skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: Explainable deep neural network for in-plain defect detection during additive manufacturing
PurposeThe purpose of this study is to develop a deep learning framework for additive manufacturing (AM), that can detect different defect types without being trained on specific defect data sets and can be applied for real-time process control. Design/methodology/approachThis study develops an explainable artificial intelligence (AI) framework, a zero-bias deep neural network (DNN) model for real-time defect detection during the AM process. In this method, the last dense layer of the DNN is replaced by two consecutive parts, a regular dense layer denoted (L1) for dimensional reduction, and a similarity matching layer (L2) for equal weight and non-biased cosine similarity matching. Grayscale images of 3D printed samples acquired during printing were used as the input to the zero-bias DNN. FindingsThis study demonstrates that the approach is capable of successfully detecting multiple types of defects such as cracks, stringing and warping with high accuracy without any prior training on defective data sets, with an accuracy of 99.5%. Practical implicationsOnce the model is set up, the computational time for anomaly detection is lower than the speed of image acquisition indicating the potential for real-time process control. It can also be used to minimize manual processing in AI-enabled AM. Originality/valueTo the best of the authors’ knowledge, this is the first study to use zero-bias DNN, an explainable AI approach for defect detection in AM.  more » « less
Award ID(s):
2001038
PAR ID:
10516863
Author(s) / Creator(s):
; ; ;
Publisher / Repository:
emerald insight
Date Published:
Journal Name:
Rapid Prototyping Journal
Volume:
30
Issue:
1
ISSN:
1355-2546
Page Range / eLocation ID:
49 to 59
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; (, Rapid Prototyping Journal)
    PurposePowder bed density is a key parameter in powder bed additive manufacturing (AM) processes but is not easily monitored. This research evaluates the possibility of non-invasively estimating the density of an AM powder bed via its thermal properties measured using flash thermography (FT). Design/methodology/approachThe thermal diffusivity and conductivity of the samples were found by fitting an analytical model to the measured surface temperature after flash of the powder on a polymer substrate, enabling the estimation of the powder bed density. FindingsFT estimated powder bed was within 8% of weight-based density measurements and the inferred thermal properties are consistent with literature findings. However, multiple flashes were necessary to ensure precise measurements due to noise in the experimental data and the similarity of thermal properties between the powder and substrate. Originality/valueThis paper emphasizes the capability of Flash Thermography (FT) for non-contact measurement of SS 316 L powder bed density, offering a pathway to in-situ monitoring for powder bed AM methods including binder jetting (BJ) and powder bed fusion. Despite the limitations of the current approach, the density knowledge and thermal properties measurements have the potential to enhance process development and thermal modeling powder bed AM processes, aiding in understanding the powder packing and thermal behavior. 
    more » « less
  2. ; ; ; ; ; (, Rapid Prototyping Journal)
    This study aims to discuss the state-of-the-art digital factory (DF) development combining digital twins (DTs), sensing devices, laser additive manufacturing (LAM) and subtractive manufacturing (SM) processes. The current shortcomings and outlook of the DF also have been highlighted. A DF is a state-of-the-art manufacturing facility that uses innovative technologies, including automation, artificial intelligence (AI), the Internet of Things, additive manufacturing (AM), SM, hybrid manufacturing (HM), sensors for real-time feedback and control, and a DT, to streamline and improve manufacturing operations. Design/methodology/approachThis study presents a novel perspective on DF development using laser-based AM, SM, sensors and DTs. Recent developments in laser-based AM, SM, sensors and DTs have been compiled. This study has been developed using systematic reviews and meta-analyses (PRISMA) guidelines, discussing literature on the DTs for laser-based AM, particularly laser powder bed fusion and direct energy deposition, in-situ monitoring and control equipment, SM and HM. The principal goal of this study is to highlight the aspects of DF and its development using existing techniques. FindingsA comprehensive literature review finds a substantial lack of complete techniques that incorporate cyber-physical systems, advanced data analytics, AI, standardized interoperability, human–machine cooperation and scalable adaptability. The suggested DF effectively fills this void by integrating cyber-physical system components, including DT, AM, SM and sensors into the manufacturing process. Using sophisticated data analytics and AI algorithms, the DF facilitates real-time data analysis, predictive maintenance, quality control and optimal resource allocation. In addition, the suggested DF ensures interoperability between diverse devices and systems by emphasizing standardized communication protocols and interfaces. The modular and adaptable architecture of the DF enables scalability and adaptation, allowing for rapid reaction to market conditions. Originality/valueBased on the need of DF, this review presents a comprehensive approach to DF development using DTs, sensing devices, LAM and SM processes and provides current progress in this domain. 
    more » « less
  3. ; ; ; ; (, Journal of Applied Research in Higher Education)
    Blessinger, Patrick (Ed.)
    PurposeThis study examines the impact of participation in a STEM Enrichment Summer Bridge Program, funded by the NSF Houston-Louis Stokes Alliance for Minority Participation, on undergraduate student success outcomes, particularly for under-represented students. Design/methodology/approachThe study uses propensity score matching and logistic regression analysis to examine the effects of participation in the STEM enrichment program on graduation and retention in STEM after matching on baseline socio-demographic and pre-college characteristics. FindingsThe analysis found that program participation had a significant effect on increasing both the graduation rates and retention of under-represented minority students in STEM fields. In addition, results indicated that program participation had a particularly strong impact for Pell-eligible students in terms of course grades. Research limitations/implicationsData obtained for this study were limited to a single Hispanic-serving/Asian-serving institution, and therefore are not necessarily representative of the graduation and retention trends of the larger population of underrepresented minority (URM) students across the nation. Originality/valueThis study uniquely adds to the existing body of literature surrounding the retention of URM students in STEM fields by accounting for baseline variables, such as pre-college academic achievement and socio-demographic characteristics, that could lead to bias in estimating results. Specifically, this study addresses limitations of previous studies by comparing participants and non-participants of the STEM enrichment program who are matched on a selection of baseline characteristics. 
    more » « less
  4. ; ; (, International Journal of Mentoring and Coaching in Education)
    PurposeThe purpose of this study was to examine the experiences of multiple campus teams as they engaged in the assessment of their science, technology, engineering and mathematics (STEM) mentoring ecosystems within a peer assessment dialogue exercise. Design/methodology/approachThis project utilized a qualitative multicase study method involving six campus teams, drawing upon completed inventory and visual mapping artefacts, session observations and debriefing interviews. The campuses included research universities, small colleges and minority-serving institutions (MSIs) across the United States of America. The authors analysed which features of the peer assessment dialogue exercise scaffolded participants' learning about ecosystem synergies and threats. FindingsThe results illustrated the benefit of instructor modelling, intra-team process time and multiple rounds of peer assessment. Participants gained new insights into their own campuses and an increased sense of possibility by dialoguing with peer campuses. Research limitations/implicationsThis project involved teams from a small set of institutions, relying on observational and self-reported debriefing data. Future research could centre perspectives of institutional leaders. Practical implicationsThe authors recommend dedicating time to the institutional assessment of mentoring ecosystems. Investing in a campus-wide mentoring infrastructure could align with campus equity goals. Originality/valueIn contrast to studies that have focussed solely on programmatic outcomes of mentoring, this study explored strategies to strengthen institutional mentoring ecosystems in higher education, with a focus on peer assessment, dialogue and learning exercises. 
    more » « less
  5. ; (, Rapid Prototyping Journal)
    PurposeSurface quality and porosity significantly influence the structural and functional properties of the final product. This study aims to establish and explain the underlying relationships among processing parameters, top surface roughness and porosity level in additively manufactured 316L stainless steel. Design/methodology/approachA systematic variation of printing process parameters was conducted to print cubic samples based on laser power, speed and their combinations of energy density. Melt pool morphologies and dimensions, surface roughness quantified by arithmetic mean height (Sa) and porosity levels were characterized via optical confocal microscopy. FindingsThe study reveals that the laser power required to achieve optimal top surface quality increases with the volumetric energy density (VED) levels. A smooth top surface (Sa < 15 µm) or a rough surface with humps at high VEDs (VED > 133.3 J/mm3) can serve as indicators for fully dense bulk samples, while rough top surfaces resulting from melt pool discontinuity correlate with high porosity levels. Under insufficient VED, melt pool discontinuity dominates the top surface. At high VEDs, surface quality improves with increased power as mitigation of melt pool discontinuity, followed by the deterioration with hump formation. Originality/valueThis study reveals and summarizes the formation mechanism of dominant features on top surface features and offers a potential method to predict the porosity by observing the top surface features with consideration of processing conditions. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email