Transmission of Vibration Loads between Interacting Shafts
Lead Research Organisation:
Imperial College London
Department Name: Mechanical Engineering
Abstract
This project aims to investigate how vibrations are transmitted between different shafts in a jet engine, known as cross-shaft coupling. Most modern jet engines have split compressors each driven by its own turbine, with the speed of rotation increasing at each stage for improved aerodynamic efficiency. This allows for a more compact design for the same thrust. However, current modelling techniques do not accurately predict the transmission of vibrations from one rotor to another, due to the complex interaction of shafts, rolling-element bearings (REB) and squeeze film dampers (SFD). Previous work has concentrated on developing models of the individual components and this has not been straight forward, so incorporating them into a system level model is expected be extremely challenging, because there could be sub-harmonic, instabilities or even chaotic motion. Validation of the model will be performed against a test rig which will be developed in a parallel activity. The ultimate aim is to deliver fully validated lightweight models of the relevant components, and to provide insight into the mechanisms involved in these phenomena .
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| EP/N509206/1 | 01/10/2015 | 30/09/2021 | |||
| 1790892 | Studentship | EP/N509206/1 | 26/09/2016 | 25/09/2020 | Alexander Haslam |
| Description | This project aims to ultimately understand how vibrations are transmitted between two different rotors in dual-shaft systems, where the rotors are connected with inter-shaft bearings. This arrangement is very common in aero-engines but also arises in other industries. Although the fundamentals of rotor dynamics are well-understood, there is not the same level of understanding for dual-shaft systems. A detailed study was therefore carried out, and it was shown how to characterise the mode shapes and how the frequencies are effected by the inter-shaft bearing. This work was presented at a conference in 2018. Although complex rolling-element bearing (REB) models have been developed in the tribology community, these are not typically used to model the bearings in rotor-dynamic models for industrial applications, where they are often just assumed to be linear springs. However, this is known to be a simplification, and is not capable of explaining phenomena such as chaotic rotor responses. A systematic parameter study was carried out focussing on the impact of a bearing on the non-linear dynamics of a rotor. Unlike existing papers which purely analyse the response of such systems, physical explanations for the behaviour were also provided. This highlighted the relative importance of the different mechanisms in the bearing on the rotor response which will be a useful resource for the development of future rotor-bearing models. This work has been written as journal paper, which has now been published. A test rig known as ARES, consisting of a rotor mounted on bearings, has been built and commissioned. This has been used to provide validation data for Rolls Royce, but has since be reconfigured to amplify the influence of the bearings. Some non-rotating tests have been carried out to validate the rotor model, and rotating tests will begin shortly. In order to ensure the bearing is accurately modelled, a separate bearing loading test rig was designed and built in 2019. This was successfully used to identify the unknown parameters, and the model agreed very well with the experimental results. From the numerical model of the ARES test rig, it is expected some interesting non-linear phenomena will be observed, which have not been investigated in an experimental setting before. It is hoped that this improved understanding will allow Rolls Royce to better control such complex phenomena in their engines in the future. |
| Exploitation Route | This project only considered bearings loading in the radial direction to simplify the analysis. However, in many industrial rotating machinery, the bearing must also transmit axial loads; in jet engines, the bearings must transmit the thrust to the airframe. Therefore it would be interesting to extend the analysis to look at the influence of such axial loading. The current test rig could be modified for that purpose. As previously stated, there will not be enough time to investigate the effect of the SFD any further in this project. However, the same overall strategy could be taken as with the REB analysis carried out in this project. It is important that consideration is paid to the implementation of the bearing model and analysis techniques developed in this project into the software used by Rolls Royce. Initial discussions have been made with the relevant stakeholders at Imperial and Rolls Royce. |
| Sectors | Aerospace, Defence and Marine,Energy,Manufacturing, including Industrial Biotechology,Transport |
| Description | The study on the rotordynamics of dual-rotor systems turned out to formalise much of the intuition developed by the engineers at Rolls Royce. Some other similar small-scale numerical studies have been carried out for Rolls Royce, looking at some other phenomena relating to bearings. These have already provided some design rules of thumb to minimise vibration problems, and the intention is that this work will eventually be written up as another conference paper. Since rolling-element bearings are used in a very large variety of rotating machinery, any improvement in understanding has a large potential impact in the long-run. If the vibration behaviour of the rotor-bearing system can be better predicted, then the system can be designed to inherently minimise any potentially damaging vibration, leading to a potentially improvement in reliability. This would also reduce the reliance on expensive experimental testing. For aero-engines, this has further benefit. If excessive vibration can be avoided at the design stage, it may be possible to remove vibration suppression devices from future engine designs. This could reduce the overall weight of the system, thereby improving the power-to-weight ratio and increasing the efficiency of the aircraft. |
| First Year Of Impact | 2018 |
| Sector | Aerospace, Defence and Marine |
| Impact Types | Economic |
| Title | ARES MATLAB toolbox |
| Description | In order to provide validation data for Rolls Royce, I wrote a MATLAB toolbox for post-processing all the data from the ARES test rig, so they were able to compare the results directly with their model. This saved a lot of duplication of effort, and was fully automated unlike the ad-hoc solutions written by other less-experienced MATLAB coders in the group. This toolbox will continue to be used when I complete my tests on the same test rig. |
| Type Of Material | Data analysis technique |
| Year Produced | 2019 |
| Provided To Others? | No |
| Impact | This was successfully used by Rolls Royce to validate their rotor modelling tools, and they passed the milestone. |
| URL | https://github.com/ImperialCollegeLondon/ARES-MATLAB |
| Title | HarmLAB: Harmonic Balance toolbox |
| Description | Harmonic balance is a widely used method for analysing non-linear vibrations in structures. Although the method is quite mature, there are few open-source implentions available. I therefore wrote a toolbox for the MATLAB environment, which included the more advanced version of the method. This is currently stored on GitHub, and I hope to eventually make it publicly available. This has been written in a general way so can be applied to a variety of problems in other fields such as electronic engineering. |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2018 |
| Provided To Others? | No |
| Impact | I used this toolbox to run the simulations for my journal paper. It has been used the predict the response of a digital twin model of the test rig. |
| URL | https://github.com/alxhslm/HarmLAB |
| Title | ModeLAB: Modal Analysis MATLAB toolbox |
| Description | We routine carry out hammer tests in the lab to determine the vibration properties of different components or assemblies. Currently there is no formalised way for processing and visualising this data. I have formalised my scripts into a single generic toolbox which I have applied used to analyse many different components. This will be shared with the rest of the group once documented, saving duplication of effort. |
| Type Of Material | Data analysis technique |
| Year Produced | 2018 |
| Provided To Others? | No |
| Impact | I have used this tool to analyse the test rig at Imperial, and the results will go into my thesis, and eventually an experimental journal paper. |
| URL | https://github.com/alxhslm/ModeLAB |
| Title | ROLADYN: Rotor modelling toolbox |
| Description | In order to model the test rig, I created a general rotor modelling toolbox. It can represent any configuration of discs/shafts/bearings and will be a useful tool for future Rotordynamics PhDs in the group. |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2017 |
| Provided To Others? | No |
| Impact | This model is currently being compared against the experimental data to predict the response of the rig. This will eventually be presented in another journal paper. An early version of this toolbox was used to generate data for my first conference paper. |
| URL | https://github.com/alxhslm/ROLADYN |
| Description | Rolls Royce |
| Organisation | Rolls Royce Group Plc |
| Country | United Kingdom |
| Sector | Private |
| PI Contribution | A test rig has been designed and manufactured for this project. In addition to being used for the actual research, it is being used to provide validation data of Rolls Royce software. Some other similar small-scale numerical studies have been carried out for Rolls Royce, looking at some other phenomena relating to bearings. |
| Collaborator Contribution | Rolls Royce have provided financial backing to build the test rig for this project. They have also played a key role in steering the research focus to aid their designs, and provided some technical feedback on the research, such as reviewing conference and journal papers. |
| Impact | Over the 25 year history of the VUTC, many papers have been produced. However, in terms of the specific output from this project, a conference paper was written on fundamental rotordyamics, which received some input from Rolls Royce. They have also supported much of the design of the test rig. |
| Start Year | 2016 |