position of accelerometers 2

[kashyap]

An accelerometer is fit on the bearing housing at an angle of 45 deg to the horizontal plane, and measures an acceleration of 0.3g(rms). How can i calculate the values in horiz. and vertical planes? Further, is there any advantage in mounting the accelerometer in an inclined plane (45deg)?

 

[ctolbert]

Kashyap:

The following website has some information on mounting the accelerometer. I never used the 45* angle method. I machined off the vertical and horizontal plane on the housing and mounted the contacts flush with the surface.

I used these on 1200HP+ dc mill motors though....which gave me good repeatable results. It however, may not be right for you.

http://www.ctconline.com/

Carl

 

[ctolbert]

One more time:

http://www.ctconline.com/

Carl

 

[kverspoor]

kashyap:
There's no trigonometric way to divide one reading at 45° into horizontal and vertical components. I don't believe that there is any advantage to using only sensing in the 45° direction. More commonly, of course, people use vertical and horizontal as "standard."

regards.. kv

 

[machinedoc]

I prefer two reading so that one is in load zone
machinedoctor

 

[PdMDoc]

Accelerometer mounting in this angle is cheap! This type of monitoring is requires the actual amplitude to be multiplied by the sin and cos of the 45 degrees which is 0.707. You will have to take the amplitude reading and multiply it times this value to determine the actual Horizontal (X) and Vertical (Y) values. I would recommend using an accelerometer in each horizontal or vertical plane to determine the actual vibration. This application will allow you to determine if the horizontal or vertical position is not performing. The vertical plane is usually the most rigid for horizontal mounted pumps or similar equipment.

A major turbine manufacturer only has a probe located at the 45 degree plane. This way there are fewer instruments to calibrate and fewer costs. Again I would recommend using the manufacturers installed instrumentation (i.e. least cost) augmented with your own predictive maintenance program.

 

[llangford]

I agree with the information in the above posts. However, I have had a few instances of machines that were belt- or gear-driven at angles close to 45°. On these machines, checking the vibration at the centerline of the load has helped find worn belts, pulleys and gears because the major vibration would show up best in this load direction.
I prefer horizontal and vertical for most applications. And axial is added for troubleshooting.

 

[ronfrend]

Just a quick point to note here is that if you have (in the theoretical world) a true harmonic vibration then you will see the same vibration level no matter which radial direction you are taking the measurement.

However, we know that differing causes of vibration give a resultant that is often direction specific. As almost everyone here has stated, it is best to take the readings in 2 radial directions but much more important is to be repeatable - make sure you always take the readings at the same place. Personally, I always try and take one reading in the load zone (thanks Ilangford) and at least one other as close to the perpendicular to the first reading as possible.

The practice of taking readings at 45' appears to have started with the gas turbine and centrifugal compressor community. This was done so that true orbits could be taken of the shaft using the built in displacement probes without having to worry about the gravitational effect on the shaft. This works excellently for checking alignments with vibration as well as looking for shaft whirl, rub or even a rotating stall on a centrifugal compressor element.

My point, Kashnap, is don't worry too much about convention when you are taking readings with a portable vibration analyzer but use your engineering experience and good old common sense.