Machine learning meets fracture mechanics: prediction of fatigue crack propagation
Department of Mechanical Engineering,
Facility of Science and Engineering,
Abstract This study presents the capability and applicability for fatigue crack propagation evaluation using a neural network system. Fatigue crack propagation includes several governing laws and computation methods which are elastic stress field, stress intensity factors, Paris’ law and criterion of crack propagation direction. Crack propagation phenomenon simultaneously occurs with these laws and methods. Many crack propagation paths are computed as training data in 2-dimension using s-version finite element method. Three training levels are examined. In 1st training level, crack position vector, propagation direction vector and stress intensity factors are learned. In 2nd training level, crack position vector, crack propagation direction vector, six stress components around crack tip are learned. In last training level, only crack position vector and crack propagation direction vector are trained. Last level requires only geometrical configuration to predict crack propagation. In order to simplify the problem, all of parameters are completely determined as constant values. The simulation can represent curved crack path by incremental crack propagation computations. As a result of comparisons between the each level, the 1st level exhibits the highest reproducibility. If the same computation results as the 1st level, the last level requires 10 times training data and time to get convergence of the neural network system. We’d like to discuss the results of several learning examples and show the applicability of the neural network technology to the actual engineering problem with high accuracy.
Biodata Professor Yoshitaka Wada received the B.Eng. and the M. Eng. in Mechanical Engineering from Tokyo University of Science in 1993 and in 1995 respectively, and Ph. D. in Quantum Engineering and Systems Science from University of Tokyo, Japan, in 1998. He worked at the University of Tokyo (1997-2000), the Research Institute of Science and Technology (2000-2002) and the Tokyo University of Science, Suwa (2002-2012) before joining Kindai University.
Notably, he was involved in the development of an earthquake simulation system for Earth Simulator, which was the world’s fastest supercomputer at that time and now he works at the research committee in development of notable software system for exascale (1018 floating point operations per second) supercomputer which will be next generation world’s fastest supercomputer in the future. In the fracture mechanics field, he developed a fully automatic crack propagation system using a reliable modeling technique and the s-version finite element method. The system was applied to many engineering crack propagation problems and evaluated appropriateness of several maintenance codes and standards for nuclear power plants. His research work is in computational mechanics and fracture mechanics where he specifically focuses on the development of a simulation system using automatic mesh generation and application of machine learning to practical engineering problems.
Fuel economy standards and labels: The way towards an energy efficient automotive transportation
T M Indra Mahlia
University of Technology Sydney,
Abstract The road transport and particularly the passenger cars are responsible for increasing the share of transport energy consumption and harmful emissions level growth. Automobiles are considered as the main energy consumer in the transportation sector. In order to reduce energy consumption in this sector, the policymaker should consider implementing fuel economy standards and labels for motor vehicles. The first step towards developing fuel economy standards and labeling is to create a test procedure for the vehicles. The test procedure is a well-defined protocol of the laboratory test to evaluate the energy performance of the vehicles. The test procedure is the technical foundation for all related programs namely; fuel economy standards, fuel economy labels, and incentive programs. This study developed a critical review on fuel economy testing procedure, fuel economy standards, fuel economy labels and incentive programs around the world and to propose the program for policymakers in Malaysia. Hopefully, the program can be implemented as soon as possible on the way towards an energy efficient automotive transportation and as a preparation to be a developed nation.
Biodata Professor Mahlia obtained his Ph.D. from the Department of Mechanical Engineering, University of Malaya. He was appointed as a Professor in 2009 at the same department. At the moment, he is a Distinguished Professor at the University of Technology Sydney, Adjunct Professor at the University of Indonesia and World Class Visiting Professor Programme under the Ministry of Research, Technology and Higher Education, Republic of Indonesia. His publications include more than 350 journal articles, proceedings and also research reports, which 210 are WOS and more than 240 are Scopus indexed. His research interests include Energy Systems, Energy Engineering, Energy and Environment, Energy Storage, Bioenergy and Energy Policy. He also involved as a researcher in 5 $AUD Millions project funded by various sources. He is one of Highly Cited Researcher for 2017 and 2018 by Clarivate Analytics.
Severe plastic deformation under high pressure for producing enhanced mechanical and functional properties
Department of Materials Science and Engineering, Faculty of Engineering, Kyushu University, Fukuoka, Japan
WPI, International Institute for Carbon-Neutral Energy Research (WPI-I2CNER), Kyushu University, Fukuoka, Japan
Abstract Processing through severe plastic deformation (SPD) is effective for grain refinement to the submicrometer and/or nanometer range in bulk metallic materials. When the SPD process is performed under high pressure as high-pressure torsion (HPT) and high-pressure sliding (HPS), its applicability is further promoted: (1) second phase particles in the metal matrix are fragmented to a fine dispersion of nanosized particles; (2) dissolution of the second phase particles may occur; (3) consolidation of powders at lower temperatures and thus alloying is attained through solid-state reaction; (4) fabrication of metal-matrix composites is feasible without sintering process; (5) nanostructure control is achieved through subsequent combination with annealing or aging; (6) pressure- and/or strain-induced phase transformation occurs. With such peculiar features, it is possible to enhance not only mechanical properties such as strength and ductility but also functionality of materials such as hydrogen storage capability, electrical conductivity, superconductivity, and many others.
Biodata Professor Zenji Horita is a Professor in the Department of Materials Science and Engineering in Kyushu University. He is the Director of International Research Center on Giant Straining for Advanced Materials. He is also one of the Principal Investigator in the International Institute of Carbon-Neutral Energy Research. Professor Horita is among the world key researchers in the Severe Plastic Deformation research area. He was awarded The Medal of Honor from the Emperor of Japan (Purple Ribbon) for his research accomplishments.
Towards Environmentally-Friendly Friction Brakes
School of Mechanical Engineering,
University of Leeds, UK
Abstract All road vehicles, including electric and hybrid electric vehicles, will be required to carry friction brakes for the foreseeable future. However there are major environmental issues with conventional friction brakes which will be of increasing concern as automotive engineers struggle to meet stringent new targets on emissions. This presentation focusses on 3 important environmental issues connected with friction brakes: 1) reduction of the weight of the current cast iron brake rotor by the use of composite or coated light alloy rotors; 2) mitigation of nuisance noise levels from brakes using an automated squeal reduction strategy; 3) minimization of particulate wear debris emissions from friction brakes. In each case, a fundamental approach is taken including the use of both computational analysis and experimental test. Examples from the presenter’s own recent research background will be used to demonstrate how friction brakes can be made more environmentally-friendly and therefore more acceptable for future vehicles.
Biodata Professor David Barton works at the School of Mechanical Engineering, University of Leeds. He received his BSc in Mechanical Engineering from Bristol University in 1974, MSc in Applied Mechanics from Manchester University in 1978, and his PhD also from Manchester University in 1981. His research interest includes high strain rate response of materials and structures, tribological interfaces in automotive and biomedical engineering, and properties and processing of polymers and composites. To date, he has supervised over 40 PhD students and 190 papers published in journal and at conferences.
Malaysian automotive industry in the era of IR4.0: Opportunities, challenges and the way forward
Malaysia Automotive, Robotics and IoT Institute (MARii)
Biodata Dato’ Madani Sahari is the CEO of the Malaysia Automotive Robotics and IoT Institute (MARii), previously known as the Malaysia Automotive Institute (MAI). MARii is the focal point, coordinating centre and think tank towards enhancing the competitiveness of the automotive industry and overall mobility, through the adoption of Robotics and the Internet of Things (IoT), under the custodian of the Ministry of International Trade and Industry (MITI). He has played an active role in shaping the landscape of the automotive industry through his involvement in the formulation of the National Automotive Policy and under his leadership, MARii has expanded its scope to cater for the ever increasing demands for Robotics and IoT adoption in Malaysia.
Since its establishment in 2010, the Malaysia Automotive Institute (now MARii) has developed numerous programmes and centres and excellence, equipped with the facilities to assist the industry, academia and relevant stakeholders. The existing facilities, programmes and experience within the MARii network will allow immediate access for industry players to develop their capabilities in Advanced Manufacturing and Advanced IT towards Industry 4.0 compliance.
He graduated from the University of Lorraine (formerly Nancy-Université), France, with a degree in Industrial Technology and a Masters in Quality Engineering, in addition to being a Certified Quality Engineer. Dato’ Madani carries with him 25 years as an executive manager with national and international automotive manufacturers in the areas of strategic collaboration, project development and manufacturing.
Modeling Damage and Fracture Processes for Reliability Aspect in Engineering Design
Mohd N Tamin
School of Mechanical Engineering,
Universiti Teknologi Malaysia, Malaysia
Abstract Engineering design is an iterative decision-making process with a methodical series of steps to ensure the reliability of the newly-designed component. In this respect, this paper describes the analysis step for establishing adequate reliability aspects of the component. The design process is illustrated by considering a newly-designed steel wire ropes for offshore applications. The dominant failure mode is dependent on the lay angle of the wire ropes, the local Hertzian contact condition (stick-slip) and the relative magnitude of the sliding displacements of the wires. In this deterministic approach to the reliability analysis, Lemaitre’s two-scale damage model is modified to calculate the fretting fatigue damage, while the energy-based wear model is employed to quantify the fretting wear damage accumulation of the drawn steel wires in the wire rope. In these models, the material damage is manifested through the degradation of the Young’s modulus of the wires, and the loss of material at the fretting contact surfaces. A series of interrupted fatigue tests of the drawn wire samples provide the test data for establishing the residual Young’s modulus. A fatigue damage calculation routine based on varying load cycle blocks is developed to efficiently calculate the evolution of the damage to separation of the critical material points of the drawn wires in the wire rope. This incremental damage calculation routine is incorporated into the commercially available FE analysis software to establish force equilibrium of the wire rope model following each increment of the damage. The damage-based fatigue model is validated using fatigue life data of a single strand (1×7) steel wire ropes. The model is examined through the prediction of the mechanics of deformation and failure processes of stranded spiral steel wire ropes.
Biodata Prof. Tamin earned his doctoral degree in Mechanical Engineering and Applied Mechanics from the University of Rhode Island, USA in 1997. He has been with the Faculty of Mechanical Engineering, Universiti Teknologi Malaysia since 1984. He had his fair share of administrative positions as the Head of the Applied Mechanics Department, Head of the Materials Engineering Department, Director of UTM Center for Composites, and the Deputy Dean (Research and Innovation) of the faculty. He is currently leading his research team with 11 doctoral candidates, a post-doctoral researcher, and a project manager at the Computational Solid Mechanics Laboratory (CSMLab) UTM, which he founded in 2006.
Prof. Tamin leads his research team on few successful research collaborations with industries. These include Intel Technology on the development of a validated methodology for reliability prediction of solder joints in BGA packages, and failure process modeling of TSV interconnects in microelectronic packages; with Kiswire (Korea) for fatigue life improvement of steel wire ropes; with Airbus (France) and Aerospace Malaysia Innovation Center (AMIC) for damage detection in FRP composite laminates using the digital image correlation (DIC) technique; and with Flextronics Semiconductor in promoting Surface Mount Technology through university academic curriculum. He has also served as an external examiner of the academic programs at several local universities. Prof. Tamin has been invited as a visiting researcher at Sophia University, Tokyo (Japan), a visiting professor at the Institut Supérieurde l’Automobileet des Transport, Nevers (France) and Dongguk University, Seoul (Korea). He is currently a visiting research professor at the University of Southampton (Malaysia Campus). Prof. Tamin is keen in promoting the university-industry collaboration.