ISO 17359 pdf download – Condition monitoring and diagnostics of machines -General guidelines

ISO 17359 pdf download – Condition monitoring and diagnostics of machines -General guidelines.
References to more detailed methods of carrying out FMEA and FMECA are given in the Bibliography.
7.4 AlternatIve maintenance tasks
lithe failure mode does not have a measurable symptom, alternative maintenance strategies may have to be applied. These include bum-in (initial testing), run to failure, corrective maintenance, preventive maintenance or modification (design out).
8 Monitoring method
8.1 Measurement technique
For the parlicular measurable parameter considered to be applicable lollowing the previous selection process. one or more measurement techniques may be appropriate. Measured parameters can be simple measurements of overall values or values averaged over time, For certain parameters, such as current. voltage, and vibration, simple measurements of overall values may not be sufficient to show the occurrence of a fault. Techniques such as spectral and phase measurement may be required to reveal changes caused by faults,
Examples of monitored parameters useful to consider for a number of machine types are given in Annex A. Examples of standards which may be useful in the identification of particular measurement methods and parameters for different machine types are included in the Bibliography.
The range and application area of International Standards relating to condition monitoring and diagnostics is shown in Annex D.
Condition monitoring systems can take many forms. They may utilize permanently installed. semi-permanent. or portable measuring instrumentation, or may involve methods such as sampling fluids or other materials for local or remote analysis.
8.2 Accuracy of monitored parameters
In most cases, the accuracy required of the parameters to be used for machine condition monitoring and diagnosis s not necessarily as absolute as the accuracy which may be required for other measurements such as performance testing.
For new and overhauled equipment, there may be a wear-in period. As a result. it is common to see a change in measured values during the first few days or weeks of operation. Therefore, time should be allotted for wear-in before acquiring baseline data or, for overhauled equipment, before re-establishing baselines.
For equipment which has been operating for a significant period, and monitored for the first time, a baseline can still be established as a trending reference point
9 Data acquisition and analysis
9.1 Measurement and trending
The general procedure for data acquisition is to take measurements and compare them to historical trends. baseline data or representative data for the same or similar machines. Management of the condition monitoring data collection procedure is often done on-line by arranging for the measurements to be taken in a scheduled acquisition sequence. Data collection can also be managed oil-line by taking measurements along a predetermined route or tour of the plant. Measurements are then scheduled to be carried out at an initial regular periodicity which is much more frequent than the expected failure mode. For many concbtion monitoring techniques. computer-based systems are available which assist in the management of data acquisition, data-collection routes, recording and trending of measurements
9.2 Quality of measurements
The quality of any measurement needs to be established. There are many causes of poor measurements. These may include:
— poor mounting of transducer:
— transducer fault:
cable fault:
— incorrect measurement range resulting in a saturated signal:
sampling rate insufficient to detect the actual rate of change of the measured parameter.
It poor quality measurements are defected or suspected, it may be necessary to repeat the measurements or
to rectify the measurement system fault.
9.3 Measurement companson to alert/alarm criteria
If the measured values are acceptable compared to the alert/alarm criteria, no action may be required other than to record the values and to continue to monitor them If the measured values are not acceptable compared to the alert/alarm criteria, then the diagnosis process should be initiated. There can be occasions when no anomalies are suspected or detected, but diagnosis and prognosis is still carried out because of a requirement for a machine health assessment decision, eg. when carrying out a Condition survey of equipment before a major shut-down. This may also be the case where other symptoms are noted which may have been outside the monitoring programme, but are detected by an alert operator such as noise, smell or visual symptoms.