Elsevier

Soil and Tillage Research

Volume 213, September 2021, 105123
Soil and Tillage Research

Determination of discrete element model parameters for a cohesive soil and validation through narrow point opener performance analysis

https://doi.org/10.1016/j.still.2021.105123Get rights and content

Highlights

  • Input parameters for discrete element method were calibrated for a cohesive soil.

  • New criteria were introduced to predict furrow profiles from particle displacement.

  • Furrow and soil disturbance parameters were predicted with 1 %–19 % relative errors.

  • Draught prediction errors were ≤31 % and 5 % for straight and bentleg openers.

Abstract

The discrete element method (DEM) is a powerful tool that can be used to predict soil disturbance and soil cutting forces to assist design optimisation of soil cutting tools. In this study, DEM input parameters were calibrated to model a cohesive soil (Black Vertosol of southern Queensland, Australia) using the hysteretic spring contact model, coupled with linear cohesion model, and nominal particle radius of 5 mm. DEM simulations were validated using experimental results for the effects of opener rake angle and cutting edge chamfer, and bentleg opener shank offset on no-tillage narrow point opener performance. Overall, DEM results closely agreed with experimental results and exhibited similar trends. By using particle displacement analysis to predict loosened furrow boundary, most predictions of furrow parameters namely furrow cross-sectional area, furrow width, and critical depth had relative errors ranging from 1 % to 19 %. Lateral soil throw was predicted with relative errors of 0.2 %–9 %, except for the straight opener with 45° rake angle (-32 %). Ridge height was over predicted in all cases due to larger DEM particles than actual soil particles used. Relative errors of 20 %, 22 %, -31 %, and -5 % in draught were recorded for the straight openers with 90° (blunt), 90° (chamfered), and 45° rake angles, and the bentleg opener, respectively. These results show that DEM and the input parameters determined to model the cohesive soil of this study can be used to reliably assess furrow opener performance.

Introduction

Optimisation of furrow opener and tillage tool design is conventionally done through repeated prototyping and experimental (soil bins and field) evaluations, which could be laborious and resource intensive (Shmulevich et al., 2007; Chen et al., 2013; Ani et al., 2018). Several analytical and numerical models for predicting soil forces and soil-tool interaction have been developed and used. The analytical models, which are generally derived from the universal earthmoving equation, are used for soil force predictions by assuming simple and approximate geometric profiles such as wedges and crescents to model soil failure patterns (Hettiaratchi and Reece, 1967; Godwin and Spoor, 1977; McKyes and Ali, 1977; Godwin and O’Dogherty, 2007). These models consider soil failure as bulk soil movement without accounting for interactions between individual soil particles. Thus, they fall short of what happens in practice, where soil failure is partly bulk and partly particle movement (Shmulevich et al., 2007; Chen et al., 2013). Finite element modelling (FEM) and computational fluid dynamics (CFD) have also been employed to numerically predict soil failure patterns and soil forces with some degree of accuracy (Fielke, 1999; Karmakar et al., 2009). However, both FEM and CFD are continuum methods, which also consider soils as bulk materials, disregarding the discrete nature of soil particles and aggregates. Furthermore, these methods fail to accurately model variations in soil structure, particle flow, and mixing and translocation of individual soil particles and aggregates (Shmulevich et al., 2007).

The discrete element method (DEM) is increasingly being used for modelling and predicting relative movement and forces among particles of granular materials (such as gravels, grains, soils, and powder) and with objects such as walls and machine parts (Cundall and Strack, 1979; Chen et al., 2013). Discrete Element Method can be applied to model the discrete nature of soil particles and aggregates using calculations based on Newton’s Second Law of motion and mechanical contact force models to define particle-particle and particle-tool interactions (Chen et al., 2013; Peng, 2014). To accurately predict particle movements and forces with DEM, accurate input parameters need to be determined through a calibration process (Ono et al., 2013). Discrete Element Method has been used to simulate operations such as tillage and furrow opening and to analyse particle movement and soil cutting forces, achieving results that closely agree with experimental observations as detailed below.

Using DEM, Tamas et al. (2013) predicted draught force with relative errors of -4 to -12 %. Relative error values below 4 % were predicted for four subsoilers by Bo et al. (2016). Chen et al. (2013) observed about 4–31% relative error between draught of experimental and DEM results. Bravo et al. (2014) found 9 % and 2.4 % errors in cultivator tool draught. DEM also closely predicted increase in draught with increasing depth with relative error values ranging from about 3 %–15 % (Li et al., 2014). Ucgul et al. (2014a) observed a linear increase in draught force against sweep tine width measured experimentally and predicted using DEM with maximum relative error of 8 %. Similarly, a non-linear increase in vertical force against width was predicted with maximum relative error of 13.7 %. Prediction of the effect of rake angle on soil forces followed similar trend and had relative error of 11.6 % and 15.2 % for draught and vertical forces, respectively. Murray (2016) estimated average relative error of 1.9 % for draught and 50.7 % for vertical force with a flat single disc opener.

Though most attention in DEM tillage and furrow opener studies has been on the prediction of soil forces, a number of studies (e.g., Tamas et al., 2013; Barr et al., 2018, 2019; Barr et al., 2020) have also looked at the mode of soil failure and soil disturbance or soil-tool interaction. In these studies, DEM was used to predict soil rapture and crack propagation. Furrow profile, soil movement, and various soil disturbance parameters have been predicted with DEM. Barr et al. (2018) predicted loosened furrow cross-sectional area, furrow width, dip area, furrow backfill, ridge height, and lateral soil throw in DEM with relative error values of 9 %, 26 %, 14 %, 0.8 %, 16 %, and 9 %, respectively. Chen et al. (2013) estimated relative error ≤3 % and 4 % in furrow cross-sectional area and width, respectively. Murray (2016) found average absolute relative errors of about 11 % and 15 % in lateral soil throw for a disc and hoe openers, respectively. Studies (e.g., Barr et al., 2018; Hang et al., 2018) have also revealed DEM’s ability predict crescent failure above critical depth and lateral failure below it for straight tine tools. DEM is able to simulate greater soil movement with wing attachment to tined soil cutting tools (Bo et al., 2016) and greater soil upheaval with low rake angles (Barr et al., 2018).

Most of the DEM simulations applied to the study of soil dynamics (e.g., Mak and Chen, 2014; Kotrocz et al., 2016; Li et al., 2016; Milkevych et al., 2018; Ucgul and Saunders, 2020) and those reviewed above have dealt with cohesionless or low cohesion soils (cohesion ≤36 kPa), with the exception of Bravo et al. (2014). Bravo et al. (2014) used DEM to model highly cohesive clay soil (Vertisol) with cohesive strength of up to 125 kPa. Therefore, more DEM studies on highly cohesive soils such as Vertosols are needed to expand the understanding of the applicability of DEM to the study of soil dynamics.

The objectives of this study were to: (1) determine suitable DEM input parameters for a Black Vertosol (Vertisol in the USDA Soil Taxonomy) of southern Queensland, Australia (Isbell, 2002), and (2) validate the DEM model of the above soil using experimental data derived from soil disturbance and cutting force measurements of selected narrow point opener geometries specially designed for no-tillage farming.

Section snippets

Determination of DEM input parameters

DEM simulations were carried out using the software EDEM, 2020.0 (Academic Edition, Version: 6.0.0) working on an HPE ProLiant XL190 r computer with Intel® Xeon® Gold 5120 CPU @ 2.20 GHz and 756GB RAM. Hysteretic spring contact model, coupled with linear cohesion model, was adopted to model both particle-particle and particle-tool interactions as suggested by Ucgul et al. (2014b).

Prediction of furrow and surface profiles

Fig. 14 presents DEM-predicted furrow profiles based on particle count and displacement analyses, DEM-predicted surface profiles from surface particle positions, and surface and furrow profiles from the field experiments reported by Aikins et al. (2021). The furrow and surface profiles created by the 45° rake angle straight tine opener (R45) and the bentleg opener (BL) in the DEM simulations had a similar shape as those measured in the field experiment. Both openers had a leading edge or foot

Conclusions

Input parameters used in the DEM simulations for describing the properties of a highly cohesive soil (Black Vertosol of southern Queensland, Australia), and soil-soil and soil-tool interactions were determined. The analysis undertaken took into account both soil cohesion and adhesion. A novel approach for comparing experimental data and DEM results of angle of repose test on cohesive soil was developed and used to calibrate soil-soil coefficients of static and rolling friction. The DEM model

Declaration of Competing Interest

The authors report no declarations of interest.

Acknowledgements

The authors are grateful to the Centre for Agricultural Engineering at the University of Southern Queensland (Toowoomba, QLD, Australia), Kwame Nkrumah University of Science and Technology (Kumasi, Ghana), CSIRO Agriculture and Food (Canberra, ACT, Australia), and CLAAS Stiftung (Harsewinkel, Germany) for financial and operational support to conduct this research. The technical assistance of Mr. Anthony Nolan of Dalby State High School, Queensland, Australia and Dr. Francis Gecenga, Mr. Richard

References (59)

  • C.J. Coetzee

    Calibration of the discrete element method and the effect of particle shape

    Powder Technol.

    (2016)
  • J.M. Fielke

    Finite element modelling of the interaction of the cutting edge of tillage implements with soil

    J. Agric. Eng. Res.

    (1999)
  • S.Y. Fu et al.

    Effects of particle size, particle/matrix interface adhesion and particle loading on mechanical properties of particulate–polymer composites

    Compos. Part B Eng.

    (2008)
  • R.J. Godwin

    A review of the effect of implement geometry on soil failure and implement forces

    Soil Tillage Res.

    (2007)
  • R.J. Godwin et al.

    Integrated soil tillage force prediction models

    J. Terramech.

    (2007)
  • R.J. Godwin et al.

    Soil failure with narrow tines

    J. Agric. Eng. Res.

    (1977)
  • C. Hang et al.

    Discrete element simulations and experiments of soil disturbance as affected by the tine spacing of subsoiler

    Biosyst. Eng.

    (2018)
  • D.R.P. Hettiaratchi et al.

    Symmetrical three-dimensional soil failure

    J. Terramech.

    (1967)
  • S. Karmakar et al.

    Experimental validation of computational fluid dynamics modeling for narrow tillage tool draft

    J. Terramech.

    (2009)
  • K. Kotrocz et al.

    Numerical simulation of soil-cone penetrometer interaction using discrete element method

    Comput. Electron. Agric.

    (2016)
  • B. Li et al.

    Modeling of soil-claw interaction using the discrete element method (DEM)

    Soil Tillage Res.

    (2016)
  • J.D.M. Marsh et al.

    Effects of heavy additions of organic residues on physical characteristics of three soil types in Queensland Australia

    Soil Tillage Res.

    (1991)
  • E. McKyes et al.

    The cutting of soil by narrow blades

    J. Terramech.

    (1977)
  • V. Milkevych et al.

    Modelling approach for soil displacement in tillage using discrete element method

    Soil Tillage Res.

    (2018)
  • I. Ono et al.

    Investigation of elemental shape for 3D DEM modeling of interaction between soil and a narrow cutting tool

    J. Terramech.

    (2013)
  • L. Qi et al.

    Simulations of soil flow properties using the discrete element method (DEM)

    Comput. Electron. Agric.

    (2019)
  • T. Roessler et al.

    DEM parameter calibration of cohesive bulk materials using a simple angle of repose test

    Particuology

    (2019)
  • I. Shmulevich et al.

    Interaction between soil and a wide cutting blade using the discrete element method

    Soil Tillage Res.

    (2007)
  • A. Solhjou et al.

    Soil translocation by narrow openers with various rake angles

    Biosyst. Eng.

    (2012)
  • Cited by (48)

    View all citing articles on Scopus
    View full text