Cancel
Showing results for 
Search instead for 
Did you mean: 

Question about Derived Frequency Spectrums in Time Data Processing

Hello,

 

currently I am post-processing a lot of time data with the Time Data Processing Tab. In the "Channel Processing Settings" I am trying to calculate vector sums of existing channels in the DerivedFS Tab. I am faced with two problems:

 

- How do I adress channels if I don't want to use the channel name? I could of course use:

Vectorsum(Dir1:+X;Dir2:+Y;Dir3:+Z), but to be more flexible if I post process more runs with different channel names on the same channel numbers I would like to use something like: Vectorsum(CH12;CH13;CH14). I've tried that and I do get results, but the number I am using doesn't seem to coincide with the numbering in the Time Data Processing representation of my channel numbers.  So, how do I get to know which channel numbers I have to use to get a certain channel name?

 

- For the function type I can choose 'Autopower' and 'Spectrum'. How does LMS determines the phase angle for 'Spectrum'? At the Vibration Tab in the Channel Processing Settings there is a 'Phase referenced spectra' Option. How is that with the Spectrum in the DerivedFS-Tab?

 

Thanks for any help  on those issues.

3 REPLIES

Re: Question about Derived Frequency Spectrums in Time Data Processing

Hi, I just discovered that the first non tacho channel is CH01. So the channel number minus the amount of tacho channels is the number to use. That then would be the answer for my first question.

Re: Question about Derived Frequency Spectrums in Time Data Processing

Siemens Phenom Siemens Phenom
Siemens Phenom

Hello,

 

In the top right corner of the Derived FS tab is a pulldown that lists the Channel number to use in the formulas.

 

Since Vector sum squares the channels (multiples the channel by it's complex conjugate), sums them and then takes the square root it does not matter if you use Spectrum or Autopower in this case.   The vector sum will not have phase. 

 

The Spectrum phase would be the same phase as calculated if you specified specrum under the Acoustic, Vibration or Other Group.  We go to the tracking point, look for any trigger, apply a window and then calculate the spectrum.  I checked this by measuring three channels and then the following 3 derived channels.  The measured data and the derived channels are identical.

 

Derived_CH.jpg

 

If I then specify channel two as the phase reference spectra under my Other group and repeat the calculation, the results are still identical and channel two has zero phase as expected for the Other data and the derived channel.  So, the derived channel does use the phase reference channel as well.

 

plot.jpg

Re: Question about Derived Frequency Spectrums in Time Data Processing

Siemens Legend Siemens Legend
Siemens Legend

Two articles that may help: