PSI - Issue 17

Andra Gabriela Stancu et al. / Procedia Structural Integrity 17 (2019) 238–245 A. G. Stancu et al./ Structural Integrity Procedia 00 (2019) 000 – 000

239

2

1. Introduction

1.1. Background

Mechanical integrity of drivetrains in complex machineries is critical to maintain, ensuring their availability and reducing unexpected downtime. Successfully applying condition-based maintenance (CBM) brings the benefits of increasing the operators’ awareness of the whole systems, hence scheduling maintenance strate gies that are proactive and targeted to resolve incipient faults before propagation. Being researched and practiced for more than 30 years [1], condition monitoring is one of the most adopted maintenance strategies to guarantee the safety operations of key facilities [2]. Predominantly, condition monitoring applies techniques to collect and translate the key system information through installed sensors, enabling continuous assessment with the help of Digital Signal Processing (DSP) and tracking of the condition of the monitored machinery. This principle has seen application widely in industrial rotating electric machines as reviewed in [1-5]. The cost reduction benefit of condition monitoring is well recognized, enhanced by the rapid development of available sensory technologies and their corresponding signal analysis techniques. Fault detection techniques and reliable monitoring methods for early fault diagnosis have been studied extensively, generating a continuous advancement in statistical analysis algorithms, as well as functional integration and performance of available monitoring products. However, there is still a continuous effort allocated to new developments which can offer better pattern analysis capabilities and lower-cost alternatives to available products and services. The type of the machine failure will dictate most of the time the way to prevent it from happening. Among well-established techniques, vibration analysis is widely used for condition monitoring and fault identification. It is supported by various commercially available accelerometers, offering wide selections of dynamic ranges, utilities and sensitivities. The studies on vibration-based condition monitoring are innumerable. To name a few, in [2] and [5], the authors reviewed vibration-based condition monitoring technologies, while in [6] the authors have explained theoretical methods and summarized applicable vibration processing techniques. While vibration monitoring is the most regularly used technique in rotating machinery applications, monitoring machines operating at very low rotational speed has been reported to negatively influence the effectiveness of vibration analysis [2], [7]. This is due to the compromised vibration signal-to-noise ratio (SNR) when under slow rotation, as well as the limitation of the sensor’s measurable frequency range. Acoustic Emission (AE) is another well-known and promising technology used for a wide range of applications. AE signal is a high frequency elastic wave emitted due to surface or internal deformation of the component being monitored. Typical AE sources could be fault mechanisms including cracks, corrosion, fibers breaking, increased friction, leakage, etc. While AE has been studied extensively for structural integrity and monitoring including bridges or other concrete structures, in recent decades an increasing trend of applying AE-based monitoring for rotating machinery can be observed [8] [9] [10]. It is reported that through appropriate signal processing, AE signals can be used to detect small-scale degradation, incipient defects in both low speed and high speed machines. On the other hand, as pointed out in [11], AE signals are more susceptible to complicated transmission paths, rapid amplitude attenuation as well as demanding in advanced processing techniques. Other monitoring technologies including but not limited to audible acoustics, temperature, infrared, oil debris, current signatures, etc., all have specific applications under certain conditions. 1.2. Condition Monitoring Techniques

1.3. Inspection and condition monitoring applicability

Planned inspections alone cannot capture the degradation mechanism from a one-time evaluation due to signal fluctuation and the coherent noise. Monitoring, on the other hand, addresses these challenges by instantaneously capturing a wide range of parameters and keeping track of their trend. Condition monitoring allows the machinery to be monitored autonomously for a long period, establishing a baseline of the normal operating conditions and giving warnings when signs of degradation are captured. It is therefore, better