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

b) Geotechnical Tests And Parameters The geotechnical parameters of the soils were obtained by physical, mechanical, chemical and mineralogical tests in the laboratory following, whenever possible, the Brazilian standards of the National Department of Transport Infrastructure (DNIT, formerly DNER). In all, 57 variables were considered, and Table 2 presents a list of the tests performed with an indication of which of the parameters obtained through them were used as the basis for the generation of soil clustering, including their respective units, as well as the standards and manuals that served as a reference for the execution of each test. Table 2: Performed tests, obtained geotechnical parameters and reference standards/manuals Test Parameters Standards Real Particle Density Real Particle Density ( δ , g/cm³) DNER-ME 093/94 Atterberg limits Liquid Limit (LL, %), Plastic Limit (PL, %), Plasticity Index (PI, %) DNER-ME 122/94; DNER-ME 082/94 Granulometric analysis (with deflocculant) %Clay, %Silt, %Sand, %Fine Sand (%FS), %Medium Sand (%MS), % Coarse Sand (%CS), %Gravel(%Grav) DNER-ME 051/94 MCT Classification Coefficients c' and d', e’ index, Immersion Mass Loss (PMI, %), Immersion Mass Loss – specimen with optimal soil moisture (PMIwo, %) DNER-ME 258/94; DNER-ME 256/94 Compaction Apparent dry specific weight ( ρ dmáx , g/cm³) and optimal moisture content (w o , %) DNIT 164/2013-ME California Bearing Ratio CBR (%) and expansion (Exp., %) at the optimal moisture content DNIT 172/2016-ME Chemical tests hydrogen potential in: water (pH H2O ) and KCl (pH KCl ), Δ pH (pH KCl - do pH H2O ), Organic Matter (M.O., g.kg -1 ), Base saturation (V, %), Aluminum saturation (S, %), Cation Exchange Capacity (CEC, cmol c dm -3 ) EMBRAPA (2017) Soil-Water Characteristic Curve Soil suction at optimal moisture contente (S wo , kPa), saturation humidity ( θ s, %), residual moisture ( θ r , %), suction obtained in the resilient module specimens (SMR, kPa) and parameters of Model’s Gitirana Jr. and Fredlund (2004) - Ψ b1, Ψ res1, Sres1, Ψ b2, Sb, Ψ res2, Sres2 ASTM D5298 – 16 Resilient Modulus Resilient Modulus medium (MR medium , MPa), regression coefficients of linear models as a function of stresses: confining stress (K1TC e K2TC) and deviator (K1TD e K2TD), Composite model parameters (K1MC, K2MC e K3MC) DNIT 134/2018-ME Permanent Deformation Maximum permanente deformation ( ε p , mm) obtained under confining stress - σ 3 (120 kPa) and deviator stress - σ d (360 kPa), Parameters of the Guimaraes Model (2009) - Ψ 1, Ψ 2, Ψ 3 and Ψ 4 DNIT 179/2018-IE The LL and PL tests were performed both with samples previously dried in the air and destroyed (standard procedure) and without previous drying, and the parameters obtained were identified as follows: without previous drying - LL1, PL1 and PI1 and, with drying and destruction: LL2, PL2 and PI2. The test of the soil-water characteristic curve was performed using the filter paper method, with the specimen compacted on the intermediate Proctor energy and considering a drying and wetting mixed trajectory. The model adjustment parameters of Gitirana Jr. and Fredlund (2004) follow equation 1, applicable to soils with bimodal behavior commonly identified in lateritic tropical soils. The coefficients of determination (R²) obtained were all very close to 1 (>0.91) showing suitability of the model to the soil behavior of this article. = 1 − 2 1+( � 1 1 ) 1 + 2 − 3 1+( � 1 2 ) 2 + 3 − 4 1+( � 2 2 ) 3 (1) Where: S = degree of saturation; = suction obtained in the laboratory test;b = suction at the air inlet point; res = residual suction and d1, d2, d3 = model's parameters. Mechanical compaction, CBR, resilient modulus (MR) and permanent deformation tests were performed on the intermediate Proctor energy. The equations that were used to model resilient behaviors and permanent soil deformation are indicated in Table 3. © 2022 Global Journals Volume XXII Issue III Version I 13 ( ) Global Journal of Human Social Science - Year 2022 B Clustering of Fine-Grained Tropical Soils using Data Science Tools Applied to their Geotechnical Properties