東華大學圖書館 |

Applying Machine Learning Algorithms in the Design and Manufacturing Process of Bioinspired Architectured Materials.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Applying Machine Learning Algorithms in the Design and Manufacturing Process of Bioinspired Architectured Materials./
作者:	Fatehi, Erfan.
面頁冊數:	1 online resource (69 pages)
附註:	Source: Masters Abstracts International, Volume: 84-10.
Contained By:	Masters Abstracts International84-10.
標題:	Mechanical properties. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30347006click for full text (PQDT)
ISBN:	9798377670346

Applying Machine Learning Algorithms in the Design and Manufacturing Process of Bioinspired Architectured Materials.
Fatehi, Erfan.

Applying Machine Learning Algorithms in the Design and Manufacturing Process of Bioinspired Architectured Materials. - 1 online resource (69 pages)

Source: Masters Abstracts International, Volume: 84-10.

Thesis (M.S.)--McGill University (Canada), 2021.

Includes bibliographical references

Precise material architectures and interfaces can generate attractive properties in materials. For example, Topologically Interlocked Materials (TIMs) are architectures that can turn brittle materials like ceramics into tough, deformable and impact-resistant material systems. This strategy has been adapted from biological materials such as nacre and tooth enamel, which have been in nature for millions of years. There are, however, some challenges in the mimicking of the biological materials and the production of TIMs. Designing TIMs for thermomechanical applications require in-depth thermomechanical finite element modelling (FEM). All possible architectures in the vast design space, containing many potential structures and configurations arranged in numerous ways, should be considered that is a computationally expensive process. This problem can be addressed by leveraging Machine learning algorithms that are mathematical models and a branch of artificial intelligence (AI).This work consists of an extensive literature review on (1) bioinspired architectured materials, especially topologically interlocked structures and (2) machine learning techniques and their applications in manufacturing, especially laser cutting of ceramic panels for the production of TIMs and (3) the designing of architectured materials using ML algorithms. Later, in this study, we propose a new approach to design TIMs, using machine learning (ML), trained with finite element modeling (FEM) data and together with a self-learning algorithm, to discover high-performance ceramics in thermomechanical environments. First, a parametric study is conducted over topologically interlocked ceramic panels. A limited number of architectures is subjected to a thermal load and is studied using ANSYS finite element package. Finally, the multilinear perceptron is used to train the machine learning model on FEM data to predict the thermomechanical performance of architectured panels (TIMs) with varied design parameters (e.g. interlocking angle and number of blocks).Overall, the developed machine learning based framework can boost the design algorithm efficiency and provides new designs for various high-temperature applications in aerospace industries. This study demonstrates that architectured ceramic panels with ML-assisted engineered patterns can demonstrate up to 30% improvement in frictional energy dissipation and 7% in the sliding distance of the tiles and 80% reduction in the strain energy that results in a higher safety factor and structural failure delay compared to plain ceramics subjected to a thermal load.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2023

Mode of access: World Wide Web

ISBN: 9798377670346Subjects--Topical Terms:

3549505
Mechanical properties.
Index Terms--Genre/Form:

542853
Electronic books.

Applying Machine Learning Algorithms in the Design and Manufacturing Process of Bioinspired Architectured Materials.
LDR:07024nmm a2200433K 4500 001 2356183
005 20230612071838.5
006 m o d
007 cr mn ---uuuuu
008 241011s2021 xx obm 000 0 eng d
020 $a 9798377670346
035 $a (MiAaPQ)AAI30347006
035 $a (MiAaPQ)McGill_zp38wj79q
035 $a AAI30347006
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Fatehi, Erfan. $3 3696664
245 1 0 $a Applying Machine Learning Algorithms in the Design and Manufacturing Process of Bioinspired Architectured Materials.
264 0 $c 2021
300 $a 1 online resource (69 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Masters Abstracts International, Volume: 84-10.
500 $a Advisor: Akbarzadeh, Abdolhamid.
502 $a Thesis (M.S.)--McGill University (Canada), 2021.
504 $a Includes bibliographical references
520 $a Precise material architectures and interfaces can generate attractive properties in materials. For example, Topologically Interlocked Materials (TIMs) are architectures that can turn brittle materials like ceramics into tough, deformable and impact-resistant material systems. This strategy has been adapted from biological materials such as nacre and tooth enamel, which have been in nature for millions of years. There are, however, some challenges in the mimicking of the biological materials and the production of TIMs. Designing TIMs for thermomechanical applications require in-depth thermomechanical finite element modelling (FEM). All possible architectures in the vast design space, containing many potential structures and configurations arranged in numerous ways, should be considered that is a computationally expensive process. This problem can be addressed by leveraging Machine learning algorithms that are mathematical models and a branch of artificial intelligence (AI).This work consists of an extensive literature review on (1) bioinspired architectured materials, especially topologically interlocked structures and (2) machine learning techniques and their applications in manufacturing, especially laser cutting of ceramic panels for the production of TIMs and (3) the designing of architectured materials using ML algorithms. Later, in this study, we propose a new approach to design TIMs, using machine learning (ML), trained with finite element modeling (FEM) data and together with a self-learning algorithm, to discover high-performance ceramics in thermomechanical environments. First, a parametric study is conducted over topologically interlocked ceramic panels. A limited number of architectures is subjected to a thermal load and is studied using ANSYS finite element package. Finally, the multilinear perceptron is used to train the machine learning model on FEM data to predict the thermomechanical performance of architectured panels (TIMs) with varied design parameters (e.g. interlocking angle and number of blocks).Overall, the developed machine learning based framework can boost the design algorithm efficiency and provides new designs for various high-temperature applications in aerospace industries. This study demonstrates that architectured ceramic panels with ML-assisted engineered patterns can demonstrate up to 30% improvement in frictional energy dissipation and 7% in the sliding distance of the tiles and 80% reduction in the strain energy that results in a higher safety factor and structural failure delay compared to plain ceramics subjected to a thermal load.
520 $a Des architectures et interfaces de materiaux precises peuvent generer des proprietes interessantes dans les materiaux. Par exemple, les materiaux topologiquement imbriquees (TIM) sont des architectures qui peuvent transformer des materiaux fragiles comme la ceramique en systemes solides, deformables et resistants aux chocs. Cette strategie s'inspire de materiaux biologiques tels que la nacre et l'email dentaire qui sont utilises avec succes dans la nature depuis des millions d'annees. L'imitation de ces materiaux biologiques presente cependant de multiples defis : la conception de TIM pour des applications thermomecaniques necessite une analyse par elements finis (FEM) approfondie, et l'exploration de l'espace des configurations possible est couteuse en calcul. Ce probleme peut etre resolu en tirant parti des algorithmes d'apprentissage automatique qui sont des modeles mathematiques et une branche de l'intelligence artificielle (IA).Ce travail consiste donc en une vaste revue de la litterature sur (1) les materiaux architectures bioinspires, et notamment les structures topologiquement imbriquees, (2) les techniques d'apprentissage automatique et leurs applications dans la production, en particulier la decoupe laser de panneaux ceramiques pour la production de TIM ainsi que sur (3) la conception de materiaux architectures a l'aide d'algorithmes ML. Plus tard, dans cette etude, nous proposons une nouvelle approche pour concevoir des TIM avec l'apprentissage automatique (ML), formes avec des donnees de modelisation par elements finis (FEM) et avec un algorithme d'auto-apprentissage. Ceci afin de developper des ceramiques haute performance dans des environnements sous contrainte thermomecaniques. Afin de mener ce projet a bien, une etude parametrique est tout d'abord menee sur des panneaux ceramiques imbriques topologiquement. Un nombre limite d'architectures sont concues, soumises a une charge thermique etudiee a l'aide des methode d'elements finis ANSYS. Enfin, le perceptron multilineaire est utilise pour entrainer le modele d'apprentissage automatique sur les donnees de FEM pour predire les performances thermomecaniques des panneaux architectures (TIM) avec des parametres de conception varies (tel que l'angle de verrouillage et le nombre de blocs).Dans l'ensemble, le cadre base sur l'apprentissage automatique ainsi developpe peut augmenter l'efficacite de l'algorithme de conception et fournir de nouvelles conceptions pour diverses applications a haute temperature dans les industries aerospatiales. Dans cette etude, il est demontre que les panneaux ceramiques architectures a motifs d'ingenierie concu par ML montrent jusqu'a 30% d'amelioration de la dissipation d'energie de frottement et 7% de la distance de glissement des carreaux et une reduction de 80% de l'energie de deformation. Ceci conduisant a un facteur de securite plus eleve et le delai de rupture structurelle par rapport aux ceramiques simples soumises a une contrainte thermique.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2023
538 $a Mode of access: World Wide Web
650 4 $a Mechanical properties. $3 3549505
650 4 $a Mathematical models. $3 522882
650 4 $a Recommender systems. $3 3562220
650 4 $a Crack initiation. $3 3564902
650 4 $a Energy. $3 876794
650 4 $a Ceramics. $3 651796
650 4 $a Composite materials. $3 654082
650 4 $a Crack propagation. $3 3683840
650 4 $a Design optimization. $3 3681984
650 4 $a Lasers. $3 535503
650 4 $a Hydroxyapatite. $3 669750
650 4 $a Neural networks. $3 677449
650 4 $a Flexibility. $3 3560705
650 4 $a 3-D printers. $3 3681750
650 4 $a Fractures. $3 823229
650 4 $a Algorithms. $3 536374
650 4 $a Decision trees. $3 827433
650 4 $a Expected values. $3 3563993
650 4 $a Design. $3 518875
650 4 $a Industrial engineering. $3 526216
650 4 $a Information science. $3 554358
650 4 $a Materials science. $3 543314
650 4 $a Mathematics. $3 515831
650 4 $a Mechanics. $3 525881
650 4 $a Optics. $3 517925
655 7 $a Electronic books. $2 lcsh $3 542853
690 $a 0791
690 $a 0800
690 $a 0389
690 $a 0546
690 $a 0723
690 $a 0794
690 $a 0405
690 $a 0346
690 $a 0752
710 2 $a ProQuest Information and Learning Co. $3 783688
710 2 $a McGill University (Canada). $3 1018122
773 0 $t Masters Abstracts International $g 84-10.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30347006 $z click for full text (PQDT)