Redesign of a High-Pressure Compressor Blade Accounting for Nonlinear Structural Interactions

Recent numerical developments dedicated to the simulation of rotor/stator interaction involving direct structural contacts have been integrated within the Snecma industrial environment. This paper presents the first attempt to benefit from these developments and account for structural blade/casing contacts at the design stage of a high-pressure compressor blade. The blade of interest underwent structural divergence after blade/abradable coating contact occurrences on a rig test. The design improvements were carried out in several steps with significant modifications of the blade stacking law while maintaining aerodynamic performance of the original blade design. After a brief presentation of the proposed design strategy, basic concepts associated with the design variations are recalled. The iterated profiles are then numerically investigated and compared with respect to key structural criteria such as: (1) their mass, (2) the residual stresses stemming from centrifugal stiffening, (3) the vibratory level under aerodynamic forced response and (4) the vibratory levels when unilateral contact occurs. Significant improvements of the final blade design are found: the need for an early integration of nonlinear structural interactions criteria in the design stage of modern aircraft engines components is highlighted.

On the Reduction of the Pre-Processing Effort and the Application of a Contact Meshing Approach for Complex Jet Engine Component Assemblies

A significant proportion of the work effort for finite element (FE) analysis is spent for pre-processing activities, especially for complex structural components and component assemblies. An exclusive use of hexahedron (hex) elements increases the meshing effort substantially compared to tetrahedral elements. An automated method to generate high quality hexahedral meshes for an arbitrary geometry does not exist. In this work, commercially available FE software tools for meshing were investigated with the focus on an advantageous pre-processing effort. The evaluation showed that the software package NX (Siemens PLM Software) offers robust advanced semiautomatic hex meshing capabilities. Furthermore, a Contact Meshing Approach (CMA) was elaborated to reduce the effort of the challenging and time-consuming geometry decomposition significantly. Using the example of an intermediate pressure compressor it can be shown that the pre-processing effort time can be reduced up to 75%. Due to the independent meshes, element transitions in the geometry become redundant. This results in lower total element numbers and higher mesh qualities and subsequently leads to more efficient calculations. Moreover, the increased element quality has positive effects on the result quality.

Influence of Alloy Behavior on Multiaxial Fatigue Lifing Criteria for Turbine Disk Bores

In a previous paper, a criterion for multiaxial lifing of turbine disk bores made from a nickel base super alloy, DP-718, was substantiated using results from spin pit testing of mini disks. In this paper, another turbine and compressor disk alloy, Alloy 10, exhibiting a different material behavior than DP-718 is examined from multiaxial point of view. This new alloy manufactured by powder metallurgy processing, has a coarser grain size, lower tensile yield strength, contains a much finer distribution of brittle carbides in its microstructure and shows a significant difference in failure initiation behavior under low cycle fatigue when compared to DP-718. Under the multiaxiality conditions experienced by disk bores, Alloy-10 lives seem to correlate well using an effective stress based criterion, whereas DP-718 was found to follow principal stress based criterion. Interesting differences between DP-718 and Alloy-10 in alloy and fractographic behavior under uniaxial and multiaxial conditions are discussed.

Method for Selecting Master Degrees of Freedom for Rotating Substructure

A method of selecting master degrees of freedom (DOFs) for rotating substructure is presented in this paper to obtain reduced 3D rotor models. Fixed modes of the substructure below thrice the operating frequency are analyzed. According to each mode shape, the DOFs at where main kinetic energy locates are selected as master DOFs to decrease missing of dynamic coupling. Additional DOFs may be selected based on traditional substructure method. In the stationary reference frame, frequency-dependent gyroscopic effects can be included as damping matrices changing with spin speed. Besides, by selecting appropriate substructure, localized damping and key parts of the rotor for analysis can be kept the same as the original model. A reduced model of a high pressure rotor amply demonstrated the capability of the method in reducing the model size and increasing the computational efficiency with less than two percent error.

Nonlinear Dynamic Study on Effects of Rub-Impact Caused by Oil-Rupture in a Multi-Shafts Turbine With a Squeeze Film Damper

An investigation is carried out the analysis of nonlinear dynamic behavior on effects of rub-impact caused by oil-rupture in a multi-shafts turbine system with a squeeze film damper. Main components of a multi-shafts turbine system includes an outer shaft, an inner shaft, an impeller shaft, ball bearings and a squeeze film damper. In the squeeze film damper, oil forces can be derived from the short bearing approximation and cavitated film assumption. The system equations of motion are formulated by the global assumed mode method (GAMM) and Lagrange’s approach. The nonlinear behavior of a multi-shafts turbine system which includes the trajectories in time domain, frequency spectra, Poincaré maps, and bifurcation diagrams are investigated. Numerical results show that large vibration amplitude is observed in steady state at rotating speed ratio adjacent to the first natural frequency when there is no squeeze film damper. The nonlinear dynamic behavior of a multi-shafts turbine system goes in its way into aperiodic motion due to oil-rupture and it is unlike the usual way (1T = >2T = >4T = >8T etc) as compared to one shaft rotor system. The typical routes of bifurcation to aperiodic motion are observed in a multi-shafts turbine rotor system and they suddenly turn into aperiodic motion from the periodic motion without any transition. Consequently, the increasing of geometric or oil parameters such as clearance or lubricant viscosity will improve the performance of SFD bearing.

Analysis of the Rotordynamic Response of a Centrifugal Compressor Subject to Aerodynamic Loads due to Rotating Stall

In the current industrial research on centrifugal compressors, manufacturers are showing increasing interest in the extension of the minimum stable flow limit in order to improve the operability of each unit. The aerodynamic performance of a compressor stage is indeed often limited before surge by the occurrence of diffuser rotating stall. This phenomenon generally causes an increase of the radial vibrations, which, however, is not always connected with a remarkable performance detriment. In case the operating curve has been limited by a mechanical criterion, i.e. based on the onset of induced vibrations, an investigation on the evolution of the aerodynamic phenomenon when the flow rate is further reduced can provide some useful information. In particular, the identification of the real thermodynamic limit of the system could allow one to verify if the new load condition could be tolerated by the rotordynamic system in terms of radial vibrations. Within this context, recent works showed that the aerodynamic loads due to a vaneless diffuser rotating stall can be estimated by means of test-rig experimental data of the most critical stage. Moreover, by including these data into a rotordynamic model of the whole machine, the expected vibration levels in real operating conditions can be satisfactorily predicted. To this purpose, a wide-range analysis was carried out on a large industrial database of impellers operating in presence of diffuser rotating stall; the analysis highlighted specific ranges for the resultant rotating force in terms of intensity and excitation frequency. Moving from these results, rotordynamic analyses have been performed on a specific case study to assess the final impact of these aerodynamic excitations.

Switch Triggers for Optimal Vibration Reduction via Resonance Frequency Detuning

For systems subjected to linear frequency sweep excitation, piezoelectric-based resonance frequency detuning provides vibration reduction by altering the stiffness state of the material as it passes through resonance. This vibration reduction technique applies to turbomachinery experiencing changes in rotation speed, for example on spool-up and spool-down. The peak response dynamics are determined by the system’s sweep rate, modal damping ratio, electromechanical coupling coefficient, and, most importantly, the frequency at which the stiffness state is altered. An analytical approach is employed to solve the nondimensional single degree of freedom equation of motion and is scaled to incorporate the altered system frequency following the stiffness state switch. This paper provides an extensive study over a range of sweep rates, damping ratios, and electromechanical coupling coefficients to determine the optimal frequency switch trigger that minimizes the response envelope. This switch trigger is primarily a function of the electromechanical coupling coefficient and the phase of vibration at which the switch occurs. As the coupling coefficient increases, the switch trigger decreases and is approximately linear with the square of this coupling coefficient. Furthermore, as with other state-switching techniques, the optimal frequency switch occurs when the phase of vibration is at the point of maximum displacement, or peak strain energy.

Life Fraction Hardening Applied to a Modified Theta Projection Creep Model for a Nickel-Based Super-Alloy

This paper presents the application of a life fraction hardening rule to the analytical calculation of creep in hot section components. Accurate prediction of creep is critical to assuring the mechanical integrity of heavy-duty, industrial gas turbine (IGT) hardware. The accuracy of such predictions depend upon both the creep models assumed and how those models are implemented in a finite element solution. A modified theta projection creep model for a nickel-based super alloy was presented in a previous paper as an accurate simulation of creep behavior [1]. Application of such a user defined creep law depends upon definition of a hardening rule in the form of either an explicit or an implicit integration scheme in order to calculate incremental strains during any time increment. Time hardening is the simplest and least computationally intensive of the two most common hardening rules, but does not correctly show the effect of changing stresses or temperatures. Strain hardening may provide the most accurate solution, but the creep models are too complex to invert, which results in highly iterative and computationally intensive solutions. A life fraction hardening rule has been presented in other works [2] as a compromise between time hardening and strain hardening. Life fraction hardening is presented here as a highly efficient and accurate means of calculating incremental creep strain when applied to a modified theta projection creep model. A user creep subroutine was defined using a state variable to represent the strain life fraction at any time. By using the time to tertiary creep as the denominator for the life fraction, no new material constants are needed to relate to creep failure. The start of tertiary creep is effectively considered to be a failure. Additional design insight can be provided through the inclusion of other state variables to calculate temperature margins at current conditions. Material testing with changing stress levels will be used to help validate the technique. A simplified example of the technique is presented in the paper. More accurate creep predictions allow our company to improve the structural integrity of its turbine blades and vanes.

Analysis of the Stiffness of Hirth Couplings in Rod-Fastened Rotors Based on Experimental Modal Parameter Identification

Hirth couplings are widely used to transport torque between discs in turbine machinery for the advantages of precise centering, reliable positioning and excellent structural stability. However, the stiffness of the rotor segment with Hirth couplings is less than that of the integrated structure because it is weaken in the discontinuous structure. For the rotor segment with Hirth couplings, there are two types of contact status depending on the relationship between the separate stress and the compression stress. The first type is that the whole zone of Hirth couplings is in contact, where all of the Hirth couplings make contribution to the stiffness of the rotor segment. The second type of contact status is that some zones of the Hirth couplings are separated, where the equivalent stiffness of the rotor segment is only determined by the zones which are in contact. Obtaining accurate stiffness of Hirth couplings is of great significance in rotor dynamic performance analysis. In this study, the flexural stiffness of Hirth couplings of a gas turbine was calculated by three dimensional (3D) nonlinear contact Finite Element Method (FEM), and based on which the natural frequencies of a rod-fastened rotor were investigated. The stiffness modifying method which modifies the elasticity modulus of the material in the connection segment was used to conduct modal analysis in order to evaluate the effect of the stiffness weakening of the Hirth couplings. One experimental rotor with Hirth couplings and a central tie rod was designed to verify the theoretical and numerical analysis results. According to the results, the stiffness coefficient is constant when the load factor γ is less than 1.0, which means the stiffness of the Hirth couplings segment is almost constant when the pre-tightening force is large enough to keep the whole zone of the Hirth couplings in contact. The stiffness coefficient constant is about 0.32 (specific for the investigated model). And the stiffness coefficient drops dramatically when the load factor γ exceeds 1.0, which indicates that the stiffness of the Hirth couplings segment decreases dramatically since the pre-tightening force is insufficient and the Hirth couplings are separated. The Hirth couplings segment stiffness obtained by experimental modal parameter identification turned out to be consistent with the calculated results by FEM. The modified stiffness of the Hirth couplings segment were applied on the dynamic performance analysis of a real gas turbine rotor and the accuracy of calculating results was improved.

Operational Modal Analysis of Lateral Rotordynamic Modes of Rotating Machinery

The lateral rotordynamic response of turbomachinery is typically speed dependent due to hydrodynamic lubricated bearings, seals, gyroscopic and centrifugal effects, etc. Rotordynamic tools are used to predict the behavior of the machine during operation, however validating these results is challenging. Traditional experimental modal testing techniques rely on controlled and measured excitation together with measured responses. However, during operation this is unpractical, as the actual excitation force is rarely known. Operational modal analysis (OMA) can identify the modal parameters of a system over its entire operational range from measurement of response due to some (unknown) excitation. OMA has proven successful on non-rotating structures, but has seldom been applied to rotating machinery. Three case studies are presented demonstrating the use of OMA in identifying lateral rotors modes based on measurements from existing radial proximity probes during normal production undertaken as part of commissioning campaigns. Challenges encountered in using and interpreting OMA results are discussed. The results show that proximity probe data acquired during normal operation may be used as input to OMA for the assessment of stability margins of rotating machinery, to produce experimentally derived Campbell diagrams and to identify backwards as well as forwards whirling modes.