Sub-Synchronous Turbo machinery Faults

Random Sub-synchronous Turbomachinery Faults

(It is not always Oil Whirl or Journal Bearing issues)

Unlike frequency band between (0 < f < 0.5X), sub-synchronous bandwidth (0.5X < f < 1X) could cause some confusion when interpreting vibration peaks within it or trying to imagine timing (frequency) of faults within this region.

This frequency span involves the response of excitation forces that relates DIRECTLY to rotor rotation (NOT whirling). In this region, peak of vibration signal occurs at one and part of rotor rotation.

👉 0.67X = 1 Peak per 1.5 revolution
👉 0.75X = 1 Peak per 1.33 revolution
👉 0.80X = 1 Peak per 1.25 revolution
👉 and so on..

This means (in my interpretation for several diagnostic cases) that is some sort of force that is delayed from rotor residual unbalance rotation by a constant time. This indicates some faults such as:

✔ Trapped fluid (fluid whirl inside a bore clearance slower than rotor speed)
✔ Seal defects at specific point that causes leak pulses or non-uniform turbulence (e.g. partial stall, fouling, flow restriction, etc.)
✔ Looseness or crack in stationary part (ring or labyrinth) that lowers its natural frequency below 1X slightly

🛠 All these faults are common that the rotor is the element that CONTROLS their peak amplitude. If we change rotor speed or reduce rotor residual imbalance (or even load in thermal sensitive rotors), its value will be decreased rapidly and simultaneously.

📈 If we cannot do this at certain moment, monitoring the peak will be enough unless another fault symptom interferes with its response causing (often) a nonlinear behavior. In all cases, it won't be eliminated unless we remove the source of excitation itself during maintenance activities.

💡 Notice that images represent different cases for above possibilities! Share your experience in comments. Do you experience random peaks below 1X frequency. If so, what was your interpretation!?