Machinery - Repository of the Faculty of Mechanical Engineering, University of Belgrade
Not a member yet
8397 research outputs found
Sort by
OPTIMIZATION OF STANLEY PLUM DRYING PROCESS USING SOLID FUEL BATCH DRYER WITH TRAYS
Get PDFDrying fruit is a traditional preservation method that has been used for thousands of years. This technique allows for long-term storage of fruit, keeping it fresh and usable under ordinary storage conditions. In this research, the drying process of the selected variety of Stanley plums was analyzed in a batch dryer with trays using solid fuel. The research aimed to calculate the optimal air temperatures at the inlet and outlet of the heater, as well as the temperature at the outlet of the dryer itself. The constructed dryer has double trays, allowing drying capacities from 1550 kg to 3100 kg of Stanley plums. The main focus was on achieving efficient drying with minimal heat loss. The selection of basic equipment, such as a hot water boiler, circulation pump, heat exchanger, and fan, was tailored to the specificities of the dryer to maximize energy efficiency. The heat
exchanger utilizes compact lamellar Cu-Al technology, while the fan is equipped with a filter section for a cleaner drying process. The use of solid fuels such as biomass brings significant economic benefits, reducing the costs of dried products. However, manual labor and detailed equipment inspection are necessary for each drying cycle, which is the main drawback of this method
MECHANICAL PROPERTIES OF PA12 SPECIMENS PRODUCED USING SLS TECHNOLOGY
No full textIn this study, the research focuses on tensile PA12 specimens with CAD model dimensions selected according to the ISO 527-2 standard, with bulk dimensions of 170x20x4 mm. The study utilizes a Fuse 1 (FormLabs, Summerville, MA) 3D printer that employs SLS technology. This printer is capable of producing objects with different shapes and dimensions simultaneously, provided that they are printed at a minimum distance of 5 mm apart.
The research involves producing two batches of specimens, each differing in printing orientation (i.e., horizontal and vertical). After printing, the specimens are tested using a Shimadzu AGS-X universal testing machine with a capacity of 100 kN to evaluate the mechanical characteristics of the materials. The findings reveal that horizontally printed specimens have a modulus of elasticity about 9.3% higher than the manufacturer's stated value, while vertically printed specimens show an increase of approximately 11.4%. This enhancement is likely due to the thermal and cooling cycles during printing. The difference in modulus of elasticity between the orientations is around 2%, indicating a minor impact on mechanical properties
An advanced machining robot flexible programming methodology supported by verification in a virtual environment
No full textThe solutions for robot offline programming use different programming environments to facilitate
the deployment of robots in machining tasks with the development of specific postprocessors.
There is no easy exchange of data between programming software to realise robot machining
tasks. This paper presents a flexible programming methodology using several interchange file
formats that can be easily exchanged between different CAD, CAD/CAM, or specialised robot
programming software. Integrating the standard programming based on CAD, CAD/CAM systems,
and STEP-NC protocol through different output files enables data interoperability during the
realisation of the robot machining tasks. The developed methodology is proposed for executing
programming, verification supported by a virtual environment, postprocessing, and machining by
industrial robots. It uses the software for programming machine tools and adapts them to the
specifics of robots and their programming languages using a developed postprocessor. The
presented methodology enables the programming of robots for 2.5 to 5-axis machining tasks
depending on the neutral file format used. Programming verification was realised, first, by simula
tion on configured virtual robots and developed robot native language editor and, second, by
machining on the available robots. These experiments include different contours and shapes and
file formats, which show the programming’s reliability and accuracy
Expertise on Environmental and Social Assessment of 12 SHPPs in Western Balkans - Final report
No full textM103 - The expertise is done for Deloitte Consulting co. The investor was EIB (European Investment Bank)
Environmental and Social Assessment of SHPP Brza voda 3 - Expertise
No full textM103 - The expertise is done for Deloitte Consulting co. The investor was EIB (European Investment Bank)
Prototype of SCARA-type industrial robot
No full textOblast primene industrijskih robota se brzo širi uz stalnu potražnju za poboljšanjem njihovih
funkcija, tehničkih karakteristika kao i sistema upravljanja i programiranja. Jedan od ciljeva
aktuelnih istraživanja u okviru Laboratorije za robotiku i veštačku inteligenciju je razvoj
domaćeg industrijskog robota sa mogućnošću automatizovanog programiranja na osnovu
informacija dobijenih od kamere. U okviru ovog rada je predstavljen prvi deo istraživanja koji
podrazumeva razvoj i realizaciju 4-osnog industrijskog robota SCARA konfiguracije. Profesor
Hiroši Makino sa Univerziteta Jamanaši projektovao je SCARA (Selective Compliance Assembly
Robot Arm) robota, koji predstavlja najpoznatiju konfiguraciju robota nastalu na
univerzitetima. Ovaj deo istraživanja obuhvata kinematičko modeliranje projektovanog robota,
njegovu fizičku realizaciju i razvoj sistema upravljanja i programiranja sa integracijom virtuelnog
robota. Sistem upravljanja je realizovan uz pomoć sistema upravljanja otvorene arhitekture
LinuxCNC, koji omogućava dalji razvoj i integraciju informacija sa kamere kao spoljašnjeg senzora.
Razvijeni sistem upravljanja obuhvata i model virtuelnog robota koji predstavlja digitalnu senku
realizovanog SCARA robota. Verifikacija kompletno razvijenog prototipa robota je sprovedena
kroz nekoliko primera realizovanja tehnološkog zadatka laserskog senjenja i graviranja.The field of application of industrial robots is expanding rapidly with a constant demand for improving their functions, technical characteristics as well as control and programming systems. One of the goals of current research within the Laboratory for Robotics & AI is the development of a domestic industrial robot with the possibility of automated programming based on information obtained from the camera. This paper presents the first part of the research, which includes the development and realization of a 4-axis SCARA-type industrial robot. Professor Hiroshi Makino from Yamanashi University designed the SCARA (Selective Compliance Assembly Robot Arm) robot, which representing the most famous robot configuration created at universities. This part of the research includes the kinematic modeling of the designed robot, its physical realization, and the development of the control and programming system with the virtual robot integration. The developed control system is based on the open architecture control system LinuxCNC, which
enables further development and integration of information from the camera as an external sensor. The developed control system also includes a virtual robot model that represents a digital shadow of the realized SCARA robot. The complete robotic system verification was performed through several examples of the technological task of laser cutting and engraving.Za izdavača: Dekan, dr Vladimir Popović, red. prof.
Tehnički urednici:
Prof. dr Saša Živanović, Doc. dr Miloš Pjević, Doc. dr Dušan Nedeljković, Lazar Matijašević, mast. inž. maš.
Štampa: Planeta print
Development of prediction methods required for stability assessment of river-sea ships
No full textThe natural roll frequency and the effective wave slope coefficient are two essential parameters for dynamic ship stability assessment. While these parameters may be calculated based on first principles, or even experimentally, simplified semi-empirical methods still represent a more practical approach for everyday use. However, there is a lack of adequate methods for river-sea ships. Therefore, the paper presents regression formulae for added inertia, for use in the prediction of natural roll frequency, and for effective wave slope coefficient at the natural roll frequency. The performance of the developed formulae is assessed with respect to the original data they were derived from, as well as with respect to a test dataset. Furthermore, for the prediction of natural roll frequency, a comparison is carried with an existing semi-empirical method specifically intended for inland vessels and river-sea ships, showing that the formulae proposed in the paper could lead to significant improvements in prediction capabilities
A CFD-Based Analysis of Bow Modification Influence on Ship Resistance and Energy Efficiency
No full textDifferent approaches can be applied in order to obtain reduction in the amount of greenhouse gases that are emitted by merchant ships every day. Those methods are divided into two groups, operational methods and technical methods. Regarding technical methods, improvement in ship design through bow retrofit is a possible solution for existing ships, whether by means of different bow shape (vertical stem, axe bow, bulbous bow...) or different bulb shape (Elliptical shape, Nabla shape, Delta shape). In this study, this technical method is assessed by analysing viscous, pressure and total resistance of a bulk carrier through a number of full-scale computational fluid dynamics (CFD) simulations. The simulations were conducted for the original bulbous bow, three modified models with the previously mentioned different bulb shapes and for three versions of vertical bow stem. All simulations were conducted for speeds of 12kn, 14kn and 16kn at design draft (11m) and scantling draft (12.8m). The energy efficiency of the seven different hulls are analysed based on the results of CFD self-propulsion simulations and calculated Energy Efficiency of Existing Ship Index (EEXI)
Advancing CFD validation methods for self-propelled inland waterway vessels a workshop initiative
No full textThis paper represents a workshop about developing
methodologies for validating computational fluid dynamics (CFD)
simulations concerning full-scale self-propelled inland waterway vessels.
Given the absence of previous workshops addressing this vessel category, the
primary objective is to establish validation methods through the acquisition
of self-propulsion speed and power measurements. Methodological
consistency is paramount, emphasizing the creation of comprehensive 3D
models covering vessel hull, appendages, superstructures, and propulsion
components using original ship documentation and 3D scanning technology
where necessary. Result presentation encompasses various hydrodynamic
parameters, including resistance forces, thrust, wake fraction, thrust
deduction factor, propeller performance, delivered power, shaft speed and
vessel speed evaluated under calm water and self-propulsion conditions.
Companies and institutions will be invited to join the workshop, where
participants will validate their CFD methodologies using standardized
models and analyze grid uncertainties across computational densities. The
workshop aims to establish a validation framework for CFD simulations in
inland waterway cargo vessels, with future plans for extending validation to
shallow water river trials. Collaboration and transparency are emphasized,
with respect for participants' preferences regarding result disclosure
DESIGN KNOWLEDGE V.S. GENERATIVE AI
No full textThe application of artificial intelligence (AI) is still a controversial topic in creative disciplines. Recent advances in AI field like generative adversarial networks (GAN) and large language models (LLM) brought about conditions for the development of numerous content creation systems that today are referred to as generative AI. Since architecture, as one of the creative disciplines, deals with the generation of models and documentation on the basis of which buildings are built; naturally the notion of generative AI has attracted a lot of attention.
However, the application of these technologies requires a detailed understanding of the algorithms underlying generative AI, as well as the nature of the architectural process. Most current AI algorithms are based on learning from existing digital media such as text, graphics, audio, and video. For the typical person today who consumes all content through electronic media, this seems like a universal solution for all disciplines. However, the architectural process is far more complex. First, the texts and drawings that the architect creates are not the final expression but constitute a representation of the architectural object being designed. On the other hand, design knowledge is not classical propositional knowledge, but knowledge in action that is created through the design process and which as such can never be complete and represented in some static form. Considering that the life of architectural objects is mainly limited by the failure of the object's function and not its construction, design knowledge represents a crucial link in ensuring the resilience of architectural objects.
The paper provides a fundamental analysis of existing AI deep learning algorithms, as well as a study of theories about the design process and considers the ways in which they can be interconnected