Applied Mechanics and Materials Vols. 10-12

Abstract: For spherical rotor of gyroscope and some other high speed rotor, static balance is important. The paper introduces a static balance method for spherical rotor: the position of the spherical rotor is fixed by the use of the special clamp apparatus, and then determines the unbalance mass using the electronic balance. This method has no strict requirement to the clamp’s symmetry and dimension and locating precision of the locating hole. The balance precision only depends on the precision of the electronic balance and shape error of the clamp’s locating surface, which makes the balance method has high precision and preferable operability.

Abstract: Nanoindentation has become the main method to obtain the micro- characteristic of materials by now. Nanoindentation device has the ability to make the load-displacement measurement with sub-nanometer indentation depth sensitivity, and the nanohardness of the material can be achieved by the load-displacement curve. The article discusses a new convenient nanohardness testing device, and approves its practicability and reliability by taking the indentation experiment on single-crystal aluminum.

Abstract: A piezoelectric ceramic power supply with self-feedback adjustment was developed according to piezoelectric ceramic driver’s characteristic. It uses multi dc voltages connected in series as the high voltage supply circuit, and has many advantages, such as small ripple, high precision, high driving capability, quick response and so on.

Abstract: Radiography inspection (X-ray or gamma ray) is one of the most commonly used Non-destructive Evaluation (NDE) methods. More and more digital X-ray imaging is used for medical diagnosis, security screening, or industrial inspection, which is important for e-manufacturing. In this paper, we firstly introduced an automatic welding defect inspection system for X-ray image evaluation, defect image database and applications of Artificial Neural Networks (ANNs) for NDE. Then, feature extraction and selection methods are used for defect representation. Seven categories of geometric features were defined and selected to represent characteristics of different kinds of welding defect. Finally, a feed-forward backpropagation neural network is implemented for the purpose of defect classification. The performance of the proposed methods are tested and discussed.

Abstract: In this paper, a cluster-based feature extraction from the coefficients of discrete wavelet transform is proposed for machine fault diagnosis. The proposed approach first divides the matrix of wavelet coefficients into clusters that are centered around the discriminative coefficient positions identified by an unsupervised procedure based on the entropy value of coefficients from a set of representative signals. The features that contain the informative attributes of the signals are computed from the energy content of so obtained clusters. Then machine faults are diagnosed based on these feature vectors using a neural network. The experimental results from the application on bearing fault diagnosis have shown that the proposed approach is able to effectively extract important intrinsic information content of the test signals, and increase the overall fault diagnostic accuracy as compared to conventional methods.

Abstract: This paper presents a novel method for bearing fault diagnosis based on wavelet transform and Gaussian mixture models (GMMs). Vibration signals for normal bearings, bearings with inner race faults, outer race faults and ball faults were acquired from a motor-driven experimental system. The wavelet transform was used to process the vibration signals and to generate feature vectors. GMMs were trained and used as a diagnostic classifier. Experimental results have shown that GMMs can reliably classify different fault conditions and have a better classification performance as compared to the multilayer perceptron neural networks.

Abstract: Prognosis of major components such as blades, rotors, valves of steam turbine is crucial to reducing operating and maintenance costs. Prognostic strategies can assist to detect, classify and predict developing faults, guarantee reliable, efficient and continuous operation of electric plants, and may even result in saving lives. In this paper, a recurrent neural network based strategy was developed for blade material degradation assessment and fatigue damage propagation prediction. Two Elman Neural Networks were developed for fatigue severity assessment and trend prediction correspondingly. The performance of the proposed prognostic methodology was evaluated by using blade material fatigue data collected from a material testing system. The prognostic method is found to be a reliable and robust material fatigue predictor.

Abstract: It conceived the concept of data whole-life management in the environment of enterprise integration. Facing many different data storage systems and data types, and different storage formats of one kind of data, it designs an extensible and function-open system integration platform based on J2EE technology. Through its intelligent management of data storage system, the system shields the different storage systems from its ultimate users. And through the function models with the functionality of transforming data stored in one format into the one stored in another format deployed onto it, the system can manage different storage formats and make them transparent to users and realizes Data Transformation Management. Through the universal format described in XML transformed from other different formats by function models, it makes the internal structure of one data ventilate to the out world and realizes Data Integration, as the basis of Business Integration.

Abstract: This study takes virtual instrument technology as the development platform to complete data acquisition, pre-processing, analysis and database storage for three orthogonal components of a cutting force and the corresponding cutting temperature. Simultaneously, single-factor experimentation is adopted to establish empirical formulas of these cutting state parameters for further check analysis. Hence real-time monitoring of cutting process can be implemented to represent cutting-tool wear, failure and rationality of parameter selection in cutting state.

Abstract: This paper presents a novel hybrid feature selection algorithm based on Ant Colony Optimization (ACO) and Probabilistic Neural Networks (PNN). The wavelet packet transform (WPT) was used to process the bearing vibration signals and to generate vibration signal features. Then the hybrid feature selection algorithm was used to select the most relevant features for diagnostic purpose. Experimental results for bearing fault diagnosis have shown that the proposed hybrid feature selection method has greatly improved the diagnostic performance.