A multi-objective optimization framework for aerosol jet customized line width printing via small data set and prediction uncertainty

Zhang, Haining; Choi, Joon Phil; Moon, Seung Ki; Ngo, Teck Hui

doi:10.1016/j.jmatprotec.2020.116779

Cited by 17 publications

(7 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Later, many researchers explored the underlying causal aerodynamic interactions that led to trends in line morphology using computational fluid dynamic models and achieved a better understanding of aerosol droplet generation, transportation, and impaction [ 104 , 111 , 112 , 113 , 114 , 115 , 116 ]. Additionally, data-driven process modeling can eliminate the drifting and stochastic nature of the print system and efficiently determine the optimal operating process window, which is attractive for controlling the line morphology [ 117 , 118 , 119 ].…”

Section: Multi-materials 3d Printing Methodsmentioning

confidence: 99%

Recent Advances in Multi-Material 3D Printing of Functional Ceramic Devices

et al. 2022

View full text Add to dashboard Cite

In recent years, functional ceramic devices have become smaller, thinner, more refined, and highly integrated, which makes it difficult to realize their rapid prototyping and low-cost manufacturing using traditional processing. As an emerging technology, multi-material 3D printing offers increased complexity and greater freedom in the design of functional ceramic devices because of its unique ability to directly construct arbitrary 3D parts that incorporate multiple material constituents without an intricate process or expensive tools. Here, the latest advances in multi-material 3D printing methods are reviewed, providing a comprehensive study on 3D-printable functional ceramic materials and processes for various functional ceramic devices, including capacitors, multilayer substrates, and microstrip antennas. Furthermore, the key challenges and prospects of multi-material 3D-printed functional ceramic devices are identified, and future directions are discussed.

show abstract

Section: Multi-materials 3d Printing Methodsmentioning

confidence: 99%

Recent Advances in Multi-Material 3D Printing of Functional Ceramic Devices

et al. 2022

View full text Add to dashboard Cite

show abstract

“…For example, Chang 218 and Wang et al 96 developed statistical models to investigate the influence of the main control parameters on the printed line features, and the effectiveness of the proposed models was validated by ANOVA and additional experiments. On the basis of the Gaussian process regression (GPR), Zhang et al 219 investigated the inherent contradictions between line thickness, line edge roughness and a customized line width based on the integration of derived GPR models and NSGA-III. And the optimal process parameters for a specified line width were determined under the dual conflicting objectives of maximizing line thickness and minimizing line edge roughness, and the objective of customizing line width.…”

Section: Ink Printabilitymentioning

confidence: 99%

Reviews on Machine Learning Approaches for Process Optimization in Noncontact Direct Ink Writing

Zhang

Moon

2021

ACS Appl. Mater. Interfaces

Self Cite

View full text Add to dashboard Cite

Recently, machine learning has gained considerable attention in noncontact direct ink writing because of its novel process modeling and optimization techniques. Unlike conventional fabrication approaches, noncontact direct ink writing is an emerging 3D printing technology for directly fabricating low-cost and customized device applications. Despite possessing many advantages, the achieved electrical performance of produced microelectronics is still limited by the printing quality of the noncontact ink writing process. Therefore, there has been increasing interest in the machine learning for process optimization in the noncontact direct ink writing. Compared with traditional approaches, despite machine learning-based strategies having great potential for efficient process optimization, they are still limited to optimize a specific aspect of the printing process in the noncontact direct ink writing. Therefore, a systematic process optimization approach that integrates the advantages of state-of-the-art machine learning techniques is in demand to fully optimize the overall printing quality. In this paper, we systematically discuss the printing principles, key influencing factors, and main limitations of the noncontact direct ink writing technologies based on inkjet printing (IJP) and aerosol jet printing (AJP). The requirements for process optimization of the noncontact direct ink writing are classified into four main aspects. Then, traditional methods and the state-of-the-art machine learning-based strategies adopted in IJP and AJP for process optimization are reviewed and compared with pros and cons. Finally, to further develop a systematic machine learning approach for the process optimization, we highlight the major limitations, challenges, and future directions of the current machine learning applications.

show abstract

“…Although there are strategies to reduce the number of runs, such as fractional factorial, these may result in reduced accuracy [42,43].…”

Section: Building Empirical Databasementioning

confidence: 99%

“…And GPR has been gaining popularity in machine learning due to its superiority for out of sample test performance, and uncertainty characterization and management [14,48,49]. Further, GPR provides predictive variance which allows modelling of predictive uncertainty which can be used for further optimization and decision making [13,43]. The following section discuss in detail the implementation of GPR in the proposed framework.…”

Section: Surrogate Modellingmentioning

confidence: 99%

A Data-Driven Framework to Predict Fused Filament Fabrication Part Properties Using Surrogate Models and Multi-Objective Optimisation

Zhang

Choi

Moon

2021

Preprint

Self Cite

View full text Add to dashboard Cite

In additive manufacturing (AM), due to large number of process parameters and multiple responses of interest, it is hard for AM designers to attain optimal part performance without a systematic approach. In this research, a data-driven framework is proposed to achieve the desired AM part performance and quality by predicting part properties and optimizing AM process parameters effectively and efficiently. The proposed framework encompasses efficient sampling of design space and establishing the initial experiment points. Based on established empirical data, surrogate models, are used to characterise influence of critical process parameters on responses on interest. Further, process maps can be generated for enhancing understanding on the influence of process parameters on responses of interests and AM process characteristics. Subsequently, multi-objective optimisation coupled with a multi criteria decision making technique is applied to determine an optimal design point, which maximises the identified responses of interest to meet the part functional requirements. A case study is used to validate the proposed framework for optimising an ULTEM™ 9085 fused filament fabrication part to meet its functional requirements of surface roughness and mechanical strength. From the case study, results indicate that the proposed approach is able to achieve good predictive results for responses of interest with a relatively small dataset. Further, process maps generated from the surrogate model provide a visual representation of the influence between responses of interest and critical process parameters for FFF process, which traditionally requires multiple investigations to arrive at similar conclusions.

show abstract

A multi-objective optimization framework for aerosol jet customized line width printing via small data set and prediction uncertainty

Cited by 17 publications

References 26 publications

Recent Advances in Multi-Material 3D Printing of Functional Ceramic Devices

Recent Advances in Multi-Material 3D Printing of Functional Ceramic Devices

Reviews on Machine Learning Approaches for Process Optimization in Noncontact Direct Ink Writing

A Data-Driven Framework to Predict Fused Filament Fabrication Part Properties Using Surrogate Models and Multi-Objective Optimisation

Contact Info

Product

Resources

About