1S5: Case Study: Signal Processing in Action for Clear Defect Identification

Bruno Fouille, ACOEM


A vibration analyser is manufactured with lots of efforts of R&D, new technology and new hardware. All best processors are packed in an analyser to ensure user can collect the required signals and make a precise judgement for their machine. In order to achieve this function, analyser has to be configured suitably. Its a process similar to a situation : Different lock has different key. You can not expect a standard configuration will work for monitoring all machines.

This presentation will give you insight of signal processing for envelopping process which is very popular for identification of lubrication problem in bearings as well as early identification of damage in bearings. Major topics to be covered in the presentation are:

  • Problem with raw signal when there is defect in bearing
  • Missing of shocks
  • Overview of envelopping process
  • Customising envelopping parameter and its effect
  • How to cover more revolution in measurement
  • Kurtosis function
  • Extract from Time waveform
  • Shock finder and bearing defect factor (These are special processing by Acoem. In case it is accepted by Mobius, same will be presented otherwise it will be eleminated)
  • Selection of sensor
  • Sensor mounting

The presentation will also have a case study on identification of bearing defect on “Cement Mill – Roller Press” which is also know as HPGR – High Pressure Grinding Rolls.
This machine rotates @ 18 RPM and it is very crucial to to have early defect identification. The differentiator for this successful monitoring was customising the envelopping process.

*Understanding of Envelopping Customising
*Measurement configuration
*Identification of shocks