Probabilistic Design Tool for Small Turbine Engine High Cycle Fatigue (HCF) Life

This project will significantly improve the accuracy of high cycle fatigue (HCF) life predictions of centrifugal compressors used in small gas turbine engines. This will be done by developing a probabilistic reduced-order model HCF life prediction computercode that captures the random blade mistuning effects present in actual hardware. Mistuning effects can double or triple the stresses of an ideal bladed-disk assembly (one without blade-to-blade geometry and/or material variations). The reduced-ordermodeling technique enables a sufficiently accurate representation of the underlying physics while avoiding the prohibitively long computational times required by direct simulations involving random variables. An existing mistuning code (REDUCE) will beused to determine the vibration frequencies and amplitudes and an existing structural analysis code (ANSYS) will be used to generate the corresponding probabilistic stresses and strains required by the HCF life model. Phase II will also develop models topredict the redistribution of residual stresses induced by three types of surface enhancement processes (SEP) under thermal and probabilistic HCF loading conditions and incorporate these models into the HCF life prediction code. Fatigue life testing of anexisting impeller will be conducted to verify the accuracy of the HCF code. The traditional life prediction methodologies used in current small turbine engine design processes are seriously compromised by their inability to capture the inherently randombehavior characteristics of bladed disk assemblies due to mistuning. Consequently, either the compressor designs are overly conservative (with concomitant loss in efficiency and weight penalties) or excessive failures occur during operational use. Theprediction of mistuning effects in bladed disk assemblies is therefore of fundamental importance in achieving a robust HCF-resistant design for small gas turbine engines. The development and validation of an advanced and economical tool for theprobabilistic assessment of random mistuning effects is key to effectively accounting for these important effects during the design process, thereby producing more robust, HCF-free high performance engine components. The application of such an advancedlife prediction tool will result in higher performance engines as well as fewer operational component failures due to manufacturing and material property variances. Both military and commercial small engine applications will benefit from the exploitationof this advanced technology. This proposal was recommended by Honeywell Engines & Systems who intend to apply the analysis tool to assist them in the their