Browsing by Author "Koçak, Eyup"

Now showing 1 - 9 of 9

Evaluating Machine Learning Techniques for Fluid Mechanics: Comparative Analysis of Accuracy and Computational Efficiency
(2024) Koçak, Eyup
This study focuses on applying machine learning (ML) techniques to fluid mechanics problems. Various ML techniques were used to create a series of case studies, where their accuracy and computational costs were compared, and behavior patterns in different problem types were analyzed. The goal is to evaluate the effectiveness and efficiency of ML techniques in fluid mechanics and to contribute to the field by comparing them with traditional methods. Case studies were also conducted using Computational Fluid Dynamics (CFD), and the results were compared with those from ML techniques in terms of accuracy and computational cost. For Case 1, after optimizing relevant parameters, the Artificial Neural Network (ANN), Random Forest (RF), and Support Vector Machine (SVM) models all achieved an R² value above 0.9. However, in Case 2, only the ANN method surpassed this threshold, likely due to the limited data available. In Case 3, all models except for Linear Regression (LR) demonstrated predictive abilities above the 0.9 threshold after parameter optimization. The LR method was found to have low applicability to fluid mechanics problems, while SVM and ANN methods proved to be particularly effective tools after grid search optimization.
Citation - WoS: 1
Citation - Scopus: 1
Investigation of aerodynamic and aeroacoustic behavior of bio-inspired airfoils with numerical and experimental methods
(Sage Publications Ltd, 2024) Aylı, Ülkü Ece; Guzey, Kaan; Ayli, Ulku Ece; Koçak, Eyup; Kocak, Eyup; Aradag, Selin; 265836; 283455
This article presents numerical and experimental studies on the aerodynamic and aeroacoustic characteristics of the NACA0012 profile with owl-inspired leading-edge serrations for aeroacoustic control. The leading-edge serrations under investigation are in a sinusoidal profile with two main design parameters of wavelength and amplitude. The noise-suppressing ability of sinusoidal serrations is a function of several parameters such as amplitude, wavelength, inflow speed, angle of attack, which are examined in this study. Amplitude (A) and wavelength (& lambda;) of the serration are varied between 1.25 and 2.5, 20 < & lambda; < 60, respectively. The corresponding Reynolds numbers are between 1 and 3 x 10(5). The angle of attack for each configuration is changed between 4 & DEG; and 16 & DEG;. Forty different configurations are tested. According to the results, owl-inspired leading-edge serrations can be used as aeroacoustic control add-ons in blade designs for wind turbines, aircraft, and fluid machinery. Results show that the narrower and sharper serrations have a better noise reduction effect. Overall sound pressure level (SPL) reduces up to 20% for the configuration with the largest amplitude and smaller wavelength. The results also showed that serration amplitude had a distinct effect on aeroacoustic performance, whereas wavelength is a function of amplitude. At the smaller angle of attack values, AOA < 8 & DEG;, the lift and drag coefficients are almost the same for both clean and wavy profiles. On the other hand, typically for angle of attack values more than 12 & DEG; (after stall), when the angle of attack is increased, serration adversely affects aerodynamic performance.
Citation - WoS: 5
Citation - Scopus: 6
Machine Learning Based Developing Flow Control Technique Over Circular Cylinders
(Asme, 2023) Koçak, Eyup; Ayli, Ece; Kocak, Eyup; Türkoğlu, Haşmet; Turkoglu, Hasmet; 265836; 283455; 12941
This paper demonstrates the feasibility of blowing and suction for flow control based on the computational fluid dynamics (CFD) simulations at a low Reynolds number flows. The effects of blowing and suction position, and the blowing and suction mass flowrate, and on the flow control are presented in this paper. The optimal conditions for suppressing the wake of the cylinder are investigated by examining the flow separation and the near wake region; analyzing the aerodynamic force (lift and drag) fluctuations using the fast Fourier transform (FFT) to separate the effects of small-scale turbulent structures in the wake region. A method for stochastic analysis using machine learning techniques is proposed. Three different novel machine learning methods were applied to CFD results to predict the variation in drag coefficient due to the vortex shedding. Although, the prediction power of all the methods utilized is in the acceptable accuracy range, the Gaussian process regression (GPR) method is more accurate with an R-2(coefficient of determination) > 0.95. The results indicate that by optimizing the blowing and suction parameters like mass flowrate, slot location, and the slot configuration, up to 20% reduction can be achieved in the drag coefficient.
Numerical Analysis of Aerodynamic and Aeroacoustic Characteristics of Subsonic Rectangular Cavity with Different Aspect Ratios
(2020) Koçak, Eyup; Türkoğlu, Haşmet; Türkoğlu, Haşmet; 283455; 265836; 12941
Citation - WoS: 1
Citation - Scopus: 1
Performance Determination of Axial Wind Tunnel Fan with Reverse Engineering, Numerical and Experimental Methods
(Asme, 2022) Kocak, Eyup; Koçak, Eyup; Ayli, Ece; 283455; 265836
In today's technology, in case of the need for rehabilitation, renovation, or damage, it is necessary to recover the problems quickly with a cost-effective approach. In the case of destructive failure, or misdesign of the devices, replacing the problematic part with the new design is crucial. In order to substitute the related part with the efficient one, reverse engineering (RE) methodology is utilized. In this paper, from the perspective of engineering implementation and based on the idea of reverse engineering, axial wind tunnel fan is rehabilitated using numerical and experimental methods. The current study is focused on an axial pressurization fan placed into Cankaya University Mechanical Engineering Laboratory wind tunnel that has firm guaranteed specifications of 5.55 m(3)/s airflow capacity. The measurements performed during experiments showed that the fan provides less than 60% airflow compared with firm guaranteed specifications. In order to determine the problems of the existing fan, a reverse engineering methodology is developed, and the noncontact data acquisition method is used to form a computer aided drawing (CAD) model. A computational fluid dynamics (CFD) methodology is developed to analyze existing geometry numerically, and results are compared with an experimental study to verify numerical methodology. According to the results, the prediction accuracy of the numerical method can attain 92.95% and 96.38% for flowrate and efficiency, respectively, at the maximum error points.
Performance Evaluation of reverse engineered wind tunnel axial fan
(2020) Koçak, Eyup; Kartal, M.; Mendi, C.; Surgun, K.; Günce, C.; Kantar, E.; Koçak, Eyup; Aylı, Ece; 283455; 265836
Citation - WoS: 2
Citation - Scopus: 2
Performance Optimization of Finned Surfaces Based on the Experimental and Numerical Study
(Asme, 2023) Koçak, Eyup; Kocak, Eyup; Turkoglu, Hasmet; Türkoğlu, Haşmet; Ayli, Ece; 283455; 12941; 265836
This paper presents the findings of numerical and experimental investigations into the forced convection heat transfer from horizontal surfaces with straight rectangular fins at Reynolds numbers ranging from 23,600 to 150,000. A test setup was constructed to measure the heat transfer rate from a horizontal surface with a constant number of fins, fin width, and fin length under different flow conditions. Two-dimensional numerical analyses were performed to observe the heat transfer and flow behavior using a computer program developed based on the openfoam platform. The code developed was verified by comparing the numerical results with the experimental results. The effect of geometrical parameters on heat transfer coefficient and Nusselt number was investigated for different fin height and width ratios. Results showed that heat transfer can be increased by modifying the fin structure geometrical parameters. A correlation for Nusselt number was developed and presented for steady-state, turbulent flows over rectangular fin arrays, taking into account varying Prandtl number of fluids such as water liquid, water vapor, CO2, CH4, and air. The correlation developed predicts the Nusselt number with a relative root mean square error of 0.36%. This research provides valuable insights into the effects of varying Prandtl numbers on the efficiency of forced convection cooling and will help in the design and operation of cooling systems. This study is novel in its approach as it takes into account the effect of varying Prandtl numbers on the heat transfer coefficient and Nusselt number and provides a correlation for the same. It will serve as a valuable reference for engineers and designers while designing and operating cooling systems.
Citation - WoS: 4
Citation - Scopus: 5
Prediction of the heat transfer performance of twisted tape inserts by using artificial neural networks
(Korean Soc Mechanical Engineers, 2022) Ayli, Ece; Koçak, Eyup; Kocak, Eyup; 265836; 283455
A numerical study is undertaken to investigate the effect of twisted tape inserts on heat transfer. Twisted tapes with various aspect ratios and single, double, and triple inserts are placed inside a tube for Reynolds numbers ranging from 8000 to 12000. Numerical results show that the tube with a twisted tape and different numbers of tape is more effective than the smooth tube in terms of thermo-hydraulic performance. The highest heat transfer is achieved with the triple insert, with the highest turning number and an increment of 15 %. Then, an artificial neural network (ANN) model with a three-layer feedforward neural network is adopted to obtain the Nusselt number on the basis of four inputs for a heated tube with a twisted insert. Several configurations of the neural network are examined to optimize the number of neurons and to identify the most appropriate training algorithm. Finally, the best model is determined with one hidden layer and thirteen neurons in the layer. Bayesian regulation is chosen as the training algorithm. With the optimized algorithm, excellent precision for measuring the output is provided, with R2 = 0.97043. In addition, the optimized ANN architecture is applied to similar studies in the literature to predict the heat transfer performance of twisted tapes. The developed ANN architecture can predict the heat transfer enhancement performance of similar problems with R2 values higher than 0.93.
Citation - WoS: 7
Citation - Scopus: 7
Supervised learning method for prediction of heat transfer characteristics of nanofluids
(Korean Soc Mechanical Engineers, 2023) Ayli, Ece; Koçak, Eyup; Kocak, Eyup; 265836; 283455
This study focuses on the alication and investigation of the predictive ability of artificial intelligence in the numerical modelling of nanofluid flows. Numerical and experimental methods are powerful tools from an accuracy point of view, but they are also time- and cost-consuming methods. Therefore, using soft-computing techniques can improve such CFD drawbacks by patterning the CFD data. After obtaining the aropriate ANN and ANFIS architecture using the CFD data, many new data can be created without requiring numerical and experimental methods. In the scope of this research, the FCM-ANFIS and ANN methods are used to predict the thermal behaviour of the turbulent flow in a heated pipe with several nanoparticles. A parametric CFD study is carried out for water-TiO2, water-CuO, and water-SiO2 nanofluid through a pipe. The Reynolds number is varied between 7000 and 15000, and the nanofluid concentration is varied between 0.25 % and 4 %. The effects of using nanofluid on local values of Nusselt number and shear stress distribution were investigated. Numerical results indicate that with the increasing nanoparticle volume fraction of nanofluid, the average Nusselt number increases, but the required pumping power also increases. The obtained soft computing results demonstrate that the FCM clustering ANFIS has given better results both in training and testing when it is compared to the ANN architecture with an R-2 of 0.9983. Regarding this, the FCM-ANFIS is an excellent candidate for calculating the Nusselt number in heat transfer problems.