Machine learning for composite materials

Chen, Chun-Teh; Gu, Grace X.

doi:10.1557/mrc.2019.32

Cited by 236 publications

(122 citation statements)

References 79 publications

(99 reference statements)

Supporting

Mentioning

102

Contrasting

Unclassified

Order By: Relevance

“…Nevertheless, it is evident that the design of these systems is highly complex. Optimization techniques [40][41][42] such as machine learning can be a useful arsenal to accomplish the objective to find the best design of the desired behavior. In addition, these effects may be incorporated even at smaller scales, for the design of microscale metamaterials.…”

Section: Discussionmentioning

confidence: 99%

Tailoring the Dynamic Actuation of 3D‐Printed Mechanical Metamaterials through Inherent and Extrinsic Instabilities

Vangelatos

Zhang

et al. 2020

Adv Eng Mater

Self Cite

View full text Add to dashboard Cite

Architected materials have facilitated a significant breakthrough in the control of the mechanical behavior of structures. Most notably, metamaterials with unprecedented properties can be easily designed in large scales, exhibiting high strength, [1] tunability of their properties, [2] resilience or versatility to large deformations, [3,4] and auxetic behavior. [5] In addition, 4D printed structures can modify their shape depending on the working environment. [6] Due to their chemical composition, these structures can modify their mechanical and electrochemical properties based on external stimuli. [7] This mechanism has substantial implications on the mechanical response of the material, furnishing stiffening of the structure or malleability on demand. Furthermore, designing structures that are also adaptive to actuation has given rise to the design of flexible 3D-printed mechanisms. [8] Flexible structures can be utilized for engineering applications requiring large but recoverable deformations, continuous motion, and high precision. A characteristic category of structures encompassing these properties are soft robotics. Soft robotic mechanisms are composed of rubber materials that can be electrically actuated to sustain nonlinear deformations. [9,10] Combining these designs with much stronger fiber components leads to the design of artificial muscles, [11] imitating animal motion [12] and having a significantly higher strength than simple soft robot systems. [13-15] Nevertheless, metamaterials also exhibit remarkable properties that are not observed in conventional systems. A characteristic example is the transport of mechanical signals with the proclivity to a specific direction. This has been accomplished through the design of nonreciprocal mechanical metamaterials. [16] In regular materials, the transmission of any physical quantity, such as mechanical signals or mechanical waves, is identical, regardless of geometrical or material asymmetries and defects between any two points in space. However, nonreciprocal metamaterials can alter this behavior, giving directionality to the transport of these mechanical quantities. This is the paragon of the next-generation mechanical signal transport, [17] isolation, [18] and energy storage. [19] In addition, the combination of the architected design with the macroscopic dynamic loading conditions can provide the formation of distinct elastic vector solitons. [20] These coupled waves can have significantly different formations depending on the direction of the wave, controlling not only the direction of the signal, but also its shape. The inexorable advances in 3D printing technology have enabled the fabrication of these structures. Techniques such as fused deposition modeling (FDM), [21]

show abstract

Section: Discussionmentioning

confidence: 99%

Tailoring the Dynamic Actuation of 3D‐Printed Mechanical Metamaterials through Inherent and Extrinsic Instabilities

Vangelatos

Zhang

et al. 2020

Adv Eng Mater

Self Cite

View full text Add to dashboard Cite

show abstract

“…As a result, ML models are often considered as surrogate models that greatly accelerate the numerical simulation process at the expense of a small prediction bias. [53][54][55][56][57] Most ML models can be categorized into three groups in terms of their outputs: regression models, which give real number predictions; classification models, which give discrete class predictions; and unsupervised clustering models, which analyze the similarities of data points based on their features. Given the design information as input features, various ML techniques have been utilized to predict resultant mechanical and chemical properties, including strength, stiffness, deformation, toxicity, and stability.…”

Section: Ml-driven Materials Designmentioning

confidence: 99%

Machine Learning for Advanced Additive Manufacturing

Jin

Zhang

Demir

et al. 2020

Matter

Self Cite

164

View full text Add to dashboard Cite

“…It turns out that these imperfections are largely dictated by the settings of printing parameters, first‐layer calibration, and model geometry . With recent advances in applying artificial intelligence and machine learning to materials science and engineering problems, researchers have started using machine‐learning algorithms to classify and predict different printing defections including blob, warp, and delamination based on the settings of printing parameters . Another interesting approach in the field adopts a new slicing mechanism which splits prints into spatially locked bricks to reduce warping .…”

Section: Three Levels Of Nozzle Height For Four Corresponding Categoriesmentioning

confidence: 99%

Automated Real‐Time Detection and Prediction of Interlayer Imperfections in Additive Manufacturing Processes Using Artificial Intelligence

Jin

Zhang

2019

Advanced Intelligent Systems

Self Cite

109

View full text Add to dashboard Cite

Although fused deposition modeling (FDM) additive manufacturing technologies have advanced in the past decade, interlayer imperfections such as delamination and warping are still dominant when printing complex parts. Herein, a selfmonitoring system based on real-time camera images and deep learning algorithms is developed to classify the various extents of delamination in a printed part. In addition, a novel method incorporating strain measurements is established to measure and predict the onset of warping. Results show that the machine-learning model is capable of detecting different levels of delamination conditions, and the strain measurements setup successfully reflects and determines the extent and tendency of warping before it actually occurs in the print job. This multifunctional system can be applied to assess other manufacturing processes to realize autocalibration and prediagnosis of imperfections without human interaction.

show abstract

Machine learning for composite materials

Abstract: Abstract

Cited by 236 publications

References 79 publications

Tailoring the Dynamic Actuation of 3D‐Printed Mechanical Metamaterials through Inherent and Extrinsic Instabilities

Tailoring the Dynamic Actuation of 3D‐Printed Mechanical Metamaterials through Inherent and Extrinsic Instabilities

Machine Learning for Advanced Additive Manufacturing

Automated Real‐Time Detection and Prediction of Interlayer Imperfections in Additive Manufacturing Processes Using Artificial Intelligence

Contact Info

Product

Resources

About