A random elastoplastic framework of relationships between CBR and Young’s modulus for coarse-grained soils

Abstract

AbstractThe material quality indicator most commonly used for the quality of road materials is the California Bearing Ratio test (CBR). In addition, the CBR test provides a value commonly used to correlate with the resilient modulus (used in the design and analysis of pavement) due to its cost-effectiveness. This test measures soil resistance to penetration by a punch and compares it with the pressure measured in a standard material at the same penetration (2.5 mm or 5 mm). However, this test lacks physical explanations regarding mechanical behavior because the CBR value only compares soil penetration resistance with the pressure measured in a standard material. Another issue arises from the scattered results obtained from both equations and tests, highlighting the need for variability analysis of the CBR test to assess the effect of the different geotechnical variables on the CBR value. For this purpose, simulations using FEM (Finite Element Method) considering random soil parameters were performed for the CBR test. These FEMs included a linear elastic model and two failure criteria (Mohr–Coulomb and Drucker-Prager with a cap) and were prepared for granular soils. The evaluation shows that the increase or decrease in the CBR value is a function of the elastic modulus, yield stress, and friction angle. Moreover, the simulations expand the knowledge of the shearing mechanisms, generated stresses, displacement fields, and load sharing when the CBR test is made. From these results, a physical explanation of test results can be done. FEM simulations showed stress zones in conditions of elastic, compression, and shear behavior. These zones can explain the importance of elastic modulus, yielding stress, and friction angle in the CBR value. From numerical results, a new equation was proposed and compared with practical equations proposed by international standards and other sources to estimate the probability of underestimated values CBR according to the correlations used.