Understanding Order Analysis for Variable-Speed Machines

Vibration sensor

Optical Sensor (Laser Tachometer)

Balanset-4

Magnetic Stand Insize-60-kgf

Reflective tape

Dynamic balancer “Balanset-1A” OEM

Order analysis is a specialised vibration analysis technique built for machines that do not run at a single, steady speed. Instead of plotting amplitude against a fixed frequency axis in Hz or CPM, it plots amplitude against orders — multiples of the shaft’s instantaneous running speed. The 1st order is the vibration at exactly 1× (running speed), the 2nd order is 2× that speed, and so on. By tying the analysis to the shaft itself rather than to the clock, order analysis keeps speed-related components sharp no matter how the machine accelerates or coasts.

1. Definition: What Is an Order?

An order is a harmonic of the fundamental rotational speed. Because so many machine faults excite vibration at integer multiples of shaft speed, expressing the spectrum in orders maps each peak directly onto a physical cause. The 1st order almost always carries unbalance; the 2nd order is a classic marker of misalignment and certain looseness conditions; higher integer orders relate to gear mesh, vane or blade-pass events tied to the number of elements on the rotor. Non-integer (fractional) orders flag sub-synchronous phenomena such as oil whirl or belt defects. In short, the order axis is a diagnostic map that travels with the rotor.

2. Why Standard FFT Fails on Variable-Speed Machines

A conventional Fast Fourier Transform (FFT) samples vibration over a fixed window of time and assumes the speed is constant for that window. On a constant-speed machine this is perfect. But if the shaft speeds up or slows down while data is being collected, every speed-related component drifts across the spectrum during the capture. Its energy is smeared across many adjacent frequency bins, producing a broad, low, fuzzy hump instead of a clean line. A 1× unbalance peak that should tower over the spectrum can flatten into noise — impossible to read accurately and useless for trending. This smearing is the same mechanism behind spectral leakage, amplified by changing RPM. Order analysis was developed specifically to defeat it.

3. The Solution: Order Tracking

The enabling technique is order tracking, and it depends on a second input: a tachometer (or “tacho”) delivering a once-per-revolution pulse from the shaft. The analyser treats this pulse train — not its internal crystal clock — as the time base. Rather than sampling at fixed time intervals (say, every millisecond), it samples at fixed angular intervals (for example, every degree of rotation). This is known as resampling in the angle domain.

Two methods are common. Hardware (synchronous) sampling drives the analogue-to-digital converter directly from a phase-locked multiple of the tacho pulse, so each revolution always yields the same number of samples. Computed (software) order tracking samples at a high fixed rate, then digitally re-interpolates the record onto equal-angle steps using the recorded tacho timing. Either way, the resulting transform is expressed in orders, not Hz. If the machine changes speed, the 1× line stays put in the 1st-order bin as a tall, narrow peak — the smear vanishes. The tacho also supplies a phase reference, which is what lets the analyser build Bode and Nyquist plots during a run-up.

Key idea: order analysis locks the data acquisition to shaft angle instead of time, so speed-synchronous vibration stays sharp at every RPM.

4. Key Applications

Order analysis is indispensable wherever speed is not constant:

  • Vehicle and engine testing: resolving engine, transmission and driveline vibration across the entire RPM sweep.
  • Wind turbines: rotor speed tracks the wind continuously, so a fixed-frequency view is meaningless — order analysis is essential.
  • Run-up and coast-down analysis: capturing vibration as a machine starts or stops is a powerful way to locate critical speeds and resonances; order tracking keeps the resulting coast-down plots clean and readable.
  • Reciprocating machinery: compressors and engines whose instantaneous speed fluctuates within each cycle.
  • Heavy and mobile machinery: earth-moving equipment, mining vehicles and other variable-speed drives.

5. How Order Analysis Data Is Displayed

Results are viewed in several complementary formats:

  • Order spectrum: amplitude versus orders — like a standard FFT, but with orders on the x-axis.
  • Waterfall or cascade plot: a stacked 3-D set of order spectra showing how each order’s amplitude evolves as speed changes.
  • Bode plot: amplitude and phase of one tracked order (usually 1× or 2×) plotted against machine speed, the backbone of run-up/coast-down testing.
  • Campbell diagram: order lines overlaid on the system’s natural frequencies, so a resonance shows up wherever an order line crosses a natural-frequency line.

A tracking filter can isolate a single order in real time for trim work, and a Campbell Diagram Calculator helps predict where those crossings will occur before testing.

6. Order Analysis in Practical Field Work

On the shop floor, order analysis underpins balancing on machines that won’t hold a steady speed. A portable two-channel instrument such as the Balanset-1A uses its optical laser tachometer to lock vibration data to shaft angle, so the 1× unbalance component it measures for field balancing stays clean even on a fan or pump whose RPM drifts under load. The same tacho-referenced approach lets the analyser separate the speed-synchronous 1× peak from fixed-frequency noise such as bearing fault frequencies, giving a trustworthy reading of the heavy spot. In effect, order analysis is what turns vibration data from a variable-speed machine into something an engineer can act on — accurately diagnosing the health of any rotor that runs across a range of speeds.


← Back to Main Index

Categories: AnalysisGlossary

WhatsApp
Balanset-1A · €1975 Ask engineer