Glossary term
Order Tracking
Tachometer-referenced vibration analysis method that represents rotating-machine response by shaft order instead of only fixed frequency.
Definition
methodOrder tracking is a vibration analysis method that follows components synchronized to rotating speed by expressing spectral content in shaft orders.
Order tracking uses a tachometer, encoder or reliable speed estimate to relate vibration data to shaft angle or rotational speed. It separates synchronous components such as 1x, 2x, blade passing and gear mesh from fixed-frequency content, random vibration and non-synchronous faults during run-up, coast-down or variable-speed operation.
Order tracking is a vibration analysis method for rotating machinery that represents vibration components by shaft order instead of only fixed frequency. An order is a frequency normalized by rotational frequency:
where O is order, f is vibration frequency and f_{rot} is shaft rotational frequency. For rotational speed n in rpm:
so an order O appears at:
This matters because synchronous vibration changes frequency when speed changes. A fixed-bin FFT can smear a run-up or coast-down signal if the shaft speed changes during the record. Order tracking uses a tachometer, encoder or speed estimate so that peaks tied to shaft rotation remain aligned by order.
Engineering Role
Order tracking helps diagnose unbalance, misalignment, looseness, gear mesh, blade passing, shaft runout, rubs, torsional response and speed-dependent resonances. It is especially useful during run-up, coast-down, variable-speed operation, commissioning tests and troubleshooting when the machine does not remain at one steady speed.
A 1x order often suggests synchronous once-per-revolution forcing such as unbalance, eccentricity or bent shaft. A 2x component can indicate misalignment, looseness, ovality or other twice-per-revolution effects. Gear mesh, blade passing and vane passing appear at higher orders tied to tooth count, blade count or stage geometry. The order label is a clue, not a diagnosis by itself.
Good order tracking depends on the speed reference. Missed tach pulses, weak encoder signals, timing jitter, wrong pulses-per-revolution settings, shaft slip, torsional twist between measurement and tach location, and poor interpolation can all create misleading order maps.
Tracking Method and Reference Quality
The strongest order-tracking evidence comes from vibration channels recorded with a simultaneous tachometer, encoder or once-per-revolution reference. For an encoder with P pulses per revolution, the pulse frequency is:
where n is shaft speed in rpm. The acquisition system must sample the tach signal fast enough to detect pulses cleanly at maximum speed, and the pulse count must match the shaft feature being used as the angular reference. A wrong P value shifts every order label.
Synchronous resampling converts the vibration record from time spacing to approximately equal shaft-angle spacing before the order spectrum is calculated. The angular grid can be written as:
where N_{\theta} is the number of angular samples per revolution. This method is powerful during run-up or coast-down because a component locked to shaft angle stays at a stable order even while its physical frequency changes.
Computed order tracking estimates speed from the vibration signal itself or from another measured ridge when a tachometer is unavailable. It can be useful for screening old data, but it is weaker evidence for release work. If the assumed ridge jumps, blends with another component, crosses a resonance or disappears in noise, the computed speed trace can create false order stability.
Ramp rate sets another limit. During one analysis block, speed changes by approximately:
where \dot{n} is ramp rate and T_{block} is block duration. If the block is too long for the ramp, order peaks smear and narrow resonances can look broader or lower than they are. A defensible setup checks the order map against tach quality, ramp rate, block length, overlap, window, anti-alias filtering and repeat run-up or coast-down behavior.
Worked Example: Separate Synchronous and Fixed-Frequency Peaks
A fan is measured during a controlled speed sweep at three speeds. The analyst sees one peak that moves with speed and one peak fixed at 60\ \text{Hz}.
| Speed | Rotational frequency | Moving peak | Fixed peak |
|---|---|---|---|
| 1200\ \text{rpm} | 20\ \text{Hz} | 60\ \text{Hz} | 60\ \text{Hz} |
| 1500\ \text{rpm} | 25\ \text{Hz} | 75\ \text{Hz} | 60\ \text{Hz} |
| 1800\ \text{rpm} | 30\ \text{Hz} | 90\ \text{Hz} | 60\ \text{Hz} |
Compute the order of the moving peak:
The moving peak is a 3x synchronous component. It may be related to a three-bladed fan, a three-lobe geometry, a coupling feature or another rotating source that repeats three times per revolution.
Now compute the order of the fixed 60\ \text{Hz} peak:
The fixed peak is not locked to one shaft order. It may be electrical line frequency, a nearby machine, structural vibration at fixed frequency, or another non-synchronous source. At 1200 rpm it happens to coincide with 3x, which could mislead a fixed-speed analysis.
Engineering comment: order tracking prevents the analyst from treating every 60 Hz peak as the same mechanical fault. The tachometer reference reveals which components are synchronous with speed and which are fixed in physical frequency. A final diagnosis still needs sensor location, phase, load state, machine geometry and repeatability.
Distinction from Related Terms
Order tracking is not a tachometer. A tachometer provides speed or phase reference. Order tracking is the analysis method that uses that reference.
Order tracking is not simply an FFT. FFT spectra are frequency-based. Order tracking may use FFTs, resampling or time-frequency methods, but its output is organized by rotational order.
Order tracking is not a Campbell diagram. A Campbell diagram maps modal branches and excitation orders against speed. Order tracking analyzes measured vibration components relative to speed.
Order tracking is not a diagnosis by itself. A 1x, 2x or blade-passing order suggests possible mechanisms, but the conclusion depends on amplitude, phase, direction, operating state, machine construction and corroborating evidence.
Validation and Common Mistakes
A defensible order-tracking setup states the tachometer or encoder source, pulses per revolution, shaft reference location, sampling rate, anti-alias filtering, speed range, ramp rate, resampling method, windowing, averaging, channel calibration and uncertainty.
Common mistakes include:
- collecting a speed-changing record without a tachometer or reliable speed reference;
- confusing fixed-frequency electrical content with shaft-synchronous orders;
- using the wrong pulses-per-revolution setting;
- ignoring torsional speed variation between the tach location and the fault source;
- applying steady-speed FFT settings to a fast run-up record;
- diagnosing a fault from order number alone without phase, amplitude trend or machine geometry;
- accepting an order map without checking aliasing, clipping, sensor mounting and signal-to-noise ratio.