[ANSS-netops] Fw: Strong motion calibrations

[Bob Hutt]

Everyone,

I agree with the statement from Kinemetrics tech support that the 
step voltage does not provide very useful information on absolute 
sensitivity.  I look at it as diagnostic (verification of 
state-of-health).  However, assuming that the ratio of calibration 
input voltage to accelerometer output remains stable over time, I can 
imagine using something like the following procedure:

1.  Determine absolute sensitivity through the use of an accurate 
tilt calibration or accurate flip test.  This is best done in a 
laboratory setting for controlled conditions because an accurate tilt 
jig or cube with precise right angles must be used.

2.  At the same time, input a known accurate voltage step into the 
calibrator input and observe the output step amplitude.

3.  After the accelerometer is installed in the field, the same 
calibration input voltage step should produce the same output that 
was observed in Step 2.  That is, the ratio of output voltage to 
input voltage should remain the same over time.  If not, then 
absolute sensitivity must be determined again (repeat step 1).

If the ratio of output to cal input continues to change, there is 
probably a problem with the accelerometer.  Note that there is some 
bookkeeping required to keep track of this ratio for each individual 
accelerometer (what it should be and what it is for each step cal 
test performed).

Note that an accurate flip test with a properly designed cube in 
which to mount the accelerometer can also probably be done in the 
field, but this requires a level surface on which to put the cube (in 
which the accelerometer is mounted).  This level surface could be a 
flat plate of aluminum with three leveling feet, which would be set 
up and levelled using an accurate bubble level (carpenter's level is 
probably ok).  But this is additional stuff to carry.

         Bob

At 06:01 PM 6/17/2009, Lind S Gee wrote:



>Interesting message on the ANSS network operations mailing list.
>
>
>----- Forwarded by Lind S Gee/GD/USGS/DOI on 06/17/2009 06:01 PM -----
> 
>
>   From:       Nathan L Edwards 
> <nedwards at unr.edu> 
>
> 
>
>   To:         "anss-netops at geohazards.usgs.gov" 
> <anss-netops at geohazards.usgs.gov>
> 
>
>   Date:       06/17/2009 05:57 
> PM 
>
> 
>
>   Subject:    [ANSS-netops] Strong motion 
> calibrations
> 
>
>   Sent 
> by:    anss-netops-bounces at geohazards.usgs.gov 
>
> 
>
>
>
>
>
>
>Hey all,
>
>We are attempting to establish some procedures for calibrating strong
>motion instruments.  There seems to be varying opinions as to what can be
>claimed with these results.  With our EpiSensor stations, we initially
>planned on providing a step voltage from the calibrator and recording
>output to confirm amplitude response.  There is a similar procedure for the
>K2 with an internal EpiSensor that is referred to as a "functional
>test" (not claimed to be a calibration).  These are reproducible, however
>we are rethinking what these results really mean after these comments from
>Kinemetrics tech support:
>
>"Calibration functions [of the datalogger] are provided for transfer
>function frequency
>response measurement, and cannot be used in any electronic force
>feedback sensor [like the EpiSensor] for meaningful gain measurements. This
>is an aspect of
>all electronic force feedback sensors, not an aspect of any calibration
>source."
>
>Regarding strong motion sensors (not just the EpiSensor) my questions are:
>1.  Does the step function test reliably tell you anything useful other
>than that there's something alive at the end of the sensor cable (i.e. to
>what degree is it a true calibration)?
>2.  What are other labs doing to remotely confirm sensor
>function/calibration on their strong motion sensors?
>
>Thanks,
>Nathan Edwards
>Development Technician
>Nevada Seismological Laboratory
>University of Nevada Reno
>Reno, NV 89557-0174
>o:  775-784-1106
>m:  775-303-5899
>f:  775-784-4165
>_______________________________________________
>ANSS-netops mailing list
>ANSS-netops at geohazards.usgs.gov
>https://geohazards.usgs.gov/mailman/listinfo/anss-netops