•  
  • Watch the video of this example.

    Bearing diagnostics

    Bearing

    Rotating elements in machines are often mounted by rolling bearings. These are often the most stressed components because they separate the rotating part from the fixed part.  Improper installations, misalignments of shafts or unfavorable lubrications can lead to damages at the rolling bearings.

    This practical example shows how to discover bearing damages early and assign them to individual bearing components by using VibroMatrix.  Additionally the VibroMatrix Instruments InnoMeter, InnoAnalyzer and InnoScope are used.

    Defects of rolling bearings appear as impacts at a kHz level. That is the reason why you just take a look at the area from 1000 Hz upward. So you can filter out frequency components representing other effects.

    The measurement of broadband vibration parameters provides information on damages in the bearings. If these values increase during regular measurements, a damage emerges. Which bearing component (outer ring, inner ring, cage or rolling elements) is affected, you can determine with the envelope analysis.

    • Bearing - initial position of components
    • Bearing - position of components during rotation

    When the shaft rotates, the bearing components rotate at varying speeds. The standstill position is marked on the bearing in the picture on the right site. You can see the different fast movements of the individual components by rotation. So if you know the rotational frequency you can calculate the specific component frequencies from the dimensions of the bearing.


    The Frequencies are named as follows:

    BPFO

    Ball Pass Frequency Outer Race

    rolling element pass frequency inner race

    BPFI

    Ball Pass Frequency Inner Race

    rolling element pass frequency outer race

    BSF

    Ball Spin Frequency

    ½ × ring pass frequency on rolling element

    RPFB

    Ring Pass Frequency on Ball

    2 x ball spin frequency

    FTF

    Fundamental Train Frequency

    cage rotational frequency

     

    If there exist lines in the envelope analysis, you can identify defect components.

     

     

    Practical example

    Practical example at machine model

    The bearing diagnostics is performed with recorded raw data in this example. These data were recorded on a model machine with replaceable, already damaged bearings. The machine was measured with damages on an outer ring, an inner ring, a rolling element and an undamaged bearing too.

    Raw data were reproduced with the InnoMaster Replay. The recorded data are loaded. In the preview video you can see an overview of the measurement. Comments can be integrated in the raw data stream during the measurement operation.

    Let us open the InnoMeter. The vibration acceleration in the range of 1 – 20 kHz is measured. Thereby you can see mean value of the effective value and the absolute peak value. Parallel we put out the rotation frequency in Hz. Furthermore, we need the InnoScope. This instrument is a storage oscilloscope and shows a high-resolution sensor signal. The filter settings are assumed. In the InnoAnalyzer we choose the operating mode “envelope analysis to diagnose of rolling bearings”. After that we adapt the bandpass of the envelope and use the digital input of the InnoBeamers.

    Practical example - undamaged bearing
    • Characteristics: low rms and low peak value
    • Signal in time domain: no dominating shock signals
    • Envelope signal: No dominating magnitudes

    The wideband effective and peak values are relatively low. The time signal consists of random signals. No dominant shock signals are visible. There are also no strong lines in the envelope analysis. These are typical characteristics of an intact bearing.

    Practical example - damage to the inner ring
    Characteristics: Increased rms and peak value

    The characteristic values and spectra differ considerably.The RMS and peak values are increased, so there is definitely damage.

     

    In the time signal we see strong oscillations in the form of a double impulse sequence. This sequence arises because the defect is overrun when the inner ring moves as well as when the rolling elements move. This time signal can be used to obtain information about the defective component: To do this, we use the cursor tool to measure the distances between the shocks. This results in 4.082 ms or 19.955 ms. These are converted to 245 Hz or 50.1 Hz. These are pretty much the rollover frequency of the inner ring of 244.6 Hz or the rotation frequency of 49 Hz. So there is actually damage to the inner ring.

    • Signal in time domain: Double shocks, short distance
    • Signal in time domain: Double shocks, long distance
    • Envelope signal: Ball pass frequency of inner race (BPFI) with side bands in a distance of speed frequency

    The envelope analysis shows the facts even more clearly. We open the bearing database and search for the type of bearing used. In the InnoAnalyzer, the individual damage frequencies are automatically calculated, so only the marking for the inner ring damage needs to be activated. In addition, the 1st harmonic and sidebands are selected at the distance of the rotational frequency. The strong lines in the spectrum lie exactly on the damage frequencies of the inner ring. The bearing damage is clearly identified. The sidebands originate from the modulation of the harmonic frequency by the rotational frequency and reflect the double shocks in the time signal.

    Damage to the outer ring
    • Characteristics: Increased rms and peak value
    • Signal in time domain: Periodic shocks
    • Envelope signal: Ball pass frequency of outer race (BPFO) and multiples of it

    Higher characteristic values can also be observed. The time signal contains periodic shocks and the envelope curve analysis results in peaks at the rollover frequency of the outer ring and its multiples.

    Rolling element damage
    • Characteristics: Increased rms and peak value
    • Signal in time domain: Double shocks
    • Envelope signal: RPFB with sidebands in a distance of fundamental train frequency (FTF)

    The wideband characteristic values are increased and in the time signal double shocks occur, in this case the cage circulation frequency modulates the rolling element rollover frequency. Characteristics in the envelope curve analysis are peaks at rolling element damage frequency with sidebands at a distance of the cage frequency and higher orders thereof.

    Cage damage
    • Characteristics: Increased rms and peak value
    • Signal in time domain: Periodical shocks
    • Envelope signal: Fundamental train frequency (FTF) and multiples of it

    Again increased characteristic values, the time signal shows periodic shocks. In envelope curve analysis, lines occur at the cage rotation frequency and its multiples.

     

    Damage to rolling bearings can be detected by measuring wideband characteristics. The InnoMeter or InnoPlotter can be used in VibroMatrix for this purpose. The InnoLogger is also suitable for continuous monitoring.

     

    Using the time signal and envelope analysis, the damage can be assigned to a bearing component. The InnoScope and the InnoAnalyzer are essential instruments for this.

     

    Rolling bearing diagnostics: Convenient and reliable analysis with VibroMatrix.