Search

Your search keyword '"FU, Zhenbo"' showing total 18 results
Start Over Author "FU, Zhenbo"
18 results on '"FU, Zhenbo"'

1. NeutronStream: A Dynamic GNN Training Framework with Sliding Window for Graph Streams

Author

Chen, Chaoyi, Gao, Dechao, Zhang, Yanfeng, Wang, Qiange, Fu, Zhenbo, Zhang, Xuecang, Zhu, Junhua, Gu, Yu, and Yu, Ge

Subjects
Computer Science - Machine Learning, Computer Science - Distributed, Parallel, and Cluster Computing
Abstract

Existing Graph Neural Network (GNN) training frameworks have been designed to help developers easily create performant GNN implementations. However, most existing GNN frameworks assume that the input graphs are static, but ignore that most real-world graphs are constantly evolving. Though many dynamic GNN models have emerged to learn from evolving graphs, the training process of these dynamic GNNs is dramatically different from traditional GNNs in that it captures both the spatial and temporal dependencies of graph updates. This poses new challenges for designing dynamic GNN training frameworks. First, the traditional batched training method fails to capture real-time structural evolution information. Second, the time-dependent nature makes parallel training hard to design. Third, it lacks system supports for users to efficiently implement dynamic GNNs. In this paper, we present NeutronStream, a framework for training dynamic GNN models. NeutronStream abstracts the input dynamic graph into a chronologically updated stream of events and processes the stream with an optimized sliding window to incrementally capture the spatial-temporal dependencies of events. Furthermore, NeutronStream provides a parallel execution engine to tackle the sequential event processing challenge to achieve high performance. NeutronStream also integrates a built-in graph storage structure that supports dynamic updates and provides a set of easy-to-use APIs that allow users to express their dynamic GNNs. Our experimental results demonstrate that, compared to state-of-the-art dynamic GNN implementations, NeutronStream achieves speedups ranging from 1.48X to 5.87X and an average accuracy improvement of 3.97%., Comment: 12 pages, 15 figures

Published
2023

2. Mechanical Properties and Microstructure of Alkali-Activated Cements with Granulated Blast Furnace Slag, Fly Ash and Desert Sand.

Author

Liu, Yunpeng, Yang, Xihao, Tian, Wendi, Fu, Zhenbo, Zhao, Yimeng, Li, Binghan, Li, Shiji, Xu, Da, Yu, Shige, Yao, Zhiyu, Zhao, Tian, Ouyang, Xinfeng, Wang, Guangfei, Yu, Hai, Li, Dan, Guo, Rongxin, Wei, Yen, and Niu, Kangmin

Subjects
X-ray microanalysis, FLY ash, FLEXURAL strength, INFRARED spectroscopy, X-ray diffraction
Abstract

In this study, desert sand was used as supplementary materials in alkali-activated cements (AAC) with granulated blast furnace slag (GBFS) and fly ash (FA). For the first time, a systematic investigation was conducted on the effects of various treatment methods and contents of desert sand on the strength and microstructure of AAC. This study also analyzed the X-ray diffractometer (XRD), Scanning Electron Microscopy-Energy Dispersive X-ray Microanalysis (SEM-EDX), Mercury Intrusion Porosimetry (MIP), pH values, and Fourier-transform infrared spectroscopy (FT-IR) properties of AAC pastes containing differently treated desert sand to uncover the mechanisms by which these treatments and dosages influence mechanical properties of AAC. Untreated desert sand (DS), temperature-treated desert sand (DS-T), and ground desert sand for two different durations (20 mins and 30 mins) all exhibited some pozzolanic activity but primarily acted as fillers in the AAC pastes. Among the samples, DS-T demonstrated the highest pozzolanic activity, though it was still less than that of fly ash (FA). The optimal dosage for the modified desert sands was determined to be 10%. However, The optimal dosage of different modified desert sands is 10%. The flexural strength of DS-G30-10 reaches 6.62 MPa and the compressive strength reaches 72.3 MPa, showing the best comprehensive mechanical properties. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

3. The Influence of FA Content on the Mechanical and Hydration Properties of Alkali-Activated Ground Granulated Blast Furnace Slag Cement.

Author

Liu, Yunpeng, Fu, Zhenbo, Yang, Xihao, Zhao, Yimeng, Li, Binghan, Xu, Da, Yu, Shige, Yao, Zhiyu, Sun, Zhibin, Zhen, Zhilu, Ouyang, Xinfeng, Zong, Yangyang, Tian, Wendi, Yu, Hai, Zhao, Shuli, Wei, Yen, and Niu, Kangmin

Subjects
POROSITY, HEAT release rates, SLAG cement, HEAT of hydration, COMPRESSIVE strength, PORTLAND cement
Abstract

This study primarily investigates the effect of fly ash (FA) content on the mechanical properties and hydration performance of alkali-activated ground granulated blast furnace slag cement (AAGC) and compares the related properties with ordinary Portland cement (OPC). Additionally, we examined the hydration products; performed thermal analysis, MIP, and SEM; and determined chemically bound water and pH values of AAGC. The compressive strength of AAGC showed a retrogression phenomenon from 3 to 28 days, with the 14-day and 28-day compressive strengths of AAGC being higher than those of OPC. The AAGC with 20% FA content exhibited the highest 28-day compressive strength (75 MPa). The hydration heat release rate curve of OPC and AAGC was divided into the initial induction period, induction period, acceleration period, deceleration period, and steady period. As FA content increased, the 28-day pore volume of AAGC increased, while pH values and chemically bound water decreased. SEM images of AAGC with low FA content showed more microcracks. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

11. CFD Approach to the Research and Design of Low Emission Commercial Aircraft Engine Combustor

Author

Lin Yu-zhen, Fu Zhenbo, Zhang Man, and Li Jibao

Subjects
Engineering, business.industry, Turbulence, Diffusion flame, Flow (psychology), Flame structure, LPP, General Medicine, Computational fluid dynamics, Automotive engineering, NOx reduction, Low emission combustor, Combustor, Aerospace engineering, Air worthiness, CFD, business, Reynolds-averaged Navier–Stokes equations, Engineering(all), Test data
Abstract

The foundation of AVIC Commercial Aircraft Engine Company (ACAE) in Shanghai triggers the hot upsurge of the civil aircraft engine development and study in China. Air worthiness requirements of the engine emission brings the new challenge to the combustor research and design. With the motivation of design high performance and clean combustor, CFD is utilized as the powerful design approach. In this paper, Reynolds averaged resolve of Navier-Stokes (RANS) equations of reaction two-phase flow in an experimental low emission combustor are performed. The numerical approach uses an implicit compressible gas solver together with a Lagrangian liquid-phase tracking method and the extended coherent flamelet model for turbulence-combustion interaction. The NOx formation is modeled by the concept of post-processing, which resolves the NOx transport equation with the assumption of frozen temperature distribution. Both turbulence-combustion interaction model and NOx formation model are firstly evaluated by the comparison of experimental data published in open literatures of a lean direct injection (LDI) combustor. The test rig studied in this paper is called Low Emission Stirred Swirl (LESS) combustor, which is a two stage model combustor, fueled with liquid RP-3 kerosene and designed by Beijing University of Aeronautics and Astronautics (BUAA). The main stage of LESS combustor employs the principle of lean prevaporized and premixed (LPP) concept to reduce pollutant, and the pilot stage depends on a diffusion flame for flame stabilization. Detailed numerical results including of species distribution, turbulence performance and burning performance are qualitatively and quantitatively evaluated. Numerical prediction of NOx emission shows a good agreement with test data both at idle condition and full power condition of LESS combustor. Preliminary results of the flame structure are shown in this paper. The flame stabilization mechanism and NOx reduction effort are also discussed with in-depth analysis.

Published
2011
Full Text
View/download PDF

18. CFD Study of NOxEmissions in a Model Commercial Aircraft Engine Combustor

Author

ZHANG, Man, FU, Zhenbo, LIN, Yuzhen, and LI, Jibao

Abstract

Air worthiness requirements of the aircraft engine emission bring new challenges to the combustor research and design. With the motivation to design high performance and clean combustor, computational fluid dynamics (CFD) is utilized as the powerful design approach. In this paper, Reynolds averaged Navier-Stokes (RANS) equations of reactive two-phase flow in an experimental low emission combustor is performed. The numerical approach uses an implicit compressible gas solver together with a Lagrangian liquid-phase tracking method and the extended coherent flamelet model for turbulence-combustion interaction. The NOxformation is modeled by the concept of post-processing, which resolves the NOxtransport equation with the assumption of frozen temperature distribution. Both turbulence-combustion interaction model and NOxformation model are firstly evaluated by the comparison of experimental data published in open literature of a lean direct injection (LDI) combustor. The test rig studied in this paper is called low emission stirred swirl (LESS) combustor, which is a two-stage model combustor, fueled with liquid kerosene (RP-3) and designed by Beihang University (BUAA). The main stage of LESS combustor employs the principle of lean prevaporized and premixed (LPP) concept to reduce pollutant, and the pilot stage depends on a diffusion flame for flame stabilization. Detailed numerical results including species distribution, turbulence performance and burning performance are qualitatively and quantitatively evaluated. Numerical prediction of NOxemission shows a good agreement with test data at both idle condition and full power condition of LESS combustor. Preliminary results of the flame structure are shown in this paper. The flame stabilization mechanism and NOxreduction effort are also discussed with in-depth analysis.

Published
2012
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results