2022 Cilt 27 Sayı 1

Permanent URI for this collectionhttps://hdl.handle.net/11452/27900

Browse

Now showing 1 - 11 of 11

Classification of original and counterfeit gold matters by applying deep neural networks and support vector machines
(Bursa Uludağ Üniversitesi, 2022-02-27) Can, Yekta Said
Gold is one of the most counterfeited precious metals. The color of copper is like gold. For this reason, copper is one of the most used materials for color counterfeiting. When the chemical properties are concerned, wolfram is like gold (density of gold and tungsten are 19.30 g/ml and 19.25 g/ml, respectively), so it can be used as a chemical counterfeit. The purity of gold can be determined by X-ray, but this method is costly. The current low-cost methods of jewelers have been experimented with for counterfeit gold detection in this paper. When a gold matter is hit by a subject, the sound frequency is higher than the frequency of sound when the same experiment is done with copper. Furthermore, counterfeit gold color is brighter than real ones. The color of gold is unique, and it is called "gold yellow". In this research, by employing sound and image processing, counterfeit and original gold are differentiated. For the image processing part, first a Convolutional Neural Network (CNN)-based toolbox for segmenting the gold material is applied. Then, deep CNNs for differentiating the color of the gold and copper materials are employed. Promising results are achieved with both sound and image processing techniques.
Determination of press brake bending parameters for Hardox 400 steel
(Bursa Uludağ Üniversitesi, 2022-01-03) Aydemir, Fatih; Çakan, Betül Gülçimen; Durmuş, Ali; Çavdar, Kadir; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Makine Mühendisliği Bölümü.; 0000-0003-1739-1143; 0000-0003-2487-7344; 0000-0001-9126-0315
In the present study, a specific hydraulic press brake based test setup (HPBTS) was designed and manufactured which was used to determine the bending forces and the punch displacements of Hardox 400 steel depending on the plate thickness, plate length and mechanical properties. The setup was equipped with an image processing system to detect the change of the bending angle during forming. Prior to bending tests, tensile tests were carried out and the identified mechanical properties were compared with the ones reported by the manufacturers. The effect of the difference between the mechanical properties on the bending parameters was investigated. Air bending technique was used for bending tests. Experiments were carried out with varying plate thicknesses, plate lengths, bending angles, forming speeds and channel openings using flat materials supplied from two different manufacturers. The effect of these parameters on bending force, displacement and springback angle was revealed. It was shown that the mechanical properties of Hardox 400 steel vary according to the manufacturer. The difference between the measured mechanical properties and the ones obtained from the manufacturers had a direct influence on the bending properties.
Evaluation of acoustic performance of hemp and cotton fiber reinforced thermoplastic composites
(Bursa Uludağ Üniversitesi, 2022-12-21) Sezgin, Hande; Duru, Sena Cimilli; Candan, Cevza
In a developing and increasing world, the increment in the consumption of natural resources brings environmental problems together. The use of plant raw materials has great importance in the construction industry, which has a high consumption rate. Increasing demand to develop sustainable and environmentally friendly materials has increased the use of natural fibers in sustainable building materials (sound absorption, heat and sound insulation). The fact that hemp fiber does not have high water consumption in the production, can be grown without the need for fertilizers and pesticides, as well as leaving a rich soil structure for the next crop by destroying weeds ensures the protection of the soil and the environment. From this point of view, in this study sound insulation properties of hemp fiber reinforced composites were investigated. Polypropylene fiber was used as a matrix material. To compare the performance of hemp fiber, cotton fiber was also utilized and different constructions were produced from these materials. The composites were produced by the hot-press method. It was found that the composites made of hemp fiber showed significantly higher sound insulation than the cotton fiber reinforced composites and can be used as a sound isolation material at specific high frequencies.
Folded angle dependent modal analysis of the flat and folded plates
(Bursa Uludağ Üniversitesi, 2022-03-23) Gönenli, Can
The effect of folded angle and boundary condition differences on the natural frequency, which is one of the dynamic characteristics of isotropic thin plates, is investigated. Modal analyzes of flat and folded structures are performed with ANSYS under cantilever and two-side-fixed boundary conditions. The validation of the free vibration results of the structures is performed with Solidworks. The first five natural frequencies of each structure are obtained, and the values in terms of non-dimensional natural frequency parameters are interpreted with the tables and graphics. While both an increase and a decrease can be observed depending on the increase in the folded angle in the first five natural frequencies for the cantilever boundary condition, the effect of the increase in the folded angle on the natural frequencies in the two sided-fixed structures is much less in the higher folded angle values. It is seen that the reason for the sudden changes in the folded angle - non-dimensional natural frequency graphs is the mode shape change.
Prediction model for base shear increase due to vertical ground motion in friction pendulum isolated structures
(Bursa Uludağ Üniversitesi, 2022-01-21) Çilsalar, Hüseyin
Seismic isolation is one of the most effective seismic hazard mitigation techniques, which has been implemented in many structures. Friction pendulum isolators are one of the most popular isolation devices to achieve energy dissipation and shear resistance that depends on the effective radius and instantaneous vertical force on them. So, the horizontal response of this type of isolators is coupled with the axial load on them, which could change due to vertical acceleration during ground shaking. Thus, it makes consideration of the vertical excitation in the design and analysis phase inevitable, especially for structures in those regions where vertical ground accelerations are more pronounced. In this study a model for the prediction of increase in base shear due to vertical motion for friction pendulum isolated structures is proposed. A simple single friction pendulum isolator system is utilized with the assumption of rigid superstructure on top and subjected to a hundred earthquake ground motions with different characteristics. A multi-layer perceptron model with three layers is utilized, and a simple computer program, which is open source, is prepared to construct probability curves. With the help of the program, within the range of the considered structural parameters in the study, probability curves related to increase in base shear, maximum isolator displacement and residual isolator displacement can be constructed, and designers can have a rough estimation on modification of these parameters due to vertical ground motion.
Soliton wave generation on nonlinear transmission lines using a particle swarm optimization (PSO) algorithm
(Bursa Uludağ Üniversitesi, 2022-03-23) Aksoy, Abdullah; Yenikaya, Sibel; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Elektrik ve Elektronik Mühendisliği Bölümü.; 0000-0002-2400-989X; 0000-0002-9423-1752
In this article, analysis, designing, and generation of soliton waves are performed using nonlinear transmission lines (NLTLs). As a result of performing mathematical analysis processes, circuit parameters and values are determined. A particle swarm optimization (PSO) algorithm is used to determine the parameters and their values. The parameters obtained from the optimization are used to design a nonlinear transmission line. The circuit designed over the determined parameters is simulated with the LTspice program. Then, experiments of the created circuit design are carried out. The simulation data and the experimental implementation data are compared for the nonlinear transmission line circuit. Simulation and experiment results are found to be compatible.
A study on efficiency of perfectly matched layer (PML) for seismic soil-structure interaction
(Bursa Uludağ Üniversitesi, 2022-03-29) Genç, Abdulkadir; Kuvat, Ahmet; Sesli, Hasan
The direct method is a finite element approach used in soil-structure interaction (SSI) analysis. The main assumption of this method is the transformation of the infinite soil domain to the truncated soil domain applying different boundary conditions. The objective of this study is to examine the efficiency of The Perfectly Matched Layers (PML) for SSI analysis. For this purpose, the soil-structure problem with different planar-sized truncated soil domains with constrained boundary (CB) and non-reflecting boundary (NRB) were analyzed using 3 different soil stiffness. The results obtained from these analyses were compared with PML and fixed-base reference model results. According to the findings, if the truncated soil domain planar sizes are larger than 10 times the foundation widths for traditional boundary conditions, the effect of the soil domain size is negligible. When the foundation soil stiffness is assumed to be medium and hard, it was observed that PML results are very close to other boundary condition model results, but greater differences were observed in results between PML and other models for soft foundation soil conditions. From the parametrical study, it was concluded that PML provides superiority in terms of computational cost and practical application.
The design of control system simulator
(Bursa Uludağ Üniversitesi, 2022-01-07) Hatun, Metin; Vatansever, Fahri; Bursa Uludağ Üniversitesi/Mühendislik Fakültesi/Elektrik-Elektronik Mühendisliği.; 0000-0003-0279-5508; 0000-0002-3885-8622
Selection of a suitable controller type, determination of its coefficients and adjustment if necessary is an important issue in the design of control systems. Quickly checking whether the closed loop system with a designed controller meets some desired time/frequency response criteria will save time for the student/designer. In this paper, an interactive control system simulator is designed to perform the simulations for closed-loop systems obtained with user-defined systems and different types of controllers whose parameters can be changed. The simulator allows the user to see the time/frequency response of the closed-loop system for different systems defined as single or comparatively and check whether some desired time/frequency domain criteria are satisfied, quickly. In addition, the simulator allows the possibility to see the disturbance input response and the effect of the measurement noise altogether with the reference input response, comparatively. Thus, analysis and comparison of different controllers and also teaching of control systems will be performed easily, quickly and effectively by using the simulator.
Transition from static to dynamic fields in the presence of dyons
(Bursa Uludağ Üniversitesi, 2022-03-07) Zor, Ömer
We assumed the existence of dyons. Since the dyon is a particle with electric and magnetic charges, we simply considered that electric charge and corresponding magnetic charge have the same small velocity. Thus using this proposition and neglecting the interaction between electric and magnetic charges, we constructed symmetric microscopic Maxwell equations in the presence of dyons. Eventually we expanded the theory and obtained macroscopic Maxwell equations in vacuum using averaging process.
Treatment of Pb – naphthalene co – contamination in soils by zero valent iron nanoparticles (NZVI) and bioremediation
(Bursa Uludağ Üniversitesi, 2022-02-10) Kahraman, Bekir Fatih; Altın, Ahmet
The co-contamination of soils with heavy metals and organic contaminants is a problem that requires a different and innovative approach to remediate these kinds of sites. The different nature of these two contaminant groups makes the problem complicated. Recently, zero-valent iron nanoparticles (nZVI) application has been integrated with bioremediation to be used for the remediation of halogenated organics. However, the use of this process for the remediation of co-contaminated soils has not been investigated. In this study, nZVI application integrated with bioremediation approaches (bioattenuation and biostimulation) was applied to Pb and naphthalene spiked soils in laboratory-scale reactors. Soil Pb fractions and naphthalene concentrations were monitored for 90 days. The nZVI decreased the mobility and bioavailability of Pb significantly by reducing exchangeable and carbonate bound fractions from 66.2 % to 25.8 – 37.2 % range. The nZVI also caused naphthalene degradation in the range of 38.8 – 58.5 %. Significant biodegradation of naphthalene occurred at samples subjected to nZVI and biostimulation. The highest overall naphthalene degradation (76.7 %) was obtained from the nZVI application which was integrated with biostimulation.
Web page classification with deep learning methods
(Bursa Uludağ Üniversitesi, 2022-02-13) Kurt, Mehmet Salih; Yücel, Eylem
Today, millions of websites on the Internet are widely used to access information. For effective use of web pages with increasing numbers every day, they need to be well classified. In this study, binary and multi-class classification models have been created which can classify web pages with high accuracy. In our experiments, URLs and categories of English web pages in the Open Directory Project (ODP) were used. Training dataset was created by pulling web page texts from URL information. To our knowledge, this is the first comprehensive web page classification dataset for Turkish. In this study, Convolutional Neural Network (CNN), Long Short Term Memory (LSTM) and Gated Recurrent Unit (GRU) deep learning methods which are effective in text classification are used. Word embedding was used instead of n-gram approaches commonly used for feature extraction in text classification studies. In this study, hyper-parameter optimization was performed for deep learning models. Binary and multi-class classification models were created with the best parameters. Binary classification models were compared with the results of another study, and multi-class classification models were compared with each other. The performances of all models were examined by considering their training time and f1 scores.

Browse

Browsing by Language "en"