Articulated Splined Couplings for Future Aero Gas-Turbine Engines
Lead Research Organisation:
University of Nottingham
Department Name: Faculty of Engineering
Abstract
A new architecture of aero gas-turbine engines are in the process of being developed for Rolls-Royce. Spline couplings are a key component for the connection of a new gearbox within the engine. Therefore, an alternative design of spline coupling will be needed compared to traditional designs. This research project involves creating new design tools and approaches to which mechanisms will apply for the design of spline couplings for future aero gas-turbine engines to prevent premature failures. This involves developing a closer understanding to why spline couplings fail under certain failure mechanisms at different times of its predicted life cycle.
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| EP/P510592/1 | 01/10/2016 | 23/12/2021 | |||
| 1943752 | Studentship | EP/P510592/1 | 02/10/2017 | 30/09/2021 | Yuen-Ling Kong |
| Description | A spline coupling involves external gear teeth on a drive shaft, which mates with a sleeve containing the same amount of internal gear teeth. These are used to transfer torque from the shaft to the sleeve. In gas-turbine aeroengines spline couplings can experience large amounts of angular misalignments due to external loads during a flight cycle causing a significant amount of fretting wear. Fretting wear involves small relative movement between two surfaces causing the removal of material reducing the structural integrity of the component. The main aim is to create a design tool to aid the design process against fretting wear (and other modes of failure) for spline couplings. There are no effective design tools to design against fretting wear for a spline coupling, as current contact solving methods for a spline coupling take a significant amount of computational time to consider including fretting wear modelling for large number of cycles. Therefore, various contact solving methods were explored and improved. Initially, a fully analytical contact model was developed for a spline coupling under pure torsion, which contact pressures benchmarked well against the finite element method. However, it was found that to include fretting wear modelling, the contact solver needed to adapt to changing geometry as it wears, as well as including other types of loading and tooth geometries. Therefore, a method known as the Modified Simplex Method (MSM) was used as a contact solver that sets the contact problem as a mathematical program. This is an optimisation algorithm, which solves for linear constraints with a criterion for contact directly. The contacting surfaces of each tooth pair was discretised into a rectangular grid, where contact pressures were solved at the centre of each element. The separation between all the contacting pairs and the nominal loads and/or rigid body approaches can be used as the input. Tooth modifications and manufacturing errors were also explored based on the initial separation between potential contacting surfaces. When MSM was implemented for a spline coupling, it was found to take a significant amount of computational time due to multiple contacting surfaces. This needed to be optimised to allow for fretting wear modelling to run in the order of millions of cycles. Therefore, with help from Imperial College colleagues, an alternative contact solving method known as Conjugate Gradient Method (CGM), which solved the mathematical program using an iterative method instead of directly. This reduced computational time by 99%. In order to find the angular and parallel misalignments, a model of the shaft bending for two spline couplings connected via a shaft was developed. A system of equations was developed, and the unknowns (misalignments for both splines, shaft bending deflection and inflection) were solved using Newton's method with the contact solver. This allowed to observe how the shaft bending interaction influences the loading of the two spline couplings. For the fretting wear model, gross slip was assumed, and the slip values were determined based on the found misalignment values, which were averaged over a cycle. The modified Archard's wear law (a function of wear coefficient, contact pressure and contact slip) was used to model the wear depths across each contacting tooth pair, which allowed the initial separation (i.e., the surface geometry) to be adjusted after each iteration for the contact solver and the shaft bending model to solve for the contact pressures for both spline couplings. Linear and adaptive cycle jumping were used to optimise computational time, which allowed the user to run the fretting wear model in the order of a million cycles within a few hours. Overall, this fretting wear model allowed to explore: two spline geometries connected via a shaft; the influence of modifications, zero backlash and manufacturing errors; the evolution of rigid body approaches and loading; and the evolution of the wear scars and depths over each tooth of each spline coupling. |
| Exploitation Route | The fretting wear model for a spline coupling can be used as a design tool in sectors where spline couplings are used, not just for gas-turbine aeroengines. The model includes the options of tooth modifications, manufacturing errors, various loadings and the effect of shaft bending. This allows to alter various parameters within a design space and see how this affects the amount of fretting wear and its evolution over a number of cycles, as well as investigating tooth engagement using the contact solver. Additionally, the model could be extended to include alternative boundary conditions and include other contacting surfaces, as well as adapting the model for different types of gears. It could also be extended to include modelling of debris entrapment and plasticity under various parameters and loading conditions to improve the predictability of the fretting wear scars. Furthermore, when an effective wear law has been developed based on physical properties and not just an empirical law, this can be implemented into the model. This can be used to compare with experimental results and validate the wear law for this practical application. |
| Sectors | Aerospace, Defence and Marine,Manufacturing, including Industrial Biotechology,Transport |