Knowledge of soil stress induced by vehicle traffic is a premise for assessment of soil compaction risks, but obtaining absolute values of soil stress is not trivial because stress transducer probe readings are influenced by probe characteristics and soil properties. However, little research has quantified the magnitude of the deviations between measured and “true” stress (defined as soil stress in the absence of a probe). The objectives of this study were to use a finite element model to quantify how stress readings by cylindrical probes are influenced by probe design (material properties, dimensions), probe spacing in multi-probe set-ups, probe installation depth, and soil mechanical behaviour. We show that soil stress is typically overestimated by transducer probes. Overestimation decreased with probe diameter to height ratio, and increased with the ratio of probe to soil elastic modulus. Overestimation was smaller but dependent on soil strength under plastic soil conditions and higher but independent of soil strength under elastic soil behaviour. Probes interfere with each other when the vertical gap between adjacent probes is closer than about three times probe diameter. Moreover, we found that measured stress is influenced by soil depth and size of the loaded area. Some of the factors influencing stress readings are easily controlled (e.g. probe material properties, dimensions and spacing). However, other sources of deviations between measured and true soil stress are more difficult to account for. Our findings show that the ratio of measured to true soil stress is dependent on soil deformation behaviour as well as on relationships between probe depth, radius of loaded area and ratio of soil stress to applied surface stress implies that the ratio of measured to true stress is not constant when measuring soil stress at various depths in situ under field conditions. We therefore suggest that future studies should especially address how stress readings are affected by soil mechanical behaviour, soil depth, the size of the loaded area, and interactions among these factors.

车辆行进引起的土壤应力的知识是评估土壤压实风险的前提，但是获得土壤应力的绝对值并非易事，因为应力传感器的读数受探头特性和土壤特性的影响。但是，很少有研究量化所测应力与“真实”应力（定义为在没有探头的情况下的土壤应力）之间的偏差大小。这项研究的目的是使用有限元模型来量化圆柱探针的应力读数如何受到探针设计（材料特性，尺寸），多探针设置中的探针间距，探针安装深度以及土壤机械行为的影响。我们表明，换能器探头通常会高估土壤压力。高估随着探针直径与高度之比的降低而降低，并随探针与土壤弹性模量的比值增加。高估值较小，但取决于可塑土壤条件下的土壤强度，高估值较高，但与弹性土壤行为下的土壤强度无关。当相邻探针之间的垂直间隙小于探针直径的大约三倍时，探针会相互干扰。此外，我们发现测得的应力受土壤深度和负荷区大小的影响。可以轻松控制影响应力读数的某些因素（例如，探针材料的特性，尺寸和间距）。然而，更难以解释实测土壤压力与真实土壤压力之间偏差的其他来源。我们的研究结果表明，实测土壤应力与真实土壤应力的比率取决于土壤变形行为以及探头深度之间的关系，在野外条件下原位 因此，我们建议，未来的研究应特别关注应力读数如何受到土壤机械行为，土壤深度，加载区域的大小以及这些因素之间的相互作用的影响。