EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Machine Learning Assisted Cross-Scale Hopper Design for Flowing Biomass Granular Materials

Abdallah Ikbarieh, Wencheng Jin iD, Yumeng Zhao, Nepu Saha, Jordan L Klinger, Yidong Xia iD, Sheng Dai iD

ACS Sustainable Chemistry & Engineering · 28.04.2025 · 13 (16) · 5838-5851

DOI: 10.1021/acssuschemeng.4c08938 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACS Sustainable Chemistry & Engineering, online erschienen am 16.04.2025. Verfasst von Abdallah Ikbarieh und 6 weiteren Autor:innen. Beteiligte Einrichtung: School of Civil and Environmental Engineering, Georgia Institute of Technology, 790 Atlantic Dr, Atlanta, Georgia 30332, United States. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 91 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Bioenergy Technologies Office.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The promise of biomass-derived biofuels is often overshadowed by intricate material handling challenges such as hopper clogging and screw feeder jamming. These handling issues stem from the knowledge gap among particle-scale material properties (e.g., particle size), bulk-scale material attributes (e.g., relative density), macro-scale equipment design (e.g., hopper inclination), and flow performance (e.g., probability of clogging). This work combines physical experiments, validated numerical simulations, and data augmentation to develop a machine learning-based hopper design for flowing granular woody biomass materials. The flow behavior of granular biomass is simulated and validated against physical tests utilizing the developed smoothed particle hydrodynamics (SPH) solver and a modified hypoplastic model. A comprehensive evaluation of the flow performance, including flow rate, flow stability, and flow pattern, is conducted on an extensive data set encompassing various biomass particle sizes, moisture contents, relative densities, and hopper operating conditions. A feed-forward neural network is trained and optimized with this data set to correlate cross-scale attributes with the flow performance metrics. The results reveal promising predictive accuracy on seen and unseen data sets. Further evaluation of how various input attributes affect the predicted flow performance metrics is carried out. The results indicate that hopper opening width primarily dictates flow throughput, while relative density, wall friction, inclination angle, and hopper opening width collectively impact flow stability. Additionally, flow patterns are predominantly governed by relative density, wall friction, and inclination angle. Moreover, the clogging potential is found to be exclusively characterized by the index dedicated to flow stability. The combination of high moisture contents, dense packing, smooth wall friction, low inclination angles, and small hopper opening widths substantially elevates the risk of unstable flows and clogging. This study serves as a potent design tool for flowing milled woody biomass materials in hoppers for all stakeholders in biorefineries and equipment manufacturing.

Autor:innen und Institutionen

  • School of Civil and Environmental Engineering, Georgia Institute of Technology, 790 Atlantic Dr, Atlanta, Georgia 30332, United States.
  • Harold Vance Department of Petroleum Engineering, Texas A&M University, 245 Spence Street, College Station, Texas 77843, United States.
  • Energy and Environment Science and Technology Directorate, Idaho National Laboratory, 1955 N Fermont Avenue, Idaho Falls, Idaho 83415, United States.

ORCID-Profile: Wencheng Jin, Yidong Xia, Sheng Dai

Förderung

  • Bioenergy Technologies Office

Zitieren

APA

Ikbarieh, A., Jin, W., Zhao, Y., Saha, N., Klinger, J. L., Xia, Y., & Dai, S. (2025). Machine Learning Assisted Cross-Scale Hopper Design for Flowing Biomass Granular Materials. ACS Sustainable Chemistry & Engineering, 13(16), 5838-5851. https://doi.org/10.1021/acssuschemeng.4c08938

Vancouver

Ikbarieh A, Jin W, Zhao Y, Saha N, Klinger JL, Xia Y, et al. Machine Learning Assisted Cross-Scale Hopper Design for Flowing Biomass Granular Materials. ACS Sustainable Chemistry & Engineering. 2025;13(16):5838-5851. doi:10.1021/acssuschemeng.4c08938

BibTeX
@article{Ikbarieh2025,
  author = {Ikbarieh, Abdallah and Jin, Wencheng and Zhao, Yumeng and Saha, Nepu and Klinger, Jordan L and Xia, Yidong and Dai, Sheng},
  title = {Machine Learning Assisted Cross-Scale Hopper Design for Flowing Biomass Granular Materials},
  journal = {ACS Sustainable Chemistry & Engineering},
  year = {2025},
  volume = {13},
  number = {16},
  pages = {5838--5851},
  doi = {10.1021/acssuschemeng.4c08938}
}
Kostenlos starten →