Determining Soil Stress beneath a Tire: Measurements and Simulations. Issue 3 (24th June 2016)

• Record Type: Journal Article
• Authors: Keller, Thomas; Ruiz, Siul; Stettler, Matthias; Berli, Markus
• Is Part Of: Soil Science Society of America Journal. Volume 80:Issue 3(2016)
• Page Start: 541
• Page End: 553
• Publication Date: 2016-06-24
• Subjects: Soils -- United States -- Periodicals; Soil science -- Periodicals; Periodicals; 631.4973
• DOI: 10.2136/sssaj2015.07.0252
• Languages: English
• ISSNs: 0361-5995

Abstract

Abstract : Core Ideas: Bolling probe pressure is a function of soil mean normal stress and the soil's Poisson ratio. Vertical soil stress could be well described by the Boussinesq solution. Simulations reveal a significant impact of soil properties on mean normal stress. This study measured soil stress underneath a rolling wheel by using two different types of sensors: a load cell type probe referred to as an Arvidsson probe to read vertical soil stress and a fluid inclusion type sensor referred to as a Bolling probe to read soil mean normal stress. Measurements were compared with simulations using a semi‐empirical as well as a finite element model. The latter allowed us to consider a layered soil profile as well as an elastoplastic constitutive relationship in the simulations. In addition, the finite element model was used to quantify the ratio between the Bolling probe pressure and the soil mean normal stress. The Bolling probe pressure was found to be primarily a function of the soil's Poisson ratio, which supports findings from earlier studies. Our results showed better agreement between measurements and simulations for vertical stress (obtained from the Arvidsson probe readings) than for mean normal stress (calculated from the Bolling probe measurements). The finite element simulations revealed that soil properties had little influence on vertical stress, and the distribution with depth of the vertical stress could be well described by the classical Boussinesq solution. However, soil properties had a significant impact on the mean normal stress. The widely used semi‐empirical Fröhlich model performed poorly, which may have been because of an inconsistency in the model assumptions.