A long rotor can be perfectly balanced at one end and still shake, because its unbalance is not a single heavy spot but a twisting pair of them that a correction in one plane cannot fix. Two-plane rotor balancing solves this by working in two correction planes at once, using the influence-coefficient method to account for how a weight in each plane affects the vibration at both ends. It builds on the amplitude-and-phase vectors measured at running speed, adds trial weights, and solves for the corrections including the cross-effect between planes. This guide explains the method step by step, when a rotor truly needs two planes, and how it draws on monitoring data.
Two-Plane Rotor Balancing in one line: Two-plane rotor balancing is the procedure for correcting unbalance in two separate correction planes along a rotor at the same time, using the influence-coefficient method. It measures the running-speed vibration vectors at two bearings, adds trial weights in each plane to learn how each plane influences each measurement including the cross-effect, and then solves for the correction weights and angles in both planes. It is needed when a rotor has couple unbalance that a single correction plane cannot fix.
Single-plane balancing treats the rotor's unbalance as a single heavy spot that can be cancelled by one correction weight in one plane. That works well for a short, disc-like rotor such as a single fan wheel or a narrow impeller, where the mass is concentrated in essentially one plane and the running-speed vibration at both bearings responds together to a single correction. For those rotors, one plane is enough and adding a second would be unnecessary complication.
A longer rotor behaves differently because its unbalance can be distributed along its length rather than concentrated at one point. The most important case is couple unbalance, where there are effectively two equal heavy spots at opposite ends of the rotor and on opposite sides. This pair produces no net heavy spot, so it may not show much at low speed, but at running speed it creates a twisting, or moment, that makes the two ends of the rotor vibrate out of phase with each other. A single correction weight in one plane cannot cancel a couple, because whatever it does to one end it does the wrong thing to the other.
This is why couple unbalance demands two planes. To cancel a twisting pair of heavy spots you need to place correction weights in two planes, so that the corrections themselves form an opposing couple that neutralizes the original one. In general a real rotor has a mixture of static unbalance, which a single plane addresses, and couple unbalance, which requires the second plane, so two-plane balancing handles the general case of a rotor whose unbalance is spread along its length. Recognizing from the vibration that the two ends are moving out of phase is a strong hint that a second plane is needed.
The influence-coefficient method is the systematic way to solve a two-plane problem, and it starts with baseline measurements. With the machine at its balancing speed, the running-speed vibration vector, amplitude and phase, is recorded at each of the two measurement points, typically one near each bearing. These baseline vectors describe the unbalance condition the procedure has to correct. The keyphasor supplies the phase reference that makes each measurement a vector rather than just an amplitude, which is what allows the corrections to be computed rather than guessed.
Next the method learns how the rotor responds to weight in each plane. A known trial weight is placed at a known angular position in the first plane and the machine is run again, recording how the vector at each measurement point changes. The trial weight is removed and the process is repeated in the second plane. From these trial runs the influence coefficients are computed: each coefficient expresses how a unit of weight at a given angle in one plane changes the vibration vector at a given measurement point. Crucially this includes the cross-effect, the influence of a weight in the first plane on the vibration at the second point and vice versa, because in a two-plane system the planes are coupled and a weight in one plane affects both ends.
With the full set of influence coefficients known, the correction is a calculation rather than a trial. The coefficients relate the unknown correction weights in the two planes to the measured baseline vibration, and solving that set of relationships yields the size and angular position of the correction weight needed in each plane to drive the running-speed vibration toward zero at both points simultaneously. Because the cross-effect is included, the solution accounts for the fact that each correction weight will also affect the far end, which is exactly what a single-plane calculation cannot do. The computed weights are then installed and a verification run confirms the vibration has dropped.
Two-plane balancing runs entirely on the running-speed amplitude-and-phase vectors, which are the same quantities a machinery-monitoring system already captures. Clean, trustworthy vectors are the raw material of the whole procedure, so anything the monitoring system does to keep them accurate, such as applying slow-roll runout compensation and referencing a reliable keyphasor, directly improves the quality of the balance. If the vectors feeding the calculation are corrupted by runout or a bad phase reference, the computed corrections will be off no matter how sound the arithmetic.
A cloud monitoring platform helps by logging the running-speed vectors at each measurement point through the balancing runs, so the baseline and each trial-weight run are recorded consistently and can be compared with confidence. Merobix can trend the 1x amplitude and phase at each bearing, which gives the person doing the balancing a clear before-and-after picture and a permanent record of how the vectors responded to each trial weight. That record is useful both for computing the corrections and for confirming afterward that the machine settled where expected.
The historized vectors also make the balance durable rather than a one-time event. After a rotor is balanced, its running-speed vectors establish a baseline, and continued monitoring can watch for those vectors drifting away from it, which would signal that unbalance is returning through fouling, erosion, or a thrown weight. For remote and distributed machines this means a balance done once can be verified from afar and its longevity tracked, so a reliability engineer sees early when a machine is heading back toward needing another balance rather than discovering it only when vibration alarms return.
A short, disc-like rotor whose mass sits essentially in one plane can usually be corrected with single-plane balancing. A longer rotor whose unbalance is distributed along its length, particularly one with couple unbalance where the two ends vibrate out of phase, needs two planes. That is because a couple is a twisting pair of heavy spots that a single correction weight cannot cancel; you need weights in two planes to form an opposing couple. Out-of-phase vibration at the two bearings is a strong sign two planes are required.
The cross-effect is the influence that a correction weight placed in one plane has on the vibration measured at the other end of the rotor. In a two-plane system the planes are coupled, so a weight added in the first plane changes the vibration at both measurement points, not just the near one. The influence-coefficient method captures this by including cross coefficients, so the computed corrections account for each weight affecting the far end. Ignoring the cross-effect is what makes single-plane calculations inadequate for these rotors.
Influence coefficients describe how the rotor responds to weight. Each coefficient tells you how much the running-speed vibration vector at a particular measurement point changes per unit of weight added at a particular angle in a particular plane. They are found by measuring the baseline vibration, then adding a known trial weight in each plane and seeing how the vectors change. Once known, the coefficients let you solve directly for the correction weights and angles needed in both planes to minimize the vibration.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.