We urgently need increased quantitative knowledge about stress transmission in real soils that suffer heavy loads of agricultural machinery. 3D measurements of vertical stresses under tracked wheels were performed in situ in an annually ploughed Stagnic Luvisol continuously cropped with small grain cereals. The tests took place in the spring at field capacity when the topsoil had not been tilled for 1 1/2 years. Two Nokian ELS Radial-ply tyres (800/50R34 and 560/45R22.5) were loaded with two specific loads (30 kN and 60 kN), leading to four treatments labelled 800-30, 800-60, 560-30 and 560-60. We used rated tyre inflation pressures for traffic in the field (<= 10 km h(-1) driving speed). Seven load cells were inserted horizontally from a pit with minimal disturbance of soil at each of three depths (0.3, 0.6 and 0.9 m), covering the width of the wheeled area. The position of the wheel relative to the transducers was recorded using a laser sensor. Finally, the vertical stresses near the tyre-soil interface were measured in separate tests by 17 stress transducers across the width of the tyres. The level of maximum stress at 0.3 m depth was related to the surface-related stress expressions like the mean ground pressure and the tyre inflation pressure. The maximum stresses measured at 0.9 m depth were correlated with the wheel load (57 and 60 kPa at 60 IN load: 27 and 25 kPa at 30 kN load) and did not reflect the surface-related stress expressions. Our results show that the use of wide, low pressure tyres (within the technical opportunities available today) has no real effect on the stresses reaching deep subsoil layers. Our results further support the principle behind the elasticity theory. However, if fitting the Sohne model to stress measurements at all three depths, the stresses were underestimated at 0.3 and 0.6 m depth, and overestimated at 0.9 m depth. A fit of the model based on data only at 0.3 m depth indicates that stresses were transmitted nearly without attenuation through the 0-0.3 m soil layer, which cannot be described by the model of Sohne. We thus interpret the poor fit for the total profile as being due to differences in strength for the frequently tilled topsoil and the subsoil. Our results thus qualitatively confirm the principle of elasticity, but highlight the need to model arable soil as a two-layer system. (C) 2010 Elsevier BM. All rights reserved.
