Dynamic Balancing Machine

What is a Dynamic Balancing Machine

A dynamic balancing machine is a precision instrument that rotates a workpiece at a defined balancing speed and measures the centrifugal effect of any unevenly distributed mass. Every real rotor carries some unbalance: the geometric center of mass does not lie exactly on the axis defined by its journals, and the principal axis of inertia is slightly tilted relative to that geometric axis. When the part spins, this error produces a rotating force and a rotating couple that load the bearings, generate vibration and noise, and shorten machine life. The balancing machine quantifies that error and tells the operator the magnitude and angle of correction needed in each chosen plane.

The term dynamic distinguishes the machine from a static balancer. Static balancing only locates the single heavy spot that makes a non-rotating rotor roll under gravity, and it can correct mass in one plane. Dynamic balancing requires rotation because couple unbalance, the kind that produces a rocking moment, is invisible at rest and only reveals itself as a force when the part turns. A dynamic machine therefore measures in two planes at once and resolves the result into two correction weights, one at each end of the rotor. Any rotor whose axial length is significant compared with its diameter must be balanced dynamically.

The physics dates to the late nineteenth and early twentieth centuries. The Canadian inventor Henry Martinson filed the first balancing patent in 1870, and around the turn of the century Akimoff in the United States and Stodola in Switzerland worked to apply such methods industrially as steam turbines and electric motors demanded smoother rotation. Building on a 1907 patent by Franz Lawaczek, the German firm Carl Schenck of Darmstadt developed the first industrial two-plane balancing machine and acquired exclusive worldwide rights in 1915, and it remains a reference name in production balancing today. The introduction of electronic instrumentation in the mid twentieth century, and microprocessor measuring units from the 1980s onward, transformed the machine from a mechanical null-seeking device into a calibrated instrument that prints a correction in grams and degrees within a single run.

The scale of application is broad. Production balancing machines handle rotors from a few grams, such as dental-drill turbines and miniature armatures, up to tens of thousands of kilograms, such as large turbine and generator rotors and paper-mill rolls; catalog machines from suppliers like CIMAT and CEMB span roughly 50 kg to 20,000 kg or more in a single product family. Service speeds during balancing range from a few hundred to several thousand revolutions per minute, chosen relative to the machine's support resonance rather than the rotor's final operating speed.

Why it matters commercially: residual unbalance is the single largest source of avoidable vibration in rotating machinery. Excess vibration accelerates bearing and seal wear, loosens fasteners, cracks welds through fatigue, and radiates noise. A correctly balanced rotor to the right ISO 21940-11 grade can extend bearing life several-fold and is a prerequisite for meeting the broadband vibration limits that machine buyers impose, and the same residual-unbalance signature is later tracked in service by a condition monitoring system. The balancing machine is therefore both a quality gate on the production line and a diagnostic tool in the maintenance shop.

Balancing Machine Types

Balancing machines are classified along three axes that buyers must specify together: rotor orientation (horizontal or vertical), the number of correction planes (single-plane or two-plane), and whether the support is hard-bearing or soft-bearing. The first two are covered here; the bearing principle has its own chapter because it drives accuracy and workflow. The table below summarizes the orientation and plane combinations and the rotor shapes each suits best.

Horizontal machines hold the rotor on two support cradles with its axis horizontal, the rotor resting on its own journals or on supplied roller carriages. This is the dominant configuration for general-purpose work, because most industrial rotors are shaft-mounted parts whose length exceeds their diameter and which therefore require true two-plane dynamic balancing. Horizontal machines accept the widest range of part sizes and are the natural choice for crankshafts, cardan shafts, electric-motor rotors, gearbox shafts, pump shafts, and turbine spools.

Vertical machines mount the rotor on a vertical spindle and rotary fixture. They are preferred for disc-shaped parts that have no shaft of their own, such as flywheels, clutch plates, brake drums, fan and blower wheels, and grinding wheels, where clamping to a vertical table is mechanically simpler than supporting on horizontal cradles. A vertical single-plane machine handles thin discs that need only static correction; a vertical two-plane machine handles parts with enough axial height to develop couple unbalance.

Single-plane versus two-plane is the most consequential choice. A single-plane (static) machine corrects mass in one plane only and is valid where the part is thin enough that couple unbalance is negligible, the common rule of thumb being a diameter-to-width ratio above roughly 7 to 1 combined with a modest service speed. A two-plane (dynamic) machine measures and corrects in two planes and is mandatory for everything else, because a part that is statically perfect can still carry a couple that rocks the bearings at speed. When in doubt, specify two-plane: a two-plane machine can always perform a single-plane correction, but not the reverse.

A further sub-classification is by automation level. General-purpose or universal machines accept a wide family of rotors with manual loading and operator-applied correction, and a single universal machine commonly spans a 100-to-1 weight ratio from its largest to smallest acceptable rotor. Dedicated or special-purpose machines are built around one part family (for example crankshafts or armatures) and often integrate measurement with automatic correction by drilling, milling, or grinding to maximize cycle time and repeatability in high-volume production.

Hard-Bearing vs Soft-Bearing Principles

The single most important architectural decision in a balancing machine is whether the rotor supports are hard-bearing or soft-bearing. The names refer to the stiffness of the suspension relative to the balancing speed, and the choice determines whether the machine measures force or displacement, whether it needs trial-weight calibration for each new rotor, and how robust it is on the shop floor. The table below contrasts the two architectures on the parameters that matter to a buyer.

Hard-bearing machines use rigid support structures whose own resonance frequency sits well above the balancing speed, so the rotor is balanced below resonance. In this regime the support barely moves and the machine measures the centrifugal force the unbalance transmits into the bearings. That force is a predictable function of the rotor's mass and the axial positions of its correction planes and journals, which means a hard-bearing machine can be calibrated permanently from a few dimensions entered into the measuring unit. The operator mounts a new rotor type, enters its geometry, and reads a correct unbalance value on the very first run with no trial weights. This single-run capability, combined with mechanical robustness and long-term stability, is why hard-bearing machines dominate production balancing.

Soft-bearing machines use a compliant suspension, typically a pendulum or low-stiffness flexure, whose resonance frequency is well below the balancing speed, so the rotor is balanced above resonance. Here the suspension swings freely under the centrifugal force, and the machine measures the amplitude of that vibration, typically with a pickup such as an accelerometer or velocity sensor, rather than force. Because the relationship between vibration amplitude and unbalance depends on the dynamics of each particular rotor and suspension combination, the machine cannot read unbalance directly: each new rotor type must first be characterized by adding known trial weights to establish a baseline, which adds setup runs. The advantage is very high sensitivity to small unbalances, which makes soft-bearing machines well suited to extremely light rotors and to certain flexible-rotor work.

The practical consequence for a buyer is workflow. For a job shop or production line that balances many different rotor types in short runs, the hard-bearing machine's permanent calibration removes the trial-weight setup penalty and is almost always the right answer. For a laboratory balancing a small number of very light or very precise rotors where ultimate sensitivity matters more than throughput, a soft-bearing machine can be justified. Modern measuring units, such as Schenck's CAB 920, run either architecture and present the operator with the same correction output in grams and degrees, so the bearing choice is about the mechanics and the calibration model, not the user interface.

A related concept is rigid versus flexible rotor behavior. ISO 21940-11 applies to rotors that behave rigidly at their service speed, meaning they do not deform appreciably under unbalance forces. Rotors that operate above their first bending critical speed deform, and balancing them requires multi-plane and multi-speed methods covered separately by ISO 21940-12. Most production balancing machines target rigid-rotor work; specifying flexible-rotor capability is a distinct and more demanding requirement.

Unbalance Types, Correction, and Standards

To specify a balancing machine you must first understand what it is correcting. ISO 21940-11 distinguishes four fundamental unbalance conditions, defined by the relationship between the rotor's principal axis of inertia and its shaft (rotation) axis. The four conditions determine whether single-plane or two-plane balancing is required.

Static unbalance shifts the center of mass off the axis while keeping the principal inertia axis parallel to it; it produces a single rotating force and can be corrected in one plane. Couple unbalance leaves the center of mass on the axis but tilts the principal axis so the two ends are heavy on opposite sides; it produces a rocking moment and is invisible at rest, requiring two-plane correction. Quasi-static and dynamic unbalance are combinations of the two; dynamic unbalance is the general real-world case and the reason most rotors need a two-plane machine. A dynamic machine measures the combined effect and resolves it mathematically into one correction weight per plane.

Correction is the physical step of adding or removing mass at the calculated angle in each plane. Mass removal methods include drilling, milling, grinding, and laser ablation, used on cast or machined rotors where material can be safely taken out. Mass addition methods include bolting or welding correction weights, riveting balance clips onto fan blades, and applying balancing putty or epoxy on lighter parts. High-volume dedicated machines integrate the correction tool directly, so the part is measured and drilled in one fixture; universal machines report the correction and the operator applies it offline.

Tolerances come from the balance quality grade G defined in ISO 21940-11, the current revision of the long-used ISO 1940-1. The grade is the product of the permissible specific unbalance and the service angular velocity, expressed in millimeters per second, so a numerically smaller grade is a finer balance. The permissible residual unbalance follows from the formula Uper = 9549 x G x m / n in gram-millimeters, where m is rotor mass in kilograms and n is maximum service speed in revolutions per minute. The table in Chapter 5 lists representative grades and their applications. For rotors balanced in their own bearings rather than on a machine, ISO 21940-13 governs field (in-situ) balancing procedures and tolerances.

The machine itself is graded by ISO 21940-21 (formerly ISO 2953), which defines acceptance tests using a calibrated proving rotor. The two headline metrics are the minimum achievable residual unbalance, Umar, the smallest unbalance the machine can reliably indicate, and the unbalance reduction ratio, URR, the fraction of a deliberate initial unbalance the machine removes in a single correction step. These are the numbers a buyer should demand on an acceptance certificate, because they describe what the machine can actually do on real parts, not just its catalog capacity.

Key Specification Parameters

Balancing-machine spec sheets list many numbers, but only a handful drive the selection decision: rotor weight range, maximum rotor diameter and length (swing and journal span), balancing speed range, drive type and power, minimum achievable residual unbalance (Umar), unbalance reduction ratio (URR), number of planes, and the measuring-unit features. The first comparison table below decodes the key specifications across machine sizes; the second maps balance quality grades to applications.

Rotor weight range is the span from the smallest to the largest rotor a machine can balance with rated accuracy; a universal machine typically covers a 100-to-1 ratio, so the minimum rotor matters as much as the maximum. Balancing a rotor far below the machine's design minimum sacrifices accuracy because the unbalance force falls toward the machine's noise floor. Maximum diameter (swing) and journal span set the physical envelope: the swing is the largest rotor diameter that clears the bed, and the journal-distance range must bracket your bearing-to-bearing dimension.

Balancing speed is selected relative to the support resonance, not the rotor's service speed; for rigid-rotor work the result is independent of speed within the rated range, so a moderate speed that develops adequate force without excess windage is preferred. Drive type and power follow the rotor: belt drive for small precise parts, universal-joint or end drive for large high-inertia parts, with drive power sized to overcome bearing friction and aerodynamic drag at speed.

Minimum achievable residual unbalance (Umar) is the machine's resolution floor, often expressed as specific unbalance in gram-millimeters per kilogram of rotor mass; a value near 0.1 g·mm/kg is representative of a good production machine, and it caps how fine a balance grade you can actually reach. Unbalance reduction ratio (URR), the fraction removed per correction step, determines how few runs a rotor needs: a 95 percent URR lets a rotor entering at 20 times tolerance reach tolerance in one correction. Demand both Umar and URR on the acceptance certificate, verified with a traceable proving rotor per ISO 21940-21.

The second table connects the balance quality grade to real machinery, so you can pick the grade your parts require before sizing the machine. Choosing a grade that is unnecessarily fine forces a more expensive machine and longer cycle times; choosing one too coarse leaves damaging vibration in the field.

Selection Decision Factors

To turn the preceding chapters into a purchase, work through the decision sequence below. The most common specification errors are sizing the machine to the largest rotor while ignoring the smallest, and accepting a catalog capacity without an Umar and URR guarantee on a proving rotor. These eight steps can serve as a fixed RFQ template.

1. Rotor envelope: Fix the weight range (both maximum and minimum), the maximum diameter (swing), the journal-to-journal span, and the rotor length. Remember that a universal machine spans roughly 100 to 1 in weight, so a single machine may not cover both your heaviest and lightest parts at full accuracy.
2. Number of planes: Decide single-plane (static) versus two-plane (dynamic) from the rotor's diameter-to-width ratio and service speed. When in doubt choose two-plane, because it can always do single-plane work, and most shaft-mounted rotors need it.
3. Bearing architecture: Choose hard-bearing for production and mixed-rotor work to get permanent calibration and single-run results; reserve soft-bearing for very light or specialized flexible rotors where ultimate sensitivity outweighs throughput.
4. Required balance grade: Set the ISO 21940-11 grade (G 6.3, G 2.5, G 1.0, and so on) from the application, then verify the machine's Umar is fine enough to reach it. Compute the permissible unbalance with Uper = 9549 x G x m / n for your real rotor.
5. Drive type and power: Belt drive for small precise rotors with a clean cylindrical surface; universal-joint or end drive for large, high-inertia, or shaft-end-driven parts. Consider a dual universal-joint-plus-belt machine if the part mix is wide.
6. Correction method and automation: Decide whether the machine only measures (operator corrects offline) or integrates automatic drilling, milling, or grinding for high-volume single-part-family production. Match the correction method to the rotor material.
7. Acceptance metrics: Demand a documented Umar and URR measured on a traceable proving rotor per ISO 21940-21 at acceptance, and confirm the supplier provides or specifies the proving rotor used.
8. Measuring unit and tooling: Evaluate the measuring electronics (for example Schenck CAB-class units), the rotor-setup memory, reporting and SPC outputs, plus the cost and lead time of mandrels, adapters, and roller carriages, which often add materially to the quoted price.

One last commonly overlooked dimension is serviceability and re-verification. A balancing machine must be re-checked periodically against its proving rotor to confirm Umar and URR have not drifted, so factor in calibration service availability, proving-rotor ownership, spare drive belts and pickups, and measuring-unit software support. Established suppliers such as Schenck RoTec, Hofmann, CEMB, CIMAT, Hines Industries, and IRD Balancing maintain service networks and proving-rotor programs; confirming local calibration support before purchase prevents a production line from sitting idle when annual re-verification comes due.