Powerful Interpolation Method Demonstrated on Experimental Data From Fast Response Pressure Transducers in a Turbofan Engine

Author(s):  
Felix Kern ◽  
Stefan Bindl ◽  
Marcel Stößel ◽  
Svenja Aberle ◽  
Reinhard Niehuis

Interpolation is a common procedure in scientific practice. In many cases, linear interpolation is tacitly applied in order to present measurements or numerical results. The application of sophisticated interpolation methods is rather rare, even though they can help to extract more information from a measurement, especially if the resolution is relatively low. This paper presents an interpolation method and a validation procedure, which is demonstrated on experimental data from a low pressure compressor of a twin spool turbofan engine. As a research test vehicle the Larzac 04 turbofan engine is used, which is equipped with extensive instrumentation and control systems. This publication focuses on the high speed pressure transducers mounted directly above the tip of the first LPC rotor blades. Besides the installation of those sensors in a jet engine, the processing, the storage, and the analysis of the data is challenging, as the sampling rate is extraordinary high (1 MHz). Those sampling rates result in a high circumferential resolution (1 MHz ≙ 2.69 samples per mm at 90% rel. spool speed) compared to the relatively low geometrical resolution of the nine axial sensors (0.23 samples per mm). Therefore, sophisticated interpolation methods are applied on the data points in axial direction, aiming to get the best possible result from the acquired data. A major part of this paper is the discussion and validation of several different interpolation methods. Especially with the Akima sub-spline interpolation, very promising results were obtained. Finally, some of the results from the fast response pressure transducers are presented. The casing wall pressure distribution and some flow phenomena at the design operating point are compared to the corresponding near stall conditions.

2002 ◽  
Vol 124 (3) ◽  
pp. 784-790 ◽  
Author(s):  
Jorge L. Parrondo-Gayo ◽  
Jose´ Gonza´lez-Pe´rez ◽  
Joaquı´n Ferna´ndez-Francos

An experimental investigation is presented which analyzes the unsteady pressure distribution existing in the volute of a conventional centrifugal pump with a nondimensional specific speed of 0.48, for flow-rates from 0% to 160% of the best-efficiency point. For that purpose, pressure signals were obtained at 36 different locations along the volute casing by means of fast-response pressure transducers. Particular attention was paid to the pressure fluctuations at the blade passage frequency, regarding both amplitude and phase delay relative to the motion of the blades. Also, the experimental data obtained was used to adjust the parameters of a simple acoustic model for the volute of the pump. The results clearly show the leading role played by the tongue in the impeller-volute interaction and the strong increase in the magnitude of dynamic forces and dipole-like sound generation in off-design conditions.


2005 ◽  
Vol 6 (1) ◽  
pp. 85-93 ◽  
Author(s):  
H Nakamura ◽  
I Asano ◽  
M Adachi ◽  
J Senda

The Pitot tube flowmetering technique has been used to measure pulsating flow from a vehicle engine exhaust. In general, flowmetering techniques that utilize differential pressure measurements based on Bernoulli's theory are likely to show erroneous readings when measuring an average flowrate of pulsating flow. The primary reason for this is the non-linear relationship between the differential pressure and the flowrate; i.e. the flowrate is proportional to the square root of the differential pressure. Therefore, an average of the differential pressure does not give an average of pulsating flow. In this study, fast response pressure transducers have been used to measure the pulsating pressure. Then the pulsating differential pressure is converted to the flowrate while keeping the pulsation unaveraged. An average flowrate is then calculated in the flowrate domain in order to maintain linearity before and after averaging. The peak amplitude of a pulsation measured here was about 1800 L/min at an average flowrate of 70 L/min when the engine ran at idle speed. This measurement has been confirmed by measuring the pulsation with a gas analyser. The results show a large amount of back and forth gas movement in the exhaust tube. This magnitude of pulsation can cause as much as five times higher erroneous results with the pressure domain averaging when compared to a flowrate domain averaging.


Author(s):  
O. Schennach ◽  
R. Pecnik ◽  
B. Paradiso ◽  
E. Go¨ttlich ◽  
A. Marn ◽  
...  

The current paper presents the results of numerical and experimental clocking investigations performed in a high-pressure transonic turbine with a downstream vane row. The objective was a detailed analysis of shock and wake interactions in such a 1.5 stage machine while clocking the vanes. Therefore a transient 3D-Navier Stokes calculation was done for two clocking positions and the three dimensional results are compared with Laser-Doppler-Velocimetry measurements at midspan. Additionally the second vane was equipped with fast response pressure transducers to record the instantaneous surface pressure for 20 different clocking positions at midspan.


2011 ◽  
Vol 50-51 ◽  
pp. 564-567
Author(s):  
Yun Feng Yang ◽  
Xiao Guang Wei ◽  
Zhi Xun Su

Image interpolation is used widely in the computer vision. Holding edge information is main problem in the image interpolation. By using bilinear and bicubic B-spline interpolation methods, a novel image interpolation approach was proposed in this paper. Firstly, inverse distance weighted average method was used to reduce image’s noise. Secondly, edge detection operator was used to extract image's edges information. It can help us to select different interpolation methods in the image interpolation process. Finally, we selected bilinear interpolation approach at non-edge regions, and bicubic B-spline interpolation method was used near edges regions. Further more, control vertexes were computed from pixels with calculation formula which has been simplified in the B-spline interpolation process. Experiments showed the interpolated image by the proposed method had good vision results for it could hold image's edge information effectively.


Author(s):  
Meera Day Towler ◽  
Tim Allison ◽  
Paul Krueger ◽  
Karl Wygant

This investigation studies fast-response pressure measurements as an indicator of the onset of surge in a single-stage centrifugal compressor. The objective is to determine an online monitoring approach for surge control that does not rely on surge margin relative to maps from predictions or factory testing. Fast-response pressure transducers are installed in the suction piping, inducer, diffuser, and discharge piping. A speed line is mapped, and high-speed pressure data are collected across the compressor map. The compressor is driven into surge several times to collect pressure data between during surge and between surge events. Following testing, these data are post-processed via filtration and statistical analyses. It is determined that, when taken together, the mean and range of the standard deviation of the time signal for multiple time steps can be used to determine whether the compressor’s operating point is approaching surge for the conditions tested.


Author(s):  
R. Amirante ◽  
L. A. Catalano ◽  
A. Dadone ◽  
V. Lombardo

The aim of this paper is to investigate the use of fast-response pressure transducers for measuring the instantaneous pressure in different sections of a common-rail diesel injection system, both for a single injection and for multiple injections. The influence of the pressure transducer onto the measured pressure is evaluated numerically by comparing the pressure history computed without the pressure transducer and that computed with the presence, and thus with the disturbance, of this sensor. A new electric circuit is proposed in substitution of the standard electronic central unit, which allows to modify the injection parameters and to perform injections on a test rig, as done in the automotive applications. Experimental results are provided both for a single injection and for multiple injections, to demonstrate the capabilities of the proposed test bench for the unijet injectors.


Author(s):  
R. F. White ◽  
D. F. D'Arcy

The effects of a rapid loss of coolant caused, for example, by the fracture of a pipe in the primary coolant circuit play an important part in the safety assessment of a water-cooled nuclear reactor. In particular, it is essential to know the rate of decrease of coolant density in the core and the rate of discharge of coolant into the containment space. It is uneconomic to obtain this information from experiments on each new reactor. Consequently computational methods, supported by experiments on basic elements of a coolant circuit, are being developed which will be applicable to a wide variety of reactors. An experiment is described in which fundamental information on fluid flow dynamics was obtained. In the experiment, a flow of two-phase water at high pressure was established in a straight pipe 15 ft long and 0–621 in i.d. At the downstream end of the pipe a rupture disc was suddenly burst. The pressure versus time histories at three locations along the pipe were recorded, using fast response pressure transducers and a multichannel oscilloscope. Experimental results are given for pre-blowdown pressures of 500 and 1000 lb/in2 (abs.) and steam qualities of 10–30 per cent. The mass flux was 0·6 times 106 lb/ft2 h, and the flow regime was annular. From the recorded pressure traces, the speed of the head of the rarefaction wave travelling upstream into the fluid was measured and also the pressure at critical flow conditions as functions of the pre-blowdown conditions. These measurements are compared with values calculated from steam tables.


1996 ◽  
Vol 118 (1) ◽  
pp. 41-51 ◽  
Author(s):  
M. Ubaldi ◽  
P. Zunino ◽  
G. Barigozzi ◽  
A. Cattanei

Detailed flow measurements were taken in a centrifugal turbomachine model to investigate the aerodynamic influence of the vaned diffuser on the impeller flow. The model consists of an unshrouded centrifugal impeller with backswept blades and a rotatable vaned diffuser, which enables a continuous variation of the vaned diffuser location with respect to the measuring points. Phase-locked ensemble-averaged velocity components have been measured with hot-wire probes at the impeller outlet for 30 different relative positions of the probe with respect to the diffuser vanes. The data also include the distribution of the ensemble-averaged static pressure at the impeller front end, taken by means of miniature fast response pressure transducers flush-mounted at the impeller stationary casing. By circumferentially averaging the results obtained for the different circumferential probe locations, the periodically perturbed impeller flow has been split into a relative steady flow and a stator-generated unsteadiness. The results for the different probe positions have also been correlated in time to obtain instantaneous flow field images in the relative frame, which provide information on the various aspects of the diffuser vane upstream influence on the relative flow leaving the impeller.


1997 ◽  
Vol 119 (3) ◽  
pp. 634-645 ◽  
Author(s):  
B. V. Marathe ◽  
B. Lakshminarayana ◽  
D. G. Maddock

The stator flow field of an automotive torque converter is highly unsteady due to potential and viscous interactions with upstream and downstream rotors. The objective of this investigation is to understand the influence of potential and viscous interactions of the upstream rotor on the stator surface pressure field with a view toward improvement of the stator design. Five miniature fast-response pressure transducers were embedded on the stator blade. The measurements were conducted at three locations near the leading edge and two locations near the trailing edge at the midspan location. The upstream flow field was measured using a fast response five-hole probe and is described in Part I of this paper. The experimental data were processed in the frequency domain by spectrum analysis and in the temporal-spatial domain by the ensemble-averaging technique. The flow properties were resolved into mean, periodic, aperiodic, and unresolved components. The unsteady amplitudes agreed well with the pressure envelope predicted by panel methods. The aperiodic component was found to be significant due to the rotor–rotor and rotor–stator interactions observed in multistage, multispool environment.


Author(s):  
B. V. Marathe ◽  
B. Lakshminarayana ◽  
Donald G. Maddock

The stator flow field of an automotive torque converter is highly unsteady due to potential and viscous interactions with upstream and downstream rotors. The objective of this investigation is to understand the influence of potential and viscous interactions of the upstream rotor on the stator surface pressure field with a view towards improvement of the stator design. Five miniature fast-response pressure transducers were embedded on the stator blade. The measurements were conducted at three locations near the leading edge and two locations near the trailing edge at the mid-span location. Upstream flow field was measured using a fast response five-hole probe and is described in the first part of this paper. The experimental data were processed in the frequency domain by spectrum analysis and in temporal-spatial domain by the ensemble averaging technique. The flow properties were resolved into mean, periodic, aperiodic and unresolved components. The unsteady amplitudes agreed well with the pressure envelope predicted by panel methods. Aperiodic component was found to be significant due to the rotor-rotor and rotor-stator interactions observed in multistage, multi-spool environment.


Sign in / Sign up

Export Citation Format

Share Document