Investigating the relationship between horizontal penetrometer resistance and soil precompression stress

Introduction Soil compaction is a serious concern in modern agriculture. Field traffic using machines with high axle loads is likely to compact the soil below the plough layer. Compaction of the subsoil should be avoided since soil productivity is at risk of being reduced, and because the effects are very persistent, perhaps even permanent. Precompression stress is widely applied as a border between elastic and plastic soil deformations under stress application. As long as the stress does not exceed the precompression stress, soil deformation is expected to be recoverable after the removal of stress (i.e. passage of the tire). The precompression stress is conventionally determined using the standard procedure of Casagrande from the log stress – void ratio curve resulting from confined uniaxial compression of intact soil samples taken at the field. With advances in the technology of precision agriculture, site specific management of machinery traffic is under focus by researchers. Field mapping of soil precompression stress would allow for site specific modifying the machine parameters (e.g. the tire inflation pressure) to control the applied stress below the precompression stress. For example, where the soil moisture is high, a decrease in tire inflation pressure increases the soil-tire contact area and thus decreases the severity of stress propagation in soil. However, the conventional method of estimating the precompression stress is time-consuming, labor intensive thus not suitable for mapping applications. Horizontal penetrometer is a popular device for on-the-go measuring of soil strength. Horizontal penetrometer resistance is an attractive measurement because it is relatively simple, fast and cheap, can be carried out on-the-go, thus yielding spatial information with a high resolution. A multi-tip horizontal penetrometer would also allow for discrete-depth measuring the soil strength. As penetrometer resistance and precompression stress are both measures of soil strength, it was hypothesized that they are highly correlated. Therefore, the relationship between precompression stress (σpc) and horizontal penetrometer resistance (PR) was investigated in a wide range of soil textures. This would suggest an alternative for on-the-go measurement of σpc for field mapping and site-specific management of soil trafficability.
Materials and Methods Field measurements were conducted in different soil textures in Switzerland. The clay content of the soils varied from 189 to 584 g kg-1. Horizontal penetrometer resistance was measured at 0.25 m depth at a traveling speed of 0.25 m s-1. Cylindrical core samples were taken at the local minima and maxima of PR along the transects. Cone index measurements were also performed to a depth of 0.5 m at the points of core samples. The samples were subjected to stepwise compression stresses by an Oedometer and the resulting deformation was recorded. The void ratio was calculated with particle density and bulk density at the end of each stress step. The precompression stress was estimated at the point of maximum curvature of log stress- void ratio with fitting Gompertz function. Correlation and regression analyses were performed in SAS software.
Results and Discussion The results showed that the Gompertz function explains well the void ratio versus log of stress for different soil textures. The Gompertz parameters were characterized with respect to soil physical characteristics. Precompression stress decreased with increasing soil moisture, soil clay and organic matter content. A relatively strong correlation was found between σpc and PR (R2= 0.47, RMSE= 15.4 kPa) which was significantly improved (R2= 0.59, RMSE= 13.7 kPa) with the effect of soil water content. The high scatter of the relationship between precompression stress and horizontal penetrometer resistance was discussed to be likely due to the difference between the soil failure mechanisms around a penetrating tip and under uniaxial compression. A comparison of different compaction tests (e.g. semi-confined and plate sinkage) with PR may suggest stronger correlations. A strong correlation (R2= 0.6) was also found between PR and cone index (CI). CI was found to be larger than PR for all the soils.
Conclusion It was concluded that the soil compaction characteristic (log stress versus void ratio) is strongly governed by the soil initial void ratio with an increase of precompression stress with decreasing the initial void ratio. In the range of variations tested, the relationship between PR and σpc was not affected by soil texture (clay content). The study suggests that measurement of PR can be a fast alternative for mapping of soil precompression stress by compensating for the effect of soil moisture (by e.g. a dielectric sensor). therefore, a combined horizontal penetrometer needs to be employed.