2023 - Publications Repository - Gdańsk University of Technology

Show publications from the year

Publications from the year 2023

Show all

Effect of oxidation and in vitro intestinal hydrolysis on phospholipid toxicity towards HT29 cell line serving as a model of human intestinal epithelium
- Karol Parchem
- Monika Baranowska
- Anna Kościelak
- Ilona Kłosowska-Chomiczewska
- Maria Rosario Domingues
- Adam Macierzanka
- Agnieszka Bartoszek-Pączkowska
2023 Full text FOOD RESEARCH INTERNATIONAL
Oxidation of food-derived phospholipids (PLs) can influence nutrient digestion and induce oxidative stress in gastrointestinal epithelium. In this study, hen egg yolk PL fraction was used to evaluate the effect of lipoxygenase (LOX)-induced PL oxidation on the rate of PL hydrolysis catalyzed by pancreatic phospholipase A2 (PLA2) in the presence of bile salts (BSs). Then, PL/BS solutions containing native or oxidized PLs were used in in vitro intestinal digestion to assess the effect of PL oxidation and hydrolysis on the toxicity towards HT29 cell line. Based on the obtained results, we suggest that hexanal and (E)-2-nonenal, formed by the decomposition of PL hydroperoxides, inhibited PLA2 activity. The cell exposure to simulated intestinal fluid (SIF) containing BSs decreased HT29 cell viability and significantly damaged cellular DNA. However, the genotoxic effect was reversed in the presence of all tested PL samples, while the protective effect against the BS-induced cytotoxicity was observed for native non-hydrolyzed PLs, but was not clearly visible for other samples. This can result from an overlap of other toxic effects such as lipotoxicity or disturbance of cellular redox homeostasis. Taking into account the data obtained, it was proposed that the PLA2 activity decline in the presence of PL oxidation products may be a kind of protective mechanism against rapid release of oxidized FAs characterized by high cytotoxic effect towards intestinal epithelium cells.
Effect of processing parameters on the cyclic behaviour of aluminium friction stir welded to spark plasma sintered aluminium matrix composites with bimodal micro-and nano-sized reinforcing alumina particles
- Behzad Sadeghi
- Pasquale Cavaliere
- Aleksandra Laska
- Angelo Perrone
- Gianni Blasi
- Arun Gopinathan
- M. Shamanian
- F. Ashrafizadeh
2023 MATERIALS CHARACTERIZATION
Understanding the cyclic behaviour of Alumina reinforced Aluminium composites (Al-A2O3) was of critical importance, for their further application in the different industrial sectors. The present study is focussing on the cyclic behaviour of the Al-Alumina nanocomposite produced through the combination of spark plasma sintering (SPS) method and friction stir welding (FSW). The added Alumina with total content of 10% is the combination of nano and micro-sized particles and its ratio differ for each sample. The microstructure of the SPSed samples is characterized using optical microscopy (OM), scanning electron microscopy (SEM), and energy dispersive X-ray spectroscopy (EDS). The microstructure of the processed composite samples is characterized and its mechanical behaviour is studied. Microstructural studies showed that nano sized particles of Alumina were mostly distributed along the grain boundaries and inside the grains, while micron-sized ones mostly settled on the grain boundaries. In addition, the hardness and tensile properties of the produced samples are analysed concerning the reinforcement size and the percentage of nanoparticles addition. The obtained results reveal that the mechanical and fatigue properties of the nanocomposite materials mainly depend on the material properties at the initial stage and the applied conditions of friction stir welding such as rotating speed and movement speed. The facture surface of the nanocomposites revealed a combined ductile–brittle fracture mode with finer dimples with emphasis on the pronounced role of nano-metric dispersoids.
Effect of pulse laser treatment at different process variables on mechanical behavior of carbon nanotubes electrophoretically deposited on titanium alloy
- Beata Majkowska-Marzec
- Joanna Sypniewska
- Kacper Staszewski
2023 Acta of Bioengineering and Biomechanics
Purpose: Titanium and its alloys are widely used as biomaterials for long-term implants, but they are usually surface-modified due to their weak bioactivity and wear resistance. Laser processing was used to modify the surface layer, and elemental carbon was a component of the deposited coatings. This research aims to use a combination of both methods based on preliminary electrophoretic deposition of multi-wall carbon nanotubes (MWNCTs) followed by pulse laser treatment. Carbon nanotubes were chosen due to their mechanical and chemical stability as well as their tubular shape, resulting in enhanced mechanical properties of laser-modified layers. Methods: The pulse laser power and laser scanning speed were defined as variable process parameters. The microstructure, roughness Ra, nanohardness H, Young’s modulus E, and indent depth values were measured, and the H/E, H 3 /E2 , and relative changes of all these values in comparison to MWCNTs-coated and non-coated surfaces, were calculated. Results: The obtained results show that the best mechanical properties of MWCNTs-coated and laser-treated specimens are obtained at a laser power of 900 W and laser feed of 6 mm/s. The observed relations can be explained considering processes occurring on the surface such as deposition of carbon nanotubes, melting and re-crystallization of the surface layer, formation and possible partial decomposition of titanium carbides, and associated changes in local chemical composition, phase composition, and a level of residual stresses beneath the surface. Conclusions: The developed process can substitute the time and money-consuming carbonization of titanium and its alloys.
Effect of Simultaneous Valve Closures in Hydraulic Piping Systems
- Kamil Urbanowicz
- Igor Haluch
- Anton Bergant
- Adam Deptuła
- Paweł Śliwiński
2023
The paper investigates wave interference (between pressure waves) occurring in simple hydraulic systems. Water hammer was induced by simultaneous closure of three valves located at the reservoirs of a “Y” type hydraulic system. Numerical simulations were carried out with the help of the freeware computer package Allievi enabling the reader to replicate results in a direct manner. The influence of the following quantities has been tested: Reynolds number Re (laminar and turbulent flow), the length of the pipe arms (L1, L2, L3), the constant pipe internal diameter (Di = const.) and variable internal pipe diameters Di (D1 = D2 = D3). The results of the research showed how unwanted interferences between pressure waves may occur and thus increase the possibility of piping system damage. Further numerical investigations are sought in consideration of unsteady skin friction losses and viscoelastic pipe wall effects.
Effect of soil on the capacity of viscous dampers between adjacent buildings
- Elif Cagda Kandemir
- Robert Jankowski
2023 Full text GRADEVINAR
This study investigated the seismic pounding of two adjacent buildings considering soil–structure interaction (SSI). A comprehensive parametric study of buildings with different heights was performed to reveal the pounding-involved behaviour considering the soil effect. Wavelet transform has been conducted to gain insight into the differences in the frequency contents of the impact forces between fixed- and flexible-base adjacent structures. Linear viscous dampers (LVDs) between adjacent floors were used as pounding protection measures. The required supplemental damping ratio of the LVDs was determined through optimisation analysis under different soil types to verify the effect of the SSI on structural damping. Comparative results with and without SSI showed that incorporating the SSI worsens the pounding-involved responses during earthquakes.
Effect of temperature and composition on physical properties of deep eutectic solvents based on 2-(methylamino)ethanol – measurement and prediction
- Bartosz Nowosielski
- Marzena Jamrógiewicz
- Justyna Łuczak
- Agnieszka Tercjak
- Dorota Warmińska
2023 JOURNAL OF MOLECULAR LIQUIDS
Novel deep eutectic solvents were synthesized using 2-(methylamino)ethanol as hydrogen bond donor with tetrabutylammonium bromide or tetrabutylammonium chloride or tetraethylammonium chloride as hydrogen bond acceptors. Mixtures were prepared at different molar ratios of 1:6, 1:8 and 1:10 salt to alkanolamine and then Fourier Transform Infrared Spectroscopy measurements were performed to confirm hydrogen bonds interactions between components. Moreover, thermal properties such as melting points and thermal stability of deep eutectic solvents were determined and described. Each of important physical properties, including densities, viscosities, refractive indices and sound velocities at the temperature range of 293.15–333.15 K and the pressure of 0.1 MPa, were measured and discussed. The effect of hydrogen bond acceptor to hydrogen bond donor molar ratio, anion and length of alkyl chain for each synthesized salt according to their properties was evaluated. Additionally, the experimental values of each physicochemical parameter were compared with the predicted ones calculated using models recommended in literature. The main aim of this work was to assess the suitability of existing mathematical models for predicting the physicochemical properties of novel alkanolamine-based DESs. Empirical correlations for approximating phase behaviour or flow properties for DES systems are used in many procedures for design materials used of carbon-dioxide capture purposes. The obtained results indicate that in the case of deep eutectic solvents based on 2-(methylamino)ethanol, in the absence of any experimental data, the best models for density prediction are the bonding group interaction contribution method and the group contribution model. For modelling of the refractive index it has been confirmed that a method based on the critical properties is the most satisfying. However, for the viscosity and speed of sound, the absolute average relative deviations for the methods based on critical properties exceed the measurement uncertainties found in practice. Therefore, they do not seem suitable for an accurate estimation of these properties for deep eutectic solvents based on 2-(methylamino)ethanol.
Effect of water salinity on properties of multipass underwater wet manual metal arc welded joints
- Jacek Tomków
- Dariusz Fydrych
- Jerzy Łabanowski
2023 Full text Welding in the World
The weldability of steel under the water is limited due to the influence of the environment. Water causes limited visibility, presence of the residual stresses, increasing the cooling rate, and increasing the diffusible hydrogen content in deposited metal, leading to the formation of brittle microstructures in heat-affected zone (HAZ). The paper presents the results of mechanical properties testing of S420G2+M steel welded joints made with covered electrodes in the water with salinity values: 0‰, 7.5‰, and 35‰. The non-destructive tests: ultrasonic (UT) and radiographic (RT), and destructive tests: Vickers HV10 measurements, Charpy impact, and bending and metallographic macro- and microscopic tests were performed. Moreover, the diffusible hydrogen content in deposited metal by the mercury method was measured. It was observed that water salinity has an influence on the stability of the welding arc and the properties of joints. In particular, it was found that increasing the water salinity provides to: decreasing the hardness of HAZ (from 211 HV10 to 193 HV10), increasing the impact strength (from 82.5 to 101.3 J/cm2). Additionally, a slight increase in the diffusible hydrogen content in deposited metal (up to 65 ml/100g) was observed. The number of microcracks in the joints decreased with increasing the water salinity.
Effect of wet Hydrogen Sulfide on Carbon Steels Degradation in Refinery Based on Case Study
- Juliusz Orlikowski
- Agata Jażdżewska
- İlyas Uygur
- Radosław Gospoś
- Tomasz Olczak
- Kazimierz Darowicki
2023 Full text ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING
Material degradation caused by wet hydrogen sulfide (Wet H2S) is one of the major issues in refineries. Carbon steel is the main construction material used in refineries and is subjected not only to thickness loss but the hydrogen is the major concern. H2S leads to high atomic hydrogen penetration in steel which causes degradation in the form of blistering and/or cracking (HIC/SOHIC) and stress cracking in the area of welds and heat affected zones. Mechanism is unusually dangerous due to high difficulty of threat assessment, which is mostly based on ultrasonic measurements or calculations based on the API standards. This research is based on evaluation of the real threat of wet H2S degradation based on examination of construction materials after 41 years of exploitation in refinery. The comparison of the theoretical calculation based on API standards and experience of the degradation of the real objects was characterized. Laboratory measurements include mechanical tests involving the elongation in the function of tensile stress and deformation, with analysis of material hardness and gas chromatography analysis. The results are complemented by studies of the chemical composition of the streams and analytical studies of the hydrogen content in the material obtained by the gas chromatography method. The tests showed compliance of the standard analysis of the corrosion risk with the laboratory tests performed on the real samples. Research leads to indication of the destructive methods which can be used on the materials obtained during scheduled material replacements or installation modernizations.
Effective Air Quality Prediction Using Reinforced Swarm Optimization and Bi-Directional Gated Recurrent Unit
- Sasikumar Gurumoorthy
- Aruna Kumari Kokku
- Przemysław Falkowski-Gilski
- Parameshachari Bidare Divakarachari
2023 Full text Sustainability
In the present scenario, air quality prediction (AQP) is a complex task due to high variability, volatility, and dynamic nature in space and time of particulates and pollutants. Recently, several nations have had poor air quality due to the high emission of particulate matter (PM2.5) that affects human health conditions, especially in urban areas. In this research, a new optimization-based regression model was implemented for effective forecasting of air pollution. Firstly, the input data were acquired from a real-time Beijing PM2.5 dataset recorded from 1 January 2010 to 31 December 2014. Additionally, the newer real-time dataset was recorded from 2016 to 2022 for four Indian cities: Cochin, Hyderabad, Chennai, and Bangalore. Then, data normalization was accomplished using the Min-Max normalization technique, along with correlation analysis for selecting highly correlated variables (wind direction, temperature, dew point, wind speed, and historical PM2.5). Next, the important features from the highly correlated variables were selected by implementing an optimization algorithm named reinforced swarm optimization (RSO). Further, the selected optimal features were given to the bi-directional gated recurrent unit (Bi-GRU) model for effective AQP. The extensive numerical analysis shows that the proposed model obtained a mean absolute error (MAE) of 9.11 and 0.19 and a mean square error (MSE) of 2.82 and 0.26 on the Beijing PM2.5 dataset and a real-time dataset. On both datasets, the error rate of the proposed model was minimal compared to other regression models.
Effective assessment of biopolymer-based multifunctional sorbents for the remediation of environmentally hazardous contaminants from aqueous solutions
- Shahid Nawaz
- Andleeb Tabassum
- Sara Muslim
- Tayyaba Nasreen
- Ausra Baradoke
- Tak H. Kim
- Grzegorz Boczkaj
- Teofil Jesionowski
- Muhammad Bilal
2023 CHEMOSPHERE
Persistent contaminants in wastewater effluent pose a significant threat to aquatic life and are one of the most significant environmental concerns of our time. Although there are a variety of traditional methods available in wastewater treatment, including adsorption, coagulation, flocculation, ion exchange, membrane filtration, co-precipitation and solvent extraction, none of these have been found to be significantly cost-effective in removing toxic pollutants from the water environment. The upfront costs of these treatment methods are extremely high, and they require the use of harmful synthetic chemicals. For this reason, the development of new technologies for the treatment and recycling of wastewater is an absolute necessity. Our way of life can be made more sustainable by the synthesis of adsorbents based on biomass, making the process less harmful to the environment. Biopolymers offer a sustainable alternative to synthetic polymers, which are manufactured by joining monomer units through covalent bonding. This review presents a detailed classification of biopolymers such as pectin, alginate, chitosan, lignin, cellulose, chitin, carrageen, certain proteins, and other microbial biomass compounds and composites, with a focus on their sources, methods of synthesis, and prospective applications in wastewater treatment. A concise summary of the extensive body of knowledge on the fate of biopolymers after adsorption is also provided. Finally, consideration is given to open questions about future developments leading to environmentally friendly and economically beneficial applications of biopolymers.
Effective Equations for the Optimum Seismic Gap Preventing Earthquake-Induced Pounding between Adjacent Buildings Founded on Different Soil Types
- Mahmoud Miari
- Robert Jankowski
2023 Full text Applied Sciences-Basel
The best approach to avoid collisions between adjacent structures during earthquakes is to provide sufficient spacing between them. However, the existing formulas for calculating the optimum seismic gap preventing pounding were found to provide inaccurate results upon the consideration of different soil types. The aim of this paper is to propose new equations for the evaluation of the sufficient in-between separation gap for buildings founded on different soil conditions. The double-difference formula has been taken into account in this study. The seismic gap depends on the correlation factor and on the top displacements of adjacent buildings. The correlation factor depends on the ratio of the periods of adjacent buildings (smaller period to larger period). The modification of the correlation factor has been introduced for buildings founded on five different soil types. Five soil types were taken into account in this study, as defined in the ASCE 7-10 code, i.e., hard rock, rock, very dense soil and soft rock, stiff soil, and soft clay soil. The normalized root mean square errors have been calculated for the proposed equations. The results of the study indicate that the error ranges between 2% and 14%, confirming the accuracy of the approach. Therefore, the proposed equations can be effectively used for the determination of the optimum seismic gap preventing earthquake-induced pounding between buildings founded on different soil types.
Effectiveness of various types of coating materials applied in reinforced concrete exposed to freeze–thaw cycles and chlorides
- Ginneth Millan Ramirez
- Hubert Byliński
- Maciej Niedostatkiewicz
2023 Full text Scientific Reports
This study assesses the durability of coated and uncoated concrete surfaces protected with four Different coating materials: water-soluble (BW), solvent-based (BR), mineral (MI), and epoxy (EP). The durability assessment includes evaluating the absorption rate of water, pull-of adhesion strength, and coating material thickness. Concrete samples were subjected to immersion in regular water and a 7% urea solution, followed by cyclic freezing and thawing. Furthermore, the diffusion of chloride ions in concrete was evaluated using the impressed voltage method, with the samples exposed to the aging process, immersed in a 3.5% NaCl solution. The results indicate that EP and BW coatings were significantly affected by the presence of urea and freeze–thaw cycles, resulting in a 43% and 47% reduction in pull-of adhesion strength, respectively. Notably, the MI-coated concrete samples exposed to urea solution and the freeze–thaw cycles exhibited a significant reduction in the absorption rate due to the accumulation of crystals on the coating surface, resulting in reduced porosity of the material.
Effects of Deck-Abutment Pounding on the Seismic Fragility Curves of Box-Girder Highway Bridges
- Seyyed Amirhossein Moayyedi
- Hossein Rezaei
- Afshin Kalantari
- Robert Jankowski
2023 JOURNAL OF EARTHQUAKE ENGINEERING
Earthquake-induced pounding in bridges is a complex contact phenomenon in which the dynamic responses of structures, including collisions between deck and abutments, are strongly related to structural properties and earthquake excitation. The goal of this study is to develop and compare the seismic fragility curves of overall system and individual components of regular and irregular box-girder highway bridges in two cases: with pounding and without pounding. For this purpose, four levels of altitudinal irregularity, ranging from regular to highly irregular, are considered. To extend the results for all bridges in the same class, different sources of uncertainties related to earthquakes, structural geometries, and material properties are taken into account. The analytical fragility curves have been developed based on nonlinear time history analyses in OpenSees finite element software for the cases with and without pounding effects. The process has been repeated for each two-, three-, and four-span classes at four irregularity levels. The fragility function parameters for the two cases with and without pounding have been compared for all classes considered in this study. Using fragility functions, this paper clarifies the interactive roles of irregularity and pounding between deck and abutments for seismic vulnerability of multi-span box-birder highway bridges. The results indicate that collisions often show an adverse effect on all structural components. It has also been observed that the detrimental effect of pounding on seismic fragility is more apparent in irregular bridges when compared with regular ones. In addition, the study introduces a conversion coefficient to clarify the effects of pounding on the fragility of bridge components and the overall system. This coefficient can be applied in both conventional analytical methods like static or simplified analysis and technical earthquake models like HAZUS, adjusting fragility values for pounding and irregularity effects.
Effects of Storing Flux-Cored Wires under Various Conditions
- Aleksandra Świerczyńska
- Michał Landowski
- Adrian Wolski
- Grzegorz Lentka
- Dariusz Fydrych
- Jerzy Łabanowski
2023 Full text Biuletyn Instytutu Spawalnictwa w Gliwicach
Welding processes involving the use of flux-cored wires are becoming increasingly popular, particularly in shipbuilding as well as in off-shore and civil engineering. The article presents characteristics of the welding process, its areas of application as well as advantages and disadvantages (e.g. necessity of ensuring appropriate conditions for the storage of filler metal wires). The satisfaction of quality-related requirements concerning welded joints necessitates controlling the quality of flux-cored wires as their condition (apart from welding conditions) is one of the most important factors affecting the welding process and the quality of joints. The analysis of related reference publications and individual study revealed that the storage of wires under conditions inconsistent with requirements specified by producers affects welding process stability and weld deposit properties. Visual tests (VT) tasked with assessing the quality of wire surface do not always provide sufficient information as regards the usability of filler metal wires in welding processes.
Effects of the Covid-19 travel restrictions on metropolises mobility: empirical evidence from the Tricity metropolis (Poland)
- Krzysztof Grzelec
- Katarzyna Hebel
- Romanika Okraszewska
- Olgierd Wyszomirski
2023 Full text Prace Komisji Geografii Komunikacji PTG
This paper aims to assess changes in mobility and modal shift caused by COVID-19 travel restrictions among the residents of Gdańsk Bay Metropolis (Poland). Measurement’s moments were assumed in periods differing in the level of restric- tions. The computer-assisted telephone interview (CATI) was carried out in November and December 2020. The results did not confirm the expected modal shift. However, significant changes in the number of trips between periods were observed. Restrictions in the first period of the pandemic resulted in a greater decrease in mobility than the restrictions at the end of the year. Moreover, significant associations were found between transport behaviour and place of residence. Nevertheless, pos- sible negative changes in the modal split after the pandemic ends should be counteracted by improving the quality of public transport services. As the most important attributes of public transport attracting passengers after the pandemic respondents recognized: high frequency of vehicles, low cost of travel, not overcrowded vehicles. However, some residents declared they will not use public transport regardless of service improvements. Consequently, to meet the objectives of sustainable mobility policy, it can be necessary to increase the share in the modal split of other sustainable modes of travel
Effects of thermal history on the performance of low-temperature solid oxide fuel cells with Sm0.2Ce0.8O2-δ electrolyte and LiNi0.81Co0.15Al0.04O2 electrodes
- Sea-Fue Wang
- Yi-Le Liao
- Yung-Fu Hsu
- Tsang Yu Weng
- Piotr Jasiński
2023 JOURNAL OF POWER SOURCES
In this study, low-temperature solid oxide fuel cells with an ∼560 μm thick Sm0.2Ce0.8O2−δ (SDC) electrolyte and ∼890 μm thick LiNi0.81Co0.15Al0.04O2−δ (NCAL) electrodes are constructed and characterized under three experimental conditions. The cell with an NCAL cathode pre-reduced under an H2 atmosphere at 550 °C presents the best electrochemical performance. This is ascribed to facts that the reduction reaction generating Ni–Co alloy particles on the NCAL surface and partial reoxidation of Ni–Co to Ni(Co)O under an air atmosphere during subsequent experiments increase the triple-phase-boundary area, improve the catalytic activity for the oxygen reduction reaction, and cause a low polarization resistance (0.186 Ω cm2). Also, the reduction of the Ni foam–NCAL anode triggers the formation of H+/O2− bi-ionic SDC–Li2CO3–LiOH composite electrolytes. The LiOH–Li2CO3 melt permeates and densifies the electrolyte layer; this significantly increases the electrical conductivity of the electrolyte layer to 0.291 S cm−1 and lowers the ohmic resistance of the cell (0.170 Ω cm2). The Li2CO3–LiOH phases serve as electron-blocking layers to limit the electronic conductivity of the electrolyte layer. This induces a high open-circuit voltage of 1.024 V and a high maximum power density of 611.3 mW·cm−2.
Efficient optimization approaches for microwave assisted extraction of high-quality antioxidant compounds from Salvia officinalis L.: UHPLC-HRMS differential analysis of phenolic profiles obtained by ultrasound and microwave extraction
- Hamza Moussa
- Farid Dahmoune
- Marika Mróz
- Hocine Remini
- Nabil Kadri
- Sarah Hamid
- Barbara Kusznierewicz
2023 Sustainable Chemistry and Pharmacy
The study aims to optimize MAE of total phenolic compounds (TPC) and antioxidant capacity from Salvia officinalis L. leaves using a definitive screening design (DSD) and I-optimal design. UHPLC-HRMS analysis was used to identify and compare the composition of MAE and UAE optimal extracts. The results showed that DSD and I-optimal design were successfully applied for the optimization of MAE targeting phenolics and other antioxidants from S. officinalis L. with the following optimum conditions: 60% ethanolic solvent, time of 4.75 min, power of 600 W, and L/S ratio of 50 mL g 1. The UHPLC-MS analysis results allowed the identification of more than 80 compounds, and the differential analysis indicated that the MAE yielded a higher level of 181 substance peaks, while the UAE yielded a higher level of 87 substance peaks. This study provides valuable information for selecting the appropriate extraction technique when targeting specific compounds.
Efficient parallel implementation of crowd simulation using a hybrid CPU+GPU high performance computing system
- Jakub Skrzypczak
- Paweł Czarnul
2023 SIMULATION MODELLING PRACTICE AND THEORY
In the paper we present a modern efficient parallel OpenMP+CUDA implementation of crowd simulation for hybrid CPU+GPU systems and demonstrate its higher performance over CPU-only and GPU-only implementations for several problem sizes including 10 000, 50 000, 100 000, 500 000 and 1 000 000 agents. We show how performance varies for various tile sizes and what CPU–GPU load balancing settings shall be preferred for various domain sizes among CPUs and GPUs of a high performance system with 2 Intel Xeon Silver multicore CPUs and 8 NVIDIA Quadro RTX 5000 GPUs. We then present how execution time depends on the number of agents as well as the number of CUDA streams used for parallel execution of several CUDA kernels. We discuss the design and implementation of an algorithm with CPU computational threads, GPU management threads, assignment of particular tasks to threads as well as usage of pinned memory and CUDA shared memory for maximizing performance.
Efficient uncertainty quantification using sequential sampling-based neural networks
- Pavankumar Koratikere
- Leifur Leifsson
- Sławomir Kozieł
- Anna Pietrenko-Dąbrowska
2023
Uncertainty quantification (UQ) of an engineered system involves the identification of uncertainties, modeling of the uncertainties, and the forward propagation of the uncertainties through a system analysis model. In this work, a novel surrogate-based forward propagation algorithm for UQ is proposed. The proposed algorithm is a new and unique extension of the recent efficient global optimization using neural network (NN)-based prediction and uncertainty (EGONN) algorithm which was created for optimization. The proposed extended algorithm is specifically created for UQ and is called uqEGONN. The uqEGONN algorithm sequentially and simultaneously samples two NNs, one for the prediction of a nonlinear function and the other for the prediction uncertainty. The uqEGONN algorithm terminates based on the absolute relative changes in the summary statistics based on Monte Carlo simulations (MCS), or a given maximum number of sequential samples. The algorithm is demonstrated on the UQ of the Ishigami function. The results show that the proposed algorithm yields comparable results as MCS on the true function and those results are more accurate than the results obtained using space-filling Latin hypercube sampling to train the NNs.
Ekspertyza dotycząca stanu zachowania drzewostanu. Osiedle domów jednorodzinnych.
- Maura Zaworska
2023
Badanie wpływu inwestycji na stan zachowania drzewostanu rosnącego w rejonie bezpośredniego oddziaływania prac budowlanych.