RheoCube makes virtual prototyping a reality
The science behind RheoCube
Complex fluids and materials are found in many products, such as perfumes, coatings, or paints. In many cases, their behaviors are not fully understood, so opportunities are lost to improve their performance or use them in new ways. RheoCube simulations unlock this potential by delivering a clear understanding of materials and how they interact.
Explore our areas of research
Simulation models for complex fluids and materials
We develop highly detailed simulation models. They shed light on the origins of rheological behaviour and enable scientists to create virtual equivalents of different materials. To get such deep insight into a complex fluid or material, it’s crucial to understand the meso- and micro-scale physics and structure formation within it.
Ingredient properties and interactions lead to the formation of specific microstructures and dynamics in a complex fluid. These in turn are responsible for specific rheological effects and transport properties (e.g. separation). We can therefore create highly detailed models of ingredients, which account for their mechanical and physical chemical properties and the interactions between them. Our simulations capture how structures emerge across the microscopic and mesoscopic scales. This insight builds an accurate description of a complex fluid’s rheological behavior.
Fluid flow at the macroscale (a scale we can see with our naked eye) is often simulated using computational fluid dynamics (CFD). CFD can simulate flow in e.g. mixers, pipes, rivers, or the atmosphere. However, materials being developed today have remarkable functions that emerge from a hierarchy of physical processes across broad length and time scales.
Take emulsions for example: the behavior of surfactants at interfaces is a microscopic process, while the behavior of emulsion droplets in flow occurs at the mesoscopic level. Simulating such complex fluids and materials therefore requires modeling all the way through the microscopic and mesoscopic scales. This strategy is referred to as “multiscale modeling”. The two core simulation techniques we use in multiscale modeling are Smoothed Particle Hydrodynamics (at the mesoscale), and Molecular Dynamics (at the microscale).
Smoothed Particle Hydrodynamics (SPH)
Our technique of choice for complex fluid simulations on the meso-scale is Smoothed Particle Hydrodynamics (SPH). This method solves the Navier-Stokes equations on a ‘moving grid’ of fluid elements, each represented as a point, with a certain mass corresponding to the material density. Hence this method is mesh-free and is called particle-based, as each point particle represents an actual fluid volume element. This is a departure from conventionally-used Computational Fluid Dynamics (CFD) techniques, which often solve the flow equations on a fixed mesh. The advantage of the particle approach is that we can also examine other components like polymers and surfactants, each with their own chemically-specific properties assigned to blobs of fluid in a simulation.
Our team has pushed the SPH method forward, and undertaken extensive theoretical work to tailor the method to multi-component fluids. The result is ‘Continuous Smoothed Particle Hydrodynamics’, or CSPH. In SPH each blob contains only one component, but in CSPH the blobs can contain multiple components. Components shift between blobs by diffusion. Our development of CSPH is crucial not only for modeling mixing of different fluids, but also for allowing components (such as surfactants) to diffuse through a fluid at experimentally realistic low concentrations.
Molecular Dynamics (MD)
The micro-scale part of our multiscale modeling strategy is Molecular Dynamics (MD). MD lets us capture the structure of molecules, and how they configure themselves in space at the microscale. To balance computational cost with physical accuracy, we employ ‘coarse-grained’ MD, wherein molecules are represented as one or more “beads”. A bead in turn represents a group of several atoms. Thus, a molecule like Hexane may be represented by two to three beads, while a long polymer will consist of hundreds of beads. The beads may be connected into any architecture, so that more elaborate molecular shapes may be simulated (e.g. branched/ring polymers).
Interactions between beads are tuned to capture the underlying chemistry that the beads represent. These interactions can range from weak intermolecular (van der Waals) forces, to strong ionic / hydrogen bonding interactions, to reversible or permanent crosslinking in dynamic polymer networks and gels. Physical conditions like temperature, pressure, and local deformation (e.g. stretching, compressing, shearing) can all be naturally included in this modeling approach.
Coarse-grained MD is powerful for simulating structure formation at the nano-scale. Examples include:
- Phase separation;
- Assembly of surfactants at interfaces between immiscible fluids;
- Micellization of solutes or insoluble oils in aqueous environments;
- Polymer conformations, entanglement, and microphase formation;
- Growth of macromolecules from polymerisation reactions;
- Crosslinking and network formation;
The results from MD simulation inform the evolution of the physics in the larger-scale smoothed particle hydrodynamics. They may also be analysed on their own, for microscopic insight into a material.
From chemical input to model interactions
One of the challenges in modeling is figuring out how to select the right input to use for your system. Using RheoCube, we help to remove some of this complexity. One way we do so is by using Hansen Solubility Parameters (HSP) as input for the cohesive energy density of an ingredient. HSP is especially suitable as it’s not limited to a narrow group of ingredients. HSP can be measured (or calculated) for solvents, polymers, or particles and even offers insight into surfactants.
Technical grade ingredients are often a blend of molecules (isomers, additives, etc.). The formulators of those ingredients don't always know their exact composition. In this case, HSP’s can still be measured however, as average values for the blend still give a good representation of its cohesive energy density. For more information about HSP see VLCI.
The technology behind RheoCube - Supercomputing
In developing our computational models, our aim is to minimize computational effort for users. The highly detailed simulations we develop do however require the use of High Performance Computing (HPC) resources (large parallel computers). This ensures the computation involved in running virtual experiments takes a reasonable amount of time. Our team has broad expertise in parallelization and optimization of simulation code, from those on small shared-memory clusters, to state-of-the-art supercomputers. This includes the efficient use of accelerators such as graphical processing units (GPU's).
We make these HPC resources available in the cloud and on-demand, which frees organizations from the need to have such computing power on-site. It also means costs are only related to actual computing hours. The computational services we employ maintain the highest security standards, so your data is safe.