An important component within the jet engine in terms of vibration and high cycle fatigue (HCF) is the blade. This is the component where continuously higher demands on weight and loading are being made. As a consequence of this, there has been a growing interest in developing both numerical methods and instrument technology for blade HCF measurement. This growing interest has also been attributed to changing attitude within the military and aerospace industry, which has tended towards driving down costs and lengthening the engine's life span. Many development technologies have been reported. One of which, is the development of a non-intrusive system for measuring blade vibratory stress. Research in non-intrusive techniques for the measurement of blade vibration has been ongoing since the early 1970' s. The aim of which, has been to replace the conventional method, using strain gauges and slip rings, with an improved system based upon non-intrusive type instrumentation such as optical or capacitance probes. One such approach is known as tip-timing. Tip-timing is a technique used to measure blade vibration using non-contact probes located around the engine casing. Many tip-timing techniques have been developed over the years, but there still remain significant problems associated with the approach. Such problems include sensitivity to noise and the high number of probes required. The development of two tip-timing methods known as the Autoregressive (AR) method and the Two Parameter Plot (2PP) method has recently been published in the open literature. This thesis describes the work done to experimentally test these two techniques. During the course of this work, an experimental optical tip-timing test facility was built. This included purpose-built optical tip-timing instrumentation, a tip-timing data acquisition system, and a post processing system incorporated into the Cranfield University low speed compressor facility. Experimental testing of the Autoregressive method and the Two Parameter Plot method was carried out using a controlled test environment, representative of a real engine. An analysis of the two methods was conducted using data from a comprehensive range of frequencies and RPM speeds. The results were then compared with previously published numerical results and the two algorithms were evaluated in terms of replacing the conventional strain gauge method. Testing of the AR method presented some interesting findings, with acceptable results produced at low rotational RPM speeds. However, as the rotational speed was increased, the accuracy of the results deteriorated. This type of result had not be highlighted in previous work. The 2PP method performed relatively well when using data sampled from the smaller 16 Engine Order (EO) response. However, this was not repeated when using the larger 72EO data. Additionally, this type of result had not been shown in previously published work. Overall, it was concluded that the issues associated with the frequency measurements should be remedied and a technique for measuring Multiple-Degree-of-Freedom responses should be explored before tip-timing techniques can be considered as a replacement to the strain gauge approach.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:422369 |
| Date | January 2004 |
| Creators | Grant, Kelly |
| Contributors | Ivey, Paul C. |
| Publisher | Cranfield University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://dspace.lib.cranfield.ac.uk/handle/1826/4623 |
Page generated in 0.0021 seconds