Correlation Analysis of Multiple Sensors for Industrial Gas Turbine Compressor Blade Health Monitoring

Author(s):  
Brian Kestner ◽  
Tim Lieuwen ◽  
Chris Hill ◽  
Leonard Angello ◽  
Josh Barron ◽  
...  

This paper summarizes an analysis of data obtained from an instrumented compressor of an operational, heavy duty industrial gas turbine; the goal of the aforementioned analysis is to understand some of the fundamental drivers, which may lead to compressor blade vibration. Methodologies are needed to (1) understand the fundamental drivers of compressor blade vibration, (2) quantify the severity of “events,” which accelerate the likelihood of failure and reduce the remaining life of the blade, and (3) proactively detect when these issues are occurring so that the operator can take corrective action. The motivation for this analysis lies in understanding the correlations between different sensors, which may be used to measure the fundamental drivers and blade vibrations. In this study, a variety of dynamic data was acquired from an operating engine, including acoustic pressure, bearing vibration, tip timing, and traditional gas path measurements. The acoustic pressure sensors were installed on the first four compressor stages, while the tip timing was installed on the first stage only. These data show the presence of rotating stall instabilities in the front stages of the compressor, occurring during every startup and shutdown, and manifesting itself as increased amplitude oscillations in the dynamic pressure measurements, which are manifested in blade and bearing vibrations. The data that lead to these observations were acquired during several startup and shutdown events, and clearly show that the amplitude of these instabilities and the rpm at which they occur can vary substantially.

Author(s):  
Brian Kestner ◽  
Chris Hill ◽  
Leonard Angello ◽  
Josh Barron ◽  
Tim Lieuwen

This paper summarizes an analysis of data obtained from an instrumented compressor of an operational, heavy duty industrial gas turbine; the goal of the aforementioned analysis is to understand some of the fundamental drivers which may lead to compressor blade vibration. Methodologies are needed to 1.) understand the fundamental drivers of compressor blade vibration, 2.) quantify the severity of “events” which accelerate the likelihood of failure and reduce the remaining life of the blade, and 3.) proactively detect when these issues are occurring so that the operator can take corrective action. The motivation for this analysis lies in understanding the correlations between different sensors which may be used to measure the fundamental drivers and blade vibrations. In this study, a variety of dynamic data was acquired from an operating engine, including acoustic pressure, bearing vibration, tip timing, and traditional gas path measurements. The acoustic pressure sensors were installed on the first four compressor stages, while the tip timing was installed on the first stage only. These data show the presence of rotating stall instabilities in the front stages of the compressor, occurring during every startup and shutdown, and manifesting itself as increased amplitude oscillations in the dynamic pressure measurements which are manifested in blade and bearing vibrations. The data that lead to these observations were acquired during several startup and shutdown events, and clearly show that the amplitude of these instabilities and the rpm at which they occur can vary substantially.


2004 ◽  
Vol 10 (6) ◽  
pp. 433-442 ◽  
Author(s):  
Giovanni Ferrara ◽  
Lorenzo Ferrari ◽  
Leonardo Baldassarre

The rotating stall is a key problem for achieving a good working range of a centrifugal compressor and a detailed understanding of the phenomenon is very important to anticipate and avoid it. Many experimental tests have been planned by the authors to investigate the influence on stall behavior of different geometrical configurations. A stage with a backward channel upstream, a 2-D impeller with a vaneless diffuser and a constant cross-section volute downstream, constitute the basic configuration. Several diffuser types with different widths, pinch shapes, and diffusion ratios were tested. The stage was instrumented with many fast response dynamic pressure sensors so as to characterize inception and evolution of the rotating stall. This kind of analysis was carried out both in time and in frequency domains. The methodology used and the results on phenomenon evolution will be presented and discussed in this article.


Author(s):  
D. Jin ◽  
U. Haupt ◽  
H. Hasemann ◽  
M. Rautenberg

Centrifugal compressor surge at high rotational speed and reduced blade thickness can produce dangerous excitation effects which have numerous resulted in problems in the past. This paper presents an investigation of blade excitation during surge in a high performance single stage centrifugal compressor with various impeller and diffuser geometry. The blade vibration was measured using blade mounted strain gages. The flow characteristics during surge as the cause of blade excitation were simultaneously determined by fast response dynamic pressure transducers. The experiments have been performed using a radial and a backswept impeller, as well as a vanless and vaned diffusers. The rotational speed of the compressor was varied from 5,000 to 14,500 rpm. The characteristics of unsteady flow during surge, such as, the flow pattern of rotating stall and the non-periodic pressure fluctuation during surge were studied in detail. The experimental results demonstrated that, in addition to the excitation of rotating stall during surge, strong non-periodic pressure fluctuations at the beginning and the end of the surge induced dangerous blade excitations in all compressor configurations. The maximum strain values of blade vibration for all compressor versions at different rotational speeds of the compressor were measured to estimate the danger of blade excitation during surge. The results showed that the blade excitation during compressor surge with vaned diffusers is stronger than the excitation with a vanless diffuser and that the blade excitation with a radial impeller is stronger than the excitation with a backswept impeller.


Author(s):  
Hyunsu Kang ◽  
Sungjong Ahn ◽  
Kyusic Hwang ◽  
Justin Bock ◽  
Jeongseek Kang ◽  
...  

Abstract This paper describes the flow and vibrations measured in a 1.5-stage transonic research compressor tested at the Notre Dame Turbomachinery Laboratory. The compressor is a sub-scale version of a large-scale industrial gas turbine. The experiment measured the compressor performance and investigated the operability issues of stall and flow-induced blade vibrations due to buffet and flutter. The buffet was investigated at full-speed with fully-closed inlet guide vanes; the full-speed, no-load condition of gas turbines used for power generation. The flutter was investigated at part-speed conditions with partially closed guide vanes; the part-power condition where stall flutter typically occurs for aero-engines. At both of these conditions the blades operate with high incidence and moderate velocity, which can result in flow-induced vibrations. Aero-elastic simulations were performed to predict the flutter boundary. The flutter analysis predicted positive aerodynamic damping near the operating line, and a decrease in aerodynamic damping as the stall boundary was approached. No flutter was observed in the stable operating range of the compressor. The experimental campaign used blade tip timing to measure the vibrations and unsteady pressure transducers above the compressor blade. These two types of data were correlated to better understand the drivers of vibration. The paper describes the behavior of the aerodynamic drivers of buffet and flutter and the resulting vibration.


Author(s):  
Stephen Peluso ◽  
Bryan D. Quay ◽  
Jong Guen Lee ◽  
Domenic A. Santavicca

An experimental study was conducted to compare the relationship between self-excited and forced flame response in a variable-length lean premixed gas turbine (LPGT) research combustor with a single industrial injector. The variable-length combustor was used to determine the range of preferred instability frequencies for a given operating condition. Flame stability was classified based on combustor dynamic pressure measurements. Particle velocity perturbations in the injector barrel were calculated from additional dynamic pressure measurements using the two-microphone technique. Global CH* chemiluminescence emission was used as a marker for heat release. The flame’s response (i.e. normalized heat release fluctuation divided by normalized velocity fluctuation) was characterized during self-excited instabilities. The variable-length combustor was then used to tune the system to produce a stable flame at the same operating condition and velocity perturbations of varying magnitudes were generated using an upstream air-fuel mixture siren. Heat release perturbations were measured and the flame transfer function was calculated as a function of inlet velocity perturbation magnitude. For cases in this study, the gain and phase between velocity and heat release perturbations agreed for both self-excited and forced measurements in the linear and nonlinear flame response regimes, validating the use of forcing measurements to measure flame response to velocity perturbations. Analysis of the self-excited flame response indicates the saturation mechanism responsible for finite limit amplitude perturbations may result from nonlinear driving or damping processes in the combustor.


2015 ◽  
Vol 137 (5) ◽  
Author(s):  
J. Yang ◽  
G. Pavesi ◽  
S. Yuan ◽  
G. Cavazzini ◽  
G. Ardizzon

The unsteady phenomena of a low specific speed pump–turbine operating in pump mode were characterized by dynamic pressure measurements and high-speed flow visualization of injected air bubbles. Analyses were carried out on the pressure signals both in frequency and time–frequency domains and by bispectral protocol. The results obtained by high-speed camera were used to reveal the flow pattern in the diffuser and return vanes channels The unsteady structure identified in the return vane channel appeared both at full and part load condition. Furthermore, a rotating stall structure was found and characterized in the diffuser when the pump operated at part load. The characteristics of these two unsteady structures are described in the paper.


Author(s):  
Fabrice Giuliani ◽  
Hans Reiss ◽  
Markus Stuetz ◽  
Vanessa Moosbrugger ◽  
Alexander Silbergasser

The new energy mix places greater demands on power gas turbine operation; precision combustion monitoring, therefore has become a major issue. Unforeseen events such as combustion instabilities can occur and represent a danger to the integrity of the hot parts and also lead to a limitation of the output power. This is usually accompanied by an increase in maintenance costs. The enlarged off-design operating envelope of gas turbines to adapt to a fast-changing grid has made this issue even more acute, necessitating an expansion of the operating envelope into areas that were — for many engines — not foreseen in the original combustor design process. A good understanding of what happens within the gas turbine combustor is crucial. Complex and costly full-field measurements such as laboratory optical instrumentation in precision combustion diagnostics are not suitable for permanent fleet deployment. For practical and financial reasons, the monitoring should ideally be achieved with a limited amount of discrete sensors. If installed and interpreted correctly, fast response measurement chains could lead to a better gas turbine combustion management, possibly yielding considerable savings in terms of operating and maintenance costs. The firm Meggitt Sensing System (MSS), assisted by Combustion Bay One (CBOne), initiated an applied research programme dedicated to this topic — with MSS providing the instrumentation and CBOne providing the facility and test conditions. The objective was to investigate realistic combustion phenomena in a precisely controlled and reproducible way and to document the individual readings of the heat-resistant fast pressure transducers mounted on the combustor casing, as well as the accelerometers mounted on the outer surface of the machine. Particular attention was paid to the correlation between these two types of sensor readings. This paper reports on the monitoring of the flame using piezoelectric dynamic pressure sensors and accelerometers in a number of different situations that are relevant to the safe and efficient operation of gas turbines. Discussed are single events such as flame ignition, lean blow-out and flash-back, as well as longer test sequences observing the effect of warming-up or the presence of flame instability. The measurement chains and processing techniques are discussed in detail. The atmospheric test rig used for this purpose and the different testing configurations required for each of these situations are also illustrated in detail. The results and recommendations for their implementation in an industrial context conclude this paper.


Author(s):  
Alexander Wiedermann ◽  
Dirk Frank ◽  
Ulrich Orth ◽  
Markus Beukenberg

Test rig results and their comparison with computational analyses of a highly-loaded 11-stage compressor for a newly developed industrial gas turbine will be presented in this paper. The scope of the tests has been validation of aerodynamic and mechanical features of the bladed flow path to meet both the demands for single- and dual-shaft operation of the gas turbine. The test was carried out in three phases using extensive instrumentation. In phase 1 the front stages have been tested, and in phase 2 the test of the full 11-stage compressor was performed including numerous aerodynamic and structural check-outs. Vane and blade vibration modes were measured in all rows with numerous strain gauges using a telemetry system and Tip Timing, which additionally was applied to the front stage rotors. Concerning the mechanical design, finite element predictions of the vibration modes of all blades and vanes were carried out in the design phase to guarantee safe and resonance-free operation for a wide range of operational speeds which could be verified by the test data up to higher modes. Flow field computations were carried out with both a through flow solver and full 3-D viscous multistage solver based on Denton’s TBLOCK, where all rotor and stator flow fields had been solved simultaneously and compared with experiments. The effects of tip clearance and stator cavities on compressor performance have been taken into account by the computational analysis. Effects of inlet distortion were examined in phase 3. Comprehensive comparisons of computed and measured results will be presented. The extensive instrumentation gave also insight into flow details as vane pressure distributions and total pressure profiles in span wise direction. It will be shown that the agreements of predicted and measured data were excellent.


Author(s):  
A. P. R. Harpin

We describe our range of high temperature (1100°C) pressure sensors capable of measuring both static pressures of several Bar as required by gas turbine and jet engines, and measuring dynamic pressure fluctuations with a total dynamic range of in excess of 100000. This is achieved by a combination of rugged sensor design and our proprietary optical interrogator. This allows operation in harsh environments, EMI immunity, and simultaneous interrogation of not only static and dynamic pressure, but also the temperature of the sensor. This allows the sensor to maintain high accuracy over a wide range of operating temperatures. To date sensors have not been able to offer operation temperatures this high whilst enabling accurate dynamic pressure readings at the locations required. Also the static pressure cannot be retrieved simultaneously in real time from the same sensor. Also the temperature coefficient of the sensor has to be taken into account by measuring the temperature the sensor is operating at. Oxsensis has addressed these issues and we will present results showing dynamic pressure and temperature and explain how we can measure the temperature of the sensor with our interrogation schemes. We will describe the form of the sensor and the test data confirming its suitability for harsh environments. We will also explain the optical interrogator performance and present simulated results. The interrogator may be realised by a slave cavity or preferably on an integrated optical platform. As these sensors are intended for hostile gas turbine and aerospace environments, we will also present data from real life engine trials that we have performed, and compare the data we obtained with existing measurement techniques. Tests on a combustor rig have tested the sensor up to 1000°C, demonstrating that using our sensors in an engine at these temperatures is a realistic prospect. We believe that the ruggedness and performance of these sensors together with our complimentary interrogators mean that they are of significant interest to instrumentation of gas turbine engines and in the future the development of sophisticated engine feedback and emission control schemes, both in land based and aerospace environments.


Pressure sensing and measurement are of utmost importance in many of the process industries and biomedical applications. The key element of the pressure sensor is diaphragm and the diaphragm design including shape and dimensions play a major role in sensitivity of pressure sensor irrespective of the type of sensor viz. capacitive, piezoresistive or piezoelectric sensor. The acoustic pressure sensors require the proper analysis of dynamic performance of the key element since the acoustic source is dynamic pressure. This paper presents the stationary and dynamic performance analysis of diaphragm for piezoelectric acoustic pressure sensor. The analysis has been done for better deflection of the diaphragm and optimized stress and strain in order to achieve maximum sensitivity. In design step, at first the diaphragm is analysed for natural frequency, modal frequencies and bandwidth of the structure since the piezoelectric resonant sensors can be used for sensing whenresonant frequency of the membrane is at least 3 to 5 times the highest applied frequency and for energy harvesting applications, when it is almost equal to the applied frequency.Hence, a comparison of shapes of diaphragm, with their fundamental and modal frequencies, deflection, and stress and strain is established. Further a resonant sensor structure is also analyzed for dynamic performance with cavity neck of different size to understand the importance of cavity neck in dynamic performance of the sensor. The circular diaphragm is found be the best choice from the point of view of maximum deflection and natural frequency and the structure with cavity neck has better bandwidth and deflection.


Sign in / Sign up

Export Citation Format

Share Document