Bulk Engineering, Technical articles, White Papers

The right tool for the job

Coupled SPH and DEM modelling of a cohesive material discharged from a conveyor belt. (a) Experiment, (b) simulation.

Dr Daniel Grasser and Dr David Bradney, from TUNRA Bulk Solids, provide an overview of the latest numerical simulation tools to solve bulk material handling problems and to optimise designs.

Bulk materials handling often includes transportation and processing of raw materials. Discrete Element Method (DEM) modelling is a numerical simulation tool used to simulate the bulk particle flow. Each particle is modelled as a discrete element and the interactions of millions of particles represent the bulk particle flow occurring in a mining application. Once calibrated, DEM is a predictive tool and provides valuable insights into optimisation potential, which would not be achievable without numerical methods. However, to use DEM as a predictive tool to assess the particle flow, calibration and a good understanding of this simulation tool are essential. 

TUNRA Bulk Solids has decades of experience in applying and calibrating DEM models. These models can be applied to predict the bulk particle flow of cohesionless (dry) and cohesive (wet, fines) materials. The particle flow behaviour such as the motion of the bulk particle flow including the particle velocity, for example inside a transfer chute, can be predicted and measured. Resulting from this, important conclusions such as design recommendations for chutes and bins, risk of blockage, recommendation on dust mitigation and strategies for increasing the particle throughput can be drawn. Moreover, options to reduce wear (qualitatively and quantitatively) can also be outlined.

DEM modelling of a chute. The colour indicates the particle velocity (Blue=slow, red=fast).
DEM modelling of a chute. The colour indicates the particle velocity (Blue=slow, red=fast). Image: TUNRA Bulk Solids

From this, the service life can be prolonged and unscheduled maintenance prevented; hence, the efficiency of the bulk materials handling operations can be increased. Importantly, applying a calibrated DEM model can be significantly cheaper and faster than experimental trials, and reduces the risks commonly associated with field tests.

Calibration and validation

Appropriate calibration of the Discrete Element Method (DEM) model is essential. In addition to qualitative site observations, a validation based on quantifiable measurements is crucial. Examples of calibration tests used at TUNRA Bulk Solids include the angle of repose, dynamic angle of repose, dynamic adhesion drop test, shear box and inclining wall friction test.

These tests are used as calibration and validation and represent different modes of particle flow. For example, the results of the shear box test are determined by the particle-to-particle interactions, where the results of the inclining wall friction test are mainly governed by the particle-to-surface interactions, where steeper shear angles are caused by cohesive materials and steeper wall angles by higher friction parameters.

DEM modelling of the Angle of Repose test. (a) Before, (b) during, and (c) after the test. Image: TUNRA Bulk Solids

The frictional parameters are physical parameters, while the shape and size of the particles are often simplified and upscaled due to computational restraints. For example, the fines fraction is often simplified by larger spherical particles and modelled in combination with a cohesive model. For larger rock particles, a good implementation of the shape (non-spherical) is important and cohesion models are not applied. Careful application of cohesion models and other models such as the rolling resistance (RR) model are important for a good calibration. This allows achieving meaningful results based on validated models in reasonable computational times.

Additionally, DEM can be coupled with smoothed particle hydrodynamics (SPH) in case the bulk flow behaviour of viscous and thixotropic materials needs to be predicted.

Smoothed particle hydrodynamics

SPH has found a wider application in the bulk materials handling sector in the last few years. The SPH method is a Lagrangian mesh-free approach that can be coupled with DEM modelling. This is especially useful when the bulk flow behaviour of two-phase materials, for example viscous slurries such as a mixture of water and particle fines, are modelled. The SPH-DEM coupling is a two-way approach coupling the momentum transfer between the fluid and particles. In detail, each SPH element has a mass, momentum, and a resulting energy, interacting with other elements. Therefore, it is important that the SPH elements (representing the liquid phase fraction) are at least three-times smaller than the DEM particles (representing the coarse fraction with discrete particles).

Coupled SPH and DEM modelling of an abrasive two-phase (water and sand) slurry inside the TUNRA impeller wear test. (a) Side view, (b) bottom view. Image: TUNRA Bulk Solids

In addition to the calibration techniques mentioned in the DEM section, additional calibration experiments need to be conducted for SPH-DEM. For example, quantifying the thixotropic behaviour of a cohesive slurry during mixing, under different pressures and under vibrations. Once calibrated, the SPH-DEM model can be used as a predictive tool to forecast the bulk flow behaviour of cohesive two-phase materials. This gives additional confidence when designing and reviewing transportation on conveyor belts, chutes and storage bins used to handle viscous two-phase slurries often possessing a complex flow behaviour.

Additionally, the SPH-DEM coupling can be applied to model the abrasion of wear liners transporting highly abrasive slurries (for example, water mixed with sand). Wear rates can be scaled based on experimental investigations and hence SPH-DEM can be applied as a predictive tool for service life estimations. However, it is important to highlight that currently SPH-DEM models require relatively high computational efforts, when compared to DEM models.

Computational fluid dynamics

As materials transition from solid and semi-solids (slurries) into fluid-governed interactions, other computational tools exist to provide further insight into design problems. Computational Fluid Dynamics (CFD) is used to simulate and analyse the behaviour of fluids (liquids and gases) and their interactions with solid surfaces. In the area of bulk materials handling, CFD is invaluable for modelling and predicting the behaviour of airflows and fluids within equipment such as pneumatic conveying systems and loading/discharge areas around transfer points. 

This simulation capability is particularly useful for understanding and optimising dust considerations when handling drier materials and may be used in the assessment of ventilation systems, or in fundamental design to optimise plant geometry to passively reduce dust generation.

CFD allows designers to model airflow patterns within various equipment. By simulating different configurations, engineers can identify the most effective designs to ensure proper airflow and gain insight into the dispersion of dust particles, enabling the design of more effective dust control and extraction systems. By understanding how dust behaves in different scenarios, engineers can implement measures to minimise dust spread and improve air quality within facilities.

The design of transfer points is critical for ensuring the efficient flow of materials. This is particularly relevant for transfer chutes, where abrupt changes in the trajectory of the material stream and large free-fall areas can result in significant air entrainment and dust generation. Advancements in computational resources and the integration of CFD with other simulation methodologies, such as DEM, have enhanced the ability to analyse complex bulk material flows. When combined with CFD, this integrated approach can simulate:

  • Air entrainment from falling material streams: The interaction between air and falling materials can significantly affect flow patterns and efficiency. Coupled CFD-DEM simulations can model this interaction, providing insights into how air entrainment impacts the overall flow.
  • Flow dispersion and collision: Understanding how particles disperse in a flowing system can have a significant impact on the amount of dust generated within a system. Areas of steep impact with large impact angles generally result in the largest release of dust. With CFD-DEM analysis, it is possible to better quantify the effects of different designs to minimise dust.

By leveraging CFD and CFD-DEM simulations, engineers can optimise bulk material handling systems, ensuring efficient operation, minimising dust generation, and improving safety and environmental conditions within facilities.

CFD-DEM modelling considerations

The accuracy of CFD predictions heavily depends on the quality of the input data, including fluid properties, boundary conditions, geometry discretisation, and material characteristics. Errors or approximations in these inputs can lead to inaccurate results, impacting the reliability of the conclusions drawn from the simulations. Therefore, validation with on-site behaviour and/or experimental data is typically required and recommended for complex problems to ensure the accuracy of the models.

When considering coupled DEM-CFD problems, the following additional considerations are also important: for the scale of most industry-based problems, at present the smallest DEM particle sizes may be in the order of 10 – 20 mm to permit  DEM simulations to effectively solve. In contrast, dust particles may be well under 1 mm in size, ranging down to µm in size. It is therefore important to consider the effect of this potential mismatch of scale on the interaction between the air and particles. For problems with larger materials sizes, the influence of the particles on the air becomes more closely aligned to reality, whereas, for fine dust particles, this difference becomes more significant, and may indicate that coupled models are less appropriate or may require more traditional multiphase-focused CFD techniques.

Finite element analysis

Up to this point, the discussed analysis techniques have primarily focused on understanding the flow and movement of bulk materials. However, the effect that bulk materials have on the structures they interact with is also of high importance to the safe and efficient design of bulk material handling and storage systems.Finite Element Analysis (FEA) is a computational technique used to simulate and analyse the stresses and deformations that result from loads applied to structures. Complex physical systems are analysed by dividing these structures into smaller, simpler parts called elements. This method enables detailed analysis of how different loads, including material loads, affect structures.

In the field of bulk material handling, FEA addresses a wide array of problems, including evaluating the structural capacity under bulk material-induced loads on structures like the walls of hoppers, bins, or silos. It can also be used to assess more complex dynamic phenomena, such as quaking in soils or train load-out bins. The tools available within FEA also allow for the analysis of effects such as vibration, dynamic movement, and thermal considerations. This enables an in-depth understanding of the impact that bulk materials can have on the structures they interact with.

Related stories:

The bulk material loads applied in an FEA model can originate from various sources, including hand calculations based on relevant standards and coupling with other numerical techniques like CFD, SPH, or DEM. Coupling techniques such as DEM with FEA allows for the development of complex load cases. However, it is important to note that the accuracy of FEA predictions heavily depends on the quality of the input data, including accurate material properties and boundary conditions. For techniques such as DEM, assumptions around particle stiffness can significantly affect the magnitude of boundary force predictions. Misestimations of the output from one simulation technique can lead to incorrect predictions, affecting reliability of results.

To address these challenges, it is recommended to use alternative methods such as fundamental approaches or on-site measurements to calibrate and validate the load inputs used in an FEA model. This helps to quantify potential modelling unknowns and ensure more reliable and accurate predictions. Integrating FEA with other analysis techniques provides a powerful toolset for understanding and optimising the interaction between bulk materials and the structures they affect, leading to safer and more efficient bulk material handling systems.

Overview of simulation tools

Bulk materials handling problems encompass a wide array of problem types and material conditions, and it is therefore important to select the most suitable simulation method to efficiently and effectively understand each problem. Fortunately, tools now exist to simulate bulk material through the full array of problem types; from dry, dusty material through to granular or cohesive flows, and even very wet slurry applications, and their effect can be assessed both in terms of material-material interaction as well as their impact on the structure and equipment they interact with. Table 1 provides a summary of these considerations.

Table 1

Ultimately, the most appropriate choice of simulation tool depends on the type of bulk material to be analysed, the particle size of the material being simulated, and the state in which the material is being handled, and the desired level of understanding of the problem. TUNRA Bulk Solids has decades of experience in both physical lab-based materials testing, and advanced simulation techniques, and can help navigate the challenges of selecting the most appropriate method for each unique bulk materials handling problem.  

Send this to a friend