Understanding the N+2 Method in Multi-Plane Balancing

Devices

Portable balancer & Vibration analyzer Balanset-1A

€1,975.00 + VAT (if applicable)

Parts

Vibration sensor

€90.00 + VAT (if applicable)

Parts

Optical Sensor (Laser Tachometer)

€124.00 + VAT (if applicable)

Devices

Balanset-4

€6,803.00 + VAT (if applicable)

Parts

Magnetic Stand Insize-60-kgf

€46.00 + VAT (if applicable)

Parts

Reflective tape

€10.00 + VAT (if applicable)

Devices

Dynamic balancer “Balanset-1A” OEM

€1,735.00 + VAT (if applicable)

The N+2 method is an advanced balancing procedure used for multi-plane balancing of flexible rotors. Its name describes the measurement strategy precisely: if N is the number of correction planes required, the method uses N trial-weight runs — one for each plane — plus two more runs, an initial baseline and a final verification, for a total of N+2 runs. It extends the logic of two-plane balancing to rotors needing three or more planes, a situation common in high-speed turbines, compressors, generators and long paper-machine rolls.

1. Definition: What the N+2 Method Is

A rigid rotor running below its first critical speed can be brought into tolerance with simple single- or two-plane correction, because its unbalance distribution does not change shape with speed. A flexible rotor is different: once it runs at or above a critical speed it bends, and that bending redistributes the effective unbalance along its length. Correcting it therefore demands several planes spread along the shaft, and a method that can untangle how each plane influences vibration everywhere else. The N+2 method is that systematic accounting procedure — a disciplined way to characterise the rotor completely, then solve for the best correction at every plane at once.

2. The Mathematical Foundation

The N+2 method is built on the influence coefficient method, generalised from one or two planes to many.

The Influence Coefficient Matrix

For a rotor with N correction planes and M measurement locations (typically M ≥ N), the system is described by an M×N matrix of influence coefficients. Each coefficient α_ij captures how a unit weight placed in correction plane j affects the vibration recorded at measurement location i. With four correction planes and four measurement locations, for example:

• α₁₁, α₁₂, α₁₃, α₁₄ describe how each of the four planes affects measurement location 1;
• α₂₁, α₂₂, α₂₃, α₂₄ describe the effects on measurement location 2;
• and so on for locations 3 and 4.

That produces a 4×4 matrix requiring the determination of sixteen influence coefficients. Each coefficient is a complex quantity, carrying both a magnitude and a phase angle, because the rotor’s response lags the applied force.

Solving the System

Once all coefficients are known, the balancing software solves a system of M simultaneous vector equations to find the N correction weights (W₁, W₂, … W_n) that minimise vibration at all M locations at once. This relies on vector mathematics and matrix-inversion (or least-squares) algorithms. When M exceeds N the system is overdetermined and a least-squares solution finds the correction set that gives the smallest residual vibration across all sensors — a more robust outcome in the presence of measurement noise.

3. The N+2 Procedure, Step by Step

The procedure follows a sequence that scales naturally with the number of correction planes.

Run 1 — Initial Baseline Measurement

The rotor is run at balancing speed in its initial unbalanced condition. Vibration amplitude and phase are recorded at all M locations — usually at each bearing, and sometimes at intermediate positions to catch mid-span motion. These readings establish the baseline unbalance vectors that must be corrected.

Runs 2 through N+1 — Sequential Trial-Weight Runs

For each correction plane in turn, from 1 to N:

1. Stop the rotor and attach a trial weight of known mass at a known angular position in that one plane only.
2. Run the rotor at the same speed and measure vibration at all M locations.
3. The change in vibration — the current vector minus the baseline vector — reveals how that specific plane influences each measurement location, yielding one column of the coefficient matrix.
4. Remove the trial weight before moving to the next plane (unless the deliberate “leave-in” variant is being used to save runs).

After all N trial runs, the complete M×N influence coefficient matrix is known.

Calculation Phase

The instrument solves the matrix equations to compute the required correction weights — both mass and angle — for each of the N planes.

Run N+2 — Verification

All N calculated corrections are installed permanently and a final run confirms that vibration has fallen to acceptable levels at every measurement location. If the result is not yet satisfactory, a trim balance or a further iteration is performed using the coefficients already in hand.

4. Worked Example: Four-Plane Balancing (N = 4)

For a long flexible rotor requiring four correction planes:

• Total runs: 4 + 2 = 6.
• Run 1: initial measurement at all four bearings.
• Run 2: trial weight in Plane 1, measure all four bearings.
• Run 3: trial weight in Plane 2, measure all four bearings.
• Run 4: trial weight in Plane 3, measure all four bearings.
• Run 5: trial weight in Plane 4, measure all four bearings.
• Run 6: verification with all four corrections installed.

This builds a 4×4 matrix of sixteen coefficients, which is solved to find the four optimal correction weights. The same arithmetic for a simpler job sits behind an influence coefficient calculator, which solves the single-plane case and makes the underlying vector method easy to see before scaling up.

5. Advantages of the N+2 Method

The approach offers several important benefits for multi-plane work:

• Systematic and complete: every correction plane is tested independently, giving a full characterisation of the rotor-bearing system‘s response across all planes and locations.
• Captures complex cross-coupling: in flexible rotors a weight in any plane can affect vibration at every bearing; the matrix records all those interactions explicitly.
• Mathematically rigorous: it uses well-established linear-algebra techniques (matrix inversion, least-squares fitting) that give optimal solutions when the system behaves linearly.
• Flexible measurement strategy: allowing M to exceed N produces an overdetermined system that is more robust against noise.
• Industry standard for complex rotors: it is the accepted method for high-speed turbomachinery and other critical flexible-rotor applications.

6. Challenges and Limitations

Multi-plane balancing by the N+2 method also presents real difficulties:

• Increased complexity: the number of trial runs grows linearly with the planes. A six-plane balance needs eight runs, sharply increasing time, cost and machine wear.
• Measurement-accuracy demands: solving large matrices amplifies the effect of measurement errors, so high-quality instrumentation and careful technique are essential.
• Numerical stability: matrix inversion can become ill-conditioned when correction planes are too close together, when the chosen measurement locations fail to capture the rotor’s response, or when trial weights produce only marginal vibration changes.
• Time and cost: each extra plane adds another run, extending downtime and labour; for critical equipment this must be weighed against the gain in balance quality.
• Requires advanced software: solving N×N systems of complex vector equations is well beyond manual calculation, so specialised multi-plane balancing software is mandatory.

7. When to Use the N+2 Method

The method is appropriate when:

• The rotor is genuinely flexible: it operates above its first — and possibly its second or third — critical speed.
• The rotor is long and slender: a high length-to-diameter ratio means significant shaft bending in service.
• Two-plane balancing has proved insufficient: earlier two-plane attempts failed to reach an acceptable result.
• Multiple critical speeds must be traversed during normal operation.
• The equipment is high-value: critical turbines, compressors or generators where comprehensive balancing is justified.
• Vibration is severe at intermediate locations, between the end bearings, signalling mid-span unbalance that end-plane correction cannot reach.

8. Alternative: Modal Balancing

For the most flexible rotors, modal balancing can outperform the conventional N+2 approach. Rather than minimising vibration at specific speeds, modal balancing targets specific vibration modes one at a time, exploiting the rotor’s mode shapes to achieve a result with fewer trial runs. The trade-off is that it demands an even deeper grasp of rotor dynamics and more sophisticated analysis. In practice the two philosophies are often blended — modal insight guides where the planes go, and the influence-coefficient solution refines the masses.

9. Best Practices for Success

Planning

• Choose the N correction-plane locations carefully — widely spaced, accessible, and ideally aligned with the rotor’s mode-shape antinodes, since a weight placed at a node has little effect on that mode.
• Select M ≥ N measurement locations that adequately capture the rotor’s vibration behaviour.
• Plan for thermal-stabilisation time between runs.
• Prepare trial weights and installation hardware in advance.

Execution

• Hold operating conditions — speed, temperature, load — absolutely consistent across all N+2 runs.
• Use trial weights large enough to produce a clear, measurable response, typically a 25–50% change in vibration.
• Take several measurements per run and average them to suppress noise.
• Document every trial weight’s mass, angle and radius.
• Verify phase-measurement quality, because phase errors are magnified in large matrix solutions.

Analysis

• Review the influence coefficient matrix for anomalies or unexpected patterns.
• Check the matrix condition number — high values warn of numerical instability.
• Confirm the calculated corrections are physically reasonable, neither absurdly large nor negligibly small.
• Consider simulating the expected final result before committing the corrections.

10. Practical Field Application and the Balanset-1A

Most flexible-rotor balancing on critical machines is done in situ at operating speed, where the rotor actually bends, rather than on a low-speed balancing machine. A portable two-channel analyser such as the Balanset-1A supplies the building blocks the N+2 method needs: synchronised 1× amplitude-and-phase measurement at each bearing, automatic computation of influence coefficients from the trial-weight runs, and verification of the residual unbalance after the corrections are installed. For two-plane jobs the instrument runs the full influence-coefficient solution directly; for more planes its single- and two-plane measurements serve as the disciplined per-plane data that a multi-plane solver combines. Because the work happens in the machine’s own bearings, the captured response includes the real support stiffness and thermal state the rotor runs in.

11. Integration with Other Techniques

The N+2 method can be combined with complementary approaches:

• Speed-stepped balancing: repeat the N+2 measurements at several speeds to optimise balance across the whole operating range, not just one speed.
• Hybrid modal–conventional: use modal analysis to inform correction-plane selection, then apply the N+2 method to size the weights.
• Iterative refinement: perform the full N+2 balance, then reuse a reduced set of influence coefficients for quick trim balancing as conditions drift in service.

---

← Back to Main Index