EVALUATION OF DIFFERENT METHODS FOR ESTIMATING THE ROLLING RESISTANCE OF AGRICULTURAL TRACTORS BASED ON BEKKER'S SOIL PARAMETERS.

The general objective of this study was to develop a device to obtain Bekker's soil parameters (k_c, k_Φ and n) in-situ. These parameters could be used to estimate the rolling resistance of agricultural tractors. The specific objective comprises a comparison among four methods for estimating rolling resistance of an agricultural tractor. In order to compare the different methods and find out the most appropriate conditions for using them, a field experiment in which all relevant variables were controlled for further simulation was conducted by using a 2WD tractor. Using the soil parameters obtained in field, a comparison was carried out with the results from the different methods. Bekker's soil parameters are obtained in this study using an ellipse shaped sinkage plate. The field measurements of Bekker's soil parameters and rolling resistance were achieved in sandy, sandy loam and silty clay soil under the same conditions. Cohesion and soil internal frication angle were obtained in laboratory using direct shear test. Three models were developed by multiple linear regressions to get Bekker's soil parameters as a function of soil cohesion, soil internal frication angle, cone index and soil texture index. Using the field measurements and predicted Bekker's soil parameters, tractor rolling resistance was obtained by the help of four methods and then compared to the measured for Nasr tractor. The rolling resistances calculated from Bekker's method (Method 3 and Method 4) were the closest to the measured. Lower absolute relative errors of 5.9, 6.0 and 8.5% for method 3 were found when the tractor was run on sandy, sandy loam and silty clay soils, respectively. Also, there is a clear effect of soil texture index on tractor rolling resistance with linear relationship with coefficients of determination (R²) ranged from 0.766 to 0.998 based on the method of obtaining rolling resistance.