A simulation model to predict agricultural tractor-semi-trailer combination traction performance under different operating conditions

Abstract A computational model was developed to evaluate the performance and design parameters of a tractor–semi-trailer combination operating under different soil conditions. The model predicts key operational parameters including draft force, drawbar power, tractive efficiency, and fuel consumption, while also generating geometric and structural design parameters for the semi-trailer such as loading box dimensions, axle load distribution, suspension configuration, and weight transfer to the tractor drawbar. Soil–tire interaction was represented using traction and rolling resistance relationships derived from established agricultural traction equations. The model was implemented in a Python-based graphical user interface that allows users to evaluate tractor–semi-trailer performance for different operating conditions including soil strength, wheel slip, forward speed, loading capacity, tractor drivetrain configuration, and tire type. Soil conditions were represented using cone index values ranging from 450 to 1800 kPa corresponding to soft sandy, tilled, firm, and hard soil. Model validation was conducted using standardized tractor performance data obtained from official tractor test reports for four agricultural tractors: John Deere 7810, John Deere 7930, New Holland G190, and New Holland TS120. Simulations were performed under identical operating conditions (12% slip and 6 km h⁻ 1 forward speed) to evaluate model scalability across tractors with different power ratings and geometric configurations. Predicted drawbar power requirements ranged from approximately 31 to 105 HP depending on soil strength and tractor characteristics. Comparison with standardized tractor test data showed that predicted power demand represented between approximately 62% and 74% of available drawbar power for most tractors under firm soil conditions. Statistical evaluation indicated stable model performance with a coefficient of variation of approximately 9.7% and a correlation coefficient of R 2 ≈ 0.84 between predicted and reference drawbar power values. The results demonstrate that the developed model provides realistic predictions of tractor–semi-trailer performance and can be used as a decision-support tool for evaluating transport operations and assisting in the design and selection of tractor–trailer combinations under varying soil and loading conditions.