Improvement and verification of pressure-sinkage model in homogeneous soil

Abstract: Vehicles mobility performance relies on the sustentation of the soil. The sinkage and driving resistance are closely related to soil pressure-bearing characteristics. There are many soil pressure-sinkage theories that allow the designers of o?-road vehicles to understand and predict vehicle mobility performance over the soft soil. In this paper, these theories were firstly divided into 3 categories in accordance with the mechanism and terramechanics after a brief introduction. Then their characteristics, applicability and meaning of equations were analyzed in detail. Bekker proposed the pressure-sinkage model based on the previous research and got soil subsidence parameters by loading plate bearing. The method is simple and still widely used today, but it is difficult in obtaining in-suit bearing test curve. American scholar Karafiath et al. predicted soil subsidence pressure curve by means of soil mechanics method, which is ideal and does not agree with the actual soil pressure stress value, especially for sandy compacted soil. British scholar Reece et al. described soil subsidence pressure curve combined with the ultimate bearing capacity concept in soil mechanics and Bekker's methods. It is relatively accurate as it has a comparative system theory. On this basis, in order to achieve an accurate prediction of soil pressure-subsidence curve by soil parameters, a new plate pressure-sinkage model was proposed combined with the exponential model and soil mechanics methods. Using the mechanical parameters in Bekker's literature about clayed soil, we found that the error between the ultimate bearing capacity and the corresponding pressure values calculated by Bekker's equation was about from 5% to 21%, and the error between the sinkage index value calculated and the actual value was about from 7% to 36%. Using the relevant research results from Zhuang Jide et al, it showed a very high coefficient between the predicted curve by parameters calculation and the actual curve obtained from the tests in sandy soil. The correlation coefficient between the subsidence curve with the improved model and the curve by calculating in the sandy soil test was 0.9998. The error between the ultimate bearing capacity and the corresponding pressure values calculated by Bekker's equation was about 9.26%, and the error between the sinkage index value calculated and the actual value was about 2.04%. We also used the soil test data of paddy soil in College of Engineering, Nanjing Agricultural University for validation. The deformation index changed with the moisture content. Curve coincidence degree was better comparing the field measured one with the one from the improved model, in which the sinkage index value ranged from 0.32 to 0.52. Therefore, the model is valid and stable for sticky and frictional soils. It is predictable that the load curve can be constructed using the mechanical parameters of soil, which provides a new theoretical idea for solving pressure-sinkage problem. It should be mentioned that, these models are applicable for homogeneous soil but not for layered soil, such as southern paddy fields. Also we find that the sinkage index value of bearing-pressure model is stable, and whether it can be used as soil confined nature eigenvalue is worthy of further discussion. Also, although the improved model has ideal effects in these examples, its specific soil type and boundary conditions still need to be validated.