All Member Open Forum

 View Only

  • 1.  Machine Vibration Field Monitoring

    Posted 03-25-2021 11:21 AM

    Hi all. We are working on putting some bearing mounted vibration probes on some very rough running equipment at our facility.  They are decanters for those of you that are familiar with them.  We expect that 1" in/sec of total vibration is our threshold we no longer want to operate the machine.  With Peak to Peak, we see instantaneous spikes much higher than 1", RMS seems to stay much below. 

    I'm curious what other facilities use for "general" health of machine vibration.  RMS or Peak to Peak.   We still do our wide spectrum vibration analysis separate of these new probes on the machine.  The probes intent is to let us see any potential issues in the machine that will alert us to a problem.  Using a peak to peak setting is giving us a lot of unwanted noise, but also may be more accurate.  RMS gives us a generally more consistent reading. 

    Our vibration analysts recommend peak to peak, this is what they use for analysis, but I'm starting to wonder if they are still thinking from an analysts point of view and not a overall view.   The RMS seems to be a better method to gauge machine health from a high level. 

    Please reply and let me know what kind of systems you use for local/data historian reading. 

    Thank you ,

    Kyle 



    ------------------------------
    Kyle Burnham
    Maintenance Reliability Engineer
    Grain Processing Corp
    Muscatine IA
    ------------------------------


  • 2.  RE: Machine Vibration Field Monitoring

    Posted 03-26-2021 08:45 AM
    Kyle,

    The most effective analysis parameter to be used should be based on what data you want to collect and record. If the machine is known to have a 'normal' high amplitude vibration at the frequency associated with its turning speed, then why  not adjust the analysis parameter to 'filter out' the high amplitude vibration by raising the FFT's Fmin to a point above the 1xTS peak and below the 1xBPFO value? That way, you can see any anomalies that would be hidden in the noise floor of a standard FFT?

    Also, use velocity displayed in ips-pk as your unit of measure for amplitude. You can set alarm levels for specific frequency bands that will inform you of known and anticipated faults. I have used this technique very successfully on high amplitude machines, like decanters, shaker screens, and other machines designed to have a high amplitude vibration.

    Please feel free to reach out to me with any questions you might have or put me in contact with your provider to coach them.

    jim.carrel@reliabilitystrategies.net

    Go forward!

    Jim

    ------------------------------
    Jim Carrel, CRL
    Reliability Coach
    Reliability Management Strategies
    jim.carrel@reliabilitystrategies.net
    206-892-8074
    ------------------------------

    Original Message


  • 3.  RE: Machine Vibration Field Monitoring

    Posted 03-26-2021 11:35 AM
    Kyle,

    Have you done a FMEA on the equipment?  Do you know the failure modes you are trying to detect?
    If the machine normally has a lot of energy, than defects can easily be masked if they have lower energy at early detection points on the P-F curve.
    As Jim said, use specific analysis techniques to determine your common faults, bearing, gear, belt, misalignment, unbalance, etc.  If you want an overall for peace of mind or to detect something unknown, then trend both until you find an unknown or uncommon failure mode and see which, peak or rms, provided you insight into that failure mode.
    Make sure you understand if you vibration output is giving true peak or derived peak (1.4*rms).  If derived, then you might as well use rms.
    When you choose 1.0 ips as your starting baseline, how did you get to that number?  It should already have units of rms or peak.
    You can also check with the OEM to see what they recommend.

    ------------------------------
    Mark Wolka
    Sr. Product Eng
    NSK Corporation
    Ann Arbor, MI USA
    ------------------------------

    Original Message


  • 4.  RE: Machine Vibration Field Monitoring

    Posted 03-26-2021 11:43 AM
    Im trying to measure the vibration of the machine itself in displacement.  We are not using this system to determine issues with bearings/belts ect.  These machines are very old (and manufacturer is out of business), and in the past used what we call as a "earth quake" switch that will kick the machine out at 5g of vibration.   The 1" ips i am using as a baseline is from a newer machine that operates the same way.   

    We use wide spectrum vibration analysis seperate of this system I am trying to setup.  This system is to kick the machine out if it becomes unsafe.

    ------------------------------
    Kyle Burnham
    Maintenance Reliability Engineer
    Grain Processing Corp
    Muscatine IA
    ------------------------------

    Original Message


Related Content