Repeated Load Californian Bearing Ratio test (RL-CBR) is a geotechnical laboratory test similar to the well-known CBR test with cyclic loading. The RL-CBR test was developed at the Netherlands University to study unbound granular materials resilient behaviour by estimating their equivalent modulus. In this case, three model parameters must be determined by Finite Element Method (FEM) and non-linear multidimensional regression between equivalent modulus as a response variable and Poisson's ratio , mean stress under plunger and recoverable plunger penetration as explanatory variables. The purpose of this study is to show how FEM inputs affect the obtained model parameters. The estimation procedure consists of the simulation of RL-CBR test varying elastic materials characteristics and plunger penetrations. For each combination, nodal reactions under plunger are captured and mean stress under plunger is calculated. A non-linear multidimensional regression allows deriving equivalent modulus equation by the least-squares method which consists of minimizing the quadratic error between real and estimated modulus. In practice, these model parameters depend also on mould characteristics and specimen-mould contact nature. In this paper, the influence of these parameters on the model's parameters' values is studied to determine the adequate set of parameters for carrying out an accurate estimation of equivalent modulus' parameters.