- Professor Harold A. Scheraga full text
- Anthony Chukwuemeka Nwachukwu and Szymon Winczewski Application of the neural networks for developing new parameterization of the Tersoff potential for carbon abstract | full text
- Magdalena Stanik, Szymon Caban, Marlena Klonecka, Mateusz Buczak and Piotr Wiśniewski Numerical studies of internal flow in different types of filters abstract | full text
- Anjali Chopra and Szymon Winczewski Influence of addition of carbon nanotubes on rheological properties of selected liquid lubricants - a computer simulation study abstract | full text
h Professor Harold A. Scheraga
hAnthony Chukwuemeka Nwachukwu and Szymon Winczewski Application of the neural networks for developing new parameterization of the Tersoff potential for carbon
Abstract: Penta-graphene (PG) is a 2D carbon allotrope composed of a layer of pentagons having sp2 - and sp3 - bonded carbon atoms. A study carried out in 2018 has shown that the parameterization of the Tersoff potential proposed in 2005 by Ehrhart and Able (T05 potential) performs better than other potentials available for carbon, being able to reproduce structural and mechanical properties of the PG. In this work, we tried to improve the T05 potential by searching for its parameters giving a better reproduction of the structural and mechanical properties of the PG known from the ab initio calculations. We did this using Molecular Statics (MS) simulations and Neural Network (NN). Our test set consisted of the following structural properties: the lattice parameter a; the interlayer spacing h; two lengths of C-C bonds, d1 and d2 respectively; two valence angles, θ1 and θ2, respectively. We also examined the mechanical properties by calculating three elastic constants, C11 , C12 and C66, and two elastic moduli, the Young's modulus Eand the Poisson's ratio ν. We used MS technique to compute the structural and mechanical properties of PG at T = 0 K. The Neural Network used is composed of 2 hidden layers, with 20 and 10 nodes for the first and second layer, respectively. We used an Adams optimizer for the NN optimization and the Mean Squared Error as the loss function. We obtained inputs (about 80 000 different sets of potential parameters) for the Molecular Statics simulation by using randomly generated numbers. The outputs from these simulations became the inputs to our Neural Network. The Molecular Statics simulations were done with LAMMPS while the Neural Network and other computations were done with Python, Pytorch, Numpy, Pandas, GNUPLOT and Bash scripts. We obtained a parameterization which has a slightly better accuracy (lower relative errors of the calculated structural and mechanical properties) than the original parameterization.
hMagdalena Stanik, Szymon Caban, Marlena Klonecka, Mateusz Buczak and Piotr Wiśniewski Numerical studies of internal flow in different types of filters
Abstract: The quality of ambient air attracts considerable, widespread interest. Over the last decades, air purification has become an integral part of HVAC systems, process engineering, automotive and respiratory protection. Efficient separation of micro- and nano- particles is solidly linked with the development of new, sophisticated filtrating materials, as well as generating and validating mathematical models of such porous structures. The paper regards the numerical modeling of various filters. The presented work aims to validate four virtual filtrating materials –the fiberglass HEPA filter, the paper filter used in the automotive industry, knitted wire mesh and polyurethane foam. The pressure drop obtained for the filters under investigation was examined. The CFD results were validated against the data available in the literature. The agreement of the results of numerical and experimental studies proves the suitability of the proposed methods. At the same time, the simplifications employed in the simulations leave room for further improvement in future works.
hAnjali Chopra and Szymon Winczewski Influence of addition of carbon nanotubes on rheological properties of selected liquid lubricants - a computer simulation study
Abstract: This work is motivated by the improvement of anti-friction properties of lubricants by addition of CNTs proved experimentally in literature. In particular, a methodology is developed to compute the shear viscosity of liquid lubricants (Propylene Glycol) based on Molecular Dynamics simulation. Non-Equilibrium molecular dynamics (NEMD) approach is used with a reactive force field ReaxFF implemented in LAMMPS. The simulations are performed using the canonical (NVT) ensemble with the so-called SLLOD algorithm. Couette flow is imposed on the system by using Lees-Edwards periodic boundary conditions. Suitable parameters such as simulation time and imposed shear velocity are obtained. Using these parameters, the influence of addition of 27 wt% CNTs to Propylene Glycol on its viscosity is analyzed. Results show that 3.2 million time-steps with a 0.1 fs time-step size is not sufficient for the system to reach equilibrium state for such calculations. With the available computational resources, a shear velocity of 5 × 10−5 Å/fs was observed to give viscosity value with approximately 43% error as compared to the experimental value. Moreover, the lubricant exhibited a shear thinning behaviour with increasing shear rates. CNTs enhanced the lubricant's viscosity by 100-190% depending upon the averaging method used for calculation.