Hello, the power spectrum of the turbulent boundary layer in our boundary layer wind tunnel has changed. There are many peaks in high frequency region, which did not appear before. In this region the vortex disspation should occur. I used the same method, CTA probes and low pass filtering and Fast-Fourier transform. Do you have any ideas what is causing the strange end of the spectrum? Thanks
Your spectra are very strange. You have two identical PSD distributions which seem to be shifted in frequency range and amplitude depending on which filter you apply. This looks to me as a problem of data treatment, not the wind tunnel.
Hello, you have two possibilities:
1 you have a problem of electric noise (for example supply harmonic, 50 Hz in France) on your sensors.
2 you have acoustic harmonics due to noise generated by your blower (low frequencies in the bandwidth of your sensors and alliasing for high frequencies).
Can you explain how you obtained your example of power spectrum ?
best regards
Fabrice De Oliveira
This example was measured with the MiniCTA device from Dantec Dynamics. The spectral density was calculated from velocity signal using Labview FFT. The blue spectrum was measured with low pass filter 1 kHz and sampling rate 5 kHz and the red spectrum was measured with l. p. filter 10 khz and rate 30 kHz. The device was properly grounded. It looks like the low pass filter is not working.
Hi Petr,
I just have a little doubt about the spectra you show, since they look to me to be identical to each other. Did I understand correctly that they come from 2 different measurements?
Yes, they come from different measurements with different settings. I agree, that they look the same, but I include more examples. The main problem is the strange end at high frequencies and why is this happening.
Thanks for the additional examples.
I am asking questions about the spectra because depending on which one you show, you have either peaks around 100-1000Hz, or peaks above 1000Hz. So firstly there is something wrong there, and I can't understand or answer your main problem unless it is clear which frequencies are true, and which ones are physically likely to be present in your flow. I am guessing that your wind tunnel installation is designed for low to moderate Reynolds numbers? We are not talking about compressible flows here, right? If so, then I cannot imagine what could produce such a concentration of fluctuations at such very high frequencies as >1kHz.
It could be possible that your hot wire support is vibrating, but still, very unlikely at such high frequencies. The best advice I could give is to check whether your equipment is properly functioning. Maybe try to measure something of a known high frequency and see what is the response of your sensor. Also try a measurement without the built-in low pass filter and perform the filtering afterwards. I can ask a colleague here who knows the Dantec mini-CTA very well.. I personally never used these new black-box devices.
If you still think this is physical, then think about what are the physical scales you are dealing with? How does the velocity and the boundary layer thickness scale with these frequencies?
But, as I said, it looks to me that this is a technical problem with your measurements, not an issue of the flow in your wind tunnel.
Hello Saskia, do you mean analog or digital filtering of the signal? Speaking about analog low pass filter, I have the possibility to use 300 Hz, 1 kHz, 3 kHz and 10 KHz built-in filter. I forgot to mention that I use the Streamline CTA anemometer, marked "Str" in the graph, "PSD" is from miniCTA.
Today I have found that spectra for heights ca. 400 mm and less above the tunnel floor look good, so it might be a property of the boundary layer. I will check it :)
Hello Petr, This problem occur because some of the data might have repeated at some interval. I have earlier faced the same problem but when i cross examine my data i found the fault. So i request you to kindly go through your data.
Regards
Shailesh Jha
I have found the problem, the CTA signal amplification was too low. I doubled the gain and now the spectra are OK.
- Badges
- Science topic
- Similar topics
- Mathematics
- Algebra
- Matrix