Uji Triaksial Terkonsolidasi Tak Terdrainase pada Tanah Lokasi PDAM Tirtanadi Medan Marelan dengan Metode Elemen Hingga Teknik Mesin

Abstract

Triaxial test is a test done to obtain the parameters of soil such as shift angle ( rp) and cohesion (c). This test has been widely used to do a testing in the projects or research (education). In this study, constant strain triaxial test was done through consolidatedundrained method (CU tests) with 3 (three) samples of soil. The result of this test will show, among other things, soil parameter, strain-stress curve, Mohr circle diagram, stress path, p-q graphic, and then the data above are reversely predicted by using the Finite Element Method program with Mohr-Coulomb and Soft Soil modeling. The result of laboratory test showed that the corifining stress of test material-] was 100 kN/m2 , the soil failed at deviator stress of 153 kN/m2 and strain 0. 085, the corifining stress of test material-2 was 200 kN/m2 , the soil failed at deviator stress of 177 kN/m2 and strain 0.085, and the corifining stress of test material-3 was 300 kN/m2 , the soil failed at deviator stress of 196 kN/m2 and strain 0.11. For reversed prediction of Mohr-Coulomb model, test material-] failed at deviator stress of 133 kN/m2 and strain 0.052, test material-2 failed at deviator stress of 136 kN/m2 and strain 0. 025, and test material-3 failed at deviator stress of 138 kN/m2 and strain 0. 04 while for reversed prediction of Soft Soil model, test material-I failed at deviator stress of 115 kN/m2 and strain 0. 0719, tests material-2 failed at deviator stress of 106 kN/m2 and strain 0. 0380, and test material-3 failed at deviator stress of 103 kN/m2 and strain 0.513. The result of laboratory test of test material-3 also showed the unnatural curve because it intersected with test material-2, in which with higher corifining stress, its deviator stress had to be higher too. The result of laboratory test if compared to that of the Finite Element Method modeling showed a very much difference. Soil failure at lower strain occurred in the response of Mohr-Coulomb modeling compared to that of laboratory test. This case was caused by the limitations of parameters of the Finite Element program used, while in the laboratory test, it was caused by many factors such as human errors, less perfect molding of test material samples, and less perfect condition of tools. The response of Soft Soil model showed that when loaded, the soil could survive the high strain compared to the response of laboratory test. The existence of the parameters of modified compression index and modified expansion index provides a clear picture that soil is expected to have an elastic behavior. A very high stress occurred in the Soft Soil model.