Individual Project. Experimental Investigation of the Resonance Properties of a Turbomachine-Pipeline System

Transcription

1 Individual Project Title: Experimental Investigation of the Resonance Properties of a Turbomachine-Pipeline System Names: Ali Parhizkar (CJ43E0) Iason Verganelakis (DVWTJ1) Mostafa Al Nukta (EYRSNQ) Soleyman Banai Saghar (MUSN1O) Date: 03/06/2014 Instructor: Dr. Hegedűs Ferenc 0

2 Table of Contents Acknowledgement:... 2 Introduction:... 3 Measurement System:... 3 Results:... 6 Amplitude-frequency (2D) Diagrams:... 6 A-f-Q Waterfall Diagrams:... 7 Conclusion: Appendix:

3 Acknowledgement: We would like to thank Dr. Hegedűs for his constant advices and consultations which through them it was possible for us to carry out the project and achieve the required results. We would also like to thank Mr. Zsolt Szeitz for his constant care and presence at the laboratory whom without we couldn t carry out the project safely. 2

4 Introduction: The project was the experimental investigation of resonance properties of a turbomachine-pipeline system. The aim was to measure different measurement points applying different volume flow rates and rotational speed and using Fourier transform (FFT) to get the amplitudes and frequencies. The system consists of an electrical motor, a centrifugal pump, water tank, throttle valve and a pipeline system. A sensor is also place at the delivery side of the pump which is connected to a computer system with software named LabView using FFT. At each working point the software gives an A-f diagram (amplitude-frequency). MATLAB software was later on used to plot these diagrams for different working points and also to find the highest amplitude for each one of these working points and also plot them in a separate diagram which then gives the resonance frequency. Measurement System: The measuring system consists of an electrical motor, a centrifugal pump, a tank and a pipeline system. The following sketch show the whole measurement system. The water comes from a tank (reservoir) to the radial pump (P) which is connected to the motor (M). It enters the pump at the suction side (I) and leaves the pump at the pressure side (II) which then flows through a pipeline system to a metering orifice (MO) and back to the reservoir through a throttle valve (TV). The throttle valve is used to control the volume flow rate. A computer system is also connected through a sensor at the delivery side of the pump which then using FFT transforms to A-f diagrams. Figure 1: simple sketch of the measurement system 3

5 The following figures show the actual system in the lab: Centrifugal pump Electrical Motor Figure 2: the motor and the radial pump Manometer Reservoir Throttle valve (a) (b) Figure 3: (a): the manometers (b): the reservoir and the throttle valve 4

6 Rotational speed controller Computer connected to the system (a) (b) Figure 4: (a): controller (b): computer system connected to the system After turning on the system, a starting rotational speed (n) was set using the controller which then using the same speed different volume flow rates (Q) were set as well controlled by the throttle valve. For each working point (each Q and n) the computer gave a corresponding diagram. The rotational speed range was from 500 to 1500 rpm and ΔQ depending on each rotational speed can be 5 mmhg or 10 mmhg. For example with the rotational speed of n=1000 rpm the maximum volume flow rate was 123 mmhg so ΔQ was chosen as 10 mmhg. For each of these working points the computer software gave an amplitude-frequency diagram. Rotational speed was checked to be constant using Jaquet indicator (tachometer) at each volume flow rate (after each change of Q). Figure 5: Jaquet indicator and the connection to the motor for checking rev. speed 5

7 Results: Amplitude-frequency (2D) Diagrams: After measuring and setting all the working points, the data and diagrams were saved which then using MATLAB codes they could be plotted based on setting the required Q and n. At first the 2D diagrams were programmed to be plotted for each Q and n. For example the diagram for n=500 rpm and Q=20mmHg was the following: Figure 6: A-f diagram for n=500 rpm and Q=20 mmhg And the diagram for n=1500 rpm and Q=200 mmhg is the following: Figure 7: A-f diagram for n=1500 rpm and Q=200 mmhg 6

8 MATLAB codes used for plotting the 2D diagrams selecting different n and Q are as follows. Typing in the specific Q and n gives the corresponding diagram: Figure 8: MATLAB codes for plotting 2D diagrams A-f-Q Waterfall Diagrams: These codes were only able to show the amplitude-frequency diagrams for one working point which means for one Q value and n value. In order to compare the results, 3D diagrams (waterfall diagrams) can be shown. In the 3D diagrams the amplitude-frequency diagram can be shown with a given revolution number while the volume flow rate is changing. It can also be done when the volume flow rate (Q) is taken constant and n is changing. In this report after consideration of both case it was decided to n=const. for each case. As shown is the following page the example waterfall diagrams for a given revolution number are shown with 3 axes: Amplitude: A Frequency: f Volume flow rate: Q This shows the amplitude change with different volume flow rates in a constant rotational speed. The diagrams in the next page are some examples of these 3D diagrams. 7

9 Figure 9: waterfall diagram for n=500 rpm Figure 10: waterfall diagram for n=1000 rpm The waterfall diagrams are very important since it is possible to understand the behavior of the flow from these diagrams. To understand the waterfall diagrams it is important to understand the behavior and operation of the impeller in the pump. The impeller of the pump in the system was a 7-blade impeller. There are two type of excitation in the system, first is the excitation of eccentricity of the impeller which is basically the frequency (rotational speed n) which it is proportional to. The second one is the excitation due to the blades in the impeller. It is possible to understand the flow rate characteristics due to these two excitations from the waterfall diagrams. 8

10 The impeller of a pump is shown in the following figure: Excitation caused by the blades (7 blades) Figure 11: pump impeller In the waterfall diagram the behavior of the flow and amplitude from both excitations (caused by eccentricity of impeller and the blades) can be seen. In the following diagrams the first wave which its highest amplitudes are shown with dots are the effects of the eccentricity of the impeller and the seventh wave is the effect of the blades (7 blades). Figure 12: waterfall diagram for n=1000 rpm In this diagram it can be seen that at the seventh wave there is first increase of amplitude and then it starts to decrease. This shows that there is a peak in amplitude on the seventh wave. The reason for these changes in amplitudes are due to characteristics of the fluid flow which can t be known until the impeller can be opened and further investigations and simulations are done. 9

11 More waterfall diagrams are shown in the following pictures. The amplitude change at the first wave is also shown in the amplitude-peak diagrams: Figure 13: waterfall diagram for n=1100 rpm Figure 14: amplitude-peak diagram of the first wave for n=1100 rpm This shows that the amplitudes due to eccentricity of the impeller is increasing as the volume flow rate increases. 10

12 Figure 15: waterfall diagram for n=1300 rpm The amplitude for 7 th wave increases and then slightly decreases at the end. The amplitude peaks for the first wave is shown in the following diagram: Figure 16: amplitude-peak diagram of the first wave for n=1300 rpm It can be seen that there is a peak of amplitude around Q=180 mmhg which can be said to be resonance flow rate. These behaviors can be seen in each diagrams. 11

13 Figure 17: waterfall diagram for n=1500 rpm It is shown here that on the 7 th wave first there is an increase in amplitude and then decrease in some range which after it increases again. The range which the amplitude is decreasing is around mmhg. These behaviors are due to fluid flow characteristics in the impeller and further investigations and simulations are needed to discover the reasons. The peak amplitudes due to eccentricity are shown in the following diagram: Figure 18: amplitude-peak diagram of the first wave for n=1300 rpm As it can be seen there is a peak in amplitude around mmhg and this flow rate can be considered as the resonance flow rate. The MATLAB codes for waterfall diagrams are attached to the report for further references. 12

14 Conclusion: It was not possible to reach a certain conclusion on how and what the resonance frequency is but there were certain results on how the amplitude is changing due to the change in flow rate and there were some discussions about the resonance flow rate which is the flow rate where the highest amplitude was achieved. Although some characteristics behavior of the flow were unknown it is suggested to have further investigation and maybe fluid flow simulations in the impeller where the amplitude changes are happening. To sum up the characteristics and changes in the amplitude and frequency of the system was investigated due to change in revolution speed and the volume flow rate. The changes and diagrams were plotted and shown. Overall the measurement and the project was successful and the results were interesting as they showed the behavior of the fluid flow in the system. 13

15 Appendix: The MATLAB codes for 2D drawing of the amplitude frequency for each volume flow rate and revolution speed is: function DataLoad Q=200; n=1500; FileName = strcat( 'Q',num2str(Q),'_n',num2str(n),'.txt' ); IsExist = exist(filename,'file'); if IsExist==2 fid = fopen(filename); HeaderLines = textscan(fid, '%s', 8); Data = textscan(fid, '%f %f'); X=Data{1}; Y=Data{2}; fclose(fid); end correntplot=plot(x,y); set(gca,'yscale','log'); Grid on Box on xlabel('frequency','fontsize',16); Ylabel('Amplitude','FontSize',16); end 14

16 For the 3D diagrams (waterfall diagrams) the MATLAB codes are the following: function Multi2data n=1400; figure(1); hold 'on'; counter=0; for Q=1:500 FileName = strcat( 'Q',num2str(Q),'_n',num2str(n),'.txt' ); IsExist = exist(filename,'file'); if IsExist==2 counter=counter+1; fid = fopen(filename); HeaderLines = textscan(fid, '%s', 8); Data = textscan(fid, '%f %f'); X=Data{1}; Y=Data{2}; fclose(fid); % CurrentPlot=plot(X,Y); % set(currentplot,'displayname',filename); % set(gca,'yscale','log'); Z=ones(length(X),1)*Q; CurrentPlot=plot3(Z,X,Y); set(currentplot,'displayname',filename); set(gca,'zscale','log','ydir','reverse'); end [Amax(counter) Index]=max(Y); Fmax(counter)=X(Index); Qmax(counter)=Z(Index); grid on box on xlabel('flow Rate','fontsize',14); zlabel('amplitude','fontsize',14); ylabel('frequency','fontsize',14); end CurrentPlot=plot3(Qmax,Fmax,Amax); set(currentplot,'displayname',filename,'linestyle','none','marker','.','markersiz e',15); set(gca,'zscale','log','ydir','reverse'); hold off; figure(2) plot(qmax,amax) grid on box on xlabel('flow Rate','fontsize',14); ylabel('amplitude','fontsize',14); 15