Search results

Sliding wear is a common phenomenon in the iron ore handling industry. Large-scale handling of iron ore bulk-solids causes a high amount of volume loss from the surfaces of bulk-solids-handling equipment. Predicting the sliding wear volume from equipment surfaces is beneficial for efficient maintenance of worn equipment. Recently, the discrete element method (DEM) simulations have been utilised to predict the wear by bulk-solids. However, the sensitivity of wear prediction subjected to DEM parameters has not been systemically investigated at single particle level. To ensure the wear predictions by DEM are accurate and stable, this study aims to conduct the sensitivity analysis at the single particle level.

In this research, pin-on-disc wear tests are modelled to predict the sliding wear by individual iron ore particles. The Hertz–Mindlin (no slip) contact model is implemented to simulate interactions between particle (pin) and geometry (disc). To quantify the wear from geometry surface, a sliding wear equation derived from Archard’s wear model is adopted in the DEM simulations. The accuracy of the pin-on-disc wear test simulation is assessed by comparing the predicted wear volume with that of the theoretical calculation. The stability is evaluated by repetitive tests of a reference case. At the steady-state wear, the sensitivity analysis is done by predicting sliding wear volumes using the parameter values determined by iron ore-handling conditions. This research is carried out using the software EDEM^® 2.7.1.

Numerical errors occur when a particle passes a joint side of geometry meshes. However, this influence is negligible compared to total wear volume of a wear revolution. A reference case study demonstrates that accurate and stable results of sliding wear volume can be achieved. For the sliding wear at steady state, increasing particle density or radius causes more wear, whereas, by contrast, particle Poisson’s ratio, particle shear modulus, geometry mesh size, rotating speed, coefficient of restitution and time step have no impact on wear volume. As expected, increasing indentation force results in a proportional increase. For maintaining wear characteristic and reducing simulation time, the geometry mesh size is recommended. To further reduce simulation time, it is inappropriate using lower particle shear modulus. However, the maximum time step can be increased to 187% T_R without compromising simulation accuracy.

The applied coefficient of sliding wear is determined based on theoretical and experimental studies of a spherical head of iron ore particle. To predict realistic volume loss in the iron ore-handling industry, this coefficient should be experimentally determined by taking into account the non-spherical shapes of iron ore particles.

The effects of DEM parameters on sliding wear are revealed, enabling the selections of adequate values to predict sliding wear in the iron ore-handling industry.

The accuracy and stability to predict sliding wear by using EDEM^® 2.7.1 are verified. Besides, this research accelerates the calibration of sliding wear prediction by DEM.

Bulk material-handling equipment development can be accelerated and is less expensive when testing of virtual prototypes can be adopted. However, often the complexity of the interaction between particulate material and handling equipment cannot be handled by a single computational solver. This paper aims to establish a framework for the development, verification and application of a co-simulation of discrete element method (DEM) and multibody dynamics (MBD).

The two methods have been coupled in two directions, which consists of coupling the load data on the geometry from DEM to MBD and the position data from MBD to DEM. The coupling has been validated thoroughly in several scenarios, and the stability and robustness have been investigated.

All tests clearly demonstrated that the co-simulation is successful in predicting particle–equipment interaction. Examples are provided describing the effects of a coupling that is too tight, as well as a coupling that is too loose. A guideline has been developed for achieving stable and efficient co-simulations.

This framework shows how to achieve realistic co-simulations of particulate material and equipment interaction of a dynamic nature.

The tensile test is one of the fundamental experiments used to evaluate material properties. Simulating a tensile test can be a replacement of experiments to determine mechanical parameters of a continuous material. The paper aims to discuss these issues.

This research uses a new approach to model a tensile test of a high-carbon steel on the basis of discrete element method (DEM). In this research, the tensile test specimen was created by using a DEM packing theory. The particle-particle bond model was used to establish the internal forces of the tensile test specimen. The particle-particle bond model was first tested by performing two-particle tensile test, then was adopted to simulate tensile tests of the high-carbon steel by using 3,678 particles.

This research has successfully revealed the relationships between the DEM parameters and mechanical parameters by modelling a tensile test. The parametric study demonstrates that the particle physical radius, particle contact radius and bond disc radius can significantly influence ultimate stress and Young’s modulus of the specimen, whereas they slightly impact elongation at fracture. Increasing the normal and shear stiffness, the critical normal and shear stiffness can enable the increase of ultimate stress, however, up to maximum values.

To improve the particle-particle bond model to simulate a tensile test for high-carbon steel, the damping factors for compensating energy loss from transition of particle motions and failure of bonds are required.

This work reinforces the knowledge of applying DEM to model continuous materials.

This research illustrates a new approach to model a tensile test of a high-carbon steel on the basis of DEM.

The macroscopic properties of dried sand can be correctly modelled when the accurate determination of the microscopic properties is available. The microscopic properties between the particles such as the coefficients of rolling (µ _r) and sliding (µ _s), are numerically determined in two different ways: with and without considering the fluid effect. In an earlier study, the microscopic properties were determined by discrete element method (DEM) and without considering the air effect on the macroscopic properties such as the Angle of Repose. The purpose of this paper is to recalibrate the microscopic properties through a coupling between the DEM and computational fluid dynamics (CFD).

The first step is dedicated to the calibration of the CFD-DEM model through modelling a single particle sedimentation within air, water, and silicon oil. The voidage and drag models, the grid size ratio (D/dx), the domain size ratio (W/D), and the optimum coupling interval between the CFD and DEM were investigated through comparing the CFD-DEM results with the analytical solution and experimental data. The next step is about modelling an Hourglass with the calibrated CFD-DEM model to recalibrate the µ _r and µ _s of dried sand particles.

It was concluded that the air has a minor effect on the macroscopic properties of the dried sand and the µ _r and µ _s that were obtained with the DEM can be utilized in the CFD-DEM simulation.

Utilizing the granulometry of dried quartz sand in the calibration process of the CFD-DEM method has raised the possibility of using the µ _r and µ _s for other applications in future studies.