[ANSS-netops] tilt testing

[Greg Steiner]

Sorry, I was out of town. We use the tilt test in the following fashion. 
First we set up a dead level plate. Then we level the sensor relative to 
the plate and null the output for 0.000VDC. I then place a ground right 
angle plate on the level plate and tilt the sensor by 90 degrees (rotate 
sensor slightly for maximum output)and adjust for the specified 
transduction (typically 5v/G). I will usually tilt the sensor 90 degrees 
in the opposite direction as well. For a horizontal sensor this should 
produce an equal and opposite response, If it doesn't, it indicates a 
physical tilt of the instrument was present when the instrument was 
initially leveled, and that a thin shim may be required under one side 
of the instrument, and the whole process repeated. Note that on a 
horizontal, when the instrument is tilted 90 degrees by mounting against 
the right angle plate, that it will be necessary to rotate the sensor 
slightly in order to maximize the response. When we want to do a more 
detailed linearity test, the sensor is attached to a variable angle sine 
plate and the plate is rotated in 5 degree increments. the measured 
value is compared to the calculated value and the linearity is 
determined from those figures. Most force balance instruments do not 
require this degree of testing on a routine basis, however if there are 
magnetic whiskers in the torqueing coil magnet gap or other obstructions 
this test will generally show them up better than just a 90 degree tilt 
test. obstructions are also an obvious problem when an accelerometer 
does not return to its initial zero value after a tilt test.
accelerometers may also be calibrated over a wide range of frequencies 
using a shake table.  It is also theoretically possible to introduce an 
impulse into the mechanical system record the response of the sensor and 
deconvolve the entire response of the instrument. this works well in 
quiet sites, however the results are usually severly contaminated by 
doing this test in a building where mechanical equipment in the building 
creates a large and typically not very white background noise. hope this 
helps.
Greg
-- 

Greg Steiner

 

VLF Designs

1621 Bella Vista Dr.

Jackson, Mo. 63755

 

Ph: 573-204-1286

Fax:573-204-1286

Ph: 573-388-2117

 

Email: vlf at cablerocket.com

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://geohazards.usgs.gov/pipermail/anss-netops/attachments/20090625/32d599a0/attachment.html