Procter & Gamble (United Kingdom)
GraFar - Repository of the Faculty of Civil EngineeringNot a member yet
3934 research outputs found
Sort by
Koncept opraštajućih puteva i njihov značaj u unapređenju bezbednosti saobraćaja na putevima
No full textKada su u pitanju važni društveni aspekti saobraćaja, kao jedan od aspekata koji značajno reflektuje uticaj putne infrastrukture na društvo u celini, je svakako i aspekt bezbednosti saobraćaja na putevima. Savremeno društvo kao takvo za sobom nepobitno povlači kontinuirano ulaganje u razvoj putne infrastrukture, u cilju što efikasnije razmene ljudi i dobara, ali to takođe kao negativan efekat za sobom povlači i uticaj na život i zdravlje ljudi koji se tim putevima kreću. Jedan od načina kako je ne tu bezbednost moguće uticati je i primena određenih tehničkih rešenja i alata u fazi eksploatacije puta, koja značajno mogu poboljšati bezbednost određenog putnog pravca. U radu će akcenat biti na objašnjenju koncepta „samoopraštajućih“ puteva, njihov značaj u poboljšanju
bezbednosti na putevima, kroz konkretne primere njihove primene u domaćoj i inostranoj putnoj praksi. U zaključcima rada biće navedene konkretne mere za efikasniju primenu koncepta opraštajućih puteva u domaćoj putnoj infrastrukturi
MaPLoRds: Mobile Application for Local Road Network Risk Assessment
No full textMaPLoRds is a software system developed to enable local municipality users to assess the resilience of their local road network from natural hazards in climate-changing conditions. The system analyzes the characteristics of the road network, vulnerability, and climate-related hazards, to evaluate the network’s resilience. A tool for mobile devices (mobile application) is developed to facilitate the data collection in the field and a web application is developed for data analysis, resilience, and priority assessment based on the methodology. The bilingual glossary, in Serbian and English, is produced for the developed version of the application, introducing key terms and phrases that are required for the MaPLoRds interface, both for mobile and web application. The system is developed according to the highest industry standards and by Platform Security and Personal Data Protection rules. Personal Data of the users (e.g., password) are stored in encrypted form in the central database, while all other non-sensitive data are not encrypted. In this paper, the MaPLoRds mobile application is available on https://maplords.rgf.bg.ac.r
Comparison of CNN Architectures for Earthquake Damage State Classification of Different Building Types
No full textRapid and reliable post-earthquake damage assessment of the built environment is critical for prioritising response and recovery efforts. This study investigates deep-learning strategies for classifying different building types into three damage states: slight, moderate and severe, using photos that capture both, the structure and its surrounding context. Unlike many prior works that rely on tightly cropped images of specific cracks, our dataset intentionally preserves a wider field of view, making the task more realistic yet significantly more challenging.
Two convolutional neural network (CNN) architectures are examined: ResNet-18 and ConvNeXt-Tiny. A two-stage training protocol is employed. In the first stage, before training on the original photos, a ResNet-18 model is pre-trained on subset composed exclusively of close-up patches, providing a strong damage sensitivity. Second stage uses the original training set of the photos, extended using horizontal-flip augmentation. The same protocol is then replicated for ConvNeXt-Tiny architecture.
Results show that dataset expansion and simple augmentation improve ResNet-18 recall, confirming the benefit of contextual diversity. Nevertheless, ConvNeXt-Tiny consistently outperforms ResNet-18, achieving a recall of 72% and an F-score of 0.85. The findings highlight the importance of appropriate selection of CNN architectures, especially for demanding tasks.
The proposed workflow offers a pragmatic route toward scalable, context-aware earthquake damage assessment tools, and provides a benchmark for future research on holistic post-disaster image analytics
Experimental analysis of thermal conductivity of fly ash from “Kostolac” thermal power plant
No full textFly ash, a by-product of coal combustion, represents both an environmental challenge and a potential resource for sustainable construction. This paper presents the results of laboratory investigations of the thermal conductivity of fly ash and an assessment of its possible application in geotechnical systems. The results show that moisture content has a dominant influence on thermal properties, while the effects of bulk density and curing are less significant. The measured values indicate that fly ash exhibits stable and consistent thermal behavior, suitable for engineered fills or soil-ash mixtures
Towards an advanced geotechnical modelling of block-in-matrix rock for robust tunnel design and construction
No full textBlock-in-matrix rock assemblages, also known as ’bimrocks’, represent structurally complex units consisting of hard rock blocks embedded in a soft matrix, with both components differing in geological origin, lithology, rheology and geometry. Considered geotechnically complex formations and being characterized by internal heterogeneity and multi-level spatial variability, such formations are extremely demanding to model using conventional geotechnical approaches. Underground construction, including urban tunnelling in such heterogeneous environments poses numerous challenges, with the most significant ones being high stratal disruptions, stress concentrations at the interface between blocks and matrix, and the inability to accurately predict the behaviour of the rock mass. Besides geological formation, structural analysis and geomechanical characterization, previous studies have only explored methods for probabilistic generation of various block configurations
within the rock mass. To tackle the bimrock as an underexplored issue, we present a methodology for up-to-date bimrock
ground modelling including lithological and mechanical parameter spatial variability assessment, paired with numerical simulations for the underground construction. Employing the three main modelling components – voxel based geological model, conditional random field parameter model and FEM numerical model, this study aims at introducing a novel, automated and reliable approach for block-in-matrix modelling. It presents an innovative framework for geotechnical modelling of bimrocks, with significant potential for further improvement in accuracy and functionality of the modelling components, as well as the identification of the most sensitive modelling parameters
Computer-aided ground modelling incorporating soil variability for geotechnical applications
No full textThis paper presents an end-to-end workflow for generating three-dimensional, voxel-based ground information model that seamlessly integrates geological complexity and soil-property variability into finite-element analyses. First, lithology is interpolated on a structured grid via Empirical Bayesian Kriging in ArcGIS Pro, producing a voxelized subsurface architecture. Next, spatial variability of key geotechnical parameters is characterized with conditional random fields using GSTools, utilizing borehole observations to ensure realistic site-specific conditions. All modelling components are stored and linked in NetCDF file, preserving dimensions and metadata, while allowing for the efficient data exchange. Finally, voxel-based ground information model is imported into a CutFEM solver for adaptive finite element simulations. We demonstrate the workflow on a synthetic excavation scenario, highlighting its computational efficiency and improved accuracy in predicting deformation compared to deterministic models. By automating the entire process and leveraging open data standards, this approach addresses interoperability challenges and enables engineers to incorporate uncertainty directly into numerical designs
Testing independence in high dimensions: a Kendall’s tau extension for incomplete data
No full textWe consider the problem of testing independence in high-dimensional settings with missing data. Building upon a recently proposed Kendall-based statistic, we introduce two new modifications specifically designed to accommodate incomplete observations. The proposed methods are studied from both theoretical and empirical perspectives. A comprehensive simulation study illustrates the robustness and applicability of the new approaches. The findings contribute to the development of nonparametric methods for analyzing high-dimensional and incomplete data structures
How close are opportunistic rainfall observations to providing societal benefit?
No full textMitigation of water-related hazards as well as sustainable water resources management are conditioned on accurate and detailed spatio-temporal rainfall observations. Today, water authorities like National Meteorological and Hydrological Services (NHMS) in developed countries operate observation systems consisting of meteorological stations and weather radars. These observations provide state-of-the-art precipitation products, but they remain error prone due to device-specific limitations. This has driven growing interest in opportunistic sensors (OS) of rainfall, primarily Commercial Microwave Links (CML) and Personal Weather Stations (PWS). In the Global South, where meteorological station networks are usually very sparse, OS rainfall data conceivably has an even higher potential to provide an added value. However, although numerous studies have demonstrated the capability and potential of accurate rainfall estimation by OS, no dedicated investigation has been made with regard to their application for operational monitoring and prediction. How close are OS rainfall data to providing societal benefit, e.g. by widespread integration in existing hydro-meteorological observation and prediction systems? We address this question by (1) making a review of studies that use OS rainfall data in applications (rainfall mapping, nowcasting and hydrological prediction), (2) providing a status report on the transition from research to operational usage from the perspective of EU COST Action OpenSense, and (3) discussing the challenges NHMS face in deploying OS rainfall data in operational services. We conclude that while distinct challenges still remain, in terms of both access and processing, the applicability of OS rainfall data is well scientifically supported and operation is underway in several countries
Modelling the Seismic Response of a Three-Storey Confined Masonry Building Using an Equivalent Diagonal Strut Models
No full textConfined masonry (CM) technology is extensively employed in many parts of the world for the construction of residential buildings due to its improved seismic performance compared to the unreinforced masonry (URM). This study focuses on the numerical simulation of seismic response of CM walls and a three-storey model of a CM building using Equivalent Strut Models (ESM) in OpenSees framework. Different strut and constitutive models available in the literature were used. The primary objective is to identify the most suitable model for simulating the global response of CM walls and structures. Firstly, different numerical models of the CM wall were compared against two full-scale shear compression tests, which enabled us to identify the most suitable and accurate model. The most suitable model was then used to develop a numerical model of the three-storey CM building, and the results of the simulations were compared with the experiment
Seizmički proračun i ojačanje zidane zgrade vrtića
No full textZidane konstrukcije čine veliki deo fonda postojećih objekata u Srbiji i regionu. Zidane zgrade, kao česta tipologija u građevinskom fondu, pokazuju značajnu ranjivost na seizmička dejstva usled krhkog ponašanja i nedostatka duktilnosti. U radu je prikazan primer seizmičkog ojačanja postojeće zidane zgrade, sa naglaskom na izbor odgovarajuće tehnike intervencije u cilju povećanja globalne stabilnosti i nosivosti konstrukcije. Analiza obuhvata proračun konstrukcije na seizmička dejstva primenom savremenih metodoloških pristupa. Dobijeni rezultati ukazuju na značajan doprinos izabrane metode ojačanja u pogledu povećanja sigurnosti objekta i smanjenja seizmičkog rizika