How to make a "stick probe"?

[apass]

Could someone share with me how to make a contacting vibration probe to directly measure shaft displacement on a pump? It would be used to double check readings that I am getting with non-contacting probes which I think are erroneous.

 

[byrdj]

I think IRD makes a "fish tale" for doing such. I have made my own for specific applications. For example, for 6" dia shaft 6000 rpms, I used a 3/16" X 3" X 24" oak slat. One end was cut a shallow notch to ride, with two point contact on the shaft. the notch was soaked with ligh oil and "burnt in" for a few minutes. If in a hurry, the siesmic detector can be held on the end of the slat. For more fancy, a block 1 X 1 X 2, could be grooved for the slat to be glued in an a mountin studs used.

when there is a guide to help support, a long wooden dowel wood can be used (1/2" dia 24")

I think IRD provides a calibration sheet with thiers that shows the small phase lag and attenuation, but I've never used with balancing data gathering

 

[sms]

Obviously don't know where you work, but most safety departments don't look favorably on fish tails... What makes you think the probes are not working right? If it is a Bently system and the green lights are lit on the face of the monitor, you can be pretty confident that the probes are working.

Is this a boiler feed pump? If so it is not unusual for repair shops to chrome the journals on BFP's. The result is that the probe reads the surface prep through the chrome, which result in a high vibration reading that is not real...

 

[TPL]

I'm with SMS on this one - as well as the safety aspect there's no guarantee that any fish tail with deliver the same result as an installed probe.

If you can describe why you think the non-contact probes might be giving incorrect readings and who manufactured them you will likely get some ideas on how to check for their condition.

It would also be useful to know the circumstances when this lack of confidence first developed. For example: was everyone happy with the non-contact system readings until the machine was worked on? following the work, the probes gave higher readings and were then thought to be reading wrong?

 

[apass]

Thanks for the replies, guys. I know it is a safety concern, but I work in Construction, so there are things that we can do during Startup and Commissioning that might not be tolerated in a normal plant operating environment.

These are new BFP's, and the probes are by Indikon. We are getting readings that exceed the bearing clearances, so something is not right.

byrdj, thanks for the tip on IRD. I'll look into it.

 

[byrdj]

I,ve Google "IRD Entek etc" and never found a photo of their "shaft probe". it came with thier model 544 detectors The fishtials I've used were also with BFPT and especially the coupling spool piece. My vintage of startups only had cap mounted siesmic, then later shaft rider types.

As for safety, I've never had any problems at the plants i've visited. My personal concern was after taken some very high magnitude readings on the spool piece, the unit was shut down and the coupling was found cracked half way around! (minutes from diasaster)

 

[sms]

I bet the probes are working fine. If the pump is turbine drive, then get it going about 200 rpm and see what the monitors read. If it is still that high level, then the problem is runout in the probe area. If it is motor drive, bump the motor or start it up and shut it down, and watch the monitor when the shaft speed slows down to around a couple hundred rpm. Again if the vib level is high, you have a shaft runout issue.

Things that can cause the probes to read high to very high include: Chrome in the probe track, residual magnetic fields (check the shaft gauss), a scratch or scratches in the probe track, a shaft that is not truely round, a bent shaft (which is a real problem that needs to be fixed), a shoulder or keyway in or near the probe track, improper grounding of the probes resulting in a ground loop, and probes mounted too close together causing cross talk.

Oh, and high vibration as well! What are the clearances and what are the levels you are reading? Depending on axially where the probes are mounted it is possible to have a vibration reading that is higher than the bearing clearance value. Another common problem is to have a stub shaft mounted to the end of the rotor, and then make the vibration readings on the stub shaft. If the stub shaft is not true then the readings will be high...