Entity

Time filter

Source Type


Duan Y.-H.,South China University of Technology | Duan Y.-H.,Guangdong Hualu Communications Technology Co | Zhang X.-N.,South China University of Technology | Zhang S.-X.,South China University of Technology | Yuan M.-M.,South China University of Technology
Journal of Harbin Institute of Technology (New Series) | Year: 2011

In this paper a series of digital image processing methods were adopted for getting separated coarse aggregates from asphalt mixture specimen using high-resolution X-ray Computed Tomography (CT) images. The existing three dimensional (3D) particles matching methods based on two dimensional (2D) continuous cross-sections were analyzed and a new 'overlap area method' was presented. After the 3D particles were extracted one by one successfully, the basic parameters of each aggregate: perimeter, area, surface area, and volume were calculated by chain code method. Finally, the 3D mass center coordinates and the sphericity index were introduced. Source


Duan Y.-H.,South China University of Technology | Duan Y.-H.,Guangdong Hualu Communications Technology Co | Zhang X.-N.,South China University of Technology | Li H.-J.,South China University of Technology | And 3 more authors.
Jianzhu Cailiao Xuebao/Journal of Building Materials | Year: 2011

The industrial CT was used for getting the cross-section images of asphalt mixture. Based on digital image processing methods, the coarse aggregates were extracted from the images, and the adhesion problem of coarse aggregates particles was resolved successfully. The judgment criterion of contact between particles was established, and a 5 pixel × 5 pixel size window was designed for contact searching along the particle boundary. After total 640 images were searched by these procedures, the data were analyzed qualitatively, and a contact degree indicator(C) was tried to establish. Four different probability density distribution functions were used to fit C data, and the validity of fit was tested by Kolmogorov-Smirnov and Chi-square methods. Finally, the lognormal distribution was determined for describing the contact characteristics of coarse aggregates particles. Source


Duan Y.,South China University of Technology | Duan Y.,Guangdong Hualu Communications Technology Co | Zhang X.,South China University of Technology | Li Z.,South China University of Technology | Wang D.,South China University of Technology
International Journal of Pavement Research and Technology | Year: 2011

Until now, voids in the Coarse Aggregate (VCA) parameter were used to judge whether the coarse aggregate skeleton is formed in asphalt mixtures. However, this method has several shortcomings and deficiencies. In this paper, the X-ray Computed Tomography (CT) is used to obtain two-dimensional (2D) images through a series of digital image processing methods to separate the aggregates successfully. A contact searching method was established to find all the contacting pixels surrounding each particle. All pixels found were stored and analyzed. The contact degree changes, along with the different depths of the specimen, and the contacting pixel values calculated by the algorithm concentrate to approximately 0-20 pixels, which equals about 0-2 mm. Three quantitative indicators, Cl, C2, and C3, were established. Their input data were fitted by 4 different probability density functions (PDFs): Normal distribution, Lognormal distribution, Gamma distribution, and Weibull distribution. The goodness of the fit was simultaneously tested through two different formal statistical tests (Kolmogorov-Smirnov and Chi-square) to provide mathematical support. Finally, the Lognormal PDF was found to be the most suitable for describing the input data distributions, and Cl was recommended as the quantitative indicator for the assessment of stone-stone contact degree. © Chinese Society of Pavement Engineering. Source


Tan Y.-Q.,Harbin Institute of Technology | Xu H.-N.,Harbin Institute of Technology | Zhou C.-X.,Dalian Maritime University | Zhang K.,Guangdong Hualu Communications Technology Co | Chen F.-C.,China Airport Construction Group Corporation of CAAC
Harbin Gongye Daxue Xuebao/Journal of Harbin Institute of Technology | Year: 2011

Based on the field data collected from a medium-scale experimental pavement system, temperature distribution characteristic of subgrade in seasonally frozen regions was studied in detail, and the prediction model for subgrade temperature distribution was proposed with regression analysis. Test data illustrated that: The temperature of subgrade changed with time, and it followed cosine distribution; Energy was stored with the heat transfer in subgrade, and it transferred nonlinear with depth. Based on the collected data, the prediction model for subgrade temperature distribution in this area was proposed. The comparison between measured and predicted temperatures indicated that this model had good applicability and accuracy for this area. Source


Duan Y.-H.,South China University of Technology | Duan Y.-H.,Guangdong Hualu Communications Technology Co | Zhang X.-N.,South China University of Technology | Li Z.,South China University of Technology | And 2 more authors.
Zhongguo Gonglu Xuebao/China Journal of Highway and Transport | Year: 2011

X-ray computed tomography (CT) was used to get the cross section images of asphalt concrete specimen. Two-dimensional profiles of aggregates were extracted by a series of digital image processing methods. The Fourier descriptors were used to represent the particle profile. Combined with the concept of equivalent diameter, the reasonable sampling intervals were determined, the value of n was qualitatively analyzed, the goodness and degradation of fit were described by coefficient of determination R 2 and graphs, the possible failure case was analyzed. The spherical harmonics transformation was used to represent the 3-D surface profile of particles, the 3-D visualization results of different spherical harmonics transformation values n ' and different sampling points were compared, changing rule of R 2 with n and sampling points was discovered. Results show that the value of n ' is more appropriate to take 25 to 30. The spherical harmonics transformation can describe the 3-D surface profile and local meso-scale characteristic of aggregate particles more accurately, and representing the aggregates by this way is beneficial to the need of computer modeling, and it is beneficial to future virtual experiment work on concrete. Source

Discover hidden collaborations