Vibration Alarm Setpoints 1

[rasmumm]

I am looking for some guidance for determining alarm setpoints. I am aware that ASME section XI has some guidance, but in my experience this is wholly inadequate to maintain a quality Predictive Program. I am also aware that much of the alarm status is based on historical component performance and baselining, but I would like some kind of standardized reference.

 

[electricpete]

I know that ISO has some good velocity limits, but I don't know the numbers.

If you are purchasing a new motor, the following limits would apply to a test rigidly mounted (NEMA MG-1).
< 1 mil (most limiting for frequencies < 1200cpm)
< 0.12ips (most limiting for frequencies 1200-24,000cpm)
< 0.18 g's. (most limiting for frequencies >24,000)
Which one is most limiting depends upon what frequency range the energy is in.

I think that those are pretty tight limits to apply to all in-service machines. I know you already know it but.... look at historical performance AND performance of sister machines.

 

[mdimond44]

Vibration limits are specific to particular machine and measurement types. Limits for relative displacement on high speed compressors will not apply to casing velocity measurements on slow speed pumps. If you are more specific in your application, I am sure you will get more input.

 

[jmack]

Setting common alarm limits, based upon mere operating speed,is clearly not the be all and end all to raising concern over machine operating condition. I have personally seen paper machine dryer bearing failures where overall magnitude readings were less than 0.05 ips (@190 rpm), and output shaft bearings on gearboxes (@260 rpm) operate normally for years at levels above 0.5 ips. Alarm level settings should be case specific as well as speed specific. It has been my experience that a much better approach is found in setting spectral band alarms, based upon statistical analysis of historical data.