Scipedia
Not a member yet
33380 research outputs found
Sort by
Geometric formulation of three-temperature radiation hydrodynamics
Get PDFThree-temperature (3T) radiation hydrodynamics models high energy-density plasma of nonlinearly coupled electron, ion, and radiation fields, finding applications in astrophysics and inertial confinement fusion. We present a geometric formulation of three-temperature radiation hydrodynamics. This is done utilizing an irreverisble portHamiltonian framework in the entropy representation. This geometric formulation separates the advection, interaction, and diffusion processes occuring into separate operators and establishes the energy-preserving interconnections between them. Structural properties such as mass, momentum and energy conservation as well as entropy production arise naturally from the geometric formulation. As an application, we briefly discuss a framework for the energy control of the 3T system within the port-Hamiltonian framework
Immersed-boundary approach based on integrated RBFs and smooth extension for solving PDEs in complex domains
Get PDFWe propose an immersed-boundary approach, based on point collocation, five-point integrated radial basis function stencils, rectangular Cartesian grids and smooth extension of the solution, for solving the two-dimensional elliptic partial differential equation in a geometrically complex domain
NONLINEAR ANALYSIS OF SOLAR FLARES OF X CLASS
No full textThe study of solar flares is definitely very important for our planet, because they have a direct impact on it. We have always wondered how solar flares, seemingly so far away from us, can affect the Earth. They are known for their effect on the ionosphere and magnetic field of our planet, thereby causing negative effects in addition to the fascinating aurora borealis. Thus, every flare that occurs on our main luminary can cause geomagnetic storms, bringing massive power outages, malfunctioning GPS systems, radio and telecommunications. In studying this issue, we decided to understand their patterns.
Thus, we studied methods of processing large volumes and presenting them in the form of graphs using MATLAB and Python programs, while using the relationship between the characteristics of the radio signal and its topology, developed by our scientist Zhanabaev Z.Ch. The uniqueness of our project is the use of only two coefficients (two-dimensional shape coefficient and information-entropy) characterizing solar flares to determine their patterns.
In conclusion, having set the goal of the project is to find the most efficient method to predict solar flare based only on the value of second flux and soft X-ray emission. We were able to make progress and discover a new to science concept of self-organization of solar flares
Developments in the use of the Bonded Particle Model to study ore fracture
No full textIn mineral processing, ore fracture is an essential first step for which the objective is to increase the exposed surface area of the valuable mineral, thereby increasing the likelihood of liberation in subsequent separation stages. This process is well known to be energy-intensive, and increasing scrutiny around sustainable practices has heightened the need to examine the efficiency of current industry approaches. Factors such as mineralogical structure and inherent weakening in the form of micro cracks are known to affect ore breakage mechanisms. However, isolating and investigating individual factors under experimental conditions is challenging and typically impractical. Numerical techniques such as the Bonded Particle Model-Discrete Element Method (BPM-DEM) have been developed as a means of investigating in isolation, the effects of different factors on ore breakage behaviour under closely controlled breakage conditions. In this work, the robustness of the BPM-DEM in predicting fracture characteristics during SILC impact breakage is evaluated. Thereafter, the BPM-DEM is used to analyse the internal mechanical response of a simulated rock specimen under impact loading commensurate with that of the SILC. The method is shown to be an insightful opportunity to study intrinsic and extrinsic rock properties during dynamic loading and breakag
Full waveform modeling in seismic exploration based on a digital geological model using spectral element method on GPU
No full textThe paper considers the solution of a three-dimensional problem of modeling of all types of seismic waves propagating in real geological media. The numerical algorithm based on the spectral element method (SEM). The main advantages of SEM (high order space discretization, explicit time integration scheme) are presented in comparison with the classical approach based on the finite element method (FEM). The features of the massively parallel implementation of the algorithm on modern MultiGPU systems (based on A100 GPU) using CUDA technology are considered. The efficiency of parallelization on hybrid systems with different SEM orders and parameters of the numerical time integration scheme is analyzed. The results of solving a three-dimensional problem of modeling the propagation of seismic waves in a heterogeneous geological media with faults and sharply varying properties of layers are presented. Analysis of the numerical convergence of SEM for dispersive waves of the Rayleigh type is performed. Local and non-local non-reflective boundary conditions on the artificial boundary of the computational region are considered. The 3D computational model is constructed using a detailed digital geological model built for one of the Arctic regions. It was converted to an unstructured hexahedral mesh to perform SEM calculations using CAE FIDESYS software. The model is further generalized for typical seismic-geological conditions of Western Siberia, so that on the basis of such modeling it is possible to conduct a wide range of studies on the possibilities of seismic exploration to study the main oil and gas reservoirs in this region. The solution was sought on a hexahedral mesh consisting of 5.5 mln spectral elements of the 5th order with a total number of SEM nodes 1.2 billion. The output results of full-wave modeling are stored in the SEG-Y format, suitable for all types of industrial seismic processing. The analysis of the obtained model seismograms and wave fields is carried out. The conclusion is made about the practical significance of the conducted research
A dynamic weighted loss function for enhancing the performance of neural networks
Get PDFIn machine learning process, hyper parameters are chosen in a way to decrease the prediction error and improve the convergence. However, the optimized hyper parameters have a limit in terms of enhancing the performance of the neural networks. In this work, the datasets used for the numerical experiments arise from the resolution of partial differential equations (PDE) defined on a spatial domain. We propose a DYNAmic WEIghted Loss (DYNAWEIL) function-based approach for neural networks that are used to learn these PDE’s solutions. This a two-step process: first we train for a few numbers of epochs in a classical way then the dynamic weighted loss function replaces the classical loss function by leveraging the information from past training error histories. To validate this method, we carry out numerical experiments with different neural networks on datasets arising on two different physics: Goldstein equation [1] and radiative transfer equation [2]. Thus, in order to demonstrate the relevance of this approach, we provide a comparison among a neural network model using a classical loss function, with and without hyper parameters optimization, and a dynamic weighted loss function for both versions
How AI is a factor of transversal innovation in Catalonia
No full textIt can be stated that Artificial Intelligence (AI) is the digital technology that has achieved one of the most important impact and social penetration level, following an exponential evolution in recent years. Despite this explosion, AI has an almost century-old history, dating back to the work of scientists and technologists who laid its foundations from the second half of the last century. AI also has a long history in Catalonia that begins, among others, with the figure of Professor Emeritus Ramon López de Mántaras and the creation of the AI Research Institute of the CSIC (IIIA-CSIC). This article takes a look at AI in Catalonia, starting with the main agents that make up the AI research and innovation ecosystem in our country. The current state of AI in Catalonia is presented based on the sectoral report published by ACCIO in May '24, and then the CATALONIA.AI strategy promoted by the Government of Catalonia and deployed from 2020 is described. The axis of business innovation that is promoted by this strategy and that materializes in the Center for Innovation in Data Technologies and Artificial Intelligence (CIDAI) is detailed, describing both qualitatively and quantitatively the most significant activities that it promotes. The article also addresses the transformative potential of AI in business from two points of view. The first one addresses the business applications of analytical AI in conjunction with other technologies in the context of Industry 4.0. The second one focuses on Generative AI, illustrating the basic principles of the foundational models and developing the multisectoral and multimodal opportunities and the impact that this variant of AI will have on the different productive and business sectors. Next, the work analyses the recurring barriers to a multisectoral implementation of AI, which, among others, refer to issues such as data availability or lack of qualified talent. In parallel to these strictly technological issues, the article also delves into the main consequences of the AI regulations that Europe has approved and how companies with products or services based on AI systems will have to adapt to them. Finally, the article gives an insight into the impact of AI on sustainability and possible ways to solve it
Una nova estratègia d’internacionalització per al segon quart del segle XXI
No full textEntre els trets més significatius que caracteritzen l’economia catalana està, sens dubte, la seva obertura exterior. Milers d’empreses -la majoria pimes- han incorporat, poc o molt, la presència en els mercats internacionals com a element indefugible de la seva estratègia empresarial. Administracions i organitzacions empresarials, en paral·lel, han estructurat programes i serveis de suport tendents a reforçar aquest procés, generant-se un ecosistema d’internacionalització amb una multiplicitat d’actors que porten a la pràctica la seva visió, les seves receptes, no sempre del tot alineades, amb l’objectiu que l’empresa les tradueixi en el seu compte de resultats.
En un escenari com l’actual, on un conjunt de megatendències estan incidint de ple en el procés d’internacionalització, tendències que no es poden afrontar de manera lineal, cal una visió holística, sistèmica, amb noves maneres de plantejar l’estratègia.
Poden les empreses seguir actuant com quan s’anunciava l’adveniment d’un món pla, de la fi de la història de Fukuyama, o han de fer front a les conseqüències d’un món multipolar en crisi permanent?
No es pot parlar de digitalització, d’intel·ligència artificial, ni d’innovació sense considerar els mercats exteriors, que són motors i destinataris d’aquests processos.
Avui, més que mai, els valors socials i ambientals estan qüestionant els fonaments del nostre sistema econòmic i les regulacions associades estan afegint complexitat a la internacionalització. Una empresa no pot obviar la sostenibilitat si vol ser capdavantera.
Si donem resposta a aquestes i d’altres variables, no serà inevitable que el sistema productiu català es plantegi una nova aproximació per afrontar la internacionalització del segon quart del segle XXI?
Es proposaran els elements clau que haurien de caracteritzar una nova estratègia guanyadora de internacionalització del teixit productiu català en aquest nou entorn
rCF LM PAEK PRESS-MOULDING PROCESS OPTIMIZATION THROUGH THE INTRODUCTION OF A WAAM MOULD CORE
No full textCarbon fibre matrix composites are consolidated in high-demanding industries such as aeronautics and aerospace. Nowadays, thermoset composites are fully integrated on several new aircrafts in production. However, thermoplastic composites are progressively emerging as a better solution due to their end-of-life recyclability and optimized manufacturing process. One of the processes used to manufacture final parts from thermoplastic composite materials is press-moulding, in which a high-temperature heating system is combined with the application of pressure at high speeds. This article studies the optimization of press-moulding process in combination with an Additive Manufacturing (AM) produced metallic mould core. The aim of this study is the manufacturing of a thermoplastic aeronautical part, that would be submitted to real-world performance test. For the press-moulding process, a high-temperature infrared heating system was designed, supported by thermal oil mould heating, this heating system was combined with a mould-pressing process by integrating the system in a commercial injection machine. Regarding the AM tooling, produced by Wire Arc Additive Manufacturing (WAAM) process, was designed to reduce tooling weight, and optimise thermal cycle. Moreover, manufacturing data from WAAM tooling was recorded and compared with that obtained from a symmetrical conventional manufacturing made mould. The analysis shows savings in terms of material and energy consumption, as well as cycle times reduction. The final part metrology study shows indistinguishable results between the press-moulding process carried out using the WAAM tooling and that manufactured through conventional methods
The Role of Proteins in Glial Cell Autophagy and Apoptosis in the Dentate Gyrus
No full textThe dentate gyrus is a unique part of the brain because it is known for housing neurogenesis in the adult brain, a process which normally stops early in development. This makes the dentate gyrus an area of great interest, especially in combating neurodegenerative diseases. This area, as well as the entire nervous system, is composed of both neurons, which send and receive signals, and glial cells, which are responsible for supporting neurons. Glial cells including astrocytes, oligodendrocytes, and ependymal cells are present in both the central nervous system and peripheral nervous system, while Schwann Cells are present only in the peripheral nervous system, and microglia are present only in the central nervous system. Apoptosis and autophagy are both processes which degrade and recycle materials. Autophagy degrades materials inside the cell, such as organelles and proteins, and uses lysosomes to carry out the process. Apoptosis consists of degrading old or damaged cells, and is known as programmed cell death. Many proteins affect these processes including Beclin-1, UC3, p62, and the BCL-2 family of proteins. Some studies have used machine learning algorithms in conjunction with simulation and statistical software to study these proteins and the processes they are involved in. Several areas of research remain unfilled, especially in the area of glial cells and the proteins involved in autophagy and apoptosis, as not many machine learning studies have examined this