Neural network modeling and analysis of the material removal process during laser machining (original) (raw)
Related papers
Prediction of Laser Hardening by Means of Neural Network
Procedia CIRP, 2013
Laser hardening is a surface treatment process characterized by a high level of performance. The resulting physical, chemical, and mechanical properties of the surface layers can be accurately designed by modifying the process parameters i.e., scanning speed, frequency and laser power. Thus, the development of the laser hardening technology requires considerable preliminary work, including the determination of the range of components that may be hardened, the selection of proper treatment conditions and the identification of optimized strategies to employ such a technology for real industrial components. The present research aimed to provide a deep understanding of the laser hardening process. The effect of process parameters i.e., the laser power, the scanning speed, the number of scans and the overlapping, has been assessed by means of a campaign of experimental tests. Thus, an attempt to predict the effect of process parameters of treated components was carried out by developing an expert system using a neural network.
Neural Network Multi Layer Perceptron Modeling For Surface Quality Prediction in Laser Machining
Application of Machine Learning, 2010
Uncertainty is inevitable in problem solving and decision making. One way to reduce it is by seeking the advice of an expert in related field. On the other hand, when we use computers to reduce uncertainty, the computer itself can become an expert in a specific field through a variety of methods. One such method is machine learning, which involves using a computer algorithm to capture hidden knowledge from data. The researchers conducted the prediction of laser machining quality, namely surface roughness with seven significant parameters to obtain singleton output using machine learning techniques based on Quick Back Propagation Algorithm. In this research, we investigated a problem solving scenario for a metal cutting industry which faces some problems in determining the end product quality of Manganese Molybdenum (Mn-Mo) pressure vessel plates. We considered several real life machining scenarios with some expert knowledge input and machine technology features. The input variables are the design parameters which have been selected after a critical parametric investigation of 14 process parameters available on the machine. The elimination of non-significant parameters out of 14 total parameters were carried out by single factor and interaction factor investigation through design of experiment (DOE) analysis. Total number of 128 experiments was conducted based on 2k factorial design. This large search space poses a challenge for both human experts and machine learning algorithms in achieving the objectives of the industry to reduce the cost of manufacturing by enabling the off hand prediction of laser cut quality and further increase the production rate and quality.
Modeling of Machining Process by Neural Network
Machining is one of the most important manufacturing processes. Machining is very complex process. In recent years, modeling techniques using neural network have attracted attention of practitioners and researchers. The learning ability of nonlinear relationship in a cutting operation without going deep into the mathematical complexity makes neural network an attractive alternative choice for many researchers to model cutting processes This technique offers a cost effective alternative in modeling of machining process. This paper discusses the basic idea and gives the concept of neural network modeling of machining process.
A novel approach to predict the surface characteristics, namely depth and surface roughness, of glass fiber reinforced plastic composite after CO 2 laser milling by using artificial neural networks is developed and optimized. The experimental data are produced using a 60 W CO 2 laser machine to perform milling of unidirectional glass fiber reinforced plastic composite. The CO 2 laser milling is performed in both parallel and perpendicular to fiber direction at five different values of energy deposition (0.066, 0.176, 0.22, 0.264, 0.308) J/mm and three different beam diameters (225, 277.398, 463.869) lm. The artificial neural network model having the architecture of 3-6-6-3, that is two hidden layers with six neurons in each layer, is found to have the best performance based on mean error value. The mean, maximum, and minimum prediction errors for this ANN model are 0.82%, 2.26%, and 0.0004%, respectively. A semiempirical model is also developed to predict the milling depth, and its response is compared with predicted depth from neural network model. The milled depth predicted using the optimized neural network model is far superior compared to the output of the semiempirical model.
Manufacturing Review, 2014
Pulsed Nd:YAG laser has high intensity and high quality beam characteristics, which can be used to produce micro-grooves and micro-turning surface on advanced engineering ceramics. The present research attempts to develop mathematical models by using response surface methodology approach for correlating the machining process parameters and the process responses during laser micro-turning of aluminum oxide (Al 2 O 3) ceramics. The process parameters such as laser average power, pulse frequency, workpiece rotating speed, assist air pressure and Y feed rate were varied during experimentation. The rotatable central composite design experimental planning has been used to design the experimentation. The performance measures considered are surface roughness (Ra) and micro-turning depth deviation. Multi-objective optimization has been carried out for achieving the desired surface roughness as well as minimum depth deviation during laser micro-turning operation. Further, an artificial neural network (ANN) model has been developed to predict the process criteria. Levenberg-Marquadt training algorithm is used for multilayer feed forward backpropagation neural network. The developed ANN model has 5-10-2 feed forward network. There are 5 neurons in the input layer, 10 neurons in the hidden layer and 2 neurons in the output layers corresponding to two output responses, respectively. The developed ANN model has been validated using data obtained by conducting additional set of experiments. It was found that the developed ANN model can predict the process criteria more accurately than response surface methodology (RSM) based developed models.
Journal of Laser Applications
Finding the optimal parameters in a laser processing application can be time-consuming given the large parameter space and various sources of error. This problem is exacerbated by day-today variation in laser beam characteristics and a large variety of materials that need to be processed. The ideal laser processing system should be "smart", meaning that it can sense changes in the environment, make proper adjustment, and predict parameters for new materials. As a step towards this goal, we propose a method to efficiently extract the areas of a large number of laserinduced damages in-situ using an automated data acquisition system that can control laser parameters, motorized stage movement, image capturing/processing, and feature extraction. The damage areas are extracted and compared with direct measurements. Damage areas are fed into an artificial neural network (ANN) for prediction. Various ANN structures and training functions are tested to create the optimal ANN for prediction. ANN predictions were found to be capable enough to accurately model and optimize the laser processing parameters that were investigated. With the capability of collecting a large amount of usable data in a short period of time, this acquisition system can be used to train sophisticated ANNs for complicated tasks such as quality control and failure prediction.
Methodology of neural network based modeling of machining processes
2010
Machining is the most important and widely used manufacturing process. As machining is very complex process, in recent years neural network based modeling has been preferred modeling of machining processes. This paper outlines and discusses the basic idea and concept of neural network modeling of machining processes. Furthermore this paper discusses the methodology of developing neural network model as well as proposing some guidelines for selecting the network training parameters and network architecture. For illustration purpose, simple neural prediction model for cutting power was developed and validated.
Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, 2010
Selection of machining parameter combinations for obtaining optimum circularity at entry and exit and hole taper is a challenging task in laser microdrilling owing to the presence of a large number of process variables. There is no perfect combination of parameters that can simultaneously result in higher circularity at entry and exit and lower hole taper. The current paper attempts to develop a strategy for predicting machining parameter settings for the generation of the maximum circularity at entry and exit and minimum hole taper. An artificial neural network (ANN) is used for process modelling of laser microdrilling of titanium nitride-alumina composite and a feed-forward back-propagation network is developed to model the machining process. The model, after proper training, is capable of predicting the response parameters as a function of five different control parameters. Experimental results demonstrate that the machining model is suitable and the optimization strategy satisfies practical requirements. The developed model is found to be unique, powerful, and flexible.