In this work, I propose a hybrid particle simulator for charged particles. The simulator consists of a physics-informed neural network, which can handle arbitrary external electric fields with continuous coordinates by solving the Poisson equation, and a graph-based algorithm that computes the interacting forces between the particles. The simulator is then applied to a set of particles inside a square domain under the influence of some external electric field. As the system evolves in time, particles will gradually leave the domain causing the particle density of the domain to change. This work aims to find a model which describes the particle density evolution of the system.

No description available.

We have performed experimental infections of Atlantic salmon (Salmo salar) from the Baltic Sea region with the protozoan parasite Spironucleus salmonicida. By this infection trial we have been able to show that the life cycle may use an alternative route of transmission. Instead of the commonly used faecal-oral route, the parasites can also be excreted directly into the surrounding water from the mucous layer of the skin or from an ulcerated skin lesion. Three different stages of the infection were identified: one intestinal, one blood stage and one tissue stage. New putative life-cycle forms of S. salmonicida cells detected by ex-vivo cell-cultures. Similar cells were also observed in imprints from skin lesions sampled during the experimental infection trials. The cells differed morphologically, from normal trophozoite cells and indicate an alternative cell-cycle when exposed to salmonid host cells.

The goal of this project was to determine the source strength of an Americium-Beryllium neutron source by irradiating metal foils of different elements. The irradiated foils were then placed on a High Purity Germanium detector and γ-decay from the decay of the neutron induced reaction products was observed and used to deduce the neutron flux. The results have been found to be in the order of magnitude of S = 2 · 105 neutrons/second, which is in fair agreement with the source strength calculated from the manufacturer’s specifications. In addition, in order to improve the reproducibility of the measurements, flexible holders have been created using 3D printers.

This work aims to plan, prepare and perform a test campaign of the Back-End Electronics(BEE) module in the phasemeter instrument on board the Laser Interferometer Space Antenna(LISA). Results from the test are analyzed to thermally characterize the BEE module as wellto evaluate how successfully an aluminum enclosure mounted on the BEE can dissipate heatloads from the BEE. In addition to the experimental tests, has a computational thermalmodel been developed in COMSOL. The experimental tests were performed in medium vacuum pressure at the Satellite TestEnvironment for Laser Applications (STELA) laboratory at Albert Einstein Institute forGravitational Physics in Hanover, Germany. Steady-state tests show that the aluminumenclosure successfully can dissipate heat-loads of 18.94 [W] from the BEE. Temperatures ofthe BEE would never exceed 55◦ C during testing which is well below the upper temperaturelimit of 75◦ C, thus resulting in a stable thermal environment. Simulations performed of theBEE model also show encouraging results. During time-dependent simulations, temperaturesof critical points from the model would never deviate more than 1.5◦ C from temperaturesof the experimental tests.

The para-subthalamic nucleus (PSTN) and amygdala are two brain regions of central focus in recent neurobiology. The PSTN is a hypothalamic brain region that acts as a relay point for sensory signals towards regions that use learned behaviour and experiences to integrate the bottom-up sensory information. One region that the PSTN sends sensory information to is the amygdala, which uses certain emotive responses to form higher cognitive function and to produce new learned behaviour. Due to the PSTN acting as the middle man of sensory integration, it connects interoception and emotion within the brain. The amygdala contains several different nuclei which can be distinguished using different molecular markers. The aim of this study is to distinguish between spatio-molecular subpopulations or anatomically identified subnuclei in the amygdala and PSTN by analysing Tac1-Cre driven GFP expression in those areas. This was tested using immunohistochemistry techniques on Cre/loxP crossed mice. Tac1 is a gene that encodes for preprotachykinin-A, which can be transformed into several other neuropeptides. The results found that Tac1-Cre driven GFP expression was especially visible and dense in the dorsal anterior amygdala and medial posterior amygdala. Scattered expression was observed in the lateral region of the amygdala. Lack of expression was further observed in the lateral and basolateral amygdalar nuclei. Tac1-Cre driven GFP positive cells were also seen in and around the PSTN. These results showed a strong expression in the central amygdalar nucleus  and some areas of the medial amygdalar nucleus, as well as throughout the extent of the PSTN.

Yes / This paper explores the relationship between news media sentiment and spillover effects in the cryptocurrency market. By employing a time-varying parameter vector autoregressive model, we initially develop measures of spillover specific to individual cryptocurrencies. Subsequently, we employ unique data on cryptocurrency-specific sentiment to assess its impact on these spillover measures using panel fixed effects regression analysis. Our findings indicate that news media sentiment plays a significant role in explaining the spillover dynamics within the cryptocurrency market. Unlike traditional assets, it appears that only positive sentiment affects the spillovers among cryptocurrencies, suggesting an asymmetric effect. Taking into account various characteristics of cryptocurrencies, we find that sentiment’s impact on spillover is more pronounced in community-based coins than in those driven by firms. An examination of news content suggests that sentiment pertaining to emotional and risk aspects of cryptocurrencies predominantly influences these spillovers. Additionally, a comparative analysis of sentiment derived from social media and traditional news sources reveals a stronger influence of the former on spillover effects. Through extensive robustness checks, our research consistently affirms the pivotal role of sentiment in driving spillovers among cryptocurrency returns, underlining the importance of sentiment analysis in understanding the dynamics of the cryptocurrency market. / The full-text of this article will be released for public view at the end of the publisher embargo on 11 Sep 2025.

The exploding foil initiator (EFI) is a high voltage detonator used to initiate explosions. It is designed to improvesafety standards by lowering the risk of accidental detonation and in doing so allows for in-line integration in, forexample, weapon systems that minimizes their complexity and thus reduces the number of possible failures. It ishighly reliable in terms of timing and avoids that functionality of the detonator deteriorates over time, which has beena problem in earlier designs of detonators. This thesis aims to develop an understanding of each part of the initiationprocess by means of computer simulations. The proportions of the electrically conducting bridge is varied in order tofind relations that optimize the design. Success depends on the simulated pressure generated inside a hexanitrostilbene(HNS) primer.The results showed that increasing the dimensions of the bridge greatly affects the pressure produced in the primer,but its proportions were still important. Unfortunately the simulations of the electrical explosion and the accelerationof the flyer suffered from convergence issues that rendered the flyer’s velocity graphs incomplete. But even with theseshortcomings, thanks to empirical data from earlier studies that analysed EFI prototypes, it was possible to makepredictions about the success of various set ups. The information gathered in this thesis should serve as a foundationfor future development of computer models of the EFI technology that will aid the production of prototypes that meetthe specific requirements.i / Den explosiva folieinitieraren (EFI) är en högspänningsdetonator som används för att initiera explosiva förlopp. Denär designad för att förbättra säkerheten genom att sänka risken för oavsiktlig detonering och till följd av detta möjligörför in-line integrering i, till exempel, vapensystem. Detta minimerar komplexiteten av sådanna system och sänker antaletmöjliga fel som kan uppstå. Denna teknolgi är högst pålitlig när det gäller timing och undviker att detonatorns funktionförsämras över tid, vilket har varit ett problem i tidigare konstruktioner av detonatorer. Denna avhandling syftar tillatt utveckla en förståelse för varje del av initieringsprocessen genom datorsimuleringar. Proportionerna av den elektrisktledande bryggan varierades för att hitta relationer som optimerar designen. Framgången beror på det simulerade trycketsom genereras inuti en primer av det explosiva materialet hexanitrostilben (HNS).Resultaten visade att ökade dimensioner på bryggan kraftigt påverkar trycket som produceras i primern, men dessproportioner var fortfarande viktiga. Tyvärr led simuleringarna av den elektriska explosionen och accelerationen avflyern av konvergensproblem som gjorde att flyerns hastighetsgrafer blev ofullständiga. Men även med dessa brister,tack vare empiriska data från tidigare studier som analyserade EFI-prototyper, var det möjligt att göra förutsägelserom framgången för olika uppsättningar. Informationen som samlats in i denna avhandling bör tjäna som en grund förframtida utveckling av dator modeller av EFI-teknologi som kommer att underlätta produktionen av prototyper somuppfyller specifika krav.ii

This thesis investigates the application of two different modeling techniques for simulating nuclear fuel rods: Constructive Solid Geometry (CSG) with Boolean operations in Serpent and mesh-based modeling in OpenFOAM. The purpose of the study was to evaluate how a mesh-based model performs compared to a CSG-based model when conducting simulations on a relatively simple geometric configuration. The reasons for using mesh-based geometry include the ability to couple Serpent with OpenFOAM, enabling simulations in both codes using the same geometry, and accommodating users who might be more comfortable with mesh-based software. By performing a direct comparison between the results from the two modeling methods, the project aimed to assess the accuracy, efficiency, and practical usability of mesh-based geometry in nuclear simulations. The results indicate that mesh-based modeling, particularly with finer mesh, can achieve a level of detail and accuracy comparable to that of CSG models. However, the simulations with finer mesh took significantly longer to complete, highlighting the need for careful consideration of mesh size depending on the specific requirements of the simulation. The study encompasses several critical aspects, including computational speed and the precision of the resulting simulation data. The analysis demonstrates that while mesh-based modeling offers high precision and flexibility, it is more computationally intensive. In contrast, CSG is faster and simpler to implement for less complex geometries, but may not adequately capture more intricate details. The conclusions suggest that mesh-based modeling can be a viable alternative for complex simulations, provided that the computational resources are available. By highlighting the potential and limitations of mesh-based modeling in comparison to traditional CSG techniques, the study provides valuable insights. Utilizing mesh-based models is highly advantageous when CFD simulations are planned, as it avoids the need to simulate one model in Serpent and another in CFD. Thus, mesh-based geometry is a promising toolkit for achieving high precision and efficiency in detailed and intricate simulations within the nuclear field.

Aurora Kinase A is a kinase involved in multiple signaling pathways and interactions during mitosis, making it an essential kinase that deregulated causes cancer diseases in affected patients. Structural research shows mainly static snapshots of possible conformations of the partly disordered protein. This is due to challenges in generating a monodisperse pure sample with high stability enough for dynamic biophysical measurements. Optimizing the production of Aurora A and validating constructs with different sequential lengths using light scattering techniques, thermal stability screening, mass characterization, mass spectrometry, and immunoassay techniques is important for future structural insights useful for drug discoveries. In this project, validation of constructs concluded that no significant difference in cleavage of His-tag, purification possibilities, monodispersity nor stability is shown by variate start residue from 118-122 to end residue 403. Expression of an Aurora Kinase A constructs with sequential length 118-403 is preferred to be executed at 18 degrees, otherwise, temperature differences during expression show no impact on produced Aurora A. Magnesium chloride has been shown to have an impact on stability where a higher concentration stabilizes Aurora Kinase A. Moreover, concentration differences of NaCl were shown to not affect the stability of Aurora A. During this project a polydisperse sample was generated and has given insights into Aurora A´s behavior in solution.