Characterization of the shape of aggregates using image analysis and machine learning classification tools

Geomechanics and Geoengineering, Journal Year: 2023, Volume and Issue: 19(4), P. 421 - 443

Published: Dec. 4, 2023

Engineering applications including pervious concrete require effective packing of aggregates to optimize strength. Size and shape distribution significantly affect the performance. Computational methods numerically represent aggregates, from image analysis, effectiveness has not been compared verified. This study aims analyse representability aspects by different computational methods. Crushed were grouped into 5 clusters, each group was milled in a Los Angeles machine for degrees (0–2000) induce morphological changes on aggregates. Aggregates ranging 30mm diameter obtained (7191 total). imageJTM, used compute dimensions factors 14 Statistical tests, Pearson's Correlations Principal Components Analysis learning classification tools, Decision-tree, Random-Forest, Naïve-Bayes, Support-Vector-Machines, K-Nearest-Neighbours Perceptron employed assess. In conclusion, no factor could be singularly aggregate particles but combination is required. Data matrix had three primary dimensions. Combination Circularity, Kumbrein-Solidity Barksdale-Shape-Factor yield best representation shape. Regression Tree method highest accuracy (0.9) classifying unmilled

Language: Английский

Comparative analysis of the expansion rate and soil erodibility factor of some gullies in Nnewi and Nnobi, Southeastern Nigeria

Scientific Reports, Journal Year: 2023, Volume and Issue: 13(1)

Published: Sept. 15, 2023

The research focused on assessing the expansion rate and soil erodibility factor (K) of specific gullies located in Nnewi Nnobi, Southeastern Nigeria. Fifteen representative were studied extensively. Grain size distribution analysis revealed that soils are composed gravel (5.77-17.67% 7.01-13.65%), sand (79.90-91.01% 82.47-88.67%), fines (2.36-4.05% 3.78-5.02%) for Nnobi respectively. cohesion internal friction angle values range from 1-5 to 2-5 kPa 29-38° 30-34° respectively, which suggests have low shear strength susceptible failure. plasticity index (PI) showed they nonplastic plastic highly liquefiable with ranging 0-10 0-9% Slope stability gave safety (FoS) 0.50-0.76 0.82-0.95 saturated condition 0.73-0.98 0.87-1.04 unsaturated both respectively indicating slopes generally unstable critically stable. erosion a fifteen-year period (2005-2020) an average longitudinal 36.05 m/yr 10.76 8.57 × 10-2 1.62 10-4 higher potentials than those Nnobi. Conclusively, area is more prone area.

Language: Английский