Global Journal of Human-Social Science, B: Geography, Environmental Science and Disaster Management, Volume 22 Issue 3

material (retained in sieve no. 10 – between 0.12% and 3.76%). Regarding the consistency indexes (LL, PL and PI) it is observed that for all soils there was an increase in values with drying. The actual density values of the grains obtained were between 2.629 – 2.699 g/cm³ which is approximate, according to Gidigasu (1976), to the quartz mineral (2.65 to 2.66 g/cm³) and the clay mineral kaolinite (2.60 to 2.68 g/cm³), which is coherent since most of the soils studied are composed of sand and clay fractions. The results of maximum dry specific mass ( ρ dmax , g/cm³) and optimum moisture (W or , %) ranged between 1.69 - 1.97 g/cm³ and between 10.6 - 19%, respectively, and showed some correspondence with the granulometry. Soil 03, for example, has a large percentage of fine sand in its composition, which is reflected in the compaction curve (low optimum moisture and high specific dry mass), while soil 08, with the highest percentage of silt, compared to the other soils, presented higher moisture and lower specific mass. Six soils presented CBR values below 20%, which is, according to the paving manual of the Brazilian Department of Transport Infrastructure (DNIT, 2006), the minimum allowed for application in subbase layers, considering the empirical sizing method. The other soils presented values higher than 20%, reaching a maximum of 35% (sample 03), which according to the mentioned guidelines could be recommended for subbase layer and none of the soils would be recommended for the base layer since CBR>60% is required. Regarding expansion, six soils showed an expansion of 0% after 96 hours of immersion in water, and the other 7 samples showed low expansion values. The values of mean MR of all soils were above 400 MPa, considered a high value for fine-grained soils, comparable to the values of boulder soils. As for permanent deformation, all soils presented low total deformation values ( ε p , mm) for all stress levels applied in the test. According to The National Pavement Sizing Method (MeDiNa), the sum of the contribution of all layers and subgrade to the sinking of the wheel track should be a maximum of 10 mm for Main Arterial Route, for example. Almost half of the soils exhibit opposite behavior in relation to the parameter Ψ 2 of the expression of permanent deformation, associated with the confining stress. Soils with negative values of Ψ 2 show a reduction in permanent deformation with increased confining stress. All values of Ψ 3 are positive indicating that the variation of the deviation stress increases permanent deformation, which is expected. The pH measurement reflects the active acidity of the soil, and the results obtained (< 5) represent soils with high acidity, which is expected for lateritic soils. Specifically in water, values between 4.4 and 4.7 were obtained indicating the presence of exchangeable aluminum (Sobral et al., 2015), which suggests possible gains in chemical stabilization processes. Δ pH is associated with the predominance of clay minerals such as kaolinite and illite, (Farias, 2012 and Camapum de Carvalho et al., 2015). The CEC presented low values (between 2.8 – 5.74 cmol c dm -3 ) indicating predominance of 1:1 clay such as kaolinite (Gidigasu, 1976; Das, 2008 and Sobral et al., 2015). Regarding Organic Matter (OM), all soils have a low content (< 15 g.kg -1 ) according to Prezotti (2013). For all soils, the percentage of base saturation (V, %) is considered "low" (< 50%) according to Prezotti, 2013). Aluminum saturation (S, %) was in the "low" class (< 50%) in soils 01, 02, 03 and 07, "middle" class (50% - 70%), in 04, 06, 08, 10, 11 and 13, and, "high" class (>70%) in soils 05, 09 and 12, according to Prezotti, (2013). Table 6: Data of the characteristic curve of the studied soils (Continuation of the database in Table 5) Soil S Wo θ s θ r SMR Ψ b1 Ψ res1 S res1 Ψ b2 S b Ψ res2 S res2 01 200 20,58 1,47 198,59 3,5 5,0 0,84 10000 0,82 26800 0,02 02 800 20,67 0,00 760,23 3,0 6,0 0,89 7500 0,85 22500 0,08 03 30 34,34 0,06 1071,43 2,5 16,0 0,70 8000 0,62 21500 0,02 04 30 17,64 0,00 487,86 4,0 5,5 0,74 10000 0,59 23000 0,03 05 300 14,35 0,26 434,80 3,4 6,5 0,78 8700 0,72 27000 0,03 06 50 29,31 0,99 372,62 4,8 6,5 0,84 18000 0,77 36000 0,10 07 6 23,60 1,01 430,23 6,0 9,0 0,89 14000 0,83 20000 0,05 08 100 22,26 0,00 485,99 3,8 10,0 0,91 12000 0,87 48000 0,07 09 400 13,49 0,26 1088,10 1,2 11,0 0,75 14000 0,72 30000 0,04 10 25 23,91 0,45 1527,01 2,1 20,0 0,74 15000 0,66 25000 0,04 11 40 33,60 0,17 1673,32 3,8 20,0 0,79 13500 0,71 22500 0,04 12 1000 17,20 0,00 1147,13 2,0 11,0 0,69 5500 0,66 24000 0,04 13 40 26,85 0,60 275,91 3,5 6,5 0,73 13800 0,65 30000 0,04 © 2022 Global Journals Volume XXII Issue III Version I 17 ( ) Global Journal of Human Social Science - Year 2022 B Clustering of Fine-Grained Tropical Soils using Data Science Tools Applied to their Geotechnical Properties 18.66% and 55.85%), and low percentage of coarse