Krzysztof Jemielniak

Tool condition monitoring using artificial intelligence methods

Abstract This paper describes an application of three artificial intelligence (AI) methods to estimate tool wear in lathe turning. The first two are “conventional” AI methods—the feed forward back propagation neural network and the fuzzy decision support system. The third is a new artificial neural network based-fuzzy inference system with moving consequents in if–then rules. Tool wear estimation is based on the measurement of cutting force components. This paper discusses a comparison of usability of these methods in practice.

Suppression of self-excited vibration by the spindle speed variation method

Abstract A method for the analysis of the influence of spindle speed variation on the course of self-excited vibration has been presented. Thisanalysis is based on the influence of self-frequency of self-excited vibration on the stability of machining, and dependence of this frequency on workpiece rotational speed. The influence of the spindle speed variation frequency and amplitude on chatter has been investigated.

Numerical simulation of non-linear chatter vibration in turning

Abstract In this paper, basic relationships and algorithms for numerical simulation of non-linear, self-excited vibrations in single degree-of-freedom cutting systems are presented. Non-linearities due to the tool leaving the cut, as well as interference between the cutting tool clearance face and cutting surface waviness, were taken into consideration. Examples of vibration simulation results are shown.

Diagnosis of tool wear based on cutting forces and acoustic emission measures as inputs to a neural network

Cutting forces and acoustic emission measures as a function of tool wear are presented for different cutting parameters and their applicability for tool condition monitoring is evaluated. The best of them, together with cutting parameters, were chosen as inputs to a feedforward, back propagation (FFBP) neural network; some training techniques were applied and their effectiveness is also evaluated. Conventional training of FFBP neural networks very soon leads to overtraining, hence to deterioration in the net response. Training of these nets depends very much on the initial weight values. A good way of finding satisfactory results is to introduce random distortions to the weight system, which efficiently push the net out of a local minimum of testing errors. An even more effective method may be to employ temporary shifts in the weights, alternately negative and positive. This has two advantages: (1) it brings the net to balance between training and testing errors and (2) it enables a great reduction in the number of hidden nodes.

Chip Form Monitoring through Advanced Processing of Cutting Force Sensor Signals

This paper draws on the activities of the CIRP Collaborative Work on “Round Robin on Chip Form Monitoring” carried out within the Scientific-Technical Committee Cutting (STC-C). This collaborative work involved the following main round robin activities: (a) generation, detection, storage and exchange of cutting force sensor signals obtained at different Laboratories during sensor-based monitoring of machining processes with variable cutting conditions yielding diverse chip forms, and (b) cutting force signal (CFS) characterization and feature extraction through advanced processing methodologies, both aimed at comparing chip form monitoring results achieved on the basis of innovative analysis paradigms.

Catastrophic Tool Failure Detection Based on Acoustic Emission Signal Analysis

Abstract Acoustic emission (AE) signal analysis is considered to be a very useful mean of on-line tool breakage detection. Many publications have proclaimed that catastrophic tool failure (CTF) causes an eminent peak in the AE signal. Therefore the magnitude of the AE RMS signal has been considered as a measure of the CTF. While strong bursts of AE signals similar to those arising from the CTF can be generated by tool engagement and disengagement in interrupted turning, this measure was found to be not always sensitive to the CTF. The aim of this paper is to present a method of the CTF detection in turning which uses symptoms other than the direct AE RMS signal value taking into considerations the likely bursts that can be generated due to interruption. The method is based on the statistical analysis of the distributions of the AE RMS signal. The kurtosis and the sum of the β distribution parameters r and s were the main measures employed. They were found to be highly sensitive to tool chipping and breakage and have given promising results with regard to CTF detection.

Tool wear monitoring using genetically-generated fuzzy knowledge bases

Fuzzy logic is an AI method that is being implemented in a growing number of different fields. One of these applications is tool wear monitoring. The construction of a fuzzy knowledge base from a set of experimental data by a human expert however, is a time consuming task, and hence, limits the expansion of the use of this AI method. Alternatively, the fuzzy knowledge base can be automatically constructed by a genetic algorithm from the same set of experimental data without requiring any human expert. This paper compares these two fuzzy knowledge base construction methods and the results obtained in a tool wear monitoring application.

Type-2 fuzzy tool condition monitoring system based on acoustic emission in micromilling

In this paper, a micromilling type-2 fuzzy tool condition monitoring system based on multiple AE acoustic emission signal features is proposed. The type-2 fuzzy logic system is used as not only a powerful tool to model acoustic emission signal, but also a great estimator for the ambiguities and uncertainties associated with the signal itself. Using the results of root-mean-square error estimation and the variations in the results of type-2 fuzzy modeling of all signal features, the most reliable ones are selected and integrated into cutting tool life estimation models. The obtained results show that the type-2 fuzzy tool life estimation is in accordance with the cutting tool wear state during the micromilling process. The information about uncertainty prediction of tool life is of great importance for tool condition investigation and crucial when making decisions about maintaining the machining quality.

Some aspects of acoustic emission signal pre-processing

The acoustic emission (AE) sensor and the pre-amplifier either built-in or connected to the sensor are the key elements in any AE based tool condition monitoring (TCM) system. This paper provides an interpretation of some common AE signal distortions and possible solutions to avoid such problems. The first two are AE signal saturation and temporary vanishing of the signal amplitude caused by overload of the pre-amplifier. The other is a result of multiple reflections of the AE wave on the different surfaces through the signal’s path. # 2001 Elsevier Science B.V. All rights reserved.

TSK fuzzy modeling for tool wear condition in turning processes: An experimental study

This paper presents an experimental study for turning process in machining by using Takagi-Sugeno-Kang (TSK) fuzzy modeling to accomplish the integration of multi-sensor information and tool wear information. It generates fuzzy rules directly from the input-output data acquired from sensors, and provides high accuracy and high reliability of the tool wear prediction over a wide range of cutting conditions. The experimental results show its effectiveness and satisfactory comparisons relative to other artificial intelligence methods.