Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #1 
I've been specifying quite a few accelerometers recently, and it seems that most of the quality related specifications (linearity, sensor noise level, base strain sensitivity) are all problems of minimal consideration nowadays. However, it seems that the transverse sensitivity is the specification that can still really screw the measurement up.

At a typical 3-10%, this is already higher than the other error contributions, but then I saw this graphic from B&K. At low frequencies, the -35dB indeed comes out to less than 2% transverse sensitivity, but of note is the transverse resonant frequency. This is at about 1/3 of the mounted resonant frequency and coincides about with the 10% upper frequency range boundary. At these frequencies, we essentially have a 50% transverse sensitivity, which is enormous. If we assume the axial and transverse vibe are about the same magnitude, the linear zone of the sensor should be significantly lower.

One could argue that by the time we get to these frequencies, we care less about the direction and absolute magnitude of vibration (preferring instead a trend approach), but it still seems quite untruthful to extend the linear range up to past the transverse resonance frequency for applications where significant transverse vibration is expected (i.e. on rotating equipment). Thoughts?


transverse_sensitivity_response.png 
 

JuddJones

Sr. Member
Registered:
Posts: 384
Reply with quote  #2 
This is not a term I have run across in any of training. What is transverse sensitivity? 
John from PA

Sr. Member
Registered:
Posts: 963
Reply with quote  #3 
Quote:
Originally Posted by Big J
This is not a term I have run across in any of training. What is transverse sensitivity? 


Picture an accelerometer mounted vertically; its axis passes 12 o'clock to 6 o'clock.  Transverse sensitivity would be sensitivity perpendicular to that axis.

See https://endevco.com/ask-the-experts/what-does-the-transverse-sensitivity-specification-mean-in-an-accelerometer-datasheet/ or https://endevco.com/news/newsletters/2012_08/f_ate.html for more details.
Walt Strong

Sr. Member
Registered:
Posts: 889
Reply with quote  #4 
I do not think you can generalize that transverse vibrations would produce that large of an amplitude without measuring the true mounting surface tangential amplitude. We probably can agree that the specification for a triaxial accelerometer has a lower F-max for the X & Y directions as compared to the Z-direction that is perpendicular to the surface. This would indicate the natural frequency of each accelerometer is dominated by the mounting and not internal construction which is the same.

Walt
John from PA

Sr. Member
Registered:
Posts: 963
Reply with quote  #5 
Quote:
Originally Posted by Curran919
 This is at about 1/3 of the mounted resonant frequency and coincides about with the 10% upper frequency range boundary.


I find this portion of your original post interesting because for many years B&K recommended you use an accelerometer no higher than 1/3 the mounted resonance for qualitative measurements and no higher than 1/5 for quantitative measurements.  I haven't used B&K accelerometers since about 1986 so that may have changed.  An example would be an accel with a 30 kHz mounted resonance; the recommendation would be to use it up to 10 kHz for a qualitative measurement and 6 kHz for a quantitative measurement. 

For the uninitiated, a qualitative measurement would be a measurement that might be trended.  In other words we seek changes in the level, as opposed to knowing the exact number with a high degree of accuracy, which would be quantitative measurement.


Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #6 

Quote:
Originally Posted by Walt Strong
I do not think you can generalize that transverse vibrations would produce that large of an amplitude without measuring the true mounting surface tangential amplitude. We probably can agree that the specification for a triaxial accelerometer has a lower F-max for the X & Y directions as compared to the Z-direction that is perpendicular to the surface. This would indicate the natural frequency of each accelerometer is dominated by the mounting and not internal construction which is the same.


Walt, this is a different problem. Actually, for years I thought that transverse sensitivity was cross-talk in a triaxial sensor. Some companies calling it cross-axis sensitivity sure doesn't help. I understand that the secondary ('transverse') axis transducers of a triax have a lower mounted natural frequency than the primary Z axis.

But this is talking about each transducer individually. Even if you had a uniaxial accel, its going to have sensitivity to vibration at 90 degrees to its axis, and the 'reed' frequency of the accel is going to be lower than the axial eigenfrequency. I believe that is what causes the mounted resonant frequency (Z) to be higher than the transverse [mounted] resonant frequency (X, Y). If you had a rigid mounting and rigid casing, I suppose the transverse resonant frequency could be lower or higher depending on the sensing geometry.

Quote:
I find this portion of your original post interesting because for many years B&K recommended you use an accelerometer no higher than 1/3 the mounted resonance for qualitative measurements and no higher than 1/5 for quantitative measurements. 

I think this rule of thumb is still 'active'. The upper limit of the frequency range is typically about 20-25% of the 'typical ideal' (i.e. stud flush mounted) mounted resonant frequency. However, this transverse resonant frequency kinda puts that into doubt. Many of the datasheet assumptions assume uniaxial vibration. If you assume vibration with similar magnitudes triaxially, My argument is that you have to decrease the upper limit of your 10% or 3dB bandwidths.

JuddJones

Sr. Member
Registered:
Posts: 384
Reply with quote  #7 
Quote:
Originally Posted by John from PA


Picture an accelerometer mounted vertically; its axis passes 12 o'clock to 6 o'clock.  Transverse sensitivity would be sensitivity perpendicular to that axis.

See https://endevco.com/ask-the-experts/what-does-the-transverse-sensitivity-specification-mean-in-an-accelerometer-datasheet/ or https://endevco.com/news/newsletters/2012_08/f_ate.html for more details.


Thank you. I figured thats what it meant. Wanted to make sure.
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.