertical vibration. The appearance and the strength of convection mech-
anisms were characterized by the dimensionless angular velocity (W ). All cases of
an occurrence of the convection were normal. The convection started to occur when
the wall friction μw > 0.26 and then reached to the maximum for μw close to μp . The
strength of convection was no longer increased for μw > μp because the particles
rotated about itself. In addition, the convection was much clearer when the vibra-
tional amplitude was increased. It must be noted that the experimental and numerical
134 T. Yachai et al.

results are correlated, but an underestimation of the results by the simulation was
observed. The contribution of this work can be considered as the basic data for the
segregation and mixing of the binary materials.

Acknowledgements This work was supported by TA/RA Scholarships from Graduate School,
Chiang Mai University and Graduate Research Assistant Scholarships from Faculty of Engineering,
Chiang Mai University.

References

ASTMD4918-97 (2007)
Chen B.: Powder Technol. 363, 575–583 (2020)
Fortini, A., Huang, K.: Phys. Rev. E 91, 032206 (2015)
Grossman, E.: Phys. Rev. E 56, 3290 (1997)
Hsiau, S., Wang, P., Tai, C.: AICHE J. 48, 1430–1438 (2002)
Hsiau, S., Liao, C., Sheng, P., Tai, S.: Exp. Fluids 51, 795–800 (2011)
Jean, M.: Comput. Method. Appl. M 177, 235–257 (1999)
LMGC90 Homepage. http://mimetics-engineering.fr/index.php/en/lmgc90-2/
Song, X., Zhang, G.: Powder Technol. 372, 40–47 (2020)
Then, H., Sekiguchi, T., Okumura, K.: Soft Matter 16, 8612–8617 (2020)
Windows-Yule, C.: New. J. Phys. 18, 033005 (2016)
Windows-Yule, C., Parker, D.: Powder Technol. 261, 133–142 (2014)
Windows-Yule, C., Rivas, N., Parker, D.: Phys. Rev. Lett. 111, 038001 (2013)
Windows-Yule, C., Weinhart, T., Parker, D., Thornton, A.: Phys. Rev. E 89, 022202 (2014)
Zeilstra, C., et al.: Powder Technol. 184, 166–176 (2008)
Zhang, F., Wang, L., Liu, C., Wu, P., Zhan, S.: Phys. Lett. A 378, 1303–1308 (2014)
Comparative Study of Radiative Flux
from Mushroom BLEVE Model

Anas Mbarki, Aziz Ettahir, and Abderraouf Guelzim

Nomenclature

VFB The total volume of the fireball (m3 )

VS Spherical volume (m3 )
VC Cylindrical volume (m3 )
FL Filing level of the vessel
td The fireball duration (s)
tlo The fireball Liftoff Time (s)
Dmax Maximum fireball diameter (m)
DFB The fireball diameter (m)
rFB The fireball radius (m)
HFB The height of the center of the fireball (m)
rS The radius of the spherical part of the mushroom (m)
rC The radius of the disk of the cylindrical part of the mushroom (m)
hC The height of the cylindrical part of the mushroom (m)
m tlo The mass of fuel involved on the liftoff time (kg)
ρ The density (kg/ m3 )
α The coefficient of proportionality or the volume ratio
L∗ The linear scale length
U∗ The linear scale velocity (the characteristic velocity owing to the buoyancy
force) (m/s)
t∗ The linear scale time
V* Characteristic volume

A. Mbarki (B) · A. Ettahir

Materials, Energy and Acoustics Team (MEAT), Mohammed V University in Rabat—EST Salé,
Rabat, Morocco
e-mail: [email protected]
A. Guelzim
Forensic Institute of the Royal Gendarmerie, Rabat, Morocco

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 135
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_14
136 A. Mbarki et al.

ΔH c Heat of combustion (kJ/kg)

M0 Mass of fuel involved (kg)
ρa The density of ambient air density (kg/ m3 )
cp,a The specific heat capacity of air (kJ/kg)
Ta The temperature of ambient air (K)
a Constant
Fr The Froude number
U0 The release velocity (m/s)
g Acceleration due to gravity (m/s)
X The distance (m)
FFB The max geometric view factor for a spherical emitter (−)
Cmax The max geometric view factor for a cylindrical emitter (−)
CV The vertical view factor for a cylindrical emitter (−)
Ch The horizontal view factor for a cylindrical emitter (−)
q The radiative flux (kW/m2 )
Es The emissive power of fire per unit surface area (W/m2 )
fh The radiative fraction of the heat of combustion (−)
M The initial mass of fuel in the fireball (mass) (kg)
τa n The atmospheric transmissivity (−)
R The fractional relative humidity
Pv The saturated vapor pressure of water at the ambient temperature (Pa)

1 Introduction

The prevention of risks associated with major industrial accidents is a necessity for
those responsible for site safety management. Since the slightest error in estimating
its dangers may have disastrous consequences on both people and structures, hence
the need to set up efficient and valid analytical models for estimating the effects of
these phenomena. This study is an attempt to model the thermal effects of a Bleve,
taking into consideration the numerous hazards that can influence this phenomenon,
namely the time and the distance from the source of the radiation, and finally the
behavior of the flaming gas cloud facing the upward entrainment of ambient air. The
mushroom Bleve model can be an alternative, making it possible to take into account
all its hazards by containing them in a single model, which varies throughout the
combustion duration.

2 Methods

During failure or total rupture of a storage tank for flammable chemicals or hydro-
carbons, the increased difference in the pressure of the products contained and the
Comparative Study of Radiative Flux from Mushroom BLEVE Model 137

ambient pressure favors the evaporation the liquid almost instantaneously; the spread
evaporated liquid forms a cloud which gradually grows following the entrainment
of the ambient air toward the interior of the cloud. The operation accentuated by the
phenomena of turbulence and ascending velocity, as observed in different experi-
ments, gives a shape resembling mushroom to the fireball in the phase of volume
growth. These phenomena are not taken into account by the dynamic model which
claims that the shape of the cloud is almost spherical and that only its size varies
with time.
In this work, we calculate the radiative flux from a proposed mathematical model,
and then compare it with that obtained by using the dynamic forecasting model,
which assumes the fireball to have a spherical geometry. This proposed mathematical
model is applied to a reservoir with a total volume of 10,796 m3 , with a filling
level of 40%, under a pressure of 15 bars, and then compared to Martinsen and
Marx’s dynamic model (Martinsen et al. 1999; INERIS 2017; Johnson and Pritchard
1990; https://www.aiche.org/resources/publications/books/guidelines-consequence-
analysis-chemical-releases).

3 Discussion of Results

3.1 The Real Fireball Geometry: “Mushroom Shape”

The model of fireball’s mushroom shape behavior is similar to the one described by
the following figure (Fig. 1); the total volume V FB is composed of a spherical volume
and a cylindrical volume denoted, respectively V S and V C .
So, we have:

VFB = VS + VC (1)

Fig. 1 Mushroom Bleve volume

138 A. Mbarki et al.

The volumes of the geometric shapes (spherical and cylindrical) obtained are
proportional to the volume of the initial fireball, considering that:

VC = α · VFB . (2)

VS = (1 − α) · VFB (3)

where α: is a dimensionless constant between 0 and 1, it is introduced to express

the proportionality between the volume of the sphere of the Martinsen and Marx
model (Martinsen et al. 1999) and the volume of the spherical part of the considered
mushroom geometry.

3.2 Modeling of the Mushroom Shape Approach

As observed in several accidents and confirmed in various recent experiments (British

Gas experiments), the mushroom shape model illustrates a more realistic represen-
tation of the true behavior of fireballs. By employing adequate equations that take
into account the fireball growth, the lift-off, the changing radiation characteristics,
the visible cloud expansion velocity, the fuel properties (boiling temperature, heat
of evaporation), the uplift velocity time scale of the fireball (Makhvilade et al. 1996,
1997, 1999a, 1999b, 2000; Makhviladze and Yakush 2005; Gostintsev 2022), the air
entrainment as well as the turbulence associated to the ascendant air flux.
During the first third of Bleve’s life, the analysis of the behavior associated with
phenomenon’s study is similar to that used in a dynamic model. The fireball has a
spherical shape.
Thereafter, during the beginning of the second third of the burning time, the
fireball stretches to take a new shape similar to that of a mushroom, formed by the
superposition of two geometric shapes, a spherical part above the cylindrical part.
The radius of this spherical part is less than that of the initial sphere; however, it
always remains proportional to it.

3.3 Determination of the Volume’s Characteristics Forming

the Mushroom and of the Volume Ratio A

The radius, the height of the cylindrical part of the mushroom, and its radius are
given by the following equations:
/
2 α
rC = · √ / · rFB (4)
3 t− 3
1−α 3
Comparative Study of Radiative Flux from Mushroom BLEVE Model 139

h C = HFB − rS (5)

√
rS = 1 − α · rFB
3
(6)

Determination of Alpha According to Makhviladze et al. (https://www.aiche.org/

resources/publications/books/guidelines-consequence-analysis-chemical-releases;
Makhvilade et al. 1996, 1997, 1999a), the characteristic volume V∗ of the fireball
is determined by equating the total combustion energy of the fuel ΔHC M0 to the
thermal energy of the cloud ρb cp (Tb − Ta )V∗ , where the subscript b corresponds to
the combustion products.
And using the dimensionless analysis, for the linear scale length and the
characteristic velocity owing to the buoyancy forces, we find:
/
V∗ = ΔHc M0 ρa cp,a Ta (7)

L ∗ = v∗1/ 3 (8)

U∗ = (L ∗ g)1/ 2 (9)

Reducing the initial Eqs. (7–9), the Froude number which describes the buoyancy
of the phenomenon can thus be expressed by:
( )2 ( )
U0 U02 ρa cp,a T 1/ 3
Fr = = (10)
U∗ g ΔHc Ma

After using the empirical relation which gives the maximum diameter of the
fireball as the mass’s function of the fuel entrained by ambient air:

Dmax = a · M0 /
1 3
(11)

Then, the explicit form of the dimensionless constant α is as follows:

[ ]
1 ( 2/ )3
α =1− · U0 a · gFr (12)
M0

where α is the constant of the ratio between the maximum diameter of the fireball
and the mass to the power 1/3.
140 A. Mbarki et al.

3.4 Radiative Flux Measurement

The numerical values are calculated from the input data proposed in Sect. 2
methodology; the equations constituting the dynamic model are illustrated in
(Eqs. 1–3).
First Step: 0 < t < td /3 We calculated the heat flux q at a certain distance from the
fire experienced by the receiver per m2 , utilizing:

q = E S · FFB · τa (13)

E s (t) = 0.0133 · f h · Hc · M 12 (14)

E s (t) = 359.347, for → 0 ≤ t ≤ 2.127

/ [ ]
FFB (X, t) = DFB (t)2 4 · HFB (t)2 + X 2 (15)

( )
1 1 2
8.664 · M / 4 t / 3
FFB (X, t) = [( )2 ]
1 1
4· 4.332 · M / 4 t / 3 + X2

/ ( )
2 2
FFB (X, t) = 3772.905 · t / 3 4 · 943.226 · t / 3 + X 2

( [/ ])
D(t) −0.09
τa (x, t) = 2.02 · RPV HFB (t)2 + x 2 − (16)
2
[/ ]−0.09
2/ 1/
τa (x, t) = 1.03 · 943.226 · t 3 + X − 30.712 · t 3
2

Second Step: td /3 < t < td The radiative flux q at a certain distance from the fire
experienced by the receiver, using:

q = qspherical + qcylindrical (17)

q = E S · (Fspherical + Fcylindrical ) · τa (18)

( )
3 t 1
E s (t) = E max · 1− , for → td < t ≤ td (19)
2 td 3
Comparative Study of Radiative Flux from Mushroom BLEVE Model 141

E s (t) = 539 − 84.48 · t, for → 2.127 < t ≤ 6.381

[ ]
1 ( 2/ )3
α =1− · U0 a · gFr (20)
M0

α = 34.25% ≈ 34%

/
• rC = 0.381 · 1
t−0.29
· rFB

/√
rC = 15.353 t − 0.29

• h C = HFB − rS = 18.570 · t − 34.388

• rS = 0.871 · rFB = 34.388 m

FMushroommax = Fspherical + Fcylindrical (21)

/ [ ]
Fspherical = DS (t)2 4 · HFB (t)2 + X 2 (22)

/
Fspherical = 4730.103 4 · (344.852 · t 2 + X 2 )
/
Fcylindrical = CH2 + CV2 (23)

⎡ ⎤
( )1 ( )1/
1 ⎣ −1 X + rC / 2 X 2 − rC2 + h 2C (X − r ) · A 2
⎦
· tan−1
C
CH = · tan − √
π X − rC A·B (X + rC ) · B
(24)
[ ( )1 /2
1 rC −1 h 2C h C · ( A − 2X · rC )
CV = · tan + √
π X X 2 − rC2 X · A·B
( )1 /2 ( )1 ]
−1 (X − r C ) · A h C −1 X − rC /2
tan − tan (25)
(X + rC ) · B X X + rC

With:

A = (X + rC )2 + h 2C (26)
142 A. Mbarki et al.

B = (X − rC )2 + h 2C (27)

( [/ ])
DS (t) −0.09
τa (X, t) = 2.02 · RPV HFB (t) + X −
2 2 (28)
2
[√ ]−0.09
τa (X, t) = 1.03 · 344.852 · t 2 + X 2 − 17.194 ∗ 2

3.5 Graphic Comparison Between the Two Models

Figures 2, 3, 4, 5, and 6 are plots of the same data calculated by the mushroom
approach compared to the measurements taken during British Gas test no. 4 and
Martinsen and Marx model, at several distance. These plots give us an overview on the
validity of the analytical predictions. who have been calculated from the mushroom
Bleve model, compared to the analytical prediction of Martinsen and Marx model
and to the measurements recorded during test no. 4 of the BG experience and this
at several times during the combustion time, depending on the distance of the target
from the tank. At the same time, it illustrates the variation of the incident thermal
radiation received as a function of distance and as a function of time. The variation
of the thermal radiation incident represented in Figs. 2, 3, 4, 5, and 6 shows a strong
similarity to the trend and a strong positive correlation almost perfect to the results
collected on the BG experiment no. 4.
The analysis of Fig. 7 gives us the following information:
The thermal dose of the spherical part during the last two thirds of the combustion
time of the Bleve is equal to the area s (104.112). While the thermal dose of the
cylindrical part during the same period of combustion of the Bleve is equal to the
area t (45.76).

Fig. 2 Radiative flux as a function of time calculated at 50 m from receiver

Comparative Study of Radiative Flux from Mushroom BLEVE Model 143

Fig. 3 Radiative flux as a function of time calculated at 75 m from receiver

Fig. 4 Radiative flux as a function of time calculated at 100 m from receiver

Fig. 5 Radiative flux as a function of time calculated at 125 m from receiver

144 A. Mbarki et al.

Fig. 6 Radiative flux as a function of time calculated at 200 m from receiver

Fig. 7 Contribution of the cylindrical part of the Bleve mushroom in the calculation of the incident
radiative flux at 50 m—measured during the simulation of BG test 4

The contribution of the cylindrical part is almost 30% of the thermal dose received
by a target moving away from 50 m from the center of the phenomenon and this during
the last two thirds of the combustion time of the Bleve, and more than 15% of the total
thermal dose received by a target exposed to the incident radiative flux throughout
the duration of the Bleve.
Comparative Study of Radiative Flux from Mushroom BLEVE Model 145

3.6 Discussion

The mushroom Bleve model has a great similarity with the dynamic model of
Martinsen. It can be seen that, the radiative fluxes are calculated as a function of
time and can be split into two parts:
• The first part at any instant t of the first third of the Bleve duration:
The two models give the same values of the radiative flux.
• The second part at any instant t is greater than the first third:
At the start of the second third, the radiative flux calculated by the dynamic model
is greater than that calculated by the mushroom Bleve model, by few milliseconds
for receivers located at a short or medium distance that does not exceed 100 m from
the center of the geometric mushroom shape. In addition, the difference between
the radiative fluxes calculated by the two models does not exceed 2.6 kw/m2 ; this
translates in a negative variance which does not exceed 6% for the radiative flux
calculated by our approach. While this duration is extended to more than one second
for receivers placed at a greater distance, and the difference between the radiative
fluxes calculated by the two models does not exceed 1.6 kw/m2 , this is transcribed
into a negative variance which does not exceed order of 7% for the radiative flux
calculated by our approach.
The ratio is quickly reversed, and the radiative flux calculated by the mushroom
Bleve model takes values higher than that calculated by the dynamic model and this
throughout the remaining existence time.
For receivers located at a short or medium distance, which does not exceed 100 m
from the center of the geometric mushroom shape, the difference between the radia-
tive fluxes calculated by the two models does not exceed 5 kw/m2 ; this translates
into a positive variance which does not exceed 24% for the radiative flux calculated
by our approach. While this variance does not exceed 1 kw/m2 , i.e., an increase of
less than 5% of the radiative flux calculated by our approach. The magnitude of
the calculation difference margin between the two models dissipates as the receiver
moves away from the radiation source.

4 Conclusion

A In this study, two models relating to the effects of radiative transfers of a fireball
generated by a Bleve were compared:
• Martinson and Marx’s dynamic model, whose fireball geometry is considered to
be spherical;
• Andour proposed model, whose fireball’s geometry is considered to have a
mushroom shape.
146 A. Mbarki et al.

The comparison of the two models was made based to the experimental data for
medium-scale fireball tests realized by Johnson and Pritchard (1990) as part of the
experiments conducted by British Gaz and cited by INERIS (INERIS 2017).
The mushroom Bleve model is an attempt in which we made sure to take into
account the real behavior of the fireball through the combustion duration following
the analysis of the media library of the tests carried out, in order to define its temporal
behavior.
This attempt will allow us to minimize the margin of error relating to the calcu-
lation of the thermal dose received by an individual relative to its distance from the
radiation source. Thus, the effective distances and the impact on structure and people
are expected to be more severe.
Although the difference is not appreciable in terms of safety, it takes into account
the geometric reality of the Bleve phenomenon throughout its lifetime.

References

Gostintsev: Theory of the Aerodynamics, Autoignition. Google Scholar. https://scholar.goo

gle.com/scholar_lookup?title=Theory%20of%20the%20aerodynamics%2C%20self%20igni
tion%2C%20and%20burnup%20of%20turbulent%20thermals%2C%20vortex%20rings%2C%
20and%20jets%20in%20a%20free%20atmosphere&journal=Khim.%20Fiz.&volume=9&
pages=1279-1290&publication_year=1982&author=Gostintsev%2CYu.%20A.&author=Sol
odovnik%2CA.%20F.&author=Lazarev%2CV.%20V. Last accessed 2022 Feb 25
Guidelines for Consequence Analysis of Chemical Releases. https://www.aiche.org/resources/pub
lications/books/guidelines-consequence-analysis-chemical-releases. Last accessed 2022 Feb 25
INERIS: OMEGA5 Le BLEVE, phénoménologie et modélisation des effets. INERIS, Verneuil-en-
Halatte (2017)
Johnson, D., Pritchard, M.: Large scale experimental study of boiling liquid expanding vapour
explosions (BLEVEs) (1990)
Makhvilade, G.M., Roberts, J.P., Yakush, S.E.: Burning regimes for the finite-duration releases of
fuel gases. In: Symposium (International) on Combustion, vol. 1, pp. 1549–1555 (1996). https://
doi.org/10.1016/S0082-0784(96)80377-8
Makhviladze, G.M., Yakush, S.E.: Modelling of formation and combustion of accidentally released
fuel clouds. Process Saf. Environ. Prot. 83, 171–177 (2005). https://doi.org/10.1205/psep.04242
Makhviladze, G.M., Roberts, J.P., Yakush, S.E.: Modelling the fireballs from methane releases. Fire
Saf. Sci. 5, 213–224 (1997)
Makhviladze, G.M., Roberts, J.P., Yakush, S.E.: Fireball during combustion of hydrocarbon fueld
releases II thermal radiation. Combust Explos Shock Waves 35, 359–369 (1999b). https://doi.
org/10.1007/BF02674465
Makhviladze, G., Roberts, J., Yakush, S.: Modelling and scaling of fireballs from single -and two-
phase hydrocarbon releases. Fire Saf. Sci. 6, 1125–1136 (2000). https://doi.org/10.3801/IAFSS.
FSS.6-1125
Makhviladze, G.M., Roberts, J.P., Yakush, S.E.: Fireball during combustion of hydrocarbon fuel
releases. I. Structure and lift dynamics. Combust. Explos. Shock Waves 35, 219–229 (1999a).
https://doi.org/10.1007/BF02674442.
Martinsen, W.E., Marx, J.D., By, P.: An improved model for the prediction of radiant heat from
fireballs quest an improved model for the prediction of radiant heat from fireballs (1999)
Applied Mechanics and Engineering
Mechanics
Belt Layer Effects on Non-pneumatic
Tire Performance by Finite Element
Analysis

Juthanee Phromjan and Chakrit Suvanjumrat

1 Introduction

The non-pneumatic tires (NPTs) are recently developed to overcome the disadvan-
tages of conventional pneumatic tires such as flat tires, tire pressure maintenance,
even though the uncomfortable ride of solid tires (Phromjan and Suvanjumrat 2018a).
The basic concept of NPT design is a combination of composite shear band and spoke
structure (Rhyen and Cron 2006). The shear band is developed to have similar primary
functions as the belt plies of a pneumatic tire which provide strength and stability to
the tire tread. Consequently, a good deal of previous research attempted to develop
the shear band of NPT. Ju et al. (2010) designed the shear band with different cellular
geometries instead of the belt layers. They studied the contact pressure profile of the
developed shear bands from the finite element analysis (FEA). There was found that
a higher negative cell angle provided a lower contact pressure. Veeramurthy et al.
(2014) found that the shear band thickness was a greater effect on rolling resistance
than the spoke thickness. It was an important parameter for design NPT with low
rolling resistance. Ju et al. (2013) studied a porous composite shear band of NPT by
finite element method (FEM). The simulation results indicated that the decrease of
elastomer volume reduced the hysteretic energy dissipation. The research, as previ-
ously mentioned, simplified the NPT model with a smooth and continuous tread
for the investigation. The authentic NPT, TWEEL by Michelin, inserted belt layers
for the increase of stiffness in its circumference. The homogenization approach was
proposed to model the shear band of TWEEL. It precisely accomplished the defor-
mation of the TWEEL model when compared against the experiment (Phromjan and

J. Phromjan · C. Suvanjumrat (B)

Department of Mechanical Engineering, Faculty of Engineering, Mahidol University, Nakhon
Pathom, Thailand
e-mail: [email protected]
Department of Mechanical Engineering, Faculty of Engineering, Laboratory of Computer
Mechanics for Design (LCMD), Mahidol University, Nakhon Pathom, Thailand

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 149
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_15
150 J. Phromjan and C. Suvanjumrat

Suvanjumrat 2020). The full-scale NPT model has also carried out the dynamic anal-
ysis (Rugsaj and Suvanjumrat 2021). However, the previous studies did not focus
to evaluate the belt layer effects on NPT performances. Therefore, it is piquancy to
clarify the effects of belt layers. This research aimed to model the full-scale TWEEL
with belt layers inside the shear band. The numbers and positions of the belt layer in
the shear band that was the composite material were varied. The hyperelastic models
which were carried out by testing the material property of each TWEEL component
were used to model material of finite element (FE) model. The vertical stiffness,
contact area, contact pressure, and safety factor of NPT were carried out to study the
belt layer effects. These performances were mainly discussed for the NPT (Phromjan
and Suvanjumrat 2021). The results from this interesting will be useful for the exten-
sion of the agricultural tires. In general, the agricultural machines which were driven
by pneumatic tires were the cause of soil compaction (Hamza and Anderson 2005).
Therefore, these useful simulation results will be a guidance tool for solving that
problem.

2 Finite Element Method

In the first stage, the computer-aided design (CAD) was performed to create the three-
dimensional (3D) model of the TWEEL NPT which was developed by Michelin in
2015. It consists of 4 main components which are the rubber tread, composite shear
band, polyurethane (PU) spokes, and steel hub as shown in Fig. 1a.

2.1 Finite Element Model

To analyze the NPT structure by FEM, the FE model of TWEEL was created
according to the CAD model. The hexagonal elements were used to create the tread

(a) (b)

Fig. 1 NPT model of a CAD model and b FE model

Belt Layer Effects on Non-pneumatic Tire Performance by Finite … 151

Table 1 Details of finite element model of each NPT component

Component Element type Number of element Average element length (mm)
Tread Solid, 8 nodes 18,304 11.48
Shear band Solid, 8 nodes 24,576 13.38
Spoke Shell, 4 nodes 11,550 11.67

Fig. 2 Cross-section of shear band

and shear band, while spokes were constructed by thick shell elements. The rigid bar
elements (RBEs) act as the steel hub of NPT are gathered at the center of TWEEL as
shown in Fig. 1b. The composite shear band was prepared following the homogenized
model which was already developed and validated by Phromjan and Suvanjumrat
(2020). The details of the FE model of each TWEEL component are presented in
Table 1. The center of the TWEEL model was fixed, while it was compressed by the
rigid flat plate for static performance analysis. The compression force was 4650 N
from the weight of the small tractor was applied by moving the rigid plate. In the
beginning, the shear band is inserted by 4 belt layers according to TWEEL as shown
in Fig. 2. Each layer was defined by the number from 1 to 4 where layer 1 was located
next to the tread and layer 4 was located next to the spoke structure. The thicknesses
were 3, 1, 1, and 2 mm for layers no. 1, 2, 3, and 4, respectively. After that, the various
NPT models with different numbers and positions of belt layer were prepared. The
descriptions of NPT models are presented in Table 2.

2.2 Finite Element Analysis

To prevent the volumetric locking, the hybrid displacement/pressure formulation was

established to analyze the incompressible analysis of the NPT model. It is a good
effective formulation for the analysis of rubberlike material which has a large strain.
The governing equation can be written by the following equation (Bathe 1997).

d t0 W = t0 Sd0t εi j (1)

where d t0 W is the incremental potential energy, t0 S is Piola–Kirchhoff stress which

is computed only from the displacement field, and d0t εi j is the incremental strain.
152 J. Phromjan and C. Suvanjumrat

Table 2 Description of belt

NPT model Belt layers inserting sequence from tread to spoke
layers for NPT models
Group 1 (4 belt layers, original)
M1-1 1234
Group 2 (3 belt layers)
M2-1 234
M2-2 134
M2-3 124
M2-4 123
Group 3 (2 belt layers)
M3-1 12
M3-2 13
M3-3 23
M3-4 34
M3-5 24
M3-6 14
Group 4 (1 belt layer)
M4-1 1
M4-2 2
M4-3 3
M4-4 4
Group (No belt layer insertion)
M5-1 –

The hyperelastic material model, Mooney–Rilvin model (Phromjan and Suvan-

jumrat 2018b), was defined to describe the mechanical properties of the NPT model.
This model is written following Eq. 2, which is a popular constitutive model.
The spoke model was the most resilient part; therefore, the generalized Maxwell’s
viscoelastic model was also used to gain more precision of results. These models are
written as follows:
1 t
t
0W = C1 (t0 J1 − 3) + C2 (t0 J2 − 3) + K ( J3 − 1)2 (2)
2 0
where Ci is hyperelastic material constants, Ji is the reduced invariants, and K is the
bulk modulus.

Σ
n
G(t) = G 0 − G i (1 − e−t/τi ) (3)
i=1

τi = ηi /E i (4)
Belt Layer Effects on Non-pneumatic Tire Performance by Finite … 153

Table 3 Constants of material property of the NPT model

Component Mooney–Rivlin’s constant Young’s modulus (GPa) Poisson’s ratio
C1 C2 K (MPa)
Tread 1.19085 0 9670.24 – –
Wall 1.19085 0 11,908.5 – –
Steel belt – – – 200 0.3
Spoke 0.5063 4.2552 42,449 – –

Table 4 Constants of
ith term τi Gi
generalized Maxwell’s
viscoelastic model of spoke 1 0.2 0.125
model (Phromjan and 2 0.02 0.125
Suvanjumrat 2020)
3 0.002 0.125

where G(t) is the shear relaxation modulus, G 0 is the initial shear modulus, G i is
ith terms of the shear modulus, t is time, τi is ith terms of the relaxation time, ηi is
the viscoelasticity, and E i is the modulus of elasticity.
The constants of material property and constants of generalized Maxwell’s
viscoelastic model of each NPT component are presented in Tables 3 and 4,
respectively.

3 Results and Discussion

The various NPT models with different numbers and positions of belt layer were
compressed by the compression force of 4,650 N in the vertical direction. When the
shear band is inserted by 4 belt layers (M1-1), the NPT model is deformed by the
vertical load as shown in Fig. 3a. Spokes in the contact region were buckled because
of supporting the vehicle’s load from a center of NPT. On the other hand, spokes
in the support region were stretched to compensate for the deformation of the shear
band. These phenomena happened only when the spoke structures work properly in
their role as resilient parts. If the shear band was removed all the belt layers out, it
would be a role as the resilient part instead. Figure 3b presents that the NPT model
is distorted when no belt layers were inserted inside the shear band. It is noted that
the belt layer’s existence is an important effect on the NPT performance. The NPT
performances by varying numbers and positions of belt layers are presented in Table
5.
154 J. Phromjan and C. Suvanjumrat

(a) (b)

Fig. 3 Simulation results of a M1-1 and b M5-1 NPT models

Table 5 Simulation results of NPT models with various numbers and positions of belt layers
NPT model Vertical stiffness Safety factor of Contact area (cm2 ) Contact pressure
(N/mm) NPT’s spoke (MPa)
Group 1
M1-1 683.82 22.22 144.83 1.064
Group 2
M2-1 512.40 16.52 161.09 1.144
M2-2 578.80 16.75 151.79 1.177
M2-3 579.19 16.75 151.79 1.177
M2-4 551.36 16.38 151.79 1.187
Group 3
M3-1 523.61 16.20 153.56 1.195
M3-2 529.02 16.24 151.13 1.192
M3-3 438.25 13.96 159.03 1.153
M3-4 475.53 15.48 162.95 1.128
M3-5 487.96 15.68 155.94 1.145
M3-6 561.80 16.57 151.79 1.181
Group 4
M4-1 483.61 15.14 150.23 1.203
M4-2 370.96 12.31 159.95 1.149
M4-3 368.11 12.32 158.97 1.122
M4-4 428.76 13.85 163.83 1.108
Group
M5-1 104.87 4.50 260.32 1.938
Belt Layer Effects on Non-pneumatic Tire Performance by Finite … 155

3.1 Effects on the Vertical Stiffness

The existence of the belt layer affected severely on the NPT deformation which is
involved in the vertical stiffness. The load capacity of NPT could be dramatically
increased by inserting proper belt layers. The simulation results indicated that the
existence of layer 1 resulted in the high vertical stiffness observed in every simulation
group. If this belt layer was eliminated from the shear band, the vertical stiffness
would be extremely reduced. It should be noted that not only the layer 1 has the
highest thickness, but it is also located next to the tread which is the first part to be
deformed by the ground. Furthermore, the existence of layers 2 or 3 insignificantly
affected the vertical stiffness of NPTs. Particularly, the average vertical stiffness was
decreased by reducing the number of belt layers.

3.2 Effects on the Safety Factor

The yield stress of TWEEL’s spokes was 6 MPa (Rugsaj and Suvanjumrat 2020).
The safety factor was observed by the stress occurrence at spokes. It was quite varied
the vertical stiffness of NPT. There was resulted in an increase of vertical stiffness
following the increase of safety factors. When all the belt layers were removed from
the shear band of NPT, the safety factor was dramatically reduced.

3.3 Effects on the Contact Area

The contact area happening under the NPT model is investigated by using a red color
contour as shown in Fig. 4. It was found that no belt layer (M5-1) appeared the contact
area higher than the original one (M1-1) about 79.74%. It should be noted that the
existence of belt layers extremely affected the contact area between NPT and road.
It was also observed that the contact area was high on the NPT model without layer
1. This is because the first layer which was located next to the tread had an important
role to prevent excessive deformation of the shear band. The contact area would
increase when the belt layers were inserted far from the NPT’s tread. However, the
existence of layer 4 resulted in a higher contact area although the vertical stiffness
was still maintained because it was thicker than a layer no. 2 and 3. The existence of
a layer no. 2 and 3 was the insignificant difference in the contact area changing.
156 J. Phromjan and C. Suvanjumrat

(a) (b)

Fig. 4 Footprint of a M1-1 and b M5-1 NPT models

3.4 Effects on the Contact Pressure

The contact pressure is defined by the ratio of the vertical load to the true contact area.
FEA results represented the contact pressure as the normal contact stress on NPT tread
(Phromjan and Suvanjumrat 2021). The simulated results, therefore, indicated that
the NPT which had the higher contact area almost had the lower contact pressure. It
should be noted that the existence of layer 1 made a higher contact pressure. However,
all the belt layers which were removed from the shear band caused the higher contact
area but caused not the lower contact pressure. It is because the more contact force
happening that the contact pressure remains a higher value. There was observed that
the maximum contact stress happened on the tread of NPT.

4 Conclusions

To study the effects of the belt layer in the shear band on the NPT performances, the
compression test of NPT models with different numbers and positions of belt layers
was simulated by FEM. The different numbers of belt layers inserting in the shear
band resulted in the difference in vertical stiffness. Therefore, it can be concluded
that belt layer 1 is the most important layer for the NPT performances. The existence
of layer 1 always brings the high vertical stiffness and safety factor. However, the
contact pressure was not increased as much, while the contact area was low. In
contrast, layers 2 and 3 were insignificant effects on the NPT performances. It was
investigated the thickness of the belt layer also affected the NPT performances. The
belt layer which has a higher thickness can increase the strength of NPTs. Moreover,
the positioning of belt layers in the shear band was found to be a severe effect on
the contact area of the NPT model. These FEA results will be a useful tool for the
design and development of NPT in many applications shortly.
Belt Layer Effects on Non-pneumatic Tire Performance by Finite … 157

Acknowledgements This work was financially supported by the Thailand Science Research and
Innovation (TSRI) and National Research Council of Thailand (NRCT) through the Royal Golden
Jubilee Ph. D. Program (Grant No. PHD/0049/2561).

References

Bathe, K.J.: Finite element procedures. Prentice Hall, London (1997)

Hamza, M., Anderson, W.: Soil compaction in cropping systems a review of the nature, causes and
possible solutions. Soil Tillage Res. 82, 121–145 (2005)
Ju, J., Ananthasayanam, B., Summers, J.D., Joseph, P.: Design of cellular shear bands of a non-
pneumatic tire-investigation of contact pressure. SAE Int. J. Passenger Cars-Mech. Syst. 3(1),
598–606 (2010)
Ju, J., Veeramurthy, M., Summers, J.D., Thompson, L.: Rolling resistance of a non-pneumatic tire
having a porous elastomer composite shear band. Tire Sci. Technol. 41(3), 1–26 (2013)
Phromjan, J., Suvanjumrat, C.: The modification of steel belt layer of airless tire for finite element
analysis. IOP Conf. Ser.: Mater. Sci. Eng. 773(012047) (2020)
Phromjan, J., Suvanjumrat, C.: Vibration effect of two different tires on baggage towing tractors. J.
Mech. Sci. Technol. 32(4), 1539–1548 (2018a)
Phromjan, J., Suvanjumrat, C.: Material properties of natural rubber solid tires for finite element
analysis. Key Eng. Mater. 775, 560–564 (2018b)
Phromjan, J., Suvanjumrat, C.: Development of solid tire model for finite element analysis of
compressive loading. Songklanakarin J. Sci. Technol. 43(1), 229–236 (2021)
Rhyen, T.B., Cron, S.M.: Development of a non-pneumatic wheel. Tire Sci. Technol. 34(3), 150–169
(2006)
Rugsaj, R., Suvanjumrat, C.: Study of geometric effects on nonpneumatic tire spoke structure using
finite element method. Mech. Based Des. Struct. Mach. 50(7), 2379-2399 (2022). https://doi.org/
10.1080/15397734.2020.1777875
Rugsaj, R., Suvanjumrat, C.: Dynamic finite element analysis of rolling non-pneumatic tire. Int. J.
Automot. Technol. 22(4), 1011–1022 (2021)
Veeramurthy, M., Ju, J., Thompson, L.L., Summers, J.D.: Optimisation of geometry and material
properties of a non-pneumatic tyre for reducing rolling resistance. Int. J. Veh. Des. 66(2), 193–216
(2014)
Analysis of Hybrid Viscoelastic Sheets
Adhered to the Hull of a GFRP Vessel
to Reduce Impact Damage

Patrick Townsend , Juan C. Suárez Bermejo ,

and Alvaro Rodríguez-Ortíz

1 Introduction

Since the world set out to protect the environment, multiple ways to do so have
developed. The maintenance of ships built in glass fiber reinforced plastic (GFRP)
generates pollution which requires high investments so that they do not harm the
surrounding environment. In the Galapagos Islands, there is a significant fleet of
GFRP planing hull vessels. As these boats sail at high speeds, they must be light
and cover long distances between islands to comply with tourist programs; this
causes severe damage to the laminates of the hull in the bow due to the impact
with the waves. Sailing at high speeds increases a phenomenon known as slamming
(Kapfenberg and Veer 2002). This is a fatigue event on the material of the hull of the
boat, which considerably reduces its useful life, making more frequent maintenance
required within the nature reserve.
Vessels built with composite materials (with different combinations of reinforcing
fibers and polymeric matrix) are more and more numerous, due to the good mechan-
ical properties of these materials. Manufacturing with composite materials involves
the placement of successive layers of reinforcing fibers (glass, carbon, polyaramid,
etc.) that are impregnated with a resin that acts as a matrix, providing adhesion
between the fibers, between the layers stacked in thickness and provides continuity
to the laminated assembly. This means in practice that the hulls are built (together
with other parts of the boat) by placing several layers, one on top of the other, until
the total thickness of the laminate is obtained according to their structural stresses.
The number of layers is associated with the structural resistance (Qin and Batra 2009;
Baucoma and Zikryb 2005) or with the other parts of the boat (decks, bulkheads,

P. Townsend (B)
Escuela Superior Politécnica del Litoral, ESPOL, Guayaquil, Ecuador
e-mail: [email protected]
J. C. Suárez Bermejo · A. Rodríguez-Ortíz
Universidad Politécnica de Madrid, UPM, Madrid, Spain

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 159
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_16
160 P. Townsend et al.

etc.). A boat made of composite materials will weigh less than a boat made of steel
or aluminum. The use of GFRP partially sacrifices some aspects of its structural
strength. The capacity to respond to impacts, fatigue events, and internal stresses is
diminished, which is directly reflected in the extension of its useful life, in service
(Belingardini and Bador 2002).
On the other hand, because they are multilayer materials, it is possible to introduce
some type of energy dissipation mechanism into the lamination sequence. In such a
way that a viscoelastic sheet that goes inside the laminate has been proven to protect
it from low energy impacts. It does not apply in cases of ballistic impacts, which
imply very high deformation rates in the material.
Slamming is a phenomenon that occurs during navigation, usually above 20 km/h.
The boat begins to rise or support itself on the surface of the water, and when the
weight of the ship and the pressure of the water become unbalanced, it causes the
ship to sink back to the initial position producing an impact on the surface of the
hull. This impact supposes a sudden contribution of energy, which can be increased
by the waves. This energy must be absorbed by the hull material, which after the
impact returns part of the energy (compliance restoration) (Blake and Shenoi 2002).
The insertion of viscoelastic sheets inside GFRP panels has been studied exten-
sively in a numerical way, and according to our own research on specimens tested
in the laboratory, the modification proposal gives results and is ready to go to ships
(Townsend and Suárez 2016, 2018). Even its adherence is applicable to the theory
of hybrid bonds in the case of these materials, in which the stresses generated at the
ends of the adhesive bond influence the adjacent layer of the bond (Dominguez and
Carral 2020).
This time, it is proposed to analyze by means of fatigue tests, the behavior of the
viscoelastic layer superimposed on the laminate to obtain the energy data that the
viscoelastic absorbs and the protection it offers. The solution to the premature aging
of the planing hull vessels and the way they are used in boats already built would
represent a great option for the tourism fleet in the Galapagos Islands and the world.

2 Viscoelastic Layers as Energy Dissipators

in the Construction of GFRP Laminates

In the Galapagos Islands, there is an important fleet of planing hull vessels between
12 and 15 m in length. They sail 4 h between Puerto Ayora and Puerto Baquerizo
Moreno at 22 knots in some cases. They carry tourists seated on the main deck on
cushions. In this area of the Pacific Ocean, the wind level and wave height have a
Beaufort scale of degree 4 half the year, and the other half with the arrival of the cold
current from Antarctica, there are several months of degree 5. The captains must be
attentive to the waves during the entire navigation because at that speed and in these
sea conditions, the impacts with the bottom of the hull are very strong. The noise of
the impact even becomes a nuisance to the human ear when reaching levels of 90 dB.
Analysis of Hybrid Viscoelastic Sheets Adhered to the Hull of a GFRP … 161

Table 1 Characteristics of
Characteristic Value
the ship at maximum load
condition Material GFRP
L (length) 13,500 [mm]
Δ (displacement) 9100 [kg]
B (breadth) 4300 [mm]
T (deadrise angle) 4.5 [°]
V (maximum speed) 22 [Kn]

The slamming phenomenon has the following specific characteristics: its

nonlinear character in which the domain of time and not frequency predominates,
and that it is completely random. The slamming pressures for the type of vessel
considered, according to the classification societies of ships, especially the Amer-
ican Bureau of Shipping (ABS), consider that they range between the order of 2.25
× 103 [kN/m2 ] to 1.10 × 103 [kN/m2 ] depending on its longitudinal position at the
bottom of the ship and by the curvature of the hull. Table 1 shows the characteristic
data of the ship selected for experimentation.
[ The hull of these vessels] corresponds to laminates by impregnation of type
WR O
0 /(0 /0 )4 /0
MI M MO
with mat-type isophtalic fabrics, at Ortoftalic, Woveng
Robing Ortoftalica. In the nomenclature, the number on the right is described as
the first layer of the laminate on the outside of the hull. Due to the type of manual
construction, there is a high percentage of porosity of up to (7%), which is reflected
in the beginning of damage due to slamming impacts during navigation. The microp-
ores line up forming microcracks, to later appear as fractures inside the ship. In most
cases, it produces water leaks. In Fig. 1, the interior connection of the hull with the
fore bulkhead of the selected ship is observed. In it, you can clearly see the breaking
line that is propagating because of the tensile stresses that occur inside due to the
deformation.
Previous research considered including viscoelastic sheets within the laminate,
that is to say, placed under the second layer of rowing fabric protecting the rest
of the layers. They are not used in the entire hull of the boat, but more critical
where the phenomenon of slamming has the greatest incidence. The impact area
occurs both to port and starboard, which directly affects the laminate of the hull,
producing microcracks that change the flexural stiffness of the structure. To achieve
this behavior, the protection layers are manufactured in very thin thicknesses with the
viscoelastic polymer, encapsulating it in cells of another much more rigid polymer,
constraining the deformation capacity of the viscoelastic polymer in its two main
directions, in the plane of the sheet. In this way, the property of the viscoelastic
material of having a Poisson’s ratio close to 0.50 is taken advantage of, that is,
the relationship between the deformations that the material experiences when it is
compressed in the thickness direction with the deformations in the two perpendicular
directions in the plane of the sheet. A material with a Poisson’s ratio of 0.5 essentially
keeps the volume constant when it is deformed in one direction.
162 P. Townsend et al.

Fig. 1 Slamming damage observed inside the vessel

When a viscoelastic is deformed by an impact without restricting its directions, it

is compressed until it can no longer absorb any more energy and then expands dissi-
pating all the elastic energy stored. On the contrary, if its deformation (expansion) is
constricted in its two main directions in the plane of the viscoelastic layer, encapsu-
lating the viscoelastic material in closed cells formed by a more rigid polymer, the
material is given a great storage capacity of energy. The more the pressures due to
impact increase, the stiffer the viscoelastic material becomes, and consequently, the
more energy it absorbs. Through this viscoelastic cell design, when the protection
layer is placed within the composite material laminate, the stiffness of the material
grows exponentially as the impact energy attempts to deform the ship’s hull. The
response is nonlinear and instantaneous, with a large amount of energy being stored
in the viscoelastic layer, which, however, does not transfer it to the rest of the lami-
nate, but rather slowly dissipates it. In this way, it is possible to reduce the damage
generated in the laminate by the impact of slamming, slowing down its progression,
and, consequently, making it possible to extend the useful life of the boat.
The present work proposes the placement of the viscoelastic sheet on the surface of
a boat hull already built and in use. Through laboratory tests with GFRP specimens,
observe the energy behavior of the impact and if the laminate is effectively protected.
This would allow the use of viscoelastic sheets to be extended and spread more widely
to improve the useful life of vessels. Figure 2 shows the difference in a section of a
ship’s hull. In (a) the viscoelastic sheet is inserted into the laminate which corresponds
to its initial conception. In (b) the proposal to adhere it to the laminate so that it does
not affect the hydrodynamics of the boat and protects from damage by slamming.
Analysis of Hybrid Viscoelastic Sheets Adhered to the Hull of a GFRP … 163

Fig. 2 Cross-section of the planing hull vessel with a viscoelastic sheet a built new b modified not
new

3 Experimental Observation of the Protection

of the Superimposed Viscoelastic Layer

For the experimental observation of the behavior of the viscoelastic sheet superim-
posed on the hull of a ship, 3 panels were manufactured with the same characteristics
of the selected boat, and the superimposed viscoelastic layer was added. The panels
to be tested for fatigue, reproducing the slamming impact, were mounted on a base
of a shaker-type vibrating equipment to deform them. This equipment, its base, the
panel and the superimposed viscoelastic sheet, can be seen in Fig. 3.
Using a load cell on the top of the vibrator to compress with the panel, the force
that deforms the panel is measured. As the vibrating equipment tries to maintain a
deformation speed, it regulates the acceleration, and therefore, the load cell fulfills
the function of registering the changes in force versus cycle. The equipment records
the accelerations and the deformation velocity at the center of the impact. Strain
gauges were also used to control the deformation of the panel at 4 [cm] of impact

Fig. 3 Impact vibrator with the test panel placed on the top
164 P. Townsend et al.

Table 2 Experimental values of the tests carried out

# Panel Frequency range [Hz] Applied cycles [n] Average strength [kJ] Amplitude [G]
A 5–10 5 × 104 50 1
B 5–10 5 × 104 40 1
C 5–10 5 × 104 30 1

for control. In Table 2, the experimental values of the developed experiments are
presented. The fatigue phenomenon was reduced on three panels at different force
conditions, equal amplitude, and acceleration.
To evaluate the energy that the panel returns as mechanical work, the principle
that a force F travels a distance × is applied. Laboratory reproductions were made for
different frequencies ω over the angle between the force vector and the displacement
vector Y, for each acceleration G at time t obtain the Eq. 1 which corresponds to the
energy returned by the panel over the shaker.
(
E = Y.ω G(t).dt (1)

When carrying out the tests, results were obtained such as those shown in Fig. 4.
In (a) it is observed that the acceleration imposed by the vibrating equipment on
panel A is rectified to maintain a regular profile curve. The peaks of highs and
lows correspond to microdamage produced during deformation due to the laminate
opposing bending. The acceleration curve corresponds to a test time of 60 [ms]. With
this imposed acceleration, it is observed in (a) that the panel impacted at 7 × 103
cycles presents microcracks around the areas of the viscoelastic in a minimal amount.
In (b) at 5 × 104 cycles of impacts made on average at 50 [KJ], microcracks have
increased without detachment of the viscoelastic layer in the matrix.
Data were taken on the force exerted by the vibrator equipment in the impact
versus cycles, and this is observed in Fig. 5. For panels A and C, there is a tendency
to decrease, being greater that of panel A. On the other hand, the panel B showed a
very complex dispersion to evaluate a trend. The results of panel B are not conclusive.

Fig. 4 Results of the tests carried out on panel A

Analysis of Hybrid Viscoelastic Sheets Adhered to the Hull of a GFRP … 165

Fig. 5 Sampling of the

applied force in the panels
versus impact cycles

Fig. 6 Energy returned by

the panels versus impact
cycles

During the tests, it was found that the force tends to regulate with the acceleration
since the vibrating equipment stabilizes in the acceleration to deform in constant
values over time. Force values are sampled because the type of load cell used does
not allow a permanent record to be kept.
When these force values are transformed to returned energy, they draw the trend
curves shown in Fig. 6. Panel A permanently loses its ability to return energy since
the measured force that is regulated in the vibrator is still at damage levels. Panel
B, product of force dispersion, does not show a trend curve. Panel C draws a trend
curve that tends to stabilize, meaning that flexural stiffness has increased due to the
level of damage reaching the threshold.

4 Conclusions

By superimposing the viscoelastic layer on the GFRP panels, the impact damage
can be effectively mitigated. Figure 6 shows that the increase in flexural stiffness
in a panel works in conjunction with the viscoelastic sheet stabilizing the damage.
This trend curve, as it stabilizes, is the increase in useful life that we wanted to
166 P. Townsend et al.

demonstrate. The generation and propagation of damage is controlled by improving

the dissipation of the energy that remains in the form of damage in the GFRP. This
approach opens up new perspectives in the design of the planing hull ship since
the superposition of the viscoelastic layers changes the way in which the stresses
due to the phenomenon of slamming within the laminate are to be distributed and
concentrated.

References

Baucoma, J., Zikryb, M.: Low-velocity impact damage progression in woven E-glass composite
systems. Compos. Part A. Appl. Sci. Manuf. 36, pp. 658–664 (2005), Section 5
Belingardini G, Bador V (2002) Low velocity impact tests of laminate glass-fiber-epoxy matrix
composite material plates International Journal of Impact Engineering, Vol 27, section 2, pp. 213–
219
Blake, J., Shenoi, R.: Strength modelling in stiffened FRP structures with viscoelastic inserts for
ocean structures. University of Southampton, UK Ocean Engineering 29, 849–869 (2002)
Dominguez, F., Carral, L.: The hybrid joints between an FRP panel and a steel panel through
tubular reinforcements: a methodology for interlaminar stress calculations. Appl. Phys. 10(11),
3962 (2020). https://doi.org/10.3390/app10113962
Kapfenberg, G., Veer, A.: Whipping loads due to aft body slamming. In: 24th Symposium on Naval
Hydrodynamics, Japan, pp. 336–337 (2002)
Qin, Z., Batra, R.: Local slamming impact of sandwich composite hulls. Int. J. Solids Struct. 46,
2011–2035 (2009)
Townsend, P., Suárez, J.C.: The effect of slamming impact on out-of-autoclave cured prepregs of
GFRP composite panels for hulls. Procedia Eng. 167, 252–326 (2016)
Townsend, P., Suárez, J.C.: Reduction of slamming damage in the hull of high-speed crafts
manufactured from composite materials using viscoelastic layers. Ocean Eng. 159, 253–267
(2018)
On the Contact Between Viscoelastic
Tube with Inner Nonuniform Coating
and Several Rigid Inserts

Kirill E. Kazakov and Svetlana P. Kurdina

1 Introduction

In industry, mechanical engineering, construction, pipes made of several layers are

often used. This is due to the fact that structures made of pipes can be exposed to
a variety of aggressive environments, must meet safety requirements, meet addi-
tional requirements, etc. Some layers of such pipes may be inhomogeneous due, for
example, to the peculiarities of the manufacturing process (Parshin and Manzhirov
2018; Parshin 2021). Ring-shaped rigid inserts can be used to connect various pipe
sections, as well as for reinforcing long sections. The presence of inserts leads to a
change in the stress–strain state of the pipe. The article is devoted to the formula-
tion and construction of an analytical solution to the problem of the interaction of
several annular inserts and a viscoelastic aging pipe with an internal inhomogeneous
coating. The solution of such a problem is reduced to the solution of a system of
integral equation with integral operators of two types (Arutyunyan and Manzhirov
1999).
The construction of a solution for such a system is carried out using a special
approach and is similar to what was done in articles (Kazakov 2018; Kazakov and
Parshin 2019; Kazakov and Kurdina 2019).

K. E. Kazakov (B)
Ishlinsky Institute for Problems in Mechanics of the Russian Academy of Sciences, pr.
Vernadskogo 101-1, Moscow, Russia
e-mail: [email protected]
S. P. Kurdina
Bauman Moscow State Technical University, ul. 2 Baumanskaya 5-1, Moscow, Russia
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 167
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_17
168 K. E. Kazakov and S. P. Kurdina

2 Mathematical Model of the Problem

Let the pipe consists of two layers: an outer viscoelastic aging layer with a thickness
of H and an inner elastic layer with a thickness of h. If r in is the inner radius of
the outer layer, and r out is its outer radius, then thickness H = r out – r in . We will
assume that inner layer thickness h is much less than the pipe inner radius r in and
the outer layer thickness H. There is a smooth contact between layers. It is believed
that the stiffness of the inner layer is less than the stiffness of the outer layer. Inner
layer has longitudinal inhomogeneity, i.e., its properties (Young’s modulus E in (z)
and Poisson’s ratio ν in (z)) depend on longitudinal coordinate z. The construction of
a solution for such a system is carried out using a special approach and is similar to
what was done in articles (Kazakov 2018; Kazakov and Parshin 2019; Kazakov and
Kurdina 2019).
At time τ start , several cylindrical inserts are placed inside such a tube with tension.
There is a smooth contact between the inner layer and the inserts. All of them can
have different external radii and widths significantly exceeding the thickness of the
inner layer of the pipe. Inserts external radii are not less than the inner radius of the
inner layer. The contact areas do not change over time and coincide with the widths
of the inserts.
The interaction scheme is shown in Fig. 1.
Generalization of the results presented in papers (Arutyunyan and Manzhirov
1999; Manzhirov and Chernysh 1990) allows us to establish that the mathematical
model of the problem is the following system of n mixed integral equations (n is
account of inserts):

E in−1 (z)h[1 − νin2 (z)] pi (z, t)

⎡
Σ (b j
( ) n
⎢ −1
+ 2π −1 1 − νout 2 −1
⎣ E out (t − tout ) Cc (rin (z − ζ )) p j (ζ, t)dζ
j=1 aj

Fig. 1 Interaction of tube with coating and three rigid inserts

On the Contact Between Viscoelastic Tube with Inner Nonuniform … 169
⎤
(t (b j
⎥
− −1
E out (τ − τout )Vout (t − tout , τ − tout ) Fc (rin−1 (z − ζ )) p j (ζ, τ )dζ dτ ⎦
τstart aj

= gi − (rin − h), ai ≤ z ≤ bi , t ≥ τstart , i = 1, 2, . . . , n. (1)

The following designations are introduced here: pi (x, t) is distribution of contact

pressures in the region of interaction between the ith insert and the pipe; ν out and
E out (t) are Poisson’s ratio and Young’s modulus of outer viscoelastic aging layer; t out
is the moment of manufacture of the outer layer; V out (t, τ ) is creep kernel of outer
layer (Arutyunyan and Manzhirov 1999; Arutyunyan 1927); it can be presented, for
example, as
[ −1
]/
Vout (t, τ ) = ∂ Cout (t, τ ) + E out (τ ) ∂τ E out (τ ),

where C out (t, τ ) is the tensile creep function, F c (y) is known kernel of the contact
problem for cylindrical tube with inner loading (Arutyunyan and Manzhirov 1999;
Manzhirov and Chernysh 1990):

(∞
Fc (y) = cos(yu)u −1 L(u)du,
0

in which following notations are introduced (I0 (u), I1 (u), K0 (u), K1 (u) are modified
Bessel functions of first and second kind)
[ ]
L(u) = S −1 (u)u kr−1 + s(kr , u)D12 (u) − u 2 kr C12 (u) ,
−1
s(r, u) = 2r −1 (1 − νout ) + r u 2 , kr = rin rout ,
S(u) = s(kr , u) + kr−1 s(1, u) − s(kr , u)B12 (u)u 2 + kr A21 (u)u 4
+ s(1, u)s(kr , u)D12 (u) − kr u 2 s(1, u)C12 (u),
A1 (u) = −I0 (ukr )K 0 (u) + I0 (u)K 0 (ukr ), B1 (u) = I1 (ukr )K 0 (u) + I0 (u)K 1 (ukr ),
C1 (u) = I1 (u)K 0 (ukr ) + I0 (ukr )K 1 (u), D1 (u) = −I1 (ukr )K 1 (u) + I1 (u)K 1 (ukr ).
(2)

Note that the resulting system includes functions describing the inner layer inho-
mogeneity, which can change rapidly. Moreover, it contains integrals with both
constant and variable limits of integration.
170 K. E. Kazakov and S. P. Kurdina

3 Mathematical Model in Dimensionless Form

Imagine the resulting dimensional system of integral Eq. (1) as dimensionless system

Σ
n
e∗ (t ∗ )D i (z ∗ ) pi∗ (z ∗ , t ∗ ) + F i j∗ ( I − V ∗ ) p j∗ (z ∗ , t ∗ ) = g i∗ ,
j=1 (3)
i = 1, 2, . . . , n, −1 ≤ z ∗ ≤ 1, t ∗ ≥ 1.

The following change of variables and functions is used in this equation (i, j =
1,2,…,n)

z ∗ = 2a i−1 (z − ηi ), ξ ∗ = 2a −1
j (ξ − η j ),

t ∗ = τstart
−1
t, τ ∗ = τstart
−1 ∗
τ, tout −1
= τstart tout , ηi∗ = 2a −1 ηi ,
g i∗ = 2a −1 [gi − (rin − h)], e∗ (t ∗ ) = E −1 E out (t − tout ),
1 1 − νin2 (z) h 2 pi (z, t) ( ) ai
D i∗ (z ∗ ) = , pi∗ (z ∗ , t ∗ ) = 1 − νout
2
,
E −1 E in (z) 1 − νout2 ai E out (t − τout ) a
(1 (t ∗
F y(z ) = y(ξ )F (z , ξ )dξ , V y(t ) = y(τ ∗ )V ∗ (t ∗ , τ ∗ )dτ ∗ ,
i j∗ ∗ ∗ ij ∗ ∗ ∗ ∗ ∗

−1 1
( )
∗
F (z , ξ ) = π
ij ∗ −1
Fc rin−1 (z − ζ ) , V ∗ (t ∗ , τ ∗ ) = τ0 Vout (t − tout , τ − tout ). (4)

Here, I is identity operator; ηi = 0.5(ai + bi ) is the midpoint of the ith insert;

a i = bi − ai is its width, a = mini=1,2,...,n a i , and E −1 is dimensional modulus; it
can be calculated, for example, as E –1 = 1/E out (0). Note that V* if Volterra integral
operator and F i j∗ are Frenholm integral operators.
For convenience, we present system (3) in operator form

e∗ (t ∗ ) D∗ (z ∗ ) p∗ (z ∗ , t ∗ ) + F ∗ (I − V ∗ ) p∗ (z ∗ , t ∗ ) = g ∗ , z ∗ ∈ [−1, 1], t ∗ ≥ 1.
(5)

Here
⎛ ⎞ ⎛ ⎞
p 1∗ (z ∗ , t ∗ ) F 11∗ F 12∗ · · · F 1n∗
⎜ p 2∗ (z ∗ , t ∗ ) ⎟ ⎜ F 21∗ F 22∗ · · · F 2n∗ ⎟
∗ ∗ ∗ ⎜ ⎟ ∗ ⎜ ⎟
p (z , t ) = ⎜ .. ⎟, F =⎜ . .. .. .. ⎟,
⎝ . ⎠ ⎝ .. . . . ⎠
p n∗ (z ∗ , t ∗ ) F n1∗ F n2∗ · · · F nn∗
⎛ ⎞ ⎛ ⎞ (6)
g 1∗ D 1 (z ∗ ) 0 · · · 0
⎜ g 2∗ ⎟ ⎜ 0 D 2 (z ∗ ) · · · 0 ⎟
⎜ ⎟ ⎜ ⎟
g ∗ = ⎜ . ⎟, D∗ (z ∗ ) = ⎜ . .. .. .. ⎟.
⎝ .. ⎠ ⎝ .. . . . ⎠
g n∗ 0 0 · · · D n (z ∗ )
On the Contact Between Viscoelastic Tube with Inner Nonuniform … 171

F ∗ is Frenholm integral operator with Hilbert–Schmidt kernel (Goursat).

In the resulting Eq. (3), the p*(z*, t*) vector function is to be determined. It is
related to the dimensional functions pi (z, t) describing contact pressures by the ratios
(4).
It is two main problems in Eq. (6):
(1) functions D i (z ∗ ) can be rapidly changing (they connect with layer properties)
(2) equation includes operators of different types (Volterra and Fredholm).

4 Solution of the Problem

If in Eq. (6) of work (Kazakov and Parshin 2019), we assume that α(t) ≡ 0, g(x) ≡ 0,
δ(t) = const, then we get equation similar to Eq. (5) from this work. Moreover,
the properties and asymptotics of the kernels included in the operators of the corre-
sponding equations are similar. Therefore, the approach to constructing a solution
will be similar to that described in Kazakov and Parshin (2019). Next, we describe
the main provisions of this approach.
First of all, it is necessary to present the desired function in a special form

p∗ (z ∗ , t ∗ ) = D−1/2∗ (z ∗ ) · P ∗ (z ∗ , t ∗ ) (7)

−1/2∗ ∗
where
√ D (z√) is diagonal √ matrix on the diagonal of which the functions
1/ D 1 (z ∗ ), 1/ D 2 (z ∗ ), …, 1/ D n (z ∗ ) are located, P ∗ (z ∗ , t ∗ ) is new unknown
function. In this case, the operator equation is transformed in the form
∗
e∗ (t ∗ ) P ∗ (z ∗ , t ∗ ) + F̃ (I − V ∗ ) P ∗ (z ∗ , t ∗ ) = D−1/2∗ g ∗ , −1 ≤ z ∗ ≤ 1, t ∗ ≥ 1.
(8)
∗
Here, F̃ is new Fredholm operator. The properties of the kernel of this operator are
the same as the properties of the kernel of the original operator F ∗ .
Secondly, it is necessary to use a special orthonormal basis constructed using
orthogonalization of the following system of linearly independent functions
√ √ √ √ √
i 1 / D 1 (z ∗ ), z ∗ i 1 / D 1 (z ∗ ), (z ∗ )2 i 1 / D 1 (z ∗ ), . . . , i 2 / D 2 (z ∗ ), z ∗ i 2 / D 2 (z ∗ ),
√ √ √ √
(z ∗ )2 i 2 / D 2 (z ∗ ), . . . , i n / D n (z ∗ ), z ∗ i n / D n (z ∗ ), (z ∗ )2 i n / D n (z ∗ ), . . . ,

where ii is orts with the ith unit coordinate. This basis can be obtained by formulas
(Szegö 1959)
172 K. E. Kazakov and S. P. Kurdina
| |
| I0,i I1,i · · · Ik,i |
| |
qki◦ (z ∗ ) i | I1,i I2,i · · · Ik+1,i |
i ∗ i◦ ∗ 1 | |
q k (z ) = √ i , qk (z ) = √ | . . . .. ||,
D (z )
i ∗ δ δ |
k−1,i k,i | .
. .
. . . . |
| 1 z ∗ · · · (z ∗ )k |
| |
| I0,i I1,i · · · Ik,i |
(1 | |
(ξ ∗ )k | I1,i J2,i · · · Ik+1,i |
| |
Ik,i = dξ ∗ , δ−1,i = 1, δk,i = | . .. . . .. ||.
D (ξ )
i ∗ | . . . |
−1 | . .
|I I ··· I |
k,i k+1,i 2k,i

Thirdly, it is necessary to find the eigenvector functions ψk (z*) (k = 0, 1, 2, …)

∗
of the operator F̃ :
∗
F̃ ψ k (z ∗ ) = γk ψ k (z ∗ ),
∗
where γk is eigenvalues of operator F̃ . Vector functions ψ k (z ∗ ) are linear
combination of basis functions q ik (z ∗ ), i.e.,

∞
Σ
∗
ψ k (z ) = ψkm
i
q im (z ∗ ).
m=0

If we represent function P ∗ (z ∗ , t ∗ ) as a series

∞
Σ
P ∗ (z ∗ , t ∗ ) = kk (t ∗ )ψ k (z ∗ ),
k=0

then it is possible to bring the system of Eq. (8) for the functional coefficients of the
decomposition of the solution to a diagonal form:
∞
Σ [ ∗ ∗ ]
e (t )kk (t ∗ ) + (I − V ∗ )γk kk (t ∗ ) ψ k (z ∗ ) = D−1/2∗ g ∗ ,
k=0 (9)
∗ ∗
−1 ≤ z ≤ 1, t ≥ 1.

After that, it is necessary to decompose the right part of (9) according to the basic
functions ψ k (z ∗ ) and equate the coefficients on the left and on the right with the
corresponding basic functions ψ k (z ∗ ).
Applying the above approach, we obtain the final formulas for the desired
dimensionless functions:
∞ ∞
1 Σ Σ
pi∗ (z ∗ , t ∗ ) = k k (t ∗
) ψkm
i
qmi◦ (z ∗ ),
D i (z ∗ ) k=0 m=0
On the Contact Between Viscoelastic Tube with Inner Nonuniform … 173

i = 1, 2, . . . , n, −1 ≤ z ∗ ≤ 1, t ∗ ≥ 1, (10)

where
Σn √
∗ i=1 ψk0
i
I0,i g i∗
kk (t ) = (I + V k ) ,
γk + e∗ (t ∗ )
(t ∗ (1
∗ ∗ (ξ ∗ )k
V k y(t ) = y(τ )Vk∗ (t ∗ , τ ∗ )dτ ∗ , Ik,i = dξ ∗ , δ−1,i = 1,
D i (ξ ∗ )
1 −1
| |
| I0,i I1,i · · · Ik,i |
| |
| I1,i I2,i · · · Ik+1,i |
i◦ ∗ 1 | |
qk (z ) = √ | . . . . |,
δk−1,i δk,i || .. .. . . .. ||
| 1 z ∗ · · · (z ∗ )k |
| |
| I0,i I1,i · · · Ik,i |
| |
| I1,i I2,i · · · Jk+1,i |
| |
δk,i = | . .. . . .. ||,
| .. . . . |
|
|I I · · · I2k,i |
k,i k+1,i

i = 1, 2, . . . , n, k = 0, 1, 2, . . . , −1 ≤ z ∗ ≤ 1, t ∗ ≥ 1, (11)

in which kernels V k (t*,τ *) are the resolvents of the V *(t*,τ *)γ k /[γ k + e*(t*)];
parameters γ k and ψkm
i
can be found from solution of the spectral problem

Σ ∞
n Σ
j ij
ψkm Fml = ψkm
i
γk , i = 1, 2, . . . , n, k, m = 0, 1, 2, . . . , (12)
j=1 l=0

ij
and K ml are calculated by the expression

(1 (1
F i j (z ∗ , ξ ∗ )qmi◦ (z ∗ )ql (ξ ∗ ) ∗ ∗
j◦
ij
Fml = dz dξ , i, j = 1, 2, . . . , n,
D i (z ∗ )D j (ξ ∗ )
−1 −1

m, l = 0, 1, 2, . . . (13)

Formulas (4) and (6) allow us to obtain dimensional contact stresses from expression
(10):
∞
Σ ∞
Σ ( )
2z − (ai + bi )
pi (z, t) = [1 − νin (z)]−1 E in (z) k̃k (t) qmi◦ ψkm
i
,
k=0 m=0
bi − ai
i = 1, 2, . . . , n, ai ≤ z ≤ bi , t ≥ τ0 , (14)
174 K. E. Kazakov and S. P. Kurdina

Fig. 2 Contact pressures in

interaction regions

where
( )
k̃k (t) = 0.5E −1 E out (t − τout )h −1 akk τ0−1 t , k = 0, 1, 2, . . . , t ≥ τ0 . (15)

In the obtained expressions (10) and (14), the functions k̃k (t), and qmi◦ (z ∗ ) are
continuous and smooth, and the functions associated with the variable properties of
the inner layer and their combinations are highlighted by separate multipliers.

5 Conclusions

Due to the representation of the solution in the form (14), in which the functions
describing the properties of the coating are allocated in a separate multiplier, it
becomes possible to perform efficient calculations even when the inhomogeneity
of the inner layer is described by a rapidly changing or discontinuous function.
Example of contact stresses distributions for such a case presented in Fig. 2. There
is considered a case when two inserts that are not far from each other are placed in
a cylinder, inhomogeneous coating of which is described by the functions D1 (z) =
D2 (z) = 0.01(1 – 0.05cos(2π z)).
Thus, a solution has been constructed to take into account the influence of the
rheological properties of the pipe, the inhomogeneity of its layers and the proximity
of inserts on the distribution of contact pressures.

Acknowledgements The study was supported by the Government program (contract #AAAA-
A20-120011690132-4).
On the Contact Between Viscoelastic Tube with Inner Nonuniform … 175

References

Arutyunyan, N.Kh.: Some Questions in the Creep Theory. Gostekhizdat, Moscow–Leningrad (1927)
(in Russian) (1952)
Arutyunyan, N.Kh., Manzhirov, A.V.: Contact Problems of the Theory of Creep. Izdat. NAN RA,
Erevan (1999) (in Russian)
Goursat, E.: Cours d’Analyse Mathématique, Tome III. Intégrales Infiniment Voisines; Équations
aux Dérivées Partielles du Second Order; Équations Intégrales; Calcul des Variations. Gauthier–
Villars, Paris
Kazakov, K.E.: Contact between a regular system of punches and a layered foundation with consid-
eration of complex surface shapes. AIP Conf. Proc. 2053, 040041 (2018). https://doi.org/10.1063/
1.5084479
Kazakov, K.E., Kurdina, S.P.: Multiple contact problem for bodies with nonuniform coatings of
variable thickness and regular systems of punches with complex base shape. J. Phys: Conf. Ser.
1203, 012019 (2019). https://doi.org/10.1088/1742-6596/1203/1/012019
Kazakov, K.E., Parshin, D.A.: On the indentation of a system of punches into a layered foundation.
Eng. Lett. 27(1), 251–255 (2019)
Manzhirov, A.V., Chernysh, V.A.: Contact problem for a layered inhomogeneous aging cylinder
reinforced by a rigid ring. J. Appl. Mech. Tech. Phys. 31, 894–900 (1990). https://doi.org/10.
1007/BF00854204
Parshin, D.A.: Technological controlling the contact pressure onto a cylindrical substrate from a
coating being formed additively on its outer surface. AIP Conf. Proc. 2340, 040005 (2021). https://
doi.org/10.1063/5.0047838
Parshin, D.A., Manzhirov, A.V.: The mechanical problems on additive manufacturing of viscoelastic
solids with integral conditions on a surface increasing in the growth process. J. Phys: Conf. Ser.
991, 012063 (2018). https://doi.org/10.1088/1742-6596/991/1/012063
Szegö, G.: Orthogonal Polynomials. American Mathematical Society, Providence (1959)
Failure Modes of Skull Bone of a Dog
Head Under Impact Loading

Reda M. Othman , M. A. Hassan , and Wael M. Khierldeen

1 Introduction

Many pets and wild animal subject to accidents that cause either skull or brain
damage. Brain injuries are devastating and, unfortunately, often fatal.There are
both primary brain injuries that are the result of a direct insult to the pets’ brain,
and secondary brain injuries that occur following the primary brain injury. Secondary
brain injuries may include bleeding from a brain blood vessel or swelling of brain
tissue. In animal as well as in human the soft tissue of the brain is surrounded and
protected with the skull casing, so it is considered the main line of defense for
the brain soft tissue from any external head trauma. Traumatic brain injuries (TBI)
and skull fractures have serious ramifications (Iboaya et al. 2019); both may lead
to permanent disability, mental infirmity, and even death especially in old pets and
human patients (Peloso et al. 2003; Syring et al. 2001). Mechanically, skull frac-
tures arise from dynamic impact load such as falling from a height, car accident,
and violent interactions. Therefore, studying the skull bone behavior under dynamic
impact load is essential to understand the mechanisms of fracture to develop tolerance
criteria and to provide fundamental data to mathematical analogs such as the finite
element model this will help to improve the design consideration of safety systems

R. M. Othman · W. M. Khierldeen (B)

Department of Mechanical Engineering Faculty of Engineering, Assuit University, Asyut 71516,
Egypt
e-mail: [email protected]
M. A. Hassan
Materials Science and Engineering Department, Egypt-Japan University of Science and
Technology, New Borg El Arab, Egypt
W. M. Khierldeen
Industrial and Manufacturing Engineering Department, Egypt-Japan University of Science and
Technology, New Borg El Arab, Egypt

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 177
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_18
178 R. M. Othman et al.

for pets especially domesticated animals like dogs, donkeys, and horses. The biome-
chanical characteristics of skull fracture have been studied widely (Distriquin et al.
2020; Nahum and Melvin 2002). Some important mechanical parameters have been
discussed such as force, sectional elastic modulus, energy absorbed, shear stress,
acceleration deformation, and bending stress. It was found that velocity, strain rate,
sampling position and inter cranial variation all have a marked effect on some of
the mechanical parameters (Jennett and MacMillan 1981). Simon R. et al., used a
clinical coma scale modified from the Glasgow Coma Scale used for humans to
predict outcomes of dog head trauma resulted from many different events, including
motor vehicle accidents, bites, kicks, or penetrating injuries such as gunshot wounds
(Platt et al. 2001). S. Chocron et al., studied Cortical bone tensile mechanical prop-
erties at quasistatic and high rates (∼300 s−1 ) were determined ex vivo using the
right femurs of 12 female baboons form different using split Hopkinson pressure bar
technique. They found that the dynamic strength results of baboon bones increase
slightly with respect to quasistatic results (Chocron et al. 2012). However, despite
all efforts done to characterize the skull fracture parameters, there is a shortage in
data describing skull crack behavior, how it propagates, and skull fracture patterns
(Motherway et al. 2009; Delye et al. 2007). In this paper, the effect of in vivo dog
head trauma under dynamic loading and at high strain rates using SHPB is presented,
the fractography fractures skulls is used with aid of CT scan and Postmortem analysis
to make a complete understand of the three types of skull fracture (depressed, fisher,
and compound elevated skull fractures). Baladi Egyptian dogs’ specie were used in
the test as they are wild tough species with strong bone structure (the samples was
taken from street dog that was about to killed by the local authorities). Baladi dogs
are the original sight hounds of the Middle East, descendants of the Egyptian Saluki,
and Pharaoh Hounds.

2 Dog Head Structure and Testing Method

2.1 Overview of Dog Head Anatomy

Dog head skull is a complicated structure formed from several bones arranged and
connected together by interconnecting joints which called sutures as shown in Fig. 1a.
Skull bones are separated into two groups forming cranial and facial bones which
have a slightly different bone thicknesses and densities from dog to dog, The dog
head (cranium) constructs from multilayers as shown in Fig. 1b, it can be ordered
from the outside to inside as follow, the scalp, muscles, skull, meninges, and then
the brain. The temporal muscles which surround the flat and Cranium bones varied
in thickness. Then the periosteum which is a fibrous membrane attaches to the skull
bone and is responsible for bone healing after injury. This study focused on the cranial
bones, which form the cranial cavity and houses the brain, two important parts of the
cranial bone are the frontal and the parietal bones, both form the upper casing for
Failure Modes of Skull Bone of a Dog Head Under Impact Loading 179

the brain, and are made mainly from flat bones. The flat bone consists of three layers
in which the outer and inner layers are compact dense cortical bones. However, the
former layer is thicker than the latter, while the middle layer (spongy, c cancellous or
trabecular bone-diploe) is sandwiched between the previous two layers (Kaminski
et al. 2019), but near the sagittal crest line there is not diploe.
Test preparation. Six clinically healthy adult (same age) male dogs were used for
this study. All animals have been checked clinically before the experiments, none of
the animals showed medical problems or neurological disorders. The animals were
housed in a clean, dry individual cage with food and water available 24 hs/day. The

Fig. 1 a Dog skull anatomy,

b cross section in frontal
bone describing the 3 main
layers
180 R. M. Othman et al.

Table 1 Tested dogs’ data

No. Head Skull cranium Bone Bone density Bone Temporal
circumference circumferences density cranial vault thickness muscle
(mm) (mm) impact site (HU) impact site thickness
(mg/cm3 ) (mg/cm3 ) (mm) impact site
(mm)
1 335.2 230.1 980 1005 4.88 10
2 339.21 222.18 1122 1037 5.1 5.62
3 371.43 214.88 932.98 1000.5 6.06 6.74
4 394.54 240 919.22 924.28 4.44 16.22
5 375.94 199.89 994.5 1007.6 6.05 8.96
6 356.24 235.51 849.52 920 6.2 11.82

experimental study was executed according to the roles of the ethical Committee of
the Faculty of Veterinary Medicine, Assuit University, Egypt. Food was withheld
12 h before the experiment and water for four hours before. General anesthesia
using diazepam (…mg/kg) followed by ketamine (5.0 mg/kg, IM) and xylazine HCl
(1.0 mg/kg, IM). Pretest computed tomography (CT) images were taken to identify
the head and skull bone properties as listed in Table 1, Hounsfield unit, mg/cm3 (is
a quantitative scale for describing radiodensity) is used in CT scans to measure in
order to determine the cranial vault.

2.2 Impact Test Apparatus

For imitating the accident rate of loading, a SHPB apparatus was used in this study.
SHPB provides a controllable impact force informant where an accurate measure-
ment for load, striking speed and displacement can be measured in few microseconds
and high Strain rate. The system consists of a striker bar fired through gas gun in which
the firing pressure can be controlled to obtain different striker bar impact speeds. A
modified viscoelastic Hopkinson pressure bars (Poly methyl methacrylate PMMA
acrylate) are used to compensate for the tested soft tissues (low impedance materials)
for low noise and smooth curves (Phan-Thien et al. 1997; Theodorakopoulos 2015).
The dog head was fixed in position by means of head holder fixation mechanism to
ensure that the bare directed to the frontal side of the skull. The head was positioned
between the incident and transmitted bars and when the striker bar hit the incident
bar a compression wave double the length of the striker bar generate and propagate
through the incident bar as shown in Fig. 2a. Two strain gage stations were attached
to the middle of the incident and transmitted bars to record the three strain waves
signals namely, εI (t), εr (t), and εt (t) which are the incident, reflected and transmitted
strain waves, respectively. The strain gage station attached on the incident and trans-
mitted bars consists of a Wheatstone bridge with four strain gages. When this wave
Failure Modes of Skull Bone of a Dog Head Under Impact Loading 181

reached the specimen part of it reflects through the incident bar and the remaining
wave splatted into two parts one consumed in the specimen and the other transmitted
to the supporting transmitted bar which was found to be too small to be considered
as shown in Fig. 2b.
The equations of the one direction wave propagation were used to calculate the
load and displacement affecting the specimen using the strain signals recorded at
the strain gage station positioned in the middle of the two bars after correcting the
dispersion and attenuation effect as shown in Fig. 2c (Othman et al. 2016).
The adult dog skull layer consists of two flat compact bones and a spongy cancel-
lous bone in the middle called the diploe. Bone thickness is not uniform throughout

Incident bar Reflected

signal wave
Incident wave

transmitted Reflected Wave

bar signal

Incident
wave

Fig. 2 a Setup of the impact test, b output signal and c processed signal
182 R. M. Othman et al.

Fig. 3 The thickness of diploe layer for dog skull at frontal bone

the skull as can be seen in the plot of Fig. 3. It can be noted that the red line refers to
the sagittal line where there is not diploe at this position (only dense cortical bone).
All previous data collected using Mimics software.
The equations of the one direction wave propagation were used to calculate the
load and displacement affecting the specimen using the strain signals recorded at
the strain gage station positioned in the middle of the two bars after correcting the
dispersion and attenuation effect as shown in Fig. 2c (Othman et al. 2016).
The adult dog skull layer consists of two flat compact bones and a spongy cancel-
lous bone in the middle called the diploe. Bone thickness is not uniform throughout
the skull as can be seen in the plot of Fig. 3. It can be noted that the red line refers to
the sagittal line where there is not diploe at this position (only dense cortical bone).
All previous data collected using Mimics software.
Sex progressive values of impact forces were directed to the frontal site of the
skull ranged from 2 to 12 kN to obtain the mechanical response of the skull bone
under dynamic load. Each tested dog was scanned within 15 min from the test using
CT scan, and with the aid of post-mortem analysis to represent the fracture surface
tomography.
Failure Modes of Skull Bone of a Dog Head Under Impact Loading 183

3 Results

The test output data showed the regularity of the behavior of the skull bone and
the impact loading where the gradual of the fracture topography was agreed with the
gradually increased impact loading. Dog 1 produced no significant data, while 2 tests
(dogs 2 and 3) showed only subcutaneous emphysema without any fracture. Three
different types of fracture were presented of the other three tests (dogs 3, 4, and 6)
depressed fracture, depressed fissure fracture, and multiple depressed fracture with
compound elevated skull fracture. Dogs (5, 6) showed brain injury that brain lacer-
ation, intracerebral air. There was a significant correlation between impact forces,
impact velocity, impact site from sagittal crest bone, bone thickness, and the fracture
behavior the load–displacement test for the dogs from 2 to 5 are shown in Fig. 4

Dog 2 Dog 3

Dog 4 Dog 5

Dog 6

Fig. 4 Load–displacement curves for tested dogs

184 R. M. Othman et al.

3.1 Discussion

The dog frontal skull bone may be considered as a flat slap without any columns
or beam intersect. The critical element in this test is that in a direct contact to the
impactor and the disk around it, because of the concentration of shear stress that is
generated in the connection zone, and due to the skull bone punching risk. At the
moment when the punching occurs, the disk element of bone physically separate
from one another, which significantly disturbs the balance of the system formed by
these elements. The force distribution model can be represented by a disk loaded
in the center through a cylindrical bar, where the force transmitted from the disk
center to disk outer boundary causing bending stress and shear stress at boundary.
This model shows a potential of shear failure accompanied by conical shear zone
attempted at the disk center as shown in Fig. 5a.
However, the skull bone is a nonhomogeneous, multilayers composite organ,
consisting of two phases fluid and solid that is classified into two materials organic
(softer) and inorganic or mineral which give bones their hard structure. The softer
element protects the stiff one from brittle cracking, while the other stiff element stops
the soft one from yielding [15 m]. To illustrate this process a more focus analysis

Fig. 5 Dog 4 a fracture model, and b load–displacement curve with the CT of the various
Failure Modes of Skull Bone of a Dog Head Under Impact Loading 185

of dog (4) which recorded a depressed fracture in which one or more fragments of
bone may press inward on the brain, damaging the brain. The brain may be exposed
to the outside. From the force displacement curve Fig. 4b, it is noted that the first
zone (A) refers to the scalp tissue deformed until retched the muscles. Second zone
(B) began with the elastic of its tissue.
At (C) the muscles began to be deformed plastically (yielding of plastic deforma-
tion) from (E) followed by sudden drop after the first peak which is due to starting
the muscles failure and its tissues layer were fully compacted but not completely
damaged. Then the curve rising again to the second peak passing with (G) with
single phase slope line due to uniform material (cortical) near sagittal line, which
represent the elasticity of bone to the maximum load which acting directly to the
bone causing starting of formation of the sheared cone, which mean that the bone
stating to lose its punching resistance at the curve peak (H), then sudden drop which
mean that after initial fracture the load carrying capacity of the skull bone drops.
Then the following peaks (M) represent the crack propagation cross the skull bone
layers in 3D directions. The same pattern can be identified for the dogs 5, 6.

4 Conclusion

Three types of Skull fractures (depressed fracture, depressed fissure fracture, and
multiple depressed with compound elevated skull fracture) have been observed. The
fracture behavior was found to be dependent on several parameters such as bone
density, impact site, and impact force magnitude. In addition, the composition of the
bone layers plays the main role of crack behavior versus impact dynamic load due to
the mechanical properties of the combination of cancellous and cortical bone layers.
The impact wave propagation dependable on the structure of the head skull layer
thickness and mechanical properties.

References

Allan Gunnarsson, C., Alexander, S.L., Weerasooriya, T.: Mechanical response and fracture of
human skull to indent loading. In: International Proceedings of the Society for Experimental
Mechanics Series. pp. 3–44 Springer Cham (2019)
Chocron, S., Nicolella, D., Nicholls, A.E., Bredbenner, T., Havill L.: Dynamic testing of old and
young baboon cortical bone with numerical validation. Eur. Phys. J. Conf. 26 (2012). https://doi.
org/10.1051/epjconf/20122603004
Delye, H., Verschueren, P., Depreitere, B., Verpoest, I., Berckmans, D., Vander Sloten, J., Van Der
Perre, G., Goffin, J.: Biomechanics of frontal skull fracture. J. Neurotrauma 24(10), 1576–1586
(2007)
Distriquin, Y., Vital, J.-M., Ella, B.: Biomechanical analysis of skull trauma and opportunity in
neuroradiology interpretation to explain the post-concussion syndrome: literature review and
case studies presentation. Eur. Radiol. Exp. 4(66) (2020). https://doi.org/10.1186/s41747-020-
00194-x open access
186 R. M. Othman et al.

Iboaya, A., Harris, J.L., Arickx, A., Nudo, R.J.: Models of traumatic brain injury in aged animals.
A clinical perspective. Neurorehabilitation Neural Repair 33(12), 975–988 (2019)
Jennett, B., MacMillan, R.: Epidemiology of head injury. Br. Med. J. (Clin. Res. Ed). 282(6258),
101–104 (1981)
Kaminski, J., Waller, B.M., Diogo, R., Hartstone-Rose, A., Burrows, A.M.: Evolution of facial
muscle anatomy in dogs. PNAS 116(29), 14677–14681 (2019)
Motherway, J.A., Verschueren, P., Van der Perre, G., Vander Sloten, J., Gilchrist, M.D.: The mechan-
ical properties of cranial bone: the effect of loading rate and cranial sampling position. J. Biomech.
42(13), 2129–2135 (2009)
Nahum, A.M., Melvin, J.W.: Accidental Injury Biomechanics and Prevention, 2nd edn. Springer,
New York (2002)
Othman, R.M., Khier, W.M., Naser, A.A.: Experimental study for dispersion and attenuation correc-
tion in viscoelastic split Hopkinson pressure bar arrangement. JES. J. Eng. Sci. 44(5), 566–576
(2016)
Peloso, P.M., von Holst, H., Kleiven, S.: The epidemiology of head injuries in Sweden from 1987
to 2000. Inj. Control Saf. Promot. 10(3), 173–180 (2003)
Phan-Thien, N., Bilston, L.E., Liu, Z.: Linear viscoelastic properties of bovine brain tissue in shear.
Biorheology 34(6), 377–385 (1997)
Platt, S.R., Radaelli, S.T., McDonnell, J.J.: The prognostic value of the modified Glasgow Coma
scale in head trauma in dogs. J. Vet. Intern. Med. 15(6), 581–584 (2001)
Syring, R.S., Otto, C.M., Drobatz, K.J.: Hyperglycemia in dogs and cats with head trauma: 122
cases (1997–1999). J. Am. Vet. Med. Assoc. 218(7) (2001). https://doi.org/10.2460/javma.2001.
218.1124
Theodorakopoulos, I.D.: Determination of the useful frequency for the calculation of the attenuation
coefficient and phase velocity in a viscoelastic Hopkinson pressure bar. Int. Appl. Mech. 51(3),
360–368 (2015)
Seismic Retrofit of RC Building
with Elastic Stage of Buckling-Restrain
Braces
Thanawat Thipprapan, Amphon Jarasjarungkiat, Panumas Saingam,
and Arthit Petchsasithon

1 Introduction

Over the past half-century, an earthquake is a natural disaster that affects serious
damage to life and property. After the Mao Lao earthquake, much of the damage
was observed in reinforced concrete (RC) structures, as presented in Lukku-
naprasit et al. (2014). Therefore, seismic design and detailing requirements for
buildings have considerably improved. In Thailand, Ministerial Regulations about
Load, Resistance, Durability of the building, and Ground supporting buildings in
Earthquake-Resistance has been published in 2021 (Ministry of Interior 2021).
Many existing RC buildings in Thailand should be retrofitted because the country
has historically been considered to have a low seismic hazard, and the majority of
current existing RC buildings were not designed with seismic resistance. If an earth-
quake occurs, the existing RC buildings may not be able to withstand earthquakes,
which has a high risk of being unable to withstand earthquakes because the structure
is not detailed the reinforcing rebar has sufficient toughness to resist earthquakes
(Lukkunaprasit et al. 2014).
At present, there are various methods of strengthening reinforced concrete build-
ings to resist earthquakes, including the extension of the column cross-section or
the use of Carbon-Fiber Reinforced Polymer (CFRP) (Triantafillou 1998). Another
method is to install a shear wall (Canbay et al. 2003), which limits drift and ensures
that the mainframe remains elastic, but imposes large floor accelerations.
Seismic retrofit of RC buildings with energy dissipation devices is an innova-
tive method to develop seismic resistance of the buildings (Saingam et al. 2020,
2021). Buckling-restrained brace (BRB) is a seismic device, composed of a tapered,
energy-dissipating steel core encased in an axially de-coupled restrainer, which is

T. Thipprapan · A. Jarasjarungkiat · P. Saingam (B) · A. Petchsasithon

Department of Civil Engineering, School of Engineering, King Mongkut’s Institute of Technology
Ladkrabang, Bangkok 10520, Thailand
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 187
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_19
188 T. Thipprapan et al.

separated by a debonding interface (Takeuchi 2018). BRB components are shown

in Fig. 1. The features of BRBs are withstanding compressive stress without lateral
buckling and being able to dissipate seismic energy that resulting in the building can
withstand earthquakes safely (Panyakapo and Ponpimonpat 2014). BRBs consist of
the following components: steel core (Fig. 2.), unbounding material, encasing mate-
rial, and outer tube. As for the steel core, there are various cross-sections such as
Flat Bar, T-Bar, or Cross, which at this point is the area that allows for yielding.
The size of the cross-section will be reduced and will be set to be in the middle of
the length of BRBs. The outer tube is defined as having higher strength than the
maximum tensile strength of the steel core according to American Institute of Steel
Construction Incorporation requirements (American Institute of Steel Construction
Incorporation requirements 2012).
This article presents the seismic strengthening of a RC building with BRBs
to improve the force resistance system of the existing RC buildings. The elastic
stage of the seismic retrofit of RC buildings with BRB is investigated. A 4-story
RC building is used as an example building. The results of the bare existing RC
building are compared to the retrofitted building to verify the presented seismic
retrofit application.

Fig. 1 Schematic of buckling-restrained brace (BRB or unbonded brace, UB) (Lai and Tsai 2002)
Seismic Retrofit of RC Building with Elastic Stage of Buckling-Restrain … 189

Fig. 2 Typical types of buckling-restrained (Lai and Tsai 2002)

2 Seismic Evaluation for Existing RC Building

This study proposes a simple method to retrofit RC buildings with BRBs. A 4-

story school reinforced concrete building in Thailand is selected as a target building.
This building was designed and constructed before the Ministerial Regulations,
Load, Resistance, Durability of the building, and Ground supporting buildings in
Earthquake-Resistance of 2021 (Ministry of Interior 2021). This affects that several
buildings, which were not designed to withstand earthquakes, are required to improve
their seismic performance. Therefore, this building had to be inspected to know the
ability to resist earthquakes and know the point of risk of failure or failure from the
earthquake, and able to strengthen at the right point.
The aforementioned building was modeled to be located in Mueang District,
Chiang Mai Province. The target area is classified as a possible area where the
building may be affected in terms of strength and a high degree of stability in the
presence of seismic vibrations. For the analysis, the model was simulated using the
MIDAS GEN program (User and‘s Guide for MIDAS, Gen manual, Getting Started,
Verification Examples 1989).
The target 4-story RC school building is shown in Fig. 3. The 1st floor is open-air
and the 2nd to 4th floors have infill walls all around. The 1st floor has a height of 4.50 m
and the 2nd –4th floor has a height of 3.50 m per floor. The steel roof has a height
190 T. Thipprapan et al.

Fig. 3 Model existing building

of 1.70 m, a building width of 10 m, and a length of 40 m. Material characteristics

considered for modeling and analysis are fc ’ = 240 ksc for grade concrete and SD40
(fy = 4000 ksc) for grade steel rebars. The equivalent static forces based on the
newest seismic design code (Ministry of Interior 2021) are applied to the target
school building. The results indicated that the 1st floor is a soft story and has a
high risk of being unable to withstand earthquakes. The detailed discussion will be
presented in the following sections.

3 Results of Seismic Evaluation for Existing RC Building

From the inspection of earthquake resistance, it was found that this building was
unable to withstand earthquakes. Ultimate compressive strength of columns on the 1st
to the 3rd floor = 1477 kN, 898 kN and 439 kN, respectively. But nominal compressive
strength of columns on the 1st to the 3rd floor = 737 kN, 709 kN and 368 kN,
respectively. Failure will occur on columns from the 1st floor to the 3rd floor as a
result of compressive strength (P) being insufficient. The lateral displacement in the
X-axis has a displacement of 48.93 mm. > H/500 = 30 mm., as shown in Fig. 4, and
a lateral displacement in the Y-axis has a displacement value of 30.93 mm. > H/500
= 30 mm., as shown in Fig. 5, which the lateral displacement is greater than the
standard specified.
It can be indicated that this school building has insufficient seismic strength and
stiffness of the building. Therefore, this study introduced the seismic retrofit of the
Seismic Retrofit of RC Building with Elastic Stage of Buckling-Restrain … 191

Fig. 4 The lateral displacement in the X-axis

Fig. 5 The lateral displacement in the Y-axis

target buildings with BRBs, which improve in terms of resistance, increase stiffness
and increase the damping of buildings.

4 Retrofit Existing RC Building with Buckling-Restrained

Braces

BRBs have several appealing qualities that are regularly considered for retrofitting
projects. Seismic retrofit is frequently challenging to implement while maintaining
192 T. Thipprapan et al.

continuous occupancy, and they frequently have a negative impact on the appearance
of the building (Takeuchi 2018).
From the inspection of the existing building structure, it was aware of the necessity
that the building’s resistance was needed to be strengthened. Then, the strength of
the BRB is required to do the seismic retrofit. The strength of 275 kN was selected
based on the local research by Prof. Dr. Pennueng Wanichchai (Wararuksajja et al.
2020). Then, quantified the number of BRBs on each floor. It is determined by the
excess strength of the columns of each floor divided by the strength of one BRB.
Sum of excess strength of columns on the 1st floor = 3019.78 kN. The required
number of BRBs is 10.98 braces. Arrange the symmetry to obtain the desired amount
at 12 braces.
Sum of excess strength of columns on the 2nd floor = 281.82 kN. The required
number of BRBs is 1.02 braces. Arrange the symmetry to obtain the desired amount
at 2 braces.
Sum of excess strength of columns on the 3rd floor = 374.29 kN. The required
number of BRBs is 1.36 sets. Arrange the symmetry to obtain the desired amount at
2 braces.
Once the number of BRB is known, then arrange the position of the BRB because
the columns of the 1st floor on the edge of the building are the part that happens
failure and to prevent torsion from occurring to the buckling. The position of BRBs
on the 1st floor will be arranged at the edge of the building on both sides, as shown
in Fig. 6.
The position of the BRBs on the 2nd and 3rd floors are arranged in the same position
at the center of the building to prevent torsion from occurring in the building.
When the positioning of the buckling-restrained braces on each floor is complete,
the next step is to determine the initial stiffness (K 0 ) in the MIDAS GEN program.
The K 0 value is determined from Eqs. (1)–(3).

Fig. 6 The position of the buckling-restrained braces

Seismic Retrofit of RC Building with Elastic Stage of Buckling-Restrain … 193
/
K 0 = P d yBRB (1)

/
δ BRB
y = L BRB θdy (2)
/( )
L BRB = L 2x,y + H 2 (3)

where P is the strength of the BRBs (kN), δy BRB is the yield displacement at the yield
point of the BRBs (m.), L BRB is the length of BRBs (m.), θdy is the yield story drift
ratio of brb (rad), which is 1/750–1/1000 rad (Takeuchi 2018; Sutcu et al. 2014), in
this article choose 1/750 rad because it gives the minimum value of K 0 , L x, y 2 is the
distance between column and column (m.), H is the height between the floors (m.)
K 0 can be calculated and defined post-yield coefficient (α1 ) as 2% of initial stiffness
(K 0 ).

5 Results of Seismic Retrofit of RC Building

with Buckling-Restrained Braces

The displacements of the before retrofit building’s displacements of the x-axis and
y-axis are shown in Figs. 7 and 9, respectively. The results are only from the elastic
stage of structural behavior. When retrofitting the existing RC building with BRBs,
the results indicated that the retrofitted building was able to withstand earthquakes
because the lateral displacements are smaller than the allowable value of 30 mm.
The strength of the column is increased while the lateral displacement values are
reduced as shown in Figs. 8 and 10, for the x-axis and y-axis, respectively. The
results indicated that the roof displacement of the x-axis and y-axis are reduced by
44.7% and 24.7%, respectively. This shows the effectiveness of the proposed retrofit
strategy.
194 T. Thipprapan et al.

Fig. 7 Lateral displacement 6

x-axis before retrofit
5

4 EQX
EQXn

Story
3 EQXp
EQY
2 EQYn
Allowable EQYp
1 Displacement
Displacement (mm)
0
-10 0 10 20 30 40 50 60

Fig. 8 Lateral displacement 6

x-axis after retrofitting with
BRB 5

4 EQX
EQXn
Story

3 EQXp
EQY
2 EQYn
Allowable
EQYp
1 Displacement

0 Displacement (mm)
-10 0 10 20 30 40 50 60

Fig. 9 Lateral displacement 6

Allowable
y-axis before retrofit
5 Displacement

4 EQX
EQXn
Story

3 EQXp
EQY
2 EQYn
EQYp
1
Displacement (mm)
0
-20 -10 0 10 20 30 40
Seismic Retrofit of RC Building with Elastic Stage of Buckling-Restrain … 195

Fig. 10 Lateral 6
displacement y-axis after
retrofitting with BRB 5

4 EQX
EQXn

Story
3 EQXp
EQY
2 EQYn

Allowable EQYp
1
Displacement
Displacement (mm)
0
-20 -10 0 10 20 30 40

6 Conclusion

The seismic retrofit of RC buildings with buckling-restrain braces is investigated. The

analysis results indicated that the existing RC building cannot withstand the seismic
force from the newest seismic design code. There were failures at the column of the
structure of the 1st floor to 3rd floor and has a lateral displacement exceeding the
standard.
After retrofitting the target RC building with buckling-restrained braces, the
results indicated that the retrofitted building was able to withstand earthquakes. The
strength of columns is increased and the lateral displacement values are reduced by
24.7–44.7%, respectively. Therefore, the retrofitted building can be resisted future
earthquakes. This verified the effectiveness of the introduced retrofit method.

References

American Institute of Steel Construction Incorporation requirements (2012)

Canbay, E., Ersoy, U., Ozcebe, G.: Contribution of reinforced concrete infills to seismic behavior
of structural systems. ACI Struct. J. 2003(100), 637–643 (2003)
Lai J.W., Tsai, K.C.: Research and application of buckling restrained braces in Taiwan. In: National
Center for Research on Earthquake Engineering, No.200, Sec. 3, Shanghai Rd., Daan Chiu, Taipei,
Taiwan 106, ROC (2002)
Lukkunaprasit, P., Ruangrassamee, A., Boonyatee, T., Chintanapakdee, C., Jankaew, K., Thana-
sisathit, N., Chandrangsu, T.: Performance of structures in the Mw 6.1 Mae Lao Earthquake in
Thailand on May 5, 2014, and implications for future construction. J. Earthq. Eng. 20, 219–242
(2015)
Ministerial regulations about load, resistance, durability of the building, and ground supporting
buildings in Earthquake-Resistance. Ministry of Interior (2021)
Panyakapo, P., Ponpimonpat, P.: In strengthening of reinforced concrete school building with
buckling restrained brace (2014)
196 T. Thipprapan et al.

Saingam, P., Sutcu, F., Terazawa, Y., Fujishita, K., Lin, P.C., Celik, O.C., Takeuchi, T.: Composite
behavior in RC Buildings Retrofitted using Buckling-Restrained Braces with Elastic Steel Frames.
Eng. Struct. 219, 110896 (2020)
Saingam, P., Matsuzaki, R., Nishikawa, K., Sitler, B., Terazawa, Y., Takeuchi, T.: Experimental
dynamic characterization of friction brace dampers and application to the seismic retrofit of RC
buildings. Eng. Struct. 242, 112545 (2021)
Sutcu, F., Takeuchi, T., Matsui, R.: In Seismic retrofit design method for RC building using buckling-
restrained braces and steel frames (2014)
Takeuchi, T.: Buckling restrained braces history design and applications. In: 9th international confer-
ence on behavior of steel structures in seismic areas Christchurch, New Zealand, Feb 14–16.
(2018)
Triantafillou, T.C.: Shear strengthening of reinforced concrete beams using epoxy-bonded FRP
composites. ACI Structural J. 95(2), 107–115 (1998)
The User ‘s Guide for MIDAS/Gen manual, Getting Started, Verification Examples (1989)
Wararuksajja, W., Leelataviwat, S., Warnitchai, P., Bing, L., Tariq, H., Naiyana1, N.: Seismic
strengthening of soft-story RC moment frames. In: The 25th National Convention on Civil
Engineering, Chonburi, Thailand. July 15–17 (2020)
Response and Damage Identification
by Long-Gauge Strain Sensor
of Beam-Like Bridges Under the Loads
of Vehicle: A Comparison
of the Analytical Model and in Situ
Measurement

Pham-Truong Son, Tien Hoang, Denis Duhamel, Gilles Foret,

Franziska Schmidt, François-Baptiste Cartiaux, and Véronique Le Corvec

1 Introduction

Most bridges in the world have been built in the twentieth century. Hence, bridge
assessment and monitoring are necessary. One of the preliminary stages of bridge
Structural Health Monitoring (SHM) is to construct a direct model that can predict
the behavior of the structure under traffic loads. Many studies have been carried
out about this problem: Ichikawa et al. (2000) found the solution of a continuous
beam subjected to a moving mass by eigenmodes superposition. Dugush et al.
(2002) used infinite polynomial series to describe the mode shapes of a contin-
uous beam with non-uniform section. In another approach, Hoang et al. (2017) used
space–time Fourier transform to calculate the dynamic response of a periodically
supported beam. Recently, Son et al. (2021) decoupled a continuous cracked beam
into several segments. By introducing transfer matrices that connect these segments,
the eigenvalue problem has been solved directly.
The presence of structural damage will modify the dynamic behavior of the beam.
Many researchers have studied the effect of cracks on the natural frequency of a
cantilever beam, and the identification of the location and the magnitude of the
cracks (Ostachowicz and Krawczuk 1991; Rizos et al. 1990). However, these methods

P.-T. Son (B) · T. Hoang · D. Duhamel · G. Foret

Laboratoire Navier, UMR 8205, Ecole Des Ponts ParisTech, Université Gustave Eiffel, CNRS, 6-8
Avenue Blaise Pascal, Champs-Sur-Marne, France
e-mail: [email protected]
F. Schmidt
Université Gustave Eiffel, IFSTTAR/MAST, 14-20 Boulevard Newton, Champs-Sur-Marne,
France
F.-B. Cartiaux · V. Le Corvec
OSMOS Group, 37 Rue La Pérouse, Paris, France

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 197
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_20
198 P.-T. Son et al.

require many sensors. A moving load is an excitation which depends not only on
time, but also on space. From the response of the beam, we can find out its modal
properties such as the natural frequencies, the mode shapes…
Nowadays, bridges can be monitored continuously thanks to advanced tech-
nology. Unlike the traditionally electrical strain gauge, a long-gauge optical strand
measures the average strain over its length. Hence, it has a higher probability to
cover the damage. Moreover, Optical Strand sensors have many advantages such
as high-sample strain measurements over long periods, high precision and dura-
bility (Cartiaux et al. 2018). The use of these sensors is not only restricted in civil
engineering, but also in aviation, pipeline transport (Ignatovich et al. 2013).
This paper presents firstly an analytical method that describes the response of a
damaged continuous beam which is subjected to a multi-axle vehicle load. Based
on the longitudinal distribution of strain on a simply supported beam, we propose
a damage index which can be used to identify and localize the damage. By using
this analytical method, a parametric study is carried out to study the dependence of
the proposed index to the properties of the damage. Finally, the analytical method is
applied to calculate the response of an existing structure, and the results are compared
to the real signals obtained during the test.

2 Formulation of the Problem

Consider a continuous beam of total length L, which has Ns spans. In the general
case, we suppose that a damage is located in the kth span. The damaged span is
divided into 3 separated beam segments: the damaged zone of length L e and two
intact segments which are on the left and on the right of the damaged segment. The
initial continuous beam can be modeled as a system of N = Ns + 2 beam segments
(Fig. 1). A vehicle of M-axles passes over the beam with constant velocity
Σ v. Let’s
denote di the distance between the ith and (i + 1)th axle and Di−1 = i−1 j=1 j is the
d
distance between the first axle and the ith axle (i ≥ 2).
Assume that the beam follows the Euler–Bernoulli beam theory and has a linear
elastic behavior. The governing equation of the r th beam segment which describes
the vertical displacement u r (xr , t) in the local coordinates xr is (Svedholm et al.
2016):

Fig. 1 A continuous beam with N spans subjected to M moving loads

Response and Damage Identification by Long-Gauge Strain Sensor … 199

∂ 4 ur ∗ ∂ ur
5
∂u r ∂ 2 ur
Er I + E I + c + ρ A = f (xr , t), 0 ≤ xr ≤ L r (1)
∂ xr4 r
∂t∂ xr4 ∂t ∂t 2

where Er I is the flexural stiffness, ρ is the mass density and A is the cross-sectional
area of the beam. In Eq. 1, the ) c(∂u r /∂t) represents the external
( familiar term
damping whereas the term Er∗ I ∂ 5 u r /∂t∂ xr4 represents the internal damping. The
force applied by the vehicle can be expressed by:

f (x, t) =P1 δ(x − vt) + P2 δ(x − vt + D1 ) + . . . + PM δ(x − vt + D M−1 )

Σ
M
= Pi δ(x − vt + Di−1 ).
i=1

where δ(x) is the Dirac distribution and x is the global coordinates.

The solution of the initial equation can be decoupled into two terms: the time-
dependent term q(t) and the mode shapes φr (xr ):

u r (xr , t) = φr (xr )q(t) (2)

Substituting Eq. 2 into Eq. 1, setting f (x, t) to be zero to obtain the free vibrations
and by separating the variables, we have:

φr(4) (x) cq̇ + ρ Aq̈

=− = λr4 (3)
φr (x) Er I ∗ q̇ + Er I q

3 Natural Frequencies and Mode Shapes

From Eq. 3, the mode shapes of the r th beam segment are the solution of the following
differential equation:

d4 φr (xr )
− λr4 φr (xr ) = 0 (4)
dxr

where φr (xr ) is the mode shape of the rth beam segment. The form of φr (xr ) can be
expressed explicitly as:

φr (xr ) = Ar sin λr xr + Br cos λr xr + Cr sinh λr xr + Dr cosh λr xr (5)

200 P.-T. Son et al.

3.1 Configuration of the Beam

BY the same concept of transfer matrices as introduced by Son et al. (2021), we search
the transfer matrix M such that X r = M X r −1 . Here X r = [Ar , Br , Cr , Dr ]T is the
vector which contains the coefficients of the mode shapes φr (xr ) for the rth beam
segment. This 4 × 4 matrix ensures the continuity conditions at the intermediate
support points and at the interface between the intact segment and the damaged
segment.

3.2 The Intermediate Supports

The intermediate supports are assumed to be infinitely rigid, so the displacement is

blocked at these points. Moreover, the slope of the beam and the bending moment
are unchanged across the support. Considering the (r − 1) th support which is found
between the (r − 1) th span and the r th span, the above conditions can be written as:

u r (0, t) = u r (L r , t) = 0,
' '
u r (0, t) = u r −1 (L r −1 , t),
'' ''
u r (0, t) = u r −1 (L r −1 , t). (6)

We obtain S(L r −1 , L r ) the transfer matrix at the intermediate support, such that:
X r = S4×4 X r −1 . This matrix can be expressed as:
⎡ ⎤
−κ.s + 2η.c −η.s − κ.c κ.sh + 2η.ch 2η.sh + κ.ch
1⎢ s c −sh −ch ⎥
S4×4 = ⎢ ⎥ (7)
⎣
2 κ.s + 2ε.c −2ε.s + κ.c −κ.sh + 2ε.ch 2ε.sh + κ.ch ⎦
−s −c sh ch

in which:

s = sin λL r −1 , c = cos λL r −1 , sh = sinh λL r −1 , ch = cosh λL r −1 ,

κ = (cos λL r − cosh λL r )/(sin λL r − sinh λL r ),
η = − sinh λL r /(sin λL r − sinh λL r ), ε = 1 − κ.
Response and Damage Identification by Long-Gauge Strain Sensor … 201

3.3 The Interface Between Intact and Damaged Beam

Segments

In the literature, a crack can be modeled by a spiral spring or by a reduction of the

bending stiffness in the concerned zone (Friswell and Penny 2002). In this paper, we
use the second model to describe the damage. We denote (E I )1 and L 1 , respectively,
the bending stiffness and the length of the segment located on the left of the interface.
By the same way, (E I )2 , L 2 are the bending stiffness and the length of the right
segment. Across the interface, the traverse displacement, the slope of the beam,
the bending moment and the shear force are unchanged. Therefore, the continuity
conditions at the interface are:

u 2 (0, t) = u 1 (L 1 , t), u '2 (0, t) = u '1 (L 1 , t),

(E I )2 u ''2 (0, t) = (E I )1 u 1 (L 1 , t), (E I )2 u 2 (0, t) = (E I )1 u ''1 (L 1 , t).
'' '''
(8)

The transfer matrix at the interface T (k, L 1 ) is such that X 2 = T (k, L 1 )X 1 :

⎡ √ √ √ √ ⎤
(1 + k)/ kc −(1 + k)/ ks (1 − k)/ kch (1 − k)/ ksh
1 ⎢ (1 + k)s (1 + k)c√ (1 − k)sh (1 − k)ch ⎥
T 4×4 = ⎢ √ √ √ ⎥
⎣
2 (1 − k)/ kc −(1 − k)/ ks (1 + k)/ kch (1 + k)/ ksh ⎦
(1 − k)s (1 − k)c (1 + k)sh (1 + k)ch

in which:
(E I )1
k2 = , (9)
(E I )2

s = sin λ1 L 1 , c = cos λ1 L 1 , sh = sinh λ1 L 1 , ch = cosh λ1 L 1 .

By assembling all these aforementioned matrices together and taking into account
the boundary conditions
√ of the continuous beam, we deduce an eigenvalue problem
to solve for ω = λr2 Er I /ρ S, which is the angular frequency of the beam. Once the
values of {ω} are found, by integrating both sides of the initial equation, we solve
the same initial differential equation for the general coordinates q(t), segment by
segment. Finally, by superposition of the modes, we have the analytical solution of
the problem.

3.4 Damage Identification Using Distributed Long-Gauge

Strain Sensors Method

For a simply supported beam, it can be shown that the maximum and the area values
of the static macro-strain follow a parabolic distribution (Wu et al. 2019). Some
202 P.-T. Son et al.

Fig. 2 Macro-strain time histories of a simply supported beam and the distribution of their area

damage indices based on this property were proposed in Wu et al. (2019); Hong,
et al. 2016). Moreover, the parabolic shape of the area values is still correct in case
where the total strain is measured by a long-gauge sensor, which is the sum of static
and dynamic strain. For illustration, it is supposed that a sequence of 10 Optical
Strands sensors of equal length l g = 1 m is placed under a simply supported beam
of 10 m length. The sensors are named as CO-1–CO-10, from left to right. Figure 2a
presents the numerical results of total macro-strain time histories obtained by these
Optical Strands, subjected to a 2-axle vehicle which passes over the beam at constant
speed v = 5 m/s. The front load and the rear load are 1 kN and 2 kN, respectively.
The distance between these two loads is 3 m. The beam is made of steel (E = 210
GPa,ρ = 7800 kg/m3 ). The Fig. 2b shows the distribution of the area of these
responses. The orange line is the regression parabola fˆ.
We propose a damage index (DI) which measures the relative error between the
approached parabola and the measured points:

|| f − f 02 ||
E G = 100 × . (10)
|| f 02 ||
/
in which: ||x||2 = x12 + x22 + . . . + xn2 is the Euclidian norm of vector x.
In order to localize the damaged zone, we propose an additional criterion, which
is based on the above damage index EG. It is calculated by the same way of the
criterion EG, but the considered point εi is excluded:
|| i ||
|| f − f i ||
E L i = EG = 100 × || i ||
i 02
(11)
|| f ||
02
Response and Damage Identification by Long-Gauge Strain Sensor … 203

3.5 Results

To illustrate the above criteria, we take the same steel beam and same vehicle load
as presented on the previous subsection. We create a damaged zone at the center of
the beam. The length of the damage varies from 0.5 m to 3.0 m and the reduction
of bending stiffness on this zone varies from 0% (intact) to 50%. As a parabola is
defined by 3 parameters, at least 4 measure points are required to determine the
damage index EG. Hence, to identify and localize the damage, at least 5 measured
points are required. In this parametric study, we use 5 Optical Strands with 2 m-length
each, covering the whole length of the beam. The response of the beam is calculated
by the current analytical model.
Figure 3 presents the dependence of EG as the function of the length and of the
reduction of bending stiffness of the damaged zone.
From Fig. 3a, we see that EG varies linearly with Δ(E I ). Figure 3b shows that
the damage index EG get its maximal value while the length of the damage is equal
to 2 m. The relation between EG and the size of the damage is also linear. The global
damage index is exceeding 5% for a damaged zone of length 1.5 m and a reduction
of bending stiffness ΔE = 10%.
To localize the damage, the local criterion EL is calculated in case the size of the
damage is 2 m, for different levels of reduction of bending stiffness. We see that this
index EL vanishes at the sensor which corresponds to the damage (Fig. 4). For a
damage of ΔE = 10%, the average value of EL is over 4%.

4 Application to Bridge Structures

The analytical method is now applied to calculate the strain response of an existing
highway overpass in France. This prestressed concrete girder bridge was built in the
1960s. The bridge has 4 spans, where each of those is composed of 10 prestressed
concrete beams. The structure has been monitored since 2011 by the Optical Strands
sensors of length 1 m (red) and 2 m (blue). Figure 5 shows the instrumented plan
of the bridge with the selected sensors. There was a calibration test on the bridge in
2019.
One of the test trucks is a 5-axle of 44 tons. From the first axle, the distances
between two consecutive axles are: d1 = 3.8 m, d2 = 3.0 m, d3 = d4 = 1.25 m. The
axle loads P1 , P2 , P3 , P4 , P5 are, respectively, 6.94 t, 10.62 t, 9.2 t, 8.64 t and 8.16 t.
The vehicle passes over the bridge in direction S–N at constant speed 50 km/h, so
it enters firstly into the Span 3, then the Span 2. In the analytical model, the bridge
is considered as a set of T-beams in parallel. The general cross-section of the T-
beam is calculated for its properties (second moment of area, area…). The density of
concrete is 2500 kg/m3 and its Young modulus is 36.6 GPa. Because the forces are
also distributed in the transverse direction, a coefficient of 0.14 is taken into account
204 P.-T. Son et al.

Fig. 3 Damage index EG in function of the properties of damage

Response and Damage Identification by Long-Gauge Strain Sensor … 205

Fig. 4 Local criterion for localizing the damage

Fig. 5 Instrumentation plan on the central spans of the bridge

for the response of the largest signal. The other signals are re-constructed from the
load distribution factor obtained for this girder bridge.
Figure 6 shows the results of the analytical method comparing with the signal
obtained from the in situ test. The solid lines present the results calculated by the
present analytical method, while the dashed lines are the measured strain during the
test. The calculated and measured signals are coherent.
206 P.-T. Son et al.

Fig. 6 Comparison of the analytical results and in situ measurement

5 Conclusion

The analytical method gives us quickly the dynamic response of the bridge structure.
It takes under 10 s to calculate the presented parametric study, whereas the same
model calculated by FEM costs more than 15 min in total. From the distribution
of the area of strain time history along the beam, the proposed index can detect the
presence of a damage and localized its position. The coherence between the simulated
results and the real signals gives an opportunity for the analytical method on modeling
of the bridge, in order to obtain rapidly the total response of the structure.

Acknowledgements This work has been developed in the partnership between OSMOS Group and
Ecole des Ponts ParisTech. The supports from the personnel of OSMOS Group are acknowledged
as well.

References

Cartiaux, F.B., Pelletier, P., Semiao, J.: Traffic and temperature effects monitoring on bridges by
optical strands strain sensors, pp. 1–8 (2018)
Dugush, Y.A., Eisenberger, M.: Vibrations of non-uniform continuous beams un-der moving loads.
J. Sound Vib. 254(5), 911–926 (2002)
Friswell, M.I., Penny, J.E.T.: Crack modeling for structural health monitoring. Struct. Heal. Monit.
1(2), 139–148 (2002)
Hoang, T., et al.: Calculation of force distribution for a periodically supported beam subjected to
moving loads. J. Sound Vib. 388, 327–338 (2017)
Hong, W. et al.: Strain-based damage-assessment method for bridges under moving vehicular loads
using long-gauge strain sensing. J. Bridg. Eng. 21(10) (2016)
Response and Damage Identification by Long-Gauge Strain Sensor … 207

Ichikawa, M., Miyakawa, Y., Matsuda, A.: Vibration analysis of the continuous beam subjected to
a moving mass. J. Sound Vib. 230(3), 493–506 (2000)
Ignatovich, S.R., et al.: Fatigue damage and sensor development for aircraft structural health
monitoring. Theor. Appl. Fract. Mech. 65, 23–27 (2013)
Ostachowicz, W.M., Krawczuk, M.: Analysis of the effect of cracks on the natural frequencies of a
cantilever beam. J. Sound Vib. 150(2), 191–201 (1991)
Rizos, P.F., et al.: Identification of crack location and magnitude in a cantilever beam from the
vibration modes. J. Sound Vib. 138(3), 381–388 (1990)
Son, P.T. et al.: Dynamic response of a cracked multi-span continuous beam subjected to a moving
multi-axle vehicle load. In: 10th International Conference on Structural Health Monitoring of
Intelligent Infrastructure, SHMII 10, 2021, no. July, pp. 1–8 (2021)
Svedholm, C., et al.: Vibration of damped uniform beams with general end conditions under moving
loads. Eng. Struct. 126, 40–52 (2016)
Wu, B., et al.: Parametric study of a rapid bridge assessment method using distributed macro-strain
influence envelope line. Mech. Syst. Signal Process. 120, 642–663 (2019)
Metallic Materials and Corrosion
Protection
Relation Between Binging Energy,
Isothermal Bulk Modulus, and Surface
Tension of Liquid Metals

Yu. N. Starodubtsev and V. S. Tsepelev

1 Introduction

Interatomic (intermolecular) forces (Israelachvili 2011) bind the atoms together and
prevent the deformation of the body. The interatomic interaction is characterized by
the cohesion energy E coh or the energy of atomization (sublimation), which is equal
to the energy required for the decomposition of a solid into independent atoms at
0 K. The vacancy formation energy E vac can also be a characteristic of interatomic
interaction. Vacancy formation energy is the energy required to take an atom from
inside the crystal, and to place it on the surface without altering the surface energy of
the crystal (Friedel 1964). The binding energies include the melting enthalpy Δl s H m
and the evaporation enthalpy Δg l H b , which determine the amount of heat required
to melt a solid at the melting point T m or to evaporate a liquid at boiling point T b .
The thermal energy RT m and RT b at the melting or boiling point is used to compare
the strength of interatomic interaction with the destructive effect of thermal motion.
Previously, the relationship between the bulk modulus, surface tension and certain
types of binding energies was demonstrated (Kaptay 2008; Iida and Guthrie 2010;
Aqra 2014; Keene 1993; Turkdogan 2002). In the present work, we investigated the
relationship between various energy quantities characterizing the interatomic bond
with the isothermal bulk modulus and surface tension of liquid metals at the melting
point. The aim of the work was to find the energy quantities that are most related to
the bulk modulus and surface tension.

Yu. N. Starodubtsev · V. S. Tsepelev (B)

Research Center for Physics of Metal Liquids, Ural Federal University, Ekaterinburg, Russia
e-mail: [email protected]
Yu. N. Starodubtsev
Gammamet Research and Production Enterprise, Ekaterinburg, Russia

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 211
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_21
212 Yu. N. Starodubtsev and V. S. Tsepelev

2 Initial Data

Data on the cohesion energy E coh (J·mol–3 ), melting enthalpy ΔH m (J·mol–3 ), evapo-
ration enthalpy ΔH b (J·mol–3 ), melting point T m (K) and boiling point T b (K), molar
volume V m (m3 ·mol−1 ) was taken from Iida and Guthrie (2015). For the experimental
vacancy formation energy E vac (J·mol–3 ) the values collected in Starodubtsev and
Tsepelev 2021a were used. We also used data on the sublimation enthalpy ΔH sub
(J·mol–3 ) (Garvin 1981), Debye temperature ΘD (K) (Stewart 1983) and surface
tension at the melting point γm (J·m−2 ) (Iida and Guthrie 2015). The isothermal bulk
modulus at the melting point K Tm (J·m−3 ) were taken from Blairs (2007); Marcus
(2017). In this work, we analyzed data for alkali, alkaline earth, rare earth, transition
metals and metals from 12 to 14 groups of 60 metals in total.

3 Results and Discussion

The relationship of different types of the binding energy is presented in Table 1.

The numbers without brackets show the ratio of the energy quantity in the column
of the table to the energy quantity in the row, which are averages for all 60 metals.
The numbers in brackets show the adjusted coefficient of determination R2 adj , which
characterizes the linear relationship between the energy quantities.
The sublimation enthalpy ΔH sub from solid at temperature 0 K coincides with the
cohesive energy E coh at 0 K. The E coh /E vac ratio is close to 3 in accordance with early
results (Gorecki 1974; Starodubtsev et al. 2020). The E coh , ΔH b and RT b quantities,
which characterize the decomposition of substance to the monatomic gas, have the
highest correlation. The ratio ΔH b /RT b corresponds to the Trouton’s rule, which
indicates a significant and almost identical change in entropy ΔS b = ΔH b /T b = 12R

Table 1 Ratios of energy quantities and the coefficients of determination R2 adj (in brackets)
E coh E vac RT b ΔH b RT m ΔH m
E coh 1 3.1 13.9 1.13 31.4 29.9
(1) (0.77) (0.96) (0.987) (0.86) (0.83)
E vac 0.32 1 5.1 0.4 10.1 9.5
(0.77) (1) (0.73) (0.84) (0.84) (0.85)
RT b 0.072 0.196 1 0.083 2.31 2.21
(0.96) (0.73) (1) (0.95) (0.80) (0.65)
Hb 0.89 2.51 12.0 1 28.7 27.1
(0.987) (0.84) (0.95) (1) (0.85) (0.86)
RT m 0.032 0.1 0.43 0.035 1 0.98
(0.86) (0.84) (0.80) (0.85) (1) (0.91)
Hm 0.035 0.11 0.45 0.037 1.02 1
(0.83) (0.85) (0.77) (0.86) (0.91) (1)
Relation Between Binging Energy, Isothermal Bulk Modulus, … 213

Fig. 1 Relationship between

the cohesive energy E coh and
the 13.9 × RT b thermal
energy parameter at
vaporization point. The
dashed line corresponds to
the equality E coh = 13.9 ×
RT b

when a liquid passes into a gas. On the contrary, during melting, the entropy change
is insignificant ΔS m = ΔH m /T m = R (Richard’s rule).
Figure 1 shows the relationship between the cohesive energy E coh and the 13.9
× RT b thermal energy parameter at vaporization point. It can be seen that all metals
can be divided into two groups: transition metals with a high boiling point and with
a faster increase in cohesion energy and all other metals with a lower boiling point.
In (Starodubtsev and Tsepelev 2021b), the kinetic energy of atomic vibrations ek
(J)

1
ek = mω2D (αa)2 , (1)
2

where ωD is the Debye frequency (s−1 )

2π kB
ωD = Θ D , (2)
h
ΘD is the Debye temperature (K), h is the Planck’s constant (J·s), a is the atomic
diameter (m), m is the atomic mass (kg), α is the relative atomic vibration amplitude,
αa is the absolute vibration amplitude, was compared with thermal energy eTm at
melting point

eTm = kB Tm . (3)

From the comparison of (1) and (3) we obtain the relative amplitude of atomic
vibrations at the melting point
/
h Tm
αm = . (4)
πaΘD 2kB m
214 Yu. N. Starodubtsev and V. S. Tsepelev

For liquid metals at the melting point, the average αm = 0.053, and the distance
between the centers of neighboring atoms increases by 0.106a. This estimate
coincides with Lindemann’s melting criterion (Lindemann 1910).
Relative amplitude of atomic vibrations at boiling point
/
Tb
αb = αm . (5)
Tm

For liquid metals at the boiling point, the average αb = 0.079, and the distance
between the centers of neighboring atoms increased by 0.158a. Consequently, an
increase in the distance between atoms by an average of 50% compared to the melting
temperature leads to a complete break in interatomic bonds and decomposition of
substance to the monatomic gas.
Dimensional analysis allows us to represent the relationship between the bulk
modulus, surface tension and the binding energy as

E
K Tm ∝ , (6)
a 3 NA
E
γm ∝ (7)
a 2 NA

where E is the molar binding energy (J·mol−3 ), N A is Avogadro constant (mol−1 ). An

analysis of relationships (6) and (7) for different molar binding energies showed that
the bulk modulus and the surface tension at melting point correlate better with the
quantities characterizing the complete breaking of interatomic bonds, i.e., with E coh ,
ΔH b and RT b , rather than with the quantities ΔH m and RT m , which characterize
the metal state at the melting point. In the future, we will investigate the relationship
between the bulk modulus, surface tension and thermal energy at boiling point RT b ,
since the boiling point can be determined most accurately by experimental methods.
Figure 2 shows the relationship between the bulk modulus K Tm and the volume
density of thermal energy at boiling point k B T b a–3 multiplied by a factor of 23.
This factor is equal to the arithmetic mean for the ratio of K Tm to k B T m a–3 for all
metals. A similar relationship for the surface tension at the melting point γm and the
surface density of thermal energy k B T b a–2 is shown in Fig. 3. Figures show a good
correlation between the elastic energy density at the melting point and the thermal
energy density at the boiling point, R2 adj = 0.86 and 0.93, respectively. A particularly
high correlation is observed for the surface, which atoms can leave at a temperature
below the boiling point.
According to (8) and (9), the bulk modulus and surface tension depend on both the
atomic size and the boiling point. To evaluate the effect of atomic size, we introduce
the reduced quantities, which are related to the unit of thermal energy at the boiling
point and are equal to the reciprocal atomic size with exponents 3 and 2, respectively.
Relation Between Binging Energy, Isothermal Bulk Modulus, … 215

Fig. 2 Relationship between

the bulk modulus K Tm and
23 × k B T b a−3 parameter.
The dashed line corresponds
to the equality K Tm = 23 ×
k B T b a−3

Fig. 3 Relationship between

the surface tension γm and 2
× k B T b a−2 parameter. The
dashed line corresponds to
the equality γm = 2 ×
k B T b a−2

K Tm
= a −3 , (8)
23kB Tb
γm
= a −2 , (9)
2kB Tb

Figure 4 shows the relationship of reduced bulk modulus K Tm × (23 × k B T b )−1

with atomic size parameter a−3 , and Fig. 5 the relationship of reduced surface tension
γm × (2 × k B T b )−1 with atomic size parameter a−2 . It can be seen that the reduced
quantities increase with decreasing the atomic size a.
Metals Zn, Cd, Hg from group 12, as well as Be in Fig. 5 show the anomalous
deviation from the regression line. If these metals are excluded, then the coefficient
of determination increases to 0.90. Figure 5 also shows that the slope of the reduced
surface tension γm × (2 × k B T b )−1 versus a−2 is steeper for transition metals, which
216 Yu. N. Starodubtsev and V. S. Tsepelev

Fig. 4 Relationship between

the reduced bulk modulus
K Tm × (23 × k B T b )−1 and
atomic size parameter a−3 .
The dashed line corresponds
to the equality K Tm × (23 ×
k B T b )−1 = a−3

Fig. 5 Relationship between

the reduced surface tension
γm × (2 × k B T b )−1 and
atomic size parameter a−2 .
The dashed line corresponds
to the equality γm × (2 ×
k B T b )−1 = a−2

have the smallest atomic size. Similar patterns, but in a more diffuse form, take place
for the reduced bulk modulus.
The ratio of γm to aK Tm compares the energy of atoms on the surface and in the
bulk and the average ratio for all metals is 0.089. Thus, the energy of atoms in the
surface layer of liquid metals at the melting point is approximately 9% of the atomic
energy inside the liquid. Figure 6 shows the relationship between the surface tension
γm and the product of the atomic size by the bulk modulus aK Tm .

4 Conclusion

In this work, we investigated the relationship of various energy quantities charac-

terizing the interatomic bond with the isothermal bulk modulus and surface tension
Relation Between Binging Energy, Isothermal Bulk Modulus, … 217

Fig. 6 Relationship between

the surface tension γm and
the 0.089 × aK Tm
parameter. The dashed line
corresponds to the equality
γm = 0.089 × (aK Tm )

of 60 liquid metals at the melting temperature. It is shown that the isothermal bulk
modulus and surface tension at melting point correlate better with the values charac-
terizing the complete breaking of interatomic bonds, i.e., with the cohesive energy
E coh , evaporation enthalpy ΔH b , and thermal energy at the boiling point RT b . With
the melting enthalpy ΔH m and thermal energy at the melting point RT m , which
characterize the state of the metal at the melting point, this relationship is weaker.
Transition metals are distinguished by a faster increase in the cohesive energy with
increasing the boiling point. Decomposition of liquid metals to the monatomic gas
occurs when the distance between atoms increases on average by 50% compared to
the melting point. The isothermal bulk modulus and surface tension at melting point
increase with decreasing atomic size. It is shown that the average energy of atoms in
the surface layer of liquid metals at the melting temperature is approximately 9% of
the energy in the atomic layer inside the liquid.

Acknowledgements The article was made within the framework of state work No. FEUZ-0836-
0020.

References

Aqra, F.: The cohesive energy density and the isothermal compressibility: their relationships with
the surface energy. Phys. B 446, 28–31 (2014)
Blairs, S.: Review of data for velocity of sound in pure liquid metals and metalloids. Int. Mater.
Rev. 52, 321–344 (2007)
Friedel, J.: Dislocations. Pergamon Press, Oxford (1964)
Garvin, D.: Thermodynamic properties of elements. Bull. Alloy Phase Diagr. 2, 261–262 (1981)
Gorecki, T.: Vacancies and changes of physical properties of metals at the melting point. Z.
Metallkunde 65, 426–431 (1974)
Iida, T., Guthrie, R.: Performance of a modified Schytil model for the surface tension of liquid
metallic elements at their melting point temperatures. Metall. Mater. Trans. B 41, 437–447 (2010)
218 Yu. N. Starodubtsev and V. S. Tsepelev

Iida, T., Guthrie, R.I.L.: The thermophysical properties of metallic liquids. University Press, Oxford
(2015)
Israelachvili, J.N.: Intermolecular and surface forces. Elsevier, Amsterdam (2011)
Kaptay, G.: Link between the semi-empirical Andrade and Schytil equations and the statistical-
mechanical Born-Green equation for viscosity and surface tension of pure liquid metals. Metall.
Mater. Trans. B 39, 387–389 (2008)
Keene, B.J.: Review of data for the surface tension of pure metals. Int. Mater. Rev. 38, 157–192
(1993)
Lindemann, F.A.: Über die Berechnung molekularer Eigenfrequenzen. Phys. z. 11, 609–612 (1910)
Marcus, Y.: On the compressibility of liquid metals. J. Chem. Thermodyn 109, 11–15 (2017)
Starodubtsev, Yu.N., Tsepelev, V.S., Wu, K.M., Kochetkova, Ye.A., Tsepeleva, N.P.: Vacancy
formation energy of metals. Key Eng. Mater. 861, 46–51 (2020)
Starodubtsev, Yu.N., Tsepelev, V.S.: Vacancy formation energy and kinetic properties of liquid
metals. Key Eng. Mater. 880, 43–48 (2021a)
Starodubtsev, Yu.N., Tsepelev, V.S.: Analysis of the kinematic viscosity and self-diffusion of liquid
metals at the melting temperature. High Temp. 59, 192–197 (2021b)
Stewart, G.R.: Measurement of low temperature specific heat. Rev. Sci. Instrum. 54, 1–11 (1983)
Turkdogan, E.T.: Thermodynamic and kinetic properties of elements and inorganic compounds
relating to the their melting and boiling point temperatures. Can. Metall. Q. 41, 151–162 (2002)
Corrosion Protection Method
with Sacrificial Anode for Steel Members
in Boundary with Ground Using Water
Swelling Rubber

J. Xu, S. Kainuma, K. Takaki, M. Yang, and T. Tomoda

1 Introduction

Many anti-corrosion methods have been applied in steel structures, such as coating
and sacrificial anodes cathodic protection (SACP), but they are not continuously
reliable in practice. The deterioration of the coating film resulting from hydrolysis
and oxidative decomposition is common (Yang et al. 2019). Besides, the quality of
the coating cannot be sufficiently secured in edge corners and imperfect areas of
connection details.
For some steel structures, corrosion is prone to occur, where the coating is vulner-
able to deterioration (Kere and Huang 2019; Jeong et al. 2019). In addition, dew and
stagnant water with chloride from antifreeze at the corner of the structure contribute
to corrosion of steel members, especially (Cramer et al. 2002). And the localized and
uniform corrosion damage of steel members in boundary with ground in an actual
bridge is shown in Fig. 1a. Apart from the coating, SACP is the most broadly utilized
strategy to diminish corrosion. However, an electrolyte is necessary, so that SACP
is preferred for protection for steel structures buried in soil or submerged in water
(Tezdogan and Demirel 2014). Its application in a dry environment is still raw and
problems exist (Suzumura and Nakamura 2004; Yang et al. 2020).
Recently, S. Kainuma fixed sacrificial anode and fiber sheet with high water
absorption and retention properties together, to achieve the anti-corrosion effect for
steel plates (Yang et al. 2019, 2020). However, the fiber sheet is difficult to fit with
steel members at the corner of the structure tightly over a long period of time. In this
study, instead of fiber sheet, water swelling rubber with higher elastic modulus was

J. Xu · S. Kainuma (B) · K. Takaki · M. Yang

Department of Civil Engineering, Faculty of Engineering, Kyushu University, Fukuoka 8190395,
Japan
e-mail: [email protected]
T. Tomoda
Transportation Division, Nippon Koei Co.,Ltd, Tokyo, Japan

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 219
R. K Agarwal (ed.), Recent Advances in Manufacturing Engineering and Processes,
Lecture Notes in Mechanical Engineering,
https://doi.org/10.1007/978-981-19-6841-9_22
220 J. Xu et al.

(a) Corrosion of steel members in (b) A SACP method using water swelling
boundary with ground rubber

Fig. 1 A SACP method with water swelling rubber applied in boundary with ground

applied in the SACP system, as shown in Fig. 1b. The water swelling rubber was
expected to act as an electrolyte carrier with high absorption and retention capacity
to prevent corrosion under different environmental conditions.

2 Experiment

2.1 CP System

The CP system applied in this study includes a Zn plate and water swelling rubber,
which are fixed together and attached to a surface-cleaned steel plate by a peek bolt
and an acrylic plate. The steel plate, acting as cathodic material, directly contacts
with the rubber and is connected with the Zn plate acting as a sacrificial anode, by
lead wires.
The water absorption mechanism of water swelling rubber is shown in Fig. 2.
The water-absorbent polymer in the rubber has a network structure, and when the
cations in the polymer are dissolved in water, osmotic pressure is generated, so that
the rubber can absorb a large amount of water. In the system, the rubber works as an
electrolyte carrier to establish the ion flow, so that the effectiveness of the CP system
is strongly associated with the absorption and retention properties of the rubber.

2.2 Water Absorption and Retention Test

Immersion and drying tests were carried out to evaluate the water absorption and
retention properties of the water swelling rubber. The appearance, enlarged view,
and properties of the rubber (40 × 40 × 6 mm) are presented in Fig. 3 and Table 1,
Corrosion Protection Method with Sacrificial Anode for Steel Members … 221

Fig. 2 Mechanism of superabsorbent polymer

respectively. In the test, the rubber was submersed in distilled water and 0.1, 3.5, and
7.0 mass% NaCl aq electrolytes, respectively. The water retention test was carried
out by allowing the rubber swollen in the water absorption test to be exposed in the
air, and the changes of weight were measured in the same manner as in the water
absorption test. The test conditions for the test are shown in Table 2.

(a) Appearance (b) Surface (×700)

(c) Cross section (×700)

Fig. 3 Appearance and enlarged view of the water swelling rubber

222 J. Xu et al.

Table 1 Properties of water swelling rubber

Density Modulus of Hardness H* Tensile Stretching Volume
(kg/dm3 ) elasticity strength HB (%) change rate
E(MPa) TB (MPa)
1.2 0.65 A40±5 4.0+ 500+ 200+
*JIS K 5253

Table 2 Condition of the water absorption and retention test

Specimen Steeping environment Steeping period Drying period Initial weight (g)
A Distilled water 60 days 60 days 12.09
B 3.5mass%NaCl aq 12.92
C 0.1mass%NaCl aq 30 days – 13.09
D 7.0mass%NaCl aq 12.76

2.3 Electrochemical Test

An electrochemical test was carried out to evaluate the anti-corrosion performance

of the CP system, as shown in Fig. 4. A steel plate without a Zn plate was set as a
controlled experiment. To simulate the environment of stagnant water and dryness at
the steel members in boundary with ground, the tests were conducted in the order of
(1) to (3) below. The test piece was placed in a chamber maintained at a temperature
of 18° C and a relative humidity of 75%.

(1) moist curing (0–120 h)

The measurement was performed by allowing the test piece to stand in a constant
temperature and humidity chamber. The measurement period was 120 h.
(2) immersion test (0–120 h)
The measurement was performed by immersing the rubber in 3.5 mass% NaCl aq.
In addition, one of the rubbers swollen with 1.0 mass% NaNO2 aq was immersed
in the immersion environment of 0.1 mass% NaCl aq. The measurement period was
120 h.
(3) moist curing (0–120 h)
The specimen submerged in (2) was taken out and conducted in the same environment
as in (1). The measurement period was also120 hrs.
The anti-corrosion current generated between the anode plate and the steel plate,
an external wire was measured and recorded by a nonresistance ammeter. Table
3 shows the electrolytes soaked in the rubber, absorption and drying periods and
conditions of immersion environment in the test.
Corrosion Protection Method with Sacrificial Anode for Steel Members … 223

(a) Three-dimensional view

(b) Side view

Fig. 4 Configuration and dimensions of the specimens (unit: mm)

Table 3 The test condition

Specimens Electrolytes Absorption Drying Immersion Sacrificial
Electrolyte Concentration period (day) period environment anode
(mass%) (day) material

Dew1.0 NaNO2 1.0 30 – 3.5mass% Zn plate

Dew5.0 5.0 NaCl aq
Dry1.0 1.0 60 40
Dry5.0 5.0
Dew1.0–0.1 1.0 30 – 0.1mass%
NaCl aq
Dew1.0-X 1.0 30 – 3.5mass% –
NaCl aq
224 J. Xu et al.

3 Results and Discussion

3.1 Water Absorption and Retention Properties of the Rubber

The time-dependent weight of the rubber in the water absorption and retention test
is shown in Fig. 5. The maximum weight of rubber immersed in distilled water is
reached about the 40th day, which increases by about 20 g. However, the weight
increase of rubber soaked in 3.5 mass% NaCl aq reaches about the 50th day, and the
weight increases by about 9 g. In distilled water, the cations of the rubber are more
easily dissolved than NaCl aq due to a larger osmotic pressure between the water
inside the polymer and the surrounding water. For NaCl aq, the water absorption of
rubber increases as the concentration decreases. Since a larger osmotic pressure is
generated in 0.1mass% NaCl aq compared to 3.5 and 7.0 mass% NaCl aq.
The weight of rubber immersed in both the distilled water and 3.5 mass% NaCl
aq decreases significantly until about 30 days after the start of drying. However, the
weight loss rate of the rubber absorbed with NaCl aq decreases after 30 days, which
takes longer to dry than distilled water. It is considered that Na+ is less likely to
dissociate and the drying rate is lower than that of distilled water due to the affinity
of the superabsorbent polymer for Na+ . Above all, it is easy for the rubber to absorb
the solution with a low ion concentration and retain electrolyte with ion.
In this method, the absorption and retention performance of the rubber, acting as
the role to provide the electrolyte, is remarkably affected by the concentration of the
immersion solution. If the concentration of the electrolyte that the rubber absorbs is
too high or too low, either performance can be significantly reduced. Considering
the balance between water absorption and retention performance, it is necessary to
select an electrolytic solution with an appropriate concentration.

30
Weight of the water−swelling rubber m / g

30
Weight of the water−swelling rubber m / g

Distilled water
3.5 mass%NaCl

25 25

20 20

Distilled water
15 0.1 mass%NaCl 15
3.5 mass%NaCl
7.0 mass%NaCl

10 10
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Time t / day Time t / day
(a) Absorbing test (b) Drying test

Fig. 5 The time-dependence of weight for the water swelling rubber

Corrosion Protection Method with Sacrificial Anode for Steel Members … 225

6
10
Dew1.0
Dew5.0
Corossion current I / µA Dry1.0
Dry5.0
4 Dew1.0−0.1
10

2
10

Moist curing Immersion test Moist curing

0
10
0 30 60 90 120 150 180 210 240 270 300 330 360
Time t / hr.

Fig. 6 The time-dependence of corrosion current for specimens

3.2 Electrochemical Test Result

The time-dependent variations of the anti-corrosion current according to environ-

mental changes are shown in Fig. 6. In Dry1.0, poor contact of the conductor occurred
near 280 h., after that data was lost. The current I of Dew1.0, Dew5.0, and Dew1.0–
0.1 have changed at the same level during 0–120 h. The current I of Dry1.0 and
Dry5.0 with dried rubber have been measured from the start of the test, and the sacri-
ficial anode anti-corrosion effect is exhibited because of a small amount of water in
the rubber.
In the immersion environment (120–240 h), the current I of the specimen
immersed in 3.5 mass% NaCl aq is larger than that in Dew 1.0-0.1 immersed in 0.1
mass% NaCl aq, because the corrosion circuit mainly depends on the surrounding
immersion environment instead of the electrolyte inside the rubber.
The current I in the second moist curing (240–360 h.) Shows the same tendency as
the first time. The current I in specimens with the swollen rubber is large, compared to
that in Dry 5.0 with dried rubber. Furthermore, the current I of the specimen remains
stable even when the environment changed from a stagnant water environment to a
dry environment, resulting from the high water retention performance of the rubber.
The corrosion conditions of steel surface in contact with the rubber after the test
are presented in Fig. 7. There is no visible corrosion on the surface of specimens (a–e)
and most surface in contact with rubber of specimen (f), which should be attributed
to the application of the rubber leading to the dissolved oxygen reduction. The black
part at the lower end of the contact surface in Fig. 8f is a rubber deposit, the enlarged
view of which is presented in Fig. 8. Corrosion can be observed in the red frame,
which results from a gap generated between the rubber and the steel plate due to the
shrink of the rubber in moist curing. However, for specimens applied with Zn plate,
corrosion does not occur, because the significant effect of SACP can be achieved,
despite the occurrence of the void.
226 J. Xu et al.

(a) Dew1.0 (b) Dew5.0 (c) Dry1.0

(d) Dry5.0 (e) Dew1.0-0.1 (f) Dew1.0-X

Fig. 7 Corrosion protection situation of the steel surface after the test

Fig. 8 Enlarged view of

specimen Dew1.0-X. (×500)

4 Conclusion

In this study, an anti-corrosion method with water swelling rubber for steel members
in boundary with ground was developed and the effectiveness of that was verified.
Corrosion Protection Method with Sacrificial Anode for Steel Members … 227

Water absorption and retention test and electrochemical test were carried out to verify
the effectiveness of the water swelling rubber.
(1) The amount of water absorbed by the rubber increases as the ion concen-
tration of the electrolyte decreases, and the rubber takes more time to dry
with the electrolyte concentration inside rubber increasing. An appropriate
electrolyte concentration should take both water absorption and retention into
consideration.
(2) If the rubber is not completely dried, the anti-corrosion circuit of the sacrificial
anode can be maintained through a small amount of water in the rubber.
(3) The effect of the sacrificial anode cathodic protection maintains due to the
high-water retention performance of the water swelling rubber, despite the
environment changing from water immersion to dryness.

References

Cramer, S.D., Covino, B.S., Jr., Bullard, S.J., Holcomb, G.R., Russel, J.H., Nelson, F.J., Laylor,
H.M., Soltesz, S.M.: Cement Concr. Compos. 24, 101–117 (2002)
Jeong, Y.S., Kim, M.J., Jeon, S.H., Ahn, J.H.: J. Korea Inst. Struct. Maint. Insp. 23, 75–83 (2019)
Kere, K.J., Huang, Q.: J. Bridge Eng. 24, 04019007 (2019)
Suzumura, K., Nakamura, S.: J. Mater. Civil Eng. 16, 1–7 (2004)
Tezdogan, T., Demirel, Y.K.: Brodogradnja. 65, 49–59 (2014)
Yang, M., Kainuma, S., Zhuang, S., Ishihara, S.: Int. J. Electrochem. Sci. 14, 9858–9875 (2019)
Yang, M., Kainuma, S., Ishihara, S., Kaneko, A., Yamauchi, T.: Const. Build. Mater. 234, 117405
(2020)