João Roberto Ferreira

Design of experiments and focused grid search for neural network parameter optimization

The present work offers some contributions to the area of surface roughness modeling by Artificial Neural Networks (ANNs) in machining processes. It proposes a method for an optimized project of a Multi-Layer Perceptron (MLP) network architecture applied for the prediction of Average Surface Roughness (Ra). The tuning method is expressed in the format of an algorithm employing two techniques from Design of Experiments (DOE) methodology: Full factorials and Evolutionary Operations (EVOP). Datasets retrieved from literature are employed to form training and test data sets for the ANN. The proposed tuning method leads to significant reduction of roughness prediction errors in machining operations in comparison to techniques currently used. It constitutes an effective option for the systematic design models based on ANN for prediction of surface roughness, filling the gap reported in the literature on this subject. We propose a systematic approach to design and optimize MLP networks.We used DOE, Evolutionary Operation and Focused Grid Search for optimization.The proposed method is compared to previous studies in machining applications.The method presents superior results for all the comparisons.

Modeling and Optimization of Multiple Characteristics in the AISI 52100 Hardened Steel Turning

This work aimed to develop a multiple response optimization procedure for the AISI 52100 hardened steel turning process. Optimizing this turning operation is important so that multiple quality characteristics are achieved simultaneously. The considered responses are: total cost, cutting time, total turning cycle time, tool life, material removal rate, and surface roughness. The adjusted process parameters were cutting speed, feed rate and depth of cut. A multi-objective optimization technique based on the Global Criterion Method and Genetic Algorithm were employed to identify the optimal settings for parameters with objective functions built through Response Surface Methodology. This two-fold approach lead up to optimized responses settled near the desired values were obtained with cutting speed = 214 m/min, feed rate = 0.088 mm/rev and depth of cut = 0.33 mm.

Multivariate Optimization of the Cutting Parameters when Turning Slender Components

The geometric features of the work piece and the cutting parameters considerably affect the quality of a finished part subjected to any machining operation owing to the imposed elastic and plastic deformations, especially when slender components are produced. This work is focused on the influence of the work piece slenderness ratio and cutting parameters on the quality of the machined part, assessed in terms of surface roughness and both geometric (run-out) and dimensional (diameter) deviations. Turning tests with coated tungsten carbide tools were performed using AISI 1045 medium carbon steel as work material. Differently from the published literature, a statistical analysis based on the multivariate one-way analysis of variance (MANOVA) was applied to the data obtained using a Box-Behnken experimental design. In order to identify the combination of parameters (slenderness ratio, cutting speed, feed rate and depth of cut) levels which simultaneously optimize the responses of interest (surface roughness, run-out and diameter deviation), a multivariate optimization method based on principal component analysis (PCA) and generalized reduced gradient (GRG) was employed.

A DOE based approach for the design of RBF artificial neural networks applied to prediction of surface roughness in AISI 52100 hardened steel turning

The use of artificial neural networks for prediction in hard turning has received considerable attention in literature. An often quoted drawback of ANNs is the lack of a systematic way for the design of high performance networks. This study presents a DOE based approach for the design of ANNs of Radial Basis Function (RBF) architecture applied to surface roughness prediction in turning of AISI 52100 hardened steel. Experimental factors are the number of radial units on the hidden layer, the algorithm employed to calculate the spread factor of radial units and the algorithm employed to calculate radial function centers. DOE is employed to select levels of factors that benefit network prediction skills. Experiments with data sets of distinct sizes were conducted and network configurations leading to high performance were identified. ANN models obtained proved capable to predict roughness in accurate, precise and affordable way. Results pointed significant factors for network design and revealed that interaction effects between design parameters have significant influence on network performance for the task proposed. The work concludes that the DOE methodology constitutes a better approach to the design of RBF networks for roughness prediction than the most common trial and error approach.

STUDY OF THE INFLUENCE OF CUTTING FLUID AMOUNT IN STEEL TURNING OF AISI 52100

Http://www.ijetmr.com©International Journal of Engineering Technologies and Management Research [49] STUDY OF THE INFLUENCE OF CUTTING FLUID AMOUNT IN STEEL TURNING OF AISI 52100 Ítalo de Abreu Gonçalves , Leonardo G. Leite , Tarcísio G. de Brito , Emerson J. de Paiva , Carlos H. de Oliveira , Jean Carlos C. Pereira *1 *1 Academic Unit of Engineering of Itabira, Federal University of Itajubá, Brazil Abstract: The steel turning AISI 52100 has been gaining prominence in industry in recent years, as it allows machined parts to have better quality without the need for furthers processes. However, to ensure the final product quality, it is important that the turning for machining procedure is well planned and prepared, so that the cutting tools have their wear minimized in the process, while putting good productivity rates and zero occurrences of reworked parts. Thus, this article will study the quality of the machined surface in the turning process using interchangeable PCBN inserts. The aim is to identify the optimal combination of the input parameters that are cutting speed (Vc), feed (f) and machining depth (ap). The response measured is the roughness parameter Ra, under the influence of cutting fluid and tool wear.