The project will be concerned with the development of multiscale modeling techniques for simulations of fluid flows in nanomaterials. Computer simulations can provide insight into such systems when they can access, both, the atomistic length scales associated with size of the nanoparticles and the micro/macro scales characteristic of the fluid flow field. Simulations using molecular dynamics can capture the atomistic details of the nanoparticle-liquid interface but due to their computational cost they cannot be extended currently to the macroscale regime of the full flow field. In turn, continuum descriptions, using the Navier-Stokes equations may capture the macro-scale behavior of the fluid flow but they fail to represent accurately the flow field at the nanoparticle surface. The multiscale approaches, on the other hand, combine the powerful features of the both descriptions, i.e., the ability to describe the macro-scale behavior of the flow as well as accurate boundary conditions around nanoparticles. Another issue to consider when studying these systems is that typical experimental setups for molecular systems are coupled to the external environment, that is, the system is open and exchanges mass, momentum, and energy with its surroundings.
Instead, standard molecular simulations are mostly performed using periodic boundary conditions with a constant number of molecules. Therefore, it is essential to develop open simulation methodologies, which, contrary to standard techniques, open up the boundaries of a molecular system and allow for exchange of energy and matter with the environment, in and out of equilibrium.
The aim of the project is to combine diverse simulation techniques that separately model effectively either atomistic, mesoscale, or continuum scales of nanomaterials in a unifed multiscale framework. We will conduct multiscale simulations of the water flow through carbon-nanotubes membranes and the flow of several organic solvents, such as liquid butane, hexane, decane, and benzene, past thiolated golden nanoparticles. To this end, we will develop and employ an open molecular dynamics technique, which will enable us to perform equilibrium molecular dynamics simulations in the grand-canonical ensemble, exchanging particles with the surrounding, as well as nonequilibrium fluid flow simulations. The flow will be introduced via an external boundary condition while the equations of motion for the bulk will remain unaltered. Using this methodology we will also study the rotational and tumbling dynamics of the melt of star polymers under shear flow. Furthermore, in this project, we will study the conformational changes of proteins exposed to the shear flow. Our open multiscale approaches will bridge the hydrodynamics from the atomic to mesoscopic scale and enable the study of physical phenomena that are beyond the scope of either atomistic or mesoscopic simulations.
Detailed implementation plan
The project is divided into four work packages (WP):
OBMD will further be developed and used to derive the partial-slip boundary condition for continuum fluid dynamics simulation of water through CNT membranes. Slip length, i.e. the parameter of the partial slip boundary condition, will be obtained by fitting the energy dissipation to its reference OBMD value. Partial-slip boundary condition through the oscillating CNT membrane will also be derived. Using it in continuum fluid dynamics simulations we aim to determine the physical mechanism that would explain the discrepancy between experimental flow rate enhancement, which grows linearly with the CNT length for all CNT lengths, and the one observed in simulations, where it reaches a plato at some CNT length. WP1 will be performed by dr. Tilen Potisk (NIC), dr. Jurij Sablić (NIC), and prof. dr. Matej Praprotnik (NIC) in collaboration with the group from UL FMF.
Flows of several organic solvents past golden particles will be studied using OBMD from WP1. Golden particles will be functionalised by alkanthiol molecules of different size, which will form arms around the metalic core. The boundary condition for the flows past the nanoparticles will be derived and its dependence on the size of golden core and the length of alkanic arms will be examined. The boundary condition will then be used in continuum fluid dynamics simulation in order to study hydrodynamic impact of the solvent on the nanoparticle and vice-versa. WP2 will be carried out by dr. Tilen Potisk (NIC) and prof. dr. Matej Praprotnik (NIC) in collaboration with the group from UL FMF.
To reach longer timescales, we will implement the OBMD algorithm, further developed in WP1 and WP2, for GPUs and test it by conducting simulations of water. Next, we will conduct the studies of solvated proteins and describe their structural and conformational changes, when exposed to the shear flow. To this end, we will use OBMD, combined with DPD and continuum hydrodynamics simulation. We will thus focus on the effect the hydrodynamic interactions have on structural and conformational changes of these macromolecules. Furthermore, OBMD will also be used to simulate the DNA molecule in several aqueous salt solutions. The effect of different salts on the macromolecule will be investigated. WP3 will be conducted by Petra Papež (NIC), Ema Slejko (NIC), and prof. dr. Matej Praprotnik (NIC).
The Monte Carlo method will be used to simulate the packed DNA. To perform simulations, we will use the oxDNA model, which is a coarse-grained model designed to capture the thermodynamic, structural, and mechanical properties of single- and double-stranded DNA. The objective is to observe the splay-density coupling on an realistic system. With this, we aim to put to test the tensorial conservation law. From simulations, we will further extract macroscopic properties, such as splay-density coupling strength, elastic constants, etc. WP4 will be conducted by dr. Tilen Potisk (NIC) and prof. dr. Matej Praprotnik (NIC) in collaboration with the group from UL FMF.
Composition of the project team
- Matej Praprotnik
- Tilen Potisk
- Daniel Svenšek
- Jurij Sablić
- Jaka Sočan
- Amaury Coste
- Petra Papež
- Ema Slejko
- Aljaž Draškovič-Bračun
- Nikolaos Ntarakas
1.07 Computer intensive methods and applications
Duration of the project
The project is funded by ARRS.