Student Posters Repository

The aim of my PhD is to computationally model acoustic instabilities within gas turbine combustion systems, with a specific emphasis on Low Frequency rumble. Instabilities in the combustion system of a jet engine can cause fatigue and failure if they are not full understood. Low frequency rumble occurs when the combustion flame burns unevenly creating hotspots or ‘entropy waves’ as the flow travels through the combustor. They are then accelerated through a nozzle creating a sound wave. Part of this sound wave is reflected back into the combustion system and at certain frequencies is amplified causing instability in the system. 

In this investigation, Direct Numerical Simulation (DNS) of a jet in a cross-flow under different streamwise pressure gradients (zero and favorable pressure gradient, hereafter ZPG and FPG) is performed. The goal is to accurately model the interaction between the inclined jet with respect to the incoming turbulent flow and to elucidate the physics behind the thermal coherent structures in cross-flow jets at different streamwise pressure gradients (ZPG vs. FPG) and velocity ratios (VR).

The Institute of Physics Belgrade (IPB) is a partner of the South East European Virtual Climate Change Center – SEEVCCC) responsible for implementation of the regional earth modeling studies, with a focus of the aerosol impacts to weather, climate and environment. IPB is participating in HPC Horizon2020 project VI-SEEM. High resolution regional climate modelling is one of the strongest climate applications in VI-SEEM. As the official sub-regional climate center of the WMO SEEVCC develops a regional Earth Modeling System (EMS) by integrating the modeling components for the atmosphere, aerosol, ocean and soil uses a multi-scale, computationally efficient atmospheric model NCEP/NMM that can be run on global and regional scales. Such model serves as a driver to Dust Regional Atmospheric Model (DREAM)

DREAM is currently used to perform investigation of ice nulceation due to dust. Dust aerosols are very efficient ice nuclei, important for heterogeneous cloud glaciation, even in regions distant from desert sources. We used DREAM dust-atmospheric modelling system with an added component for parameterizing dust-induced ice nuclei concentration (#IN) to predict conditions for cold cloud formation. For temperatures in the interval (-36°C; -5°C), we use the immersion ice nucleation parameterization developed by DeMott et al. (2015). For temperatures in the interval (-55°C; -36°C), we implement the Steinke et al. (2015) parameterization for the deposition ice nucleation. As a next step, we will use #IN as a predicted input parameter in a cloud microphysics scheme in order to improve prediction of processes related to cold cloud formation and consequences to precipitation. Further research will also consider effects of different minerals present in dust on these processes.

Simulations of turbulence past realistic, rough surfaces.

The Indian Monsoon is one of the most important circulation system in the Earth’s climate, the study of which being vitally important to an agricultural region inhabited by about 20 percent of the world’s population. The livelihoods of many people depend on the monsoonal region’s seasonal changes, which nourishes rain-fed crops for national and foreign export. Despite its significance in the economy, the monsoon continues to be poorly forecasted in weekly-monthly timescales by weather and climate models. Additionally, climate change over the past 30 years has yielded an increased frequency of extreme wet and dry spells, making farming more difficult and further pressuring scientists to better understand the physical processes driving the monsoon dynamical system. I highlight current mathematical frameworks describing the monsoon system and numerical weather prediction. Framework will include large-scale dynamics, convective cloud transports of moisture, and coupling approaches with respect to air-sea surface interactions and heat fluxes. Finally, the framework will aim to inform our understanding of the Monsoon Intra-Seasonal Oscillations (MISO) which sparks the summer Monsoon onset vortex in the Bay of Bengal.

Determining the structure of biological macromolecules by cryo-electron microscopy (cryo-EM) has, until recently, been limited to low resolution 3D density maps. Recent advances in electron detectors have now unlocked data at resolutions that rival X-ray crystallography, and many new computational processing tools have been designed to fully exploit the new technology.

Here, we show the process in determining the atomic structure for Thermus virus P23-77, a Sphaerolipovirus originally isolated from an alkaline hot spring in New Zealand. The outer shell, or capsid, of the virus has a spherical, spiked morphology with a diameter of 78 nm, and is made up of an unknown number of structural proteins that follow an icosahedral symmetry.

Using cryo-EM and single particle analysis, we have determined the structure of this bacteriophage to high resolution by applying icosahedral symmetry during image processing, potentially allowing for the identification of the protein composition of the icosahedral capsid. Imposing symmetry, however, degrades asymmetric information available in the structure. Localized and asymmetric reconstruction methodologies are required to resolve those components that do not conform to the icosahedral symmetry.