Sukru Ozsahin

Prediction of the influence of processing parameters on synthesis of Al2024-B4C composite powders in a planetary mill using an artificial neural network

Abstract In this study, an artificial neural network approach was employed to predict the effect of B4C size, B4C content, and milling time on the particle size and particle hardness of Al2024-B4C composite powders. Al2024-B4C powder mixtures with various reinforcement weight percentages (5%, 10%, and 20% B4C), reinforcement size (49 and 5 μm), and milling times (0–10 h) were prepared by mechanical alloying process. The properties of the composite powders were analyzed using a laser particle size analyzer for the particle size and a microhardness tester for the powder microhardness. The three input parameters in the proposed artificial neural network (ANN) were the reinforcement size, reinforcement ratio, and milling time. Particle size and particle hardness of the composite powders were the outputs obtained from the proposed ANN. The mean absolute percentage error for the predicted values did not exceed 4.289% for the best prediction model. This model can be used for predicting properties of Al2024-B4C composite powders produced with different reinforcement size, reinforcement ratio, and milling times.

Prediction of effect of volume fraction, compact pressure and milling time on properties of Al-Al2O3 MMCs using neural networks

An artificial neural network (ANN) model was developed to predict the effect of volume fraction, compact pressure and milling time on green density, sintered density and hardness of Al-Al2O3 metal matrix composites (MMCs). Al-Al2O3 powder mixtures with various reinforcement volume fractions of 5, 10, 15% Al2O3 and milling times (0 h to 7 h) were prepared by mechanical milling process and composite powders were compacted at various pressure (300, 500 and 700 MPa). The three input parameters in the proposed ANN were the volume fraction, compact pressure and duration of the milling process. Green density, sintered density and hardness of the composites were the outputs obtained from the proposed ANN. As a result of this study the ANN was found to be successful for predicting the green density, sintered density and hardness of Al-Al2O3 MMCs. The mean absolute percentage error for the predicted values didn’t exceed 5.53%. This model can be used for predicting Al-Al2O3 MMCs properties produced with different reinforcement volume fractions, compact pressures and milling times.

Optimization of process parameters in oriented strand board manufacturing with artificial neural network analysis

In the present work, an artificial neural network (ANN) model was developed for predicting the effects of some production factors such as adhesive ratio, press pressure and time, and wood density and moisture content on some physical properties of oriented strand board (OSB) such as moisture absorption, thickness swelling and thermal conductivity. The MATLAB Neural Network Toolbox was used for the training and optimization of the artificial neural network. The ANN model having the best prediction performance was determined by means of statistical and graphical comparisons. The results show that the prediction model is a useful, reliable and quite effective tool for predicting some physical properties of the OSB produced under different manufacturing conditions. Thus, this study has presented a novel and alternative approach to the literature to optimize process parameters in OSB manufacturing process.ZusammenfassungIn dieser Studie wurde ein künstliches neuronales Netz (ANN) entwickelt, um den Einfluss einiger Produktionsfaktoren, wie zum Beispiel Klebstoffmenge, Pressdruck, Pressdauer, Holzdichte und Holzfeuchte, auf die physikalischen Eigenschaften von OSB, wie Wasseraufnahme, Dickenquellung und Wärmeleitfähigkeit zu ermitteln. Für die Trainingsphase und Optimierung des künstlichen neuronalen Netzes wurde die MATLAB Neural Network Toolbox verwendet. Anhand statistischer und graphischer Vergleiche wurde das ANN Modell mit der besten Vorhersageleistung bestimmt. Die Ergebnisse zeigen, dass dieses Modell ein nützliches, zuverlässiges und effektives Werkzeug zur Vorhersage verschiedener physikalischer Eigenschaften von unter verschiedenen Bedingungen hergestelltem OSB ist. Somit wird in dieser Studie ein neuer und alternativer Ansatz für die Optimierung von Prozessparametern bei der OSB-Herstellung vorgestellt.

Artificial neural network-based prediction technique for coating thickness in Fe-Al coatings fabricated by mechanical milling

ABSTRACT The objective of this study was to evaluate the effect of milling time, milling speed and particle size of initial powders on the coating thickness of Fe-Al intermetallic coating by using artificial neural network (ANN). Coating morphology and cross-section microstructures were evaluated using a scanning electron microscope (SEM). It was found that an increase in the milling time provided an increase in the coating-layer thickness due to the cold welding process between particles and the steel substrate. The microstructure of the coating surface was refined by ball impacts in the milling process. As a result of this study, the ANN was found to be successful for predicting the coating thickness of Fe-Al intermetallic coatings. The correlation between the predicted values and the experimental data of the feed-forward back-propagation ANN was quite adequate. The mean absolute percentage error (MAPE) for the predicted values didn’t exceed 7.46%. The ANN model can be used for predicting the coating thickness of Fe-Al intermetallic coating produced for different milling time, milling speed and particle size.

Prediction of equilibrium moisture content and specific gravity of heat treated wood by artificial neural networks

In the present work, two artificial neural network (ANN) models were developed for modeling the effects of conditions of heat treatment process such as exposure period and temperature at equilibrium moisture content (EMC) and specific gravity (SG) at different relative humidity levels of heat treated Uludag fir (Abies bornmülleriana Mattf.) and hornbeam (Carpinus betulus L.) wood. A custom MATLAB application created with MATLAB codes and functions related to neural networks was used for the development of feed forward and back propagation multilayer ANN models. The prediction models having the best prediction performance were determined by means of statistical and graphical comparisons. The results show that the prediction models are practical, reliable and quite effective tools for predicting the EMC and SG characteristics of heat treated wood. Thus, this study presents a novel and alternative approach to the literature to optimize conditions of heat treatment process.