The P-F Interval and How Rotating Machines Fail
Why the time between a detectable fault and a functional failure decides everything in a monitoring programme.
13 min
Activities: 0 of 2
Failure is a process, not an event
Most rotating machine failures do not happen in one moment. A bearing spalls, a coupling wears, a foundation bolt loosens. Each of these starts small and grows. For a period of weeks or months, the machine still does its job, but a fault is already present and can be measured. Condition monitoring exists to use that period.
Reliability engineers describe it with the P-F curve. Point P is the potential failure: the first moment a fault can be detected with a given technique. Point F is the functional failure: the moment the machine can no longer do what the operation needs. The time between them is the P-F interval.
A rolling bearing on the P-F curve
- 1
Hidden
Sub-surface fatigue starts
Cracks begin under the raceway surface. No technique in routine use detects this yet. - 2
P (ultrasound)
High-frequency stress waves
Ultrasound and acoustic emission can detect the first surface distress and friction from poor lubrication. - 3
P (vibration)
Defect frequencies appear
Envelope analysis, then the ordinary spectrum, show the bearing defect frequencies. - 4
P (oil, heat)
Wear debris and heat
Wear particles reach the oil and the housing temperature starts to rise. Little time is now left. - 5
P (senses)
Audible noise, felt by hand
An operator can hear or feel the problem. The bearing is close to failure. - 6
F
Functional failure
The bearing seizes or the shaft drops enough to damage seals, the rotor or the casing.
Why the interval matters to the design of a programme
Each technique has its own point P on the curve. A technique that detects a fault earlier gives a longer P-F interval and more time to plan the repair. This is the reason a programme often combines several techniques.
The P-F interval also sets the inspection interval. A common rule is to inspect at no more than half of the P-F interval. With that rule, at least one inspection falls inside the interval and still leaves time to plan the work. If the P-F interval for a bearing defect on a pump is about eight weeks with routine vibration data, a monthly route is just acceptable. A route every two months could miss the fault completely.
The P-F interval is not a fixed number. It changes with load, speed, lubrication and the failure mode. A bearing on a heavily loaded slow conveyor may give months of warning. A bearing on a high-speed fan running hot with poor lubrication may give days. An advanced analyst does not use one interval for the whole plant.
Two workplace cases
Common mistakes
- Using one interval for every machine. Fix: group machines by failure mode and speed, and set intervals from the expected P-F interval of each group.
- Treating point P as the moment to act. Detection starts a planning window. The repair should be done before the end of the window, not at the first sign.
- Ignoring the time needed to act. If spares take six weeks to arrive and the P-F interval is eight weeks, a monthly route leaves almost no margin. Hold critical spares or detect earlier.
A failure mode on a pump has a P-F interval of about six weeks with route-based vibration. Which inspection interval is the best choice?
Complete the definitions
Point P is the , the first moment a fault can be detected. Point F is the . A common rule sets the inspection interval at no more than of the P-F interval.