Published on: November 30, 2022
Discover why Rocky DEM’s SPH is an efficient methodology to simulate solid-fluid applications and to solve problems that include large solids.
Computer simulation has evolved dramatically, and after years of research and development, Rocky DEM can perform simulations with hundreds of millions of particles, including flexible particles and real particle shapes. Now, Rocky is taking another huge step forward in building the power of simulation by the offering a smoothed-particle hydrodynamic tool with a fully coupled SPH-DEM, real shape representation, and multi-GPU support.
A particle-based, mesh-free (Lagrangian) method, smoothed-particle hydrodynamics was developed to provide an enhanced solution for solid-fluid applications with free surface flows.
SPH captures flow dynamics by discretizing flow into a set of fluid elements, each with mass, momentum, and energy. Each element interacts with particles, wall boundaries, and body forces via the integration of Newton’s second law. Inter-element forces derive from Navier-Stokes equations. SPH evaluates fluid quantities using a kernel function to compute smooth fields for local variables.
Rocky uses an uncommon approach to compute the force transfer from solid boundaries onto the smoothed-particle hydrodynamics elements—a DEM-style approach. The SPH element experiences normal and tangential forces. This works well with complex boundary shapes and improves computational efficiency and memory use to perform large simulations. There’s no need to construct solid boundary particles in the simulation: Rocky’s motion kernel provides full motion repertoire.
Rocky uses a resolved method for fluid-particle interaction, not relying on closure models for the fluid-particle momentum or heat transfer modeling. DEM particles are resolved by SPH elements. Both SPH and DEM use particle methods, and development of SPH for solving fluid applications is a natural outcome of flexible software development environment used by Rocky.
The interior smoothed-particle hydrodynamics elements have an associated artificial mass, density, and velocity that can satisfy the no-slip condition at the particle surface. SPH elements are placed in the interior of the DEM elements, and forces and moments applied to them are transferred to DEM particles. Indeed, you can bring your own particle shape, but the SPH elements must be smaller than the DEM particles in order for them to fit, which is a limitation of this method.
To handle large simulations, Rocky is powered by a computationally efficient GPU-based solver. As an example, a simulation was tested with 1 million SPH elements, and the GPU was only 15% loaded.
Rocky’s native implementation of SPH-DEM creates a powerful simulation software for modeling liquid-solid applications in multiple industries, including wave attenuation, wet griding, galvanization, coating, milling, etc. Below you can check out a couple of examples.
SPH-DEM coupling predicted wave behavior and design concrete blocks to protect a marina in Crete. These tetrapod-shaped rocks dissipate the incoming wave forces, and by interlocking them together, they’re less likely to become displaced. The simulation uses real-shaped DEM particles that can move.
IWF at the University of Berlin and Fraunhofer IPK used SPH-DEM to accurately predict workpiece motion and collision statistics in a centrifugal disc that contained wet and dry media. Understanding the finishing of small components with complex geometries is a major industrial challenge. Although, in this case, using real-shaped grinding particles with SPH-DEM modeling provided the insight they needed.
An SPH-DEM approach evaluated and improved a grate discharge mill. These SAG mills usually operate wet, with water and a fines mix forming a slurry phase. The design of the pulp discharge mechanism is critical for maintaining grinding efficiency, as well as minimizing wear and power draw. The complex simulation included a million triangles, 97 million smoothed-particle hydrodynamics elements, and 201,000 DEM particles. Ore was modeled as polyhedrons with PSD, and balls were modeled as spheres with PSD. SPH allowed engineers to find a solution for this complex problem.
For more information about how Rocky’s new SPH-DEM capability can help you, contact us today.
CAE Specialist at Rocky DEM - ESSS, D.Sc.
Lucilla holds a BE (Chemical) undergraduate degree, an M.Sc. in Chemical Engineering and a Ph.D. in Nuclear engineering from the Federal University of Rio de Janeiro. She focuses on applying Computational Fluid Dynamics and Discrete Element Method to solve engineering problems. She is a CAE Specialist at ESSS, responsible for technical sales development, post sales, and the support of Rocky DEM.
