Turbo machinery blade rows are designed to contain identical blades within a given blade row. However, minor differences in blade shape, structural properties, material properties, and flow field are inherently present among the blades in all-bladed disks. Collectively, these minor variations are known as “mistuning.” Mistuning coupled with flexibility of the disk can localize the vibration energy of the system to a small sector of the disk containing only a few blades. These blades could experience vibratory stresses several times higher than other blades not influenced by mistuning. Current design methods assume a rotor to be tuned (all blades identical) and do not quantitatively account for the effects of mistuning on the fatigue life of the rotor, leading to higher rates of blade failure in the field than calculated during the design process. N&R has addressed this issue by using advanced mathematical methods to analytically determine the impact of mistuning during the design process.
N&R has investigated the impact of structural mistuning of centrifugal compressors on high cycle fatigue life. A physics-based approach to accurately capture structural mistuning effects on blade stresses was developed and applied to the design process of two baseline Honeywell impellers – one splittered and the other un-splittered. A design-of-experiments technique in conjunction with response surface methodology was used to efficiently explore a 4-variable design space consisting of back plate thickness, backward curvature angle, airfoil thickness, and blade root radius. Blade stress was selected as the optimization criterion as a proxy for high-cycle fatigue life. Probabilistic stresses were calculated using a reduced order finite-element modeling mistuning code (Turbo-REDUCE) in combination with the ANSYS computer-aided design code and the FPI fast probabilistic integration code. Results for typical mistuning assumptions show that a subset of the total blade set are subject to peak stresses 60 percent greater than those without structural mistuning. This result is in agreement with experimental tests and leads to the conclusion that an entirely (or nearly so) analytic design methodology can replace the traditional practice of requiring a lengthy and expensive experimental test program to accompany the design process. The multi-code solution methodology is generic and can be applied to a broad category of mistuning problems.
- Quantified the effect of mistuning and its effect on High Cycle Fatigue (HCF) life in small engine impellers due to uncertainties in the geometric and material variables.
- Developed a methodology to perform probabilistic mistuning analysis to improve HCF life estimates.
- Provided a design-of-experiments technique to characterize the geometric uncertainties in small engine impellers.