arXiv subjects

K. S. Kolegov

Publications and source records attributed to K. S. Kolegov.

7 recordsLinked to original sources

arXiv2026

Application of Machine Learning for the Identification of 2D Colloidal Assemblies: A Case Study on Particles of Distinct Shapes

This work addresses the problem of identifying colloidal monolayer assemblies using particles of various shapes (two-dimensional coatings): spheres, ellipsoids, cuboids, and rods. The following classification of assemblies is considered: isolated particles, dimers, chains, clusters, and loops. The YOLO model was chosen as the identification method. Synthetic datasets were prepared for each of the four particle shapes to train the models. The paper discusses the application of models trained on synthetic data to experimental images. An analysis was carried out on the feasibility of using such models for recognizing configurations in real images. While recognition on artificial images is nearly perfect, tests on experimental images showed a significant deviation. The average error across all particle types was 43.1%, but a considerable spread in values is observed: from 20% for spheres to 58.5% for cuboids, indicating the algorithm's selective sensitivity to object geometry. The created datasets and trained models are freely available for use. The corresponding modules have been integrated into the previously developed information system (https://isanm.space/). To further improve prediction results, it is necessary to prepare datasets based on experimental images.

cond-mat.soft↗

arXiv2026

Identification of 2D colloidal assemblies in images: a threshold processing method versus machine learning

This paper is devoted to the problem of identification of colloidal assemblies using the example of two-dimensional coatings (monolayer assemblies). Colloidal systems are used in various fields of science and technology, for example, in applications for photonics and functional coatings. The physical properties depend on the morphology of the structure of the colloidal assemblies. Therefore, effective identification of particle assemblies is of interest. The following classification is considered here: isolated particles, dimers, chains and clusters. We have studied and compared two identification methods: image threshold analysis using the OpenCV library and machine learning using the YOLOv8 model as an example. The features and current results of training a neural network model on a dataset specially prepared for this work are described. A comparative characteristic of both methods is given. The best result was shown by the machine learning method (97% accuracy). The threshold processing method showed an accuracy of about 67%. The developed algorithms and software modules may be useful to scientists and engineers working in the field of materials science in the future.

cond-mat.soft↗

arXiv2023

Marangoni instability in oblate droplets suspended on a circular frame

We study theoretically internal flows in a small oblate droplet suspended on the circular frame. Marangoni convection arises due to a vertical temperature gradient across the drop and is driven by the surface tension variations at the free drop interface. Using the analytical basis for the solutions of Stokes equation in coordinates of oblate spheroid we have derived the linearly independent stationary solutions for Marangoni convection in terms of Stokes stream functions. The numerical simulations of the thermocapillary motion in the drops are used to study the onset of the stationary regime. Both analytical and numerical calculations predict the axially-symmetric circulatory convection motion in the drop, the dynamics of which is determined by the magnitude of the temperature gradient across the drop. The analytical solutions for the critical temperature distribution and velocity fields are obtained for the large temperature gradients across the oblate drop. These solutions reveal the lateral separation of the critical and stationary motions within the drops. The critical vortices are localized near the central part of a drop, while the intensive stationary flow is located closer to its butt end. A crossover to the limit of the plane film is studied within the formalism of the stream functions by reducing the droplet ellipticity ratio to zero value. The initial stationary regime for the strongly oblate drops becomes unstable relative to the many-vortex perturbations in analogy with the plane fluid films with free boundaries.

physics.flu-dyn↗

arXiv2022

Circulating Marangoni flows within droplets in smectic films

We present theoretical study and numerical simulation of Marangoni convection within ellipsoidal isotropic droplets embedded in free standing smectic films (FSSF). The thermocapillary flows are analyzed for both isotropic droplets spontaneously formed in FSSF overheated above the bulk smectic-isotropic transition, and oil lenses deposited on the surface of the smectic film. The realistic model, for which the upper drop interface is free from the smectic layers, while at the lower drop surface the smectic layering still persists is considered in detail. For isotropic droplets and oil lenses this leads effectively to a sticking of fluid motion at the border with a smectic shell. The above mentioned asymmetric configuration is realized experimentally when the temperature of the upper side of the film is higher than at the lower one. The full set of stationary solutions for Stokes stream functions describing the Marangoni convection flows within the ellipsoidal drops were derived analytically. The temperature distribution in the ellipsoidal drop and the surrounding air was determined in the frames of the perturbation theory. As a result the analytical solutions for the stationary thermocapillary convection were derived for different droplet ellipticity ratios and the heat conductivity of the liquid crystal and air. In parallel, the numerical hydrodynamic calculations of the thermocapillary motion in the drops were performed. Both the analytical and numerical simulations predict the axially-symmetric circulatory convection motion determined by the Marangoni effect at the droplet free surface. Due to a curvature of the drop interface a temperature gradient along its free surface always persists. Thus, the thermocapillary convection within the ellipsoidal droplets in overheated FSSF is possible for the arbitrarily small Marangoni numbers.

cond-mat.soft↗

arXiv2021

Monte Carlo simulation of particle size separation in evaporating bi-dispersed colloidal droplets on hydrophilic substrates

Colloidal droplets are used in a variety of practical applications. Some of these applications require particles of different sizes. These include medical diagnostic methods, the creation of photonic crystals, the formation of supraparticles, and the production of membranes for biotechnology. A series of earlier experiments had shown the possibility of particle separation near the contact line, dependent upon their size. A mathematical model has been developed to describe this process. Bi-dispersed colloidal droplets evaporating on a hydrophilic substrate are taken into consideration. A particle monolayer is formed near the periphery of such droplets due to the small value of the contact angle. The shape of the resulting deposit is associated with the coffee ring effect. The model takes into account both particle diffusion and transfers caused by capillary flow due to liquid evaporation. Monte Carlo simulations of such particle dynamics have been performed at several values of particle concentration in the colloidal solution. The numerical results agree with the experimental observations, in which small particles accumulate nearer to the contact line than do the large particles. However, the particles do not actually reach the contact line, but accumulate at a small distance from it. The reason for this is the surface tension acting on the particles in areas where the thickness of the liquid layer is comparable to the particle size. Indeed, the same mechanism affects the observed separation of the small and large particles.

cond-mat.soft↗

arXiv2021

Nonuniform heating of a substrate in evaporative lithography

This work is devoted to a method to generate particle cluster assemblies, and connected to evaporative lithography. Experiments are carried out using nonuniform evaporation of an isopropanol film containing polystyrene microspheres in a cylindrical cell. The local inhomogeneity of the vapor flux density is achieved by exploiting the temperature gradient. A copper rod is mounted in the central part of the bottom of the cell for further heating. The thermocapillary flow resulting from the surface tension gradient, due in turn to the temperature drop, transfers the particles that were originally at rest at the bottom of the cell. The effect of the initial thickness of the liquid layer on the height and base area of the cluster formed in the central region of the cell is studied. The velocity is measured using particle image velocimetry. A model describing the initial stage of the process is developed. The equations of heat transfer and thermal conductivity are used to define the temperature distribution in the liquid and in the cell. The fluid flow is simulated using the lubrication approximation. The particle distribution is modeled using the convection-diffusion equation. The evaporation flux density is calculated using the Hertz-Knudsen equation. The dependence of the liquid viscosity on the particle concentration is described by Mooney's formula. Numerical results show that the liquid film gradually becomes thinner in the central region, as the surface tension decreases with the increasing temperature. The liquid flow is directed to the heater near the substrate, and it transfers the particles to the center of the cell. The volume fraction of the particles increases over time in this region. The heat flow from the heater affects the geometry of the cluster for two reasons: first, the Marangoni flow velocity depends on the temperature gradient, and second, the decrease ...

cond-mat.soft↗

arXiv2018

Monte Carlo simulation of colloidal particles dynamics in a drying drop

The work is devoted to particles dynamics simulation in a colloidal drop, when it dries on a substrate and the triple-phase boundary is fixed. Experimental observations [Deegan R. D. et. al., 2000] show a ring deposition on a solid substrate after full droplet desiccation. This phenomenon of macrolevel is known as coffee ring effect. There are other experiments that show a co-effect at the microlevel. We are talking about the formation of a quasicrystalline structure on the outer part of the ring and occurrence amorphous inner zone of ring [Mar\'ın Á. G. et. al., 2011]. The goal of this work is to check the hypothesis of other authors that this phenomenon is explained by the competition between the characteristic times of diffusional displacement of particles and their transfer by a compensation flow [Mar\'ın Á. G. et. al., 2011]. Numerical calculations using a model built on the basis of such assumptions and effects did not show the formation of a quasicrystal structure. It is probably necessary to take into account additional effects in such a system, for example, surface tension.

cond-mat.soft↗