No abstract

In this paper, an innovative system for condition-based monitoring (CBM) using model-based estimation (MBE) and artificial neural network (ANN) is proposed. Fault diagnosis of deep groove ball bearings (DGBB) is a key machine element for stability of rotating machinery. MBE model is proposed to demonstrate and estimate the vibration characteristics of bearings. It is realized that it may be worth mentioning that the vibration analysis of damaged bearings at all the positions of a structure is difficult to obtain. For this purpose, methods have been discussed to get the utmost information to notify bearing faults. The ANN approach enables us to determine the effects of various parameters of the vibrations by conducting the experiments. The results point out that defect size, speed, load, unbalance, and clearance influence the vibrations significantly. Experimental simulated data using the MBE and ANN models of rotor–bearing are used to identify the damage diagnosis at a reasonable level of accuracy. The results of the experiments consist in constantly evaluating the performance of the bearing and thereby detecting the faults and vibration characteristics successfully. The effects of faults and vibration characteristics obtained using the experimental MBE and ANN are studied.

Many industries make wide use of rotor bearing systems such as high speed turbines and generators. However, the vibration of antifriction rotor–bearings is a key factor in reducing the life of the bearings; thus significantly influencing the performance and working life of the whole power plant. In earlier research on the vibration characteristics of high speed rotor–bearing systems, such as in induced draft (ID) fans, an application used in sugar cane factories, the supporting antifriction bearings were simplified as a particle on a shaft with radial stiffness and damping coefficient. However, such simplification neglects the effects of the bearing structure on the vibration performance of the rotor–bearing system. This paper demonstrates the benefits of a more holistic approach and establishes a numerical model of the stiffness of the spherical roller bearing through Buckingham's π theorem (BPT). On the basis of this model, we argue for the benefits of a new dimensional analysis (DA) technique for rotor–bearing systems. Our new DA also considers the influences of the bearing structure parameters on the vibration of rotor–bearing systems. We demonstrate the effectiveness of our approach by conducting a comparative BPT study using an ID fan, a rotor–bearing system in use in sugar cane factories. We first analyzed an ID fan using the simplified model to obtain the defect frequencies and vibration amplitude responses of the ID fan system. Subsequently the same ID fan rotor was also analyzed using our new multivariable regression analysis (MVRA) approach to verify the validity of our new and holistic BPT. The results indicate that the new method we propose in this paper for the calculation of vibration characteristics of a high speed rotor–bearing (ID fan) is credible and will save time and costs by the accurate detection of imminent bearing failure.

Diagnosis of antifriction bearings is usually performed by means of vibration signals measured by accelerometers placed in the proximity of the bearing under investigation. The aim is to monitor the integrity of the bearing components, in order to avoid catastrophic failures, or to implement condition based maintenance strategies. In particular, the trend in this field is to combine in a simple theory the different signal-enhancement and signal-analysis techniques. The experimental data based model (EDBM) has been pointed out as a key tool that is able to highlight the effect of possible damage in one of the bearing components within the vibration signal. This paper presents the application of the EDBM technique to signals collected on a test-rig, and be able to test damaged fibrizer roller bearings in different working conditions. The effectiveness of the technique has been tested by comparing the results of one undamaged bearing with three bearings artificially damaged in different locations, namely on the inner race, outer race, and rollers. Since EDBM performances are dependent on the filter length, the most suitable value of this parameter is defined on the basis of both the application and measured signals. This paper represents an original contribution of the paper.

Roller bearings are essential parts extensively used in many industries such as automobile, sugar factories, cement industries, weaving mills, chemical industries, and other process industries. The catastrophic failure of such bearings results into unplanned shutdowns, discontinuity of manufacturing process, and heavy maintenance cost. The vibration analysis of the roller bearing is a vital factor in the rotating machines because its performance significantly affects the safety and operational life of the rotating machines and subsequently entire plant. The object of this paper is to study how to predict the vibration characteristics of the rotor-bearing system by using the mathematical model. In the present research work, a empirical model for the vibration characteristics of the roller bearing has been established using FLTθ system. The new mathematical model considers the influences of the bearing variables on the vibration of the rotor system. Furthermore, a new model on bearing system is carried out by using dimensional analysis (DA) and the defect frequencies and vibration characteristics of the bearing system are obtained. The effects of speed and load along with other variables on vibration characteristics have been studied by establishing an empirical model. Experiments were conducted to validate the developed empirical model. The method proposed in this paper is based on FLTθ method of DA. The vibration characteristics thus obtained provides a complete and systematic theory and technique in this aspect.