139585 research outputs found
Sort by
Hoikkien ristiinliimattujen massiivipuuseinien mitoitusperusteet
No full textSince CLT is a relatively new material, its design principles have not yet been standardized. Manufacturers’ design guidelines are used for its dimensioning, but these guidelines leave some factors unaccounted for, which can affect the capacity of the CLT wall. Especially in the case of slender walls, factors such as humidity, temperature, initial curvature, and eccentric loads can have significant impact on the wall’s durability.
The purpose of this thesis was to investigate the behaviour of slender CLT walls under vertical loading, considering factors such as initial curvature, eccentricity of loads, the effects of humidity and temperature, and wall deflection caused by creep. The behaviour of vertically supported walls was also assessed. In addition, the impact of different factors on the wall’s vertical load capacity and lateral displacement was evaluated.
The effects of initial curvature and eccentric loads were studied by comparing the capacity of a CLT wall when the proposed minimum values for both initial curvature and eccentric loads were taken into account, against a scenario where these factors were neglected. The effects of humidity and temperature were determined by calculating the theoretical moisture differences across the wall’s surfaces in Finnish conditions and using this data to compute stresses and deflections in the wall.
By considering the proposed minimum values for initial curvature and eccentricity, it was observed that they have a clear effect on the buckling capacity of the CLT wall. In the worst-case scenario, based on Finnish conditions, the moisture and temperature effects resulted in approximately 3 MPa compressive load on the outer layer and a small deflection. Vertical supports were found to have a clear reinforcing effect on the wall’s buckling capacity. Therefore, in the case of slender walls, these factors should be considered during the design of structures.CLT:n ollessa verrattain uusi materiaali sen suunnitteluperusteita ei ole vielä standardisoitu. Sen mitoitukseen käytetään valmistajien antamia suunnitteluohjeita, mutta nämä ohjeet jättävät joitain asioita huomiotta, mikä voi vaikuttaa CLT-seinän kapasiteettiin. Varsinkin hoikkien seinien tapauksessa esimerkiksi kosteudella, lämpötiloilla, alkukaarevuudella ja epäkeskisillä kuormilla voi olla merkittäviä vaikutuksia seinän kestävyyteen.
Tämän diplomityön tarkoitus oli tutkia hoikkien CLT-seinien käyttäytymistä pystykuorman alaisuudessa, kun mitoituksessa huomioidaan alkukäyryys, kuormien epäkeskisyys, kosteuden ja lämpötilan vaikutus, sekä virumasta aiheutuva seinän kaareutuminen. Työssä arvioidaan myös pystysuunnassa tuettujen seinien käyttäytymistä. Lisäksi arvioidaan eri tekijöiden vaikutusta seinän pystykuormakapasiteettiin ja vaakasiirtymään.
Alkukäyryyden ja epäkeskisten kuormien vaikutusta tarkasteltiin vertaamalla CLT-seinän kapasiteettia, kun ehdotetut minimiarvot sekä alkukäyryydelle että epäkeskisille kuormille otettiin huomioon tilanteeseen, jossa niitä ei otettu huomioon. Kosteuden ja lämpötilan vaikutus saatiin määrittämällä teoreettiset kosteuserot seinän eri puolille Suomen olosuhteissa ja tämän perusteella laskemalla seinään muodostuneet jännitykset ja taipumat.
Ottamalla huomioon ehdotetut minimiarvot alkukäyryydelle ja epäkeskisyydelle voitiin niillä havaita olevan selvä vaikutus CLT-seinän nurjahduskapasiteettiin. Kosteuden ja lämpötilan vaikutuksesta pahimmassa mahdollisessa tilanteessa Suomen olosuhteissa saatiin noin 3 MPa:n puristuskuormitus ulkolamellille ja pieni taipuma. Pystytuennoilla havaittiin olevan selvä vahvistava vaikutus seinän nurjahduskapasiteettiin. Hoikkien CLT-seinien tapauksessa nämä asiat on siis syytä ottaa huomioon rakenteiden mitoituksessa
Effects of Nitro-Oxidative Stress on Biomolecules : Part 2—Reactive Molecular Dynamics Simulations
No full textPublisher Copyright: © 2025 by the authors.In this review article, statistical mechanisms of oxidative modification reactions in various organic compounds under the influence of reactive oxygen species (ROS) generated by cold atmospheric plasma (CAP) are investigated and analyzed based on reactive molecular dynamics (MD) simulations. As an efficient and hygienic advanced oxidation technology, CAP demonstrates tremendous potential in fields such as biomedicine and environmental protection. Through simulations, this paper provides a detailed analysis of the interaction mechanisms between ROS and components of biological tissues and environmental toxins. In this paper, we review the reactions involving four major ROS (OH radicals, O atoms, (Formula presented.) molecules, and (Formula presented.) molecules) and organic compounds, including proteins, DNA, polysaccharides, fatty acids, antibiotics, and mycotoxins. Atomic-level analysis reveals various oxidative modification reactions induced by ROS and their resulting products, including dehydrogenation reactions, bond-formation reactions, oxygen-addition reactions, and bond-cleavage reactions. Additionally, the study elucidates the role of active functional groups in various organic compounds, the presence of special elements, and the specific reactive nature of (Formula presented.). Furthermore, the influence of different ROS species and concentrations on reaction types is explored, aiming to provide a solid theoretical foundation for the application of CAP technology in biomedicine and environmental remediation.Peer reviewe
Wi-Fi-hallintamikropalvelun suorituskykytestaus ja optimointi langattomissa yritysverkoissa
No full textThis thesis investigates the requirements of a Wi-Fi management micro-service to support management of 2000 Wi-Fi access points. Existing bottlenecks standing in the way of this goal are identified and fixed. Then, the hardware resource utilization of the microservice is analyzed and opportunities for optimization are identified and addressed. The thesis encompasses the design and implementation of a systematic testing approach to evaluate and optimize the microservice's performance. The methodology involves defining test cases and requirements and simulating protocol-specific management traffic by using a load testing tool, which was extended to support the specific demands of the study. Performance data is extracted through iterative testing phases, during which bottlenecks are identified and resolved using profiling tools, hardware metrics, and log analysis. The CPU utilization of the main component of the microservice was reduced by 83% and its memory allocations by 81%.Tämä diplomityö tutkii Wi-Fi-hallintamikropalvelun vaatimuksia 2000:n tukiaseman hallintaan. Tavoitteen saavuttamista estäneet pullonkaulat tunnistettiin ja korjattiin, minkä jälkeen mikropalvelun kuluttamia resursseja analysoitiin, optimointimahdollisuudet tunnistettiin ja optimoinnit tehtiin. Työssä suunniteltiin ja toteutettiin systemaattinen testausmenetelmä mikropalvelun suorituskyvyn arvioimiseksi ja optimoimiseksi. Menetelmä sisältää testitapausten ja vaatimusten määrittämisen sekä protokollakohtaisen hallintaliikenteen simuloinnin suorituskykytestaustyökalun avulla. Työkalun ominaisuuksia laajennettiin testauksen vaatimuksiin sopiviksi. Suorituskykytiedot kerättiin iteratiivisten testausvaiheiden kautta, joissa pullonkaulat tunnistettiin ja ratkaistiin profilointityökaluilla, laitteiston metriikoilla ja lokitiedostoja analysoimalla. Mikropalvelun tärkeimmän komponentin prosessorinkäyttöä onnistuttiin vähentämään 83 % sekä muistiallokaatioiden määrää 81 %
Molekyldynamiksimuleringar av dynamisk sprickutbredning i sköra material
No full textIn this thesis, molecular dynamics is used to simulate crack propagation in brittle materials. The purpose of the study is to investigate the suitability of classical interatomic potentials for describing atomic fracture mechanisms. Three materials (silicon, iron, and nickel) have been studied that have different crystal structures, allowing for a comparative analysis. In addition, two potentials per material were analyzed to assess potential discrepancies. The study is limited to materials exposed to tensile stress caused by dynamic loading. Pristine crystal structures containing a predefined seed crack were used for the simulations.
Griffith's theory has been used to calculate the theoretical values for the critical load for each material. The theory is limited to elastic materials that fracture in a brittle manner, meaning that no plastic deformation will occur before the fracture. The theory is formulated using continuum mechanics which does not take certain atomistic fracture mechanisms into account. Two of these mechanisms are lattice trapping and dislocation emissions, which will prevent propagation to some extent and cause a higher critical load.
Previous fracture studies on the same materials and crystal orientations have exhibited brittle fractures. Most of the simulations in this study showed typical characteristics of ductile fractures. The simulation of a crack in iron along the (110) crystal axis, in particular, exhibited dislocation emission. Although most of the materials did not undergo ideal brittle fractures, plastic deformation was limited and the simulated critical loads were close to the theoretical values. Additionally, a fracture study on hexagonal ice was attempted, but was not feasible due to long simulation times and incompatibility of the ice structure and potential.
This study has shown that classical potentials are capable, to some extent, of capturing crack propagation in brittle materials. Although the simulated critical load values are consistently higher than theoretical predictions. The fracture behaviour is consistent with previous experimental and computational studies.I detta diplomarbete har molekyldynamik använts för att simulera dynamisk sprickutbredning i sköra material. Syftet med arbetet är att undersöka om klassiska potentialer kan användas för att simulera atomiska mekanismer vid utbredning av sprickor. För att ge en bred översikt har tre olika material undersökts (kisel, järn och nickel), vilka alla uppvisar olika kristallstruktur. Två potentialer per material användes för att se möjliga skillnader mellan potentialerna. Studien är begränsad till dynamiska belastningar av dragspänning på en perfekt kristall som innehåller en ursprunglig spricka som behövs för att studera sprickutbredningen.
Griffiths sprödbrottsteori har använts för att beräkna teoretiska värden för den kritiska belastningen. Teorin är begränsad till elastiska material som uppvisar sprödbrott, detta innebär att ingen plastisk deformation kommer att ske innan sprickan börjar röra sig. Denna teori baserar sig på kontinuummekanik vilket orsakar att vissa brottmekanismer som ses på atomär nivå inte tas i beaktande. Dessa mekanismer är t.ex. gitterfångning eller emission av dislokationer vilka kommer att bromsa upp sprickutbredningen och orsakar ett högre värde för den kritiska belastningen.
I tidigare studier av sprickutbredning där samma material och kristallorienteringar har använts har spröda brott kunnat ses. Trots detta har de flesta simuleringar i denna studie uppvisat egenskaper som tyder på sega brott. Detta är särskilt tydligt i simuleringar av sprickor i järn längs (110)-kristallaxeln, där en stor mängd dislokationer har observerats. Trots att materialen inte har betett sig som ideella spröda material har ingen större plastisk deformation skett och de simulerade kritiska belastningarna har varit nära de teoretiska värdena. Simuleringar av brott i hexagonal is försöktes även, men på grund av långa simuleringstider samt inkompatibilitet mellan potentialerna och is strukturerna kunde inga simuleringar köras.
Detta arbete har visat att de klassiska potentialerna relativt bra kan simulera sprickutbredning i sköra material fastän de simulerade värdena för den kritiska belastningen har varit högre än de teoretiska. De egenskaper som setts i andra experimentella eller beräkningsstudier för liknande brott har även kunnat ses i denna studie
Fotogrammetria kalliomassan luokitteluun ja virtuaalikierrokseen
No full textAccurate rock mass characterization is essential for safe design of underground structures. While traditionally done through field mapping, advances in photogrammetry now allow remote, high-resolution mapping of rock faces. This thesis explores ways to improve the accuracy of photo-grammetric models by experimenting with different scaling object configurations. A virtual tunnel tour is also developed for educational use, and digital Q-system characterization is compared to traditional field mapping.
This study evaluates how the distribution, orientation and amount of alignment distances influence the accuracy, specifically root mean square error (RMSE) of defined control distances in the photogrammetric model. Furthermore, this study compares physical and digital mapping methods from a student perspective, evaluating confidence, perceived usability, and preferences using quantitative and qualitative surveys.
This study found that scaling object layout affects model accuracy, while alignment orientation had inconsistent effects and changing the amount of alignment distances per object had little impact. Comparing physical and digital Q-mapping, confidence levels were lower in digital method, especially for parameters like joint roughness, alteration, and water reduction. Despite the lower confidence levels, digital mapping was preferred overall.Tarkka kalliomassan kartoitus on olennaista maanalaisen rakenteen turvallisen suunnittelun kannalta. Perinteisesti tämä on tehty kenttäkartoituksella, mutta fotogrammetrian kehitys on mahdollistanut kallion kartoituksen etänä. Tässä opinnäytetyössä tutkitaan tapoja parantaa fotogrammetristen mallien tarkkuutta mittaobjektien asetteluiden avulla. Lisäksi opinnäytetyössä kehitetään virtuaalinen tunnelikierros opetuskäyttöön ja verrataan digitaalista Q-luokittelua perinteiseen kenttäkartoitukseen.
Työssä arvioidaan, miten mittaobjektien asettelu, suunta ja määrä vaikuttavat fotogrammetrisen mallin tarkkuuteen, erityisesti jäännösvirhehajonnan (RMSE) näkökulmasta. Lisäksi tutkimus vertailee fyysisiä ja digitaalisia Q-kartoitusmenetelmiä opiskelijan näkökulmasta arvioiden kartoitusvarmuutta, koettua käytettävyyttä ja mieltymyksiä kvantitatiivisten and kvalitatiivisten kysymysten avulla.
Työssä havaittiin, että mittaobjektien asettelulla on vaikutusta mallin tarkkuuteen, kun taas suunnan vaikutus oli epäjohdonmukaisesti ja määrän muuttamisella ei ollut merkittävää vaikutusta. Fyysisen ja digitaalisen Q-kartoituksen vertailussa varmuus digitaalisessa menetelmässä oli alhaisempi, erityisesti parametrien kuten karheuden, muuttuneisuuden ja vedenläpäisyvyyden osalta. Alhaisemmasta kartoitusvarmuudesta huolimatta digitaalinen menetelmä koettiin kuitenkin yleisesti miellyttävämmäksi
Geospatial analysis of toponyms in geotagged social media posts
No full textPublisher Copyright: © 2025 Hiraoka et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.Place names, or toponyms, play an integral role in human representation and communication of geographic space. In particular, how people relate each toponym with particular locations in geographic space should be indicative of their spatial perception. Here, we make use of an extensive dataset of georeferenced social media posts, retrieved from Twitter, to perform a statistical analysis of the geographic distribution of toponyms and uncover the relationship between toponyms and geographic space. We show that the occurrence of toponyms is characterized by spatial inhomogeneity, giving rise to patterns that are distinct from the distribution of common nouns. Using simple models, we quantify the spatial specificity of toponym distributions and identify their core-periphery structures. In particular, we find that toponyms are used with a probability that decays as a power law with distance from the geographic center of their occurrence. Our findings highlight the potential of social media data to explore linguistic patterns in geographic space, paving the way for comprehensive analyses of human spatial representations.Peer reviewe
From invention to impact - How Finnish deep tech startups navigate the journey to product-market fit
No full textDeep tech startups often struggle with product-market fit (PMF) despite their innovative potential. The aim of this research was to understand the challenges they
face, the strategies they employ and the role of the external startup ecosystem in aiding the startups in their journey towards product-market fit. Qualitative, semi structured interviews were conducted to understand the perspectives of both startup founders and the surrounding ecosystem actors.
Although Finland has a strong and developing ecosystem, which even includes tailored support for deep tech startups, many are still struggling to find PMF. Deep tech startups often struggle with balancing business and technical development while battling resource constraints, market education and validating market needs.
Startups adopt a range of strategies to face these challenges including using a collaborative approach, customer engagement, finding alternative markets and utilising support from the surrounding ecosystem. This thesis contributes to the limited academic research on deep tech commercialisation by outlining the ways in which startups can better align their business and technology approaches to increase the chances of finding product-market fit. It also offers recommendations for both startups and those in the Finnish deep tech ecosystem members interested in increasing the success rate of PMF in Finnish deep tech startups
Learned layer-level concepts in the machine learning model on the physical layer in 6G L1
No full textAs the telecommunications industry evolves from 5G to 6G, the integration of artificial intelligence (AI) and deep learning models into communication networks is becoming increasingly vital. While AI has demonstrated significant success in this field, the complexity of their solutions comes at the expense of interpretability. Gaining insight into the inner workings of these models is crucial for building trust in stakeholders, enhancing transparency, aiding developers in improving performance, and facilitating the identification of potential issues. In light of this, this thesis will investigate the interpretability of deep learning-based receiver applications. We propose using Neural Activation Pattern (NAP) as an interpretability technique and evaluate its applicability in our use case. Since NAP technique consists of two distinct components, clustering inputs processed similarly by the neural network and identifying learned layer concepts, these aspects serve as the basis for evaluating NAP technique in our deep learning receiver. Additionally, we proposed modifications to the original technique to improve clustering results. Our enhanced approach incorporates a distribution estimation method and an alternative normalization method to address the limitations of the standard NAP technique. Experimental results show that the proposed modifications can extract a greater number of patterns that are more stable, particularly when employing kernel density estimation as the distribution estimation method and standardization as the normalization method. The computational cost of the enhanced method is significantly higher, which might be mitigated by parallelization, at the expense of resource usage, meaning there is still a need for further improving the overall effiency. In terms of interpretability, as far as signal-to-noise ratio (SNR) and velocity are concerned, our analysis reveals that the most prominent concepts are characterized by SNR values, whereas the velocity is not proven to be a key concept. These results are agreed by both the original NAP technique by Bäuerle et al. and the modified one, though in several layers, the statistics of SNR can be different between the two variants. Nonetheless, we believe that interpreting and understanding the learned concepts from NAP technique requires further research and exploration. For example, regarding robustness of clustering algorithm, as the data in the thesis is left-skewed, log-transformation or power transformation could be explored as an alternative method for standardization; whereas regarding clustering evaluation, the viability of Density-Based Clustering Validation as a better metric for comparing clustering solutions is also a possible research direction
Self-Accelerating Drops on Silicone-Based Super Liquid-Repellent Surfaces
No full textPublisher Copyright: © 2025 The Authors. Published by American Chemical Society. | openaire: EC/HE/101062409/EU//SuperElectroDesign of super liquid-repellent surfaces has relied on an interplay between surface topography and surface energy. Perfluoroalkylated materials are often used, but they are environmentally unsustainable and notorious for building up static charge. Therefore, there is a need for understanding the performance of sustainable low surface energy materials with antistatic properties. Here, we explore drop interactions with perfluoroalkyl- and silicone-based surfaces, focusing on three modes of drop-to-surface interactions. The behavior of drops rolling under gravity is compared to those subjected to lateral and normal forces under constant slide (i.e., friction) and detachment (i.e., adhesion) velocities. We demonstrate that a drop’s characteristic and dynamic mobility depends on surface chemistry, with sequential drop interactions being particularly affected. By utilizing force-and-charge instruments, we show how rolling drops are primarily governed by adhesion and its associated electrostatic effects, instead of friction. Perfluoroalkylated surfaces continuously accumulate charges, while silicone surfaces rapidly saturate. Consequently, sequentially contacting drops accumulate significant charges on the former while rapidly diminishing on the latter. The drop charge suppressing behavior of silicones enhances drop mobility despite their higher surface energy compared to perfluoroalkyls. Quantum mechanical density functional theory calculations show significant differences in surface charge distributions at the atomic level. Simulations suggest that variations in the lifetimes of surface hydroxyl ions likely drive the markedly different drop charging behaviors. Our findings demonstrate the critical role of surface chemistry and its coupled electrostatics in drop mobility, providing valuable insights for designing environmentally friendly, antistatic, super liquid-repellent surfaces.Peer reviewe
Koneoppiminen osakemarkkinoiden ennustamisessa: Ennustustarkkuuden ja kaupankäyntisuorituksen arviointi
No full textThe growing popularity of machine learning in financial market analysis increases the need to evaluate not only prediction accuracy but also the practical profitability of these models in trading scenarios. In this bachelor’s thesis, I conduct a comparative analysis between three different supervised machine learning model. The prediction models XGBoost, Random Forest, and LSTM neural network were trained using the price data of Apple’s (AAPL) stock and the S&P 500 index from the years 2018–2023. Technical indicators such as moving averages, RSI, and price momentum were used as model inputs.
The objective was to compare the models in terms of their capacity to predict next-day levels and directions, and to evaluate how these predictions translate into simulated trading returns. of the stock price and to identify which input features improve model performance. For Apple stock, Random Forest generally performed best in prediction accuracy (MSE = 22.72; MAE = 3.69), while XGBoost worked best in predicting price direction changes (Hit Accuracy = 77.7%).
For the S&P 500 index, XGBoost performed best on both price level accuracy (MSE = 1383.87) and directional accuracy (Hit Accuracy = 81.4%). The key features of each model varied, and in many cases, simple price history-based features turned out to be more significant than more complex technical indicators.
The results showed that machine learning models offer effective tools for stock price forecasting, but model selection and proper tuning of parameters play a central role in maximizing prediction accuracy. A trading simulation on the S&P 500 revealed that although XGBoost had the highest directional accuracy, Random Forest and LSTM achieved higher cumulative returns, highlighting the importance of evaluating financial models beyond predictive metricsKoneoppimisen kasvanut suosio rahoitusmarkkinoiden analysoinnissa lisää tarvetta arvioida eri mallien ennustetarkkuutta ja näiden mallien käytännön kannattavuutta kaupankäyntiskenaarioissa. Tässä kandidaatintutkielmassa teen vertailullisen analyysin kolmen eri valvotun koneoppimismallin välillä. Ennustemallit XGBoost, Random Forest ja LSTM koulutettiin Applen (AAPL) osakkeen ja S&P 500 -indeksin hintatiedoilla vuosilta 2018–2023. Mallien syötteinä käytettiin teknisiä indikaattoreita kuten liukuvia keskiarvoja, RSI:tä ja hintamomenttia.
Tavoitteena oli tutkia, miten koneoppimismallit eroavat toisistaan seuraavan päivän osakehintojen ja hintamuutoksien suunnan ennustamisen tehtävässä. Lisäksi tavoitteena oli selvittää, mitkä muuttujat ovat suorituskyvyn kannalta keskeisiä. Applen osakkeessa Random Forest -malli saavutti parhaan kokonaistarkkuuden (MSE = 22,72; MAE = 3,69), kun taas XGBoost erottui edukseen suunnanmuutosten ennustamisessa (Hit Accuracy = 77,7%).
S&P 500 indeksin kohdalla XGBoost suoriutui selvästi muita malleja paremmin sekä hintatason (MSE = 1383,87) että seuraavan päivän suunnan ennustamisen osalta (Hit Accuracy = 81,4%). Jokaisen mallin keskeisimmät piirteet vaihtelivat ja monissa tapauksissa yksinkertaiset hintahistoriaan perustuvat ominaisuudet osoittautuivat merkittävämmiksi kuin monimutkaisemmat tekniset indikaattorit.
Oikeilla parametreillä säädetyt koneoppimismallit tarjoavat hyvät mahdollisuudet osakekurssien ennustamiseen. Niiden todellinen kannattavuus riippuu tarkkuuden lisäksi myös siitä, kuinka ennusteet osuvat yhteen juuri tuottoisien markkinaliikkeiden kanssa.
S&P 500-indeksillä suoritetussa kaupankäyntisimulaatiossa havaittiin, että vaikka XGBoost saavutti korkeimman suunnan ennustustarkkuuden, Random Forest ja LSTM tuottivat korkeammat kumulatiiviset tuotot. Tämä korostaa mallien arvioinnin laajempaa näkökulmaa pelkän ennustetarkkuuden sijaan