With industrial equipment developing towards high-performance states in severe environments, the reliability of the impeller materials running under special working conditions such as high temperature, high pressure, high speed, serious corrosion, and multi-load cycles has acquired particular importance. The paper reviews materials’ fatigue reliability and failure modes in harsh service, fatigue life prediction, material response modeling, failure mode identification, and reliability estimation methods, with research accomplishment highlighted in common applications such as aero-engine turbine blades. Through contrast of different analysis models and test methods, this paper is meant to provide theoretical support and practical suggestions for engineering design and life management of impeller materials in complicated working conditions.
Introduction
Impellers have been widely used in aerospace, energy power, chemical engineering, shipbuilding, and other fields. Their operation conditions are usually combined with severe environmental environments, like high temperature, harsh thermal cycles, complex load coupling, corrosion medium erosion, and particle erosion. In such cases, impeller materials not only must have excellent mechanical properties and thermal stability but also have good fatigue resistance and failure suppression capability. Because failure could lead to severe machinery breakdowns or even complete destruction, systematic reliability analysis has been a point of interest in recent research.
Fatigue Damage Evolution and Modeling Strategies under Extreme Conditions
Under harsh service conditions, the fatigue mechanisms encountered by impeller materials are exceedingly complex. Normal fatigue like thermo-mechanical fatigue (TMF), low-cycle fatigue (LCF), and high-cycle fatigue (HCF) tend to occur simultaneously.
At high temperatures, plastic material deformation and creep interact to cause accelerated crack growth in grain boundaries or weakened areas. Furthermore, cyclic thermal loading also increases thermal fatigue damage and reduces the local fracture toughness of materials.
For life prediction, the traditional S-N curve method is well suited to predict fatigue life under a single stress amplitude but its usage will drastically reduce under complex alternating load conditions.
Fracture mechanics-based Paris model is increasingly being used to describe the crack growth behavior. Combined with finite element simulation for the computation of stress intensity factors, it is better suited for predicting crack growth life.
In addition, modern fatigue analysis has more and more used the methods of probability and statistics. With random modeling methods such as the Gamma process or Markov chain, uncertainty in fatigue life may be easily quantified. The method is particularly suitable for the treatment of high-reliability assessment dominated by uncertainties with multiple sources.
In recent years, multi-scale modeling has also been a hot topic. The method combines the material’s micro-structure (e.g., grain size and dislocation structure) with macroscopic mechanical response, making fatigue prediction more physically based and flexible.
Research Achievements on Reliability of Aero-Engine Blades
Aero-engine blades are typical high-reliability key components, operating in very hostile conditions, and therefore a hot target of fatigue studies.
Blades have to withstand high-temperature gas flow of more than 1,000°C over a long duration, with centrifugal forces due to high-speed rotation, and periodic thermal stresses and aerodynamic excitation. The cumulative effect makes blades (highly susceptible to) crack or even fracture.
Testing has proven that aero-blades mainly suffer from the following three failure mechanisms:
Low-cycle fatigue (LCF):
Repeated start-stop operation occurring frequently leads to thermal fatigue cracks at the blade root;
High-cycle fatigue (HCF):
Tip or edge resonance caused by vibration or air flow causes fatigue fracture;
Corrosion fatigue and oxidation damage:
Corrosion by operating media and oxidation of material at high temperatures degrade the surface strength, leading to micro-cracks.
Single-crystal high-temperature alloys such as CMSX-4 and DD6 find widest application in aero-engine blades to combat the above issues. These alloys have no grain boundaries, making their creep and fatigue behaviors better at high temperatures.
Concurrently, for the enhancement of corrosion resistance, blade surfaces are typically 辅以 (supplemented by) coating protection methods such as thermal barrier coatings (TBC) and aluminization to further prevent surface crack formation.
For monitoring and inspection, non-destructive testing methods such as X-ray CT, laser vibrometry, and fluorescent magnetic particle inspection are widely used in the examination of blade service conditions, thereby achieving early damage detection and warning.
Material Failure Mechanism Analysis and System-Level Reliability Evaluation Methods
Understanding the material failure mechanism is the basis of reliability analysis. Through metallographic analysis, scanning electron microscopy observation, and energy-dispersive spectroscopy, causes of crack initiation, crack propagation paths, and fracture morphology can be identified.
For example, in thermal fatigue, distributions of micro-cracks of “fishbone-like” and grain boundary oxidation are typical; for corrosion fatigue, fractures exhibit typical characteristics such as crack tip oxidation and corrosion product coverage.
Reliability assessment not only focuses on the material itself but also assesses aspects such as structural geometric flaws, manufacturing defects, and variations in working conditions.
At the engineering level, reliability analysis normally uses the Monte Carlo method (MC), response surface method (RSM), and stochastic finite element method (S-FEM), etc. By statistically calculating failure probabilities using large numbers of simulation samples, the reliability and safety margin of structures within the target life duration are estimated.
In addition, with the advent of industrial digitization, digital twin modeling of life data inversion is increasingly becoming a new direction. It can be used to real-time monitor impeller service conditions, predict remaining life, and provide maintenance suggestions.
Conclusions
Impeller material reliability analysis in given working conditions is a highly interdisciplinary and complex systematic engineering with material science, structural mechanics, thermodynamics, probability statistics, and other interdiscipline knowledge. Future research will expand in the following directions: first, multi-source coupling modeling of uncertainty to strengthen the accuracy and credibility of the assessment; second, introducing intelligent algorithms such as machine learning for prediction of fatigue life and fault diagnosis; third, establishing an Internet of Things-based service monitoring system to achieve intelligent and real-time reliability assessment of the impeller.
