Aarhus University Seal / Aarhus Universitets segl

Mathieu Lamandé

Models for prediction of soil precompression stress from readily available soil properties

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

Compaction of the subsoil is an almost irreversible damage to the soil resource. Modern machinery exerts high mechanical stresses to the subsoil, and a range of studies report significant effects on soil functions. There is an urgent need for quantitative knowledge of soil strength in order to evaluate sustainability of current field traffic. The aim of this study was to identify the most important drivers of soil precompression stress, σpc, and to develop pedotransfer functions for prediction of σpc. We revisited previously published data on σpc for a silty clay loam soil at a range of soil matric potentials. σpc was estimated from the original stress-strain curves by a novel, numerical method for estimating the stress at maximum curvature, assumingly partitioning the curve into elastic and plastic sections. Multiple regression was used to identify the drivers best describing the variation in σpc data. For the plough layer, σpc increased with bulk density (BD), which explained 77% of the variation. For the subsoil layer just beneath the ploughing depth, the model best describing σpc data included the drivers BD and pF, with pF defined as the log to the negative matric potential. The model was strongly significant with R2=0.90. The same trend was found for three subsoil layers from 0.35–0.95m depth, but the model accounted for only 16% of the variation in σpc. A model involving samples from all soil layers and including BD, pF and soil clay content accounted for 38% of the variation. This model predicted σpc to be constant at pF ~2 across soil clay contents for a given soil BD. For pF < 2, σpc was predicted to be higher for sandy soils than for soils rich in clay. In contrast, σpc increased with clay content for dryer conditions (pF > 2). Model predictions correlated well with measured data in two independent data sets from the literature. However, the predictions were approximately double those of one of the data sets. This may relate to the longer stress application used in laboratory compression tests for these data compared to the other calibration data set and to the procedure used in this study. We encourage further studies of the effect of stress application procedures in compression tests. The prediction equations established in this investigation have to be verified based on measurements of σpc for a range of soil types, soil horizons and soil moisture conditions.
Original languageEnglish
Pages (from-to)115-125
Number of pages11
Publication statusPublished - 2018

See relations at Aarhus University Citationformats

ID: 121245462