5. Vibration Isolation, Absorption, and Control

Introduction and Context

Mechanical vibration is not always undesirable. In some machines it is necessary for operation, as in vibratory conveyors, ultrasonic equipment, and material-testing systems. In most rotating and reciprocating machinery, however, vibration represents unwanted motion, fluctuating force, noise, fatigue damage, reduced accuracy, and shortened service life. The engineering problem is therefore not simply to eliminate vibration, but to control its transmission and response at acceptable cost and with adequate reliability.

Three related strategies are central to this task. Isolation reduces the transmission of vibration between a source and a receiver. Absorption introduces an auxiliary dynamic element that draws energy from the primary system, usually near a selected frequency. Control modifies system behaviour through passive, semi-active, or active means. The appropriate strategy depends on excitation frequency, operating speed, available space, environmental conditions, required bandwidth, allowable motion, maintenance capability, and consequences of failure.

This chapter develops the analytical and design principles used for vibration isolators, tuned mass absorbers, dynamic vibration absorbers, damping treatments, and modern control systems. The discussion assumes familiarity with single-degree-of-freedom vibration, frequency response, resonance, damping, and matrix methods.

Learning Objectives

After studying this chapter, the reader should be able to:

1. Distinguish force isolation from motion isolation and select suitable performance measures.

2. Derive and interpret the transmissibility of damped isolation systems.

3. Design isolators using stiffness, damping, static deflection, load capacity, and frequency-ratio criteria.

4. Explain the operation and limitations of tuned mass and dynamic vibration absorbers.

5. Compare passive, semi-active, and active vibration-control systems.

6. Evaluate damping treatments and identify practical sources of non-ideal behaviour.

7. Integrate performance, stability, robustness, cost, and safety in engineering decisions.

8. Apply these principles to an industrial vibration-control problem.

The Design Problem: Preventing Motion or Preventing Force?

A vibration-control design begins by identifying the disturbance path. A vibrating machine mounted on a flexible support may transmit force to a foundation. Conversely, a precision instrument mounted near a reciprocating machine may experience motion transmitted from the floor. These are different problems even when the same spring and damper are used.

In force isolation, the objective is to reduce the dynamic force transmitted from a vibrating source to its support. In motion isolation, the objective is to reduce the motion transmitted from a vibrating support to a protected mass. The frequency ratio is commonly written as

r = omega/omega_n,

where omega is excitation frequency and omega_n is the undamped natural frequency. For a single-degree-of-freedom system with damping ratio zeta, the force transmissibility is

T_F = sqrt[1 + (2 zeta r)^2] / sqrt[(1-r^2)^2 + (2 zeta r)^2].

For harmonic base excitation, the absolute displacement transmissibility is

T_X = sqrt[1 + (2 zeta r)^2] / sqrt[(1-r^2)^2 + (2 zeta r)^2].

The same algebraic form appears for these commonly defined ratios, although the physical interpretation differs. In practical work, the engineer must state whether the numerator represents transmitted force, absolute response, or relative motion.

Isolation begins only when transmissibility falls below unity. For the ideal force-isolation expression, this condition is approximately

r > sqrt(2).

Thus, simply making a support flexible does not guarantee isolation. Near resonance, flexibility may greatly amplify motion and transmitted force. A low natural frequency is desirable only if the machine can pass safely through the resonance region during start-up and shutdown and if static deflection, stability, and alignment remain acceptable.

Passive Isolation and the Role of Damping

A passive isolator contains no externally powered control actuator. Its behaviour is determined by its stiffness, damping, geometry, and material properties. Common devices include steel springs, elastomeric mounts, air springs, wire-rope isolators, laminated rubber bearings, and viscoelastic layers.

For a machine of mass m supported by stiffness k, the natural frequency is

omega_n = sqrt(k/m).

If damping is represented by coefficient c, the critical damping is

c_c = 2 sqrt(km),

and the damping ratio is zeta = c/c_c.

Damping has two competing effects. It reduces the resonant peak and limits motion while the system passes through resonance. At high frequency ratios, however, viscous damping can increase the force transmitted through the support because the damper directly carries a force proportional to velocity. Consequently, an isolator designed for a fixed operating frequency often uses less damping than one designed for frequent speed changes or severe transient excitation.

The transmissibility curve illustrates the principal compromise. Increasing zeta lowers and broadens the resonance peak, but it also delays the onset of highly effective high-frequency isolation. A design with very small damping may provide excellent isolation at the operating speed but experience unacceptable movement at start-up. A design with excessive damping may appear stable and controlled yet transmit too much high-frequency vibration.

Static Deflection as a Design Measure

For a vertical system, static deflection is delta_s = mg/k. Since omega_n^2 = k/m, it follows that

omega_n = sqrt(g/delta_s).

This relationship is useful because isolator selection is often specified in terms of static deflection. A lower natural frequency requires greater static deflection. The designer must therefore check available travel, spring buckling, load distribution, levelling, and geometric stability. The ideal linear model also assumes that the support remains within its elastic and geometrically stable range.

Practical Isolator Selection

Selection should consider the following sequence:

• Define source frequency, speed range, disturbance amplitude, and transient conditions.
• Determine whether the critical objective is force reduction, receiver motion reduction, or both.
• Estimate the required natural frequency and corresponding stiffness.
• Calculate static load per mount and verify allowable compression, shear, and lateral motion.
• Select damping appropriate to resonance-crossing requirements.
• Check high-frequency behaviour, mount resonances, structural flexibility, and installation constraints.
• Provide restraints or snubbers where excessive displacement could create a safety hazard.

Real isolators are not ideal point springs. Elastomers show frequency-, temperature-, amplitude-, and ageing-dependent stiffness and damping. Air springs require pressure regulation and may have rolling-lobe or membrane modes. Steel springs have low inherent damping and may need supplemental dampers. A support system can also possess rocking and torsional modes that are absent from a single-degree-of-freedom calculation.

Isolation of Rotating and Reciprocating Machines

For a machine with rotating unbalance U = me, the harmonic excitation force has amplitude F_0 = U omega^2. At low speed, the force is small; at high speed, the unbalance force increases with the square of speed. The transmitted force is determined by both the excitation amplitude and the transmissibility. Consequently, an isolator that performs well at one speed may produce a larger absolute transmitted force at another speed even when its transmissibility is lower.

A machine foundation should be treated as a system rather than as an isolated mass alone. Foundation flexibility, soil compliance, anchor stiffness, piping, electrical connections, and attached ducts may provide unintended transmission paths. Flexible mounts are ineffective if a rigid pipe bridge bypasses them. This is a frequent industrial design error: the nominal isolator is correctly selected, but the installation creates a parallel stiffness much greater than the mount stiffness.

Absorption: Tuned Mass and Dynamic Vibration Absorbers

An absorber adds a secondary mass-spring system to a vibrating primary structure. If the absorber is tuned near the disturbing frequency, energy is transferred from the primary system to the absorber, reducing the primary response. In the ideal undamped case, a primary mass m carrying harmonic force F_0 cos omega t is fitted with an absorber of mass m_a and stiffness k_a. The absorber natural frequency is

omega_a = sqrt(k_a/m_a).

When omega equals omega_a, the absorber mass moves substantially and generates an inertial force that opposes the excitation force on the primary mass. At the exact tuning frequency, the ideal primary response can theoretically be zero. This result is powerful but narrow: it assumes accurate tuning, linear behaviour, negligible friction, and a disturbance frequency that remains fixed.

The absorber mass ratio, mu = m_a/m, influences performance. A larger mass ratio generally permits a wider useful frequency range and lower peak response, but increases mass, cost, support requirements, and space. In the undamped tuned absorber, the original resonance is replaced by two nearby resonances. Slight mistuning can therefore produce large response near these new peaks.

Damped Dynamic Absorbers

Adding damping to the absorber reduces sensitivity to mistuning and limits absorber motion. It also prevents the ideal zero-response condition and may transmit more force through the auxiliary system. The design becomes an optimization problem involving primary response, absorber travel, bandwidth, damping, and robustness.

A useful design procedure is:

1. Identify the dominant excitation frequency and its stability over time.

2. Estimate the allowable primary response and available absorber mass.

3. Select a mass ratio consistent with structural and packaging constraints.

4. Tune the absorber near the dominant excitation frequency.

5. Choose absorber damping to control the split resonance peaks.

6. Compute frequency-response functions for expected mistuning and parameter variation.

7. Verify fatigue, stroke, attachment strength, and fail-safe behaviour.

The absorber must be attached to a location where the primary mode has significant motion. Attaching it near a nodal point produces little effect. For multi-mode structures, a single absorber may suppress one mode while leaving other modes unchanged or even producing undesirable interactions.

Tuned Mass Dampers in Structures and Machinery

Tuned mass dampers are used in flexible floors, bridges, towers, machine tools, vehicle bodies, and rotating equipment. Their effectiveness depends on modal participation. A design based only on the global natural frequency may fail if temperature, payload, boundary conditions, or operating conditions shift the mode. Adjustable or adaptive absorbers are therefore attractive where frequency variation is substantial.

Damping Treatments and Energy Dissipation

Damping is the mechanism by which vibration energy is dissipated. It may arise from material hysteresis, friction, viscous fluid flow, micro-slip at joints, acoustic radiation, or electromagnetic effects. In structural treatment, viscoelastic layers are bonded to a vibrating panel or used in constrained-layer constructions. When the panel bends, the viscoelastic material undergoes shear deformation and dissipates energy.

A simple loss-factor description uses complex stiffness:

K* = K(1 + i eta),

where eta is the loss factor. The exact frequency and temperature dependence must be obtained from material characterization. A material that performs well at one temperature may become too soft at a higher temperature or too stiff at a lower temperature. Treatment thickness, coverage, attachment quality, and strain distribution strongly affect the result.

Friction dampers and dry interfaces can dissipate substantial energy but may exhibit stick-slip, amplitude dependence, noise, and wear. Hydraulic dampers provide controllable energy dissipation but require seals, fluid compatibility, and maintenance. Damping should not be treated as a universal remedy: if the dominant problem is a structural resonance excited by a stable frequency, an absorber or stiffness modification may be more effective than simply adding damping.

Passive, Semi-Active, and Active Control

Passive Control

Passive systems include springs, dampers, isolators, absorbers, tuned liquid devices, constrained layers, and friction interfaces. They are valued for simplicity, low energy consumption, predictable failure modes, and low maintenance. Their principal limitation is that their properties are fixed or only weakly adjustable. A passive device must therefore be designed for a range of operating conditions rather than for a single ideal point.

Semi-Active Control

Semi-active systems alter their effective properties but do not generally inject net mechanical energy into the structure. Examples include magnetorheological dampers, controllable hydraulic valves, variable-orifice dampers, electrorheological devices, and adjustable stiffness mechanisms. A controller changes the resistance or stiffness in response to measured motion. Because the actuator primarily modulates dissipation, semi-active systems can offer adaptability with lower instability risk than fully active systems.

The control law must account for actuator limits, sensor noise, time delay, and hysteresis. A command to increase damping is useful only if the device can respond rapidly and consistently. Semi-active devices also require a suitable power supply and a safe default state.

Active Control

Active control uses sensors, a controller, and actuators capable of applying forces or moments. The measured response may be displacement, velocity, acceleration, strain, force, or a combination. The controller compares the measured state with a desired condition and generates an actuator command.

A state-space model can be written as

x_dot = Ax + Bu,

y = Cx + Du,

where x is the state vector, u is the control input, and y is the measured output. With state feedback, u = -Kx, the closed-loop dynamics become

x_dot = (A - BK)x.

The control objective is to place or shape the closed-loop poles, reduce a specified response, or optimize a cost function. In practice, not all states are measured, so an observer or estimator may be required. Robust design is essential because the model is uncertain and the actuator cannot generate unlimited force or stroke.

Active control can achieve broadband attenuation and adapt to changing disturbances, but it introduces failure modes involving sensors, software, power electronics, communication delay, saturation, and incorrect models. A mechanical system must remain safe if the controller is switched off. This principle is particularly important in safety-critical machinery.

Stability, Robustness, and Performance Trade-offs

A vibration-control design should not be judged by peak attenuation alone. Important performance measures include root-mean-square response, peak response, transmitted force, control energy, actuator stroke, settling time, acoustic output, and sensitivity to parameter variation. A controller that gives excellent nominal attenuation but becomes unstable after a small stiffness change is not a successful engineering solution.

Robustness refers to acceptable behaviour under uncertain mass, stiffness, damping, excitation frequency, temperature, ageing, and sensor error. Frequency-response analysis, sensitivity functions, Monte Carlo simulation, and experimental modal testing are useful tools. In active systems, gain and phase margins provide classical stability indicators, while robust-control methods can explicitly represent model uncertainty.

There is often a conflict between isolation and static support. Very soft mounts reduce high-frequency transmission but permit large static displacement and low-frequency motion. High damping improves transient control but may reduce high-frequency isolation. A large absorber mass increases bandwidth but adds structural weight. Active control improves adaptability but increases cost and dependence on electronics. Good design makes these compromises explicit rather than hiding them in a single performance number.

An Illustrative Industrial Case: Isolation of a Precision Grinding Machine

Background

Consider a precision grinding machine installed on a factory floor shared with heavy presses. The grinding spindle operates at a nominal speed of 3,600 rpm, corresponding to 60 Hz. Measurements show that surface-finish errors increase when nearby presses operate, and vibration is also visible during spindle start-up.

Situation and Analysis

Operational measurements identify two disturbance paths. The machine itself generates a 60 Hz rotating-unbalance component, while floor vibration from the presses contains energy between 12 and 25 Hz. The grinding machine and its support have a rigid-body vertical natural frequency near 9 Hz when mounted on proposed elastomeric isolators.

The frequency ratio for the spindle disturbance is approximately r = 60/9 = 6.67, which is within a potentially effective isolation region. The floor disturbance, however, lies closer to the support resonance and may be amplified. A design based only on the spindle frequency would therefore be inadequate.

The machine mass, centre of gravity, mount locations, and allowable levelling movement are used to calculate vertical, rocking, and lateral modes. A rigid bypass is found in the coolant pipe and is redesigned with a flexible loop. The electrical conduit is similarly supported so that it does not short-circuit the isolators.

Approach and Solution

The selected system uses four elastomeric mounts with moderate damping, supplemented by lateral restraints. The mount stiffness is chosen to maintain a support frequency below the dominant spindle frequency while keeping static compression within the manufacturer’s allowable range. Because the machine must pass through resonance at start-up, excessive damping is avoided; controlled damping is retained to limit transient movement.

An accelerometer is installed on the machine frame and another on the floor. A frequency-response comparison is used to verify isolation under several operating conditions. During commissioning, the measured support frequency is higher than predicted because of mount preload and floor flexibility. The model is updated, and the machine is re-levelled.

Outcome

The final arrangement reduces vibration transmitted from the machine to the floor and lowers the vibration entering the grinding structure from the surrounding floor. The principal improvement does not come from the mounts alone. It results from treating the complete transmission path, including piping, conduits, foundation modes, and operating transients.

Learning Points

The case demonstrates that isolation is a system-level problem. Nominal mount specifications do not guarantee performance. Frequency separation, installation details, structural modes, measurement, and commissioning are equally important. It also illustrates why a design must consider both operating frequency and the resonance-crossing region.

Common Mistakes, Risks, and Professional Considerations

A frequent error is selecting an isolator solely from static load capacity. Capacity does not establish dynamic performance. Other mistakes include ignoring rocking modes, assuming damping is always beneficial, neglecting parallel stiffness, tuning an absorber to an uncertain frequency, and failing to check absorber stroke.

Active systems introduce additional professional responsibilities. Sensor and actuator failure must be considered in hazard analysis. Software changes should be controlled and validated. Control actions must not defeat mechanical guards, overload structures, or create unsafe motion. Where machinery is governed by occupational safety, pressure-vessel, lifting, building, or machinery regulations, the vibration-control design must comply with applicable standards and documented verification procedures.

Environmental factors such as temperature, oil contamination, ultraviolet exposure, corrosion, and ageing can change isolator properties. Procurement specifications should therefore state test conditions, allowable tolerances, qualification procedures, and acceptance criteria. Field measurements should use calibrated instrumentation and documented sensor locations.

Digital Measurement and Emerging Directions

Modern vibration-control practice increasingly combines operational modal analysis, digital signal processing, finite-element modelling, and experimental validation. Order tracking separates speed-related components from broadband disturbances. Digital twins can be used to evaluate parameter changes, although their predictions remain dependent on model quality. Machine-learning methods may assist anomaly detection or estimate changing system parameters, but they should not replace physical safety limits or engineering validation.

Adaptive tuned absorbers, variable-stiffness mounts, inerter-based systems, and self-sensing materials are active areas of development. The inerter produces a force proportional to relative acceleration and can create useful dynamic properties without adding a correspondingly large physical mass. Such devices may improve performance, but their benefits must be weighed against complexity, nonlinearities, cost, and limited field experience.

Best-Practice Design Workflow

A defensible vibration-control design normally follows this sequence:

1. Define the source, receiver, disturbance spectrum, operating envelope, and consequences of vibration.

2. Identify all transmission paths, including structural, fluid, electrical, and acoustic paths.

3. Build an appropriate model, beginning with the simplest model that captures the important modes.

4. Select passive isolation, absorption, damping, semi-active, or active control according to frequency variability and risk.

5. Evaluate nominal performance and sensitivity to uncertain parameters.

6. Check static strength, dynamic fatigue, travel, stability, thermal effects, maintenance, and failure modes.

7. Validate predictions with laboratory or field measurements.

8. Commission the system under normal, transient, and fault conditions.

9. Monitor performance and revise the model when operating conditions change.

Chapter Summary

Vibration isolation reduces transmission between a source and a receiver, while absorption uses an auxiliary dynamic system to draw energy from a primary structure. Passive isolator performance depends on natural frequency, damping, transmissibility, static deflection, and installation details. Damping reduces resonance but can increase high-frequency force transmission. Tuned mass and dynamic absorbers provide strong narrowband suppression, though their performance is sensitive to mistuning, absorber stroke, and mass ratio.

Passive systems are simple and reliable; semi-active systems adapt dissipation or stiffness with comparatively low energy input; active systems provide broader control but require careful attention to stability, sensing, actuation, uncertainty, and fail-safe operation. Effective engineering decisions balance attenuation, bandwidth, cost, robustness, safety, maintenance, and structural constraints. Measurement and treatment of the complete transmission path are essential to successful industrial application.

Key Terms

Vibration isolation: Reduction of vibration transmitted between a source and a receiver.

Force isolation: Reduction of dynamic force transmitted from a vibrating machine to its support.

Motion isolation: Reduction of motion transmitted from a vibrating base to a protected system.

Transmissibility: Ratio describing transmitted response relative to excitation response.

Frequency ratio: Ratio of excitation frequency to system natural frequency.

Damping ratio: Actual damping divided by critical damping.

Static deflection: Equilibrium displacement caused by the supported weight.

Passive control: Vibration control using fixed mechanical properties and no externally powered force generation.

Tuned mass absorber: Auxiliary mass-spring system tuned to suppress a selected excitation frequency.

Dynamic vibration absorber: Secondary oscillator that transfers vibratory energy away from a primary system.

Mass ratio: Ratio of absorber mass to primary vibrating mass.

Mistuning: Difference between absorber tuning and the actual disturbance or structural frequency.

Viscoelastic damping: Energy dissipation through rate-dependent deformation of a viscoelastic material.

Loss factor: Dimensionless measure of energy dissipated relative to stored strain energy.

Semi-active control: Control that changes system properties, usually damping or stiffness, without directly supplying substantial mechanical energy.

Active control: Control using powered actuators to apply corrective forces or moments.

Robustness: Ability to maintain acceptable performance despite uncertainty and variation.

Parallel stiffness: Unintended or additional stiffness path acting alongside the intended isolator.

Actuator saturation: Condition in which an actuator reaches its maximum force, displacement, or velocity.

Fail-safe design: Design that remains acceptably safe when a component or control function fails.

Review Questions

Short Answer Questions

1. Distinguish force isolation from motion isolation.

2. Why can excessive damping reduce high-frequency isolation performance?

3. Define absorber mass ratio and explain its significance.

4. What is mistuning in a dynamic vibration absorber?

5. State two reasons why an isolator may fail to perform as predicted after installation.

Long Answer Questions

1. Derive and interpret the transmissibility of a damped single-degree-of-freedom isolation system.

2. Explain the design procedure for a tuned mass absorber, including the effects of mass ratio and damping.

3. Compare passive, semi-active, and active vibration-control systems with respect to performance, cost, reliability, and stability.

4. Discuss the relationship between static deflection, natural frequency, damping, and practical isolator selection.

5. Explain why vibration control must consider the complete mechanical transmission path rather than the isolator alone.

Analytical / Application Questions

1. A machine operates at 60 Hz on mounts having a natural frequency of 9 Hz. Assess whether the operating condition lies in a potentially effective isolation region and identify the transient risk during start-up.

2. A tuned absorber substantially reduces vibration at one operating speed but increases vibration at nearby speeds. Diagnose the likely cause and propose two design modifications.

3. An active control system gives excellent laboratory attenuation but becomes unstable after the machine is fitted with flexible piping. Explain the modelling and robustness issues involved and recommend a validation procedure.

22. Multiple Choice Questions with Answers

Question 1

For an ideal single-degree-of-freedom isolator, effective high-frequency isolation generally begins when the frequency ratio is approximately greater than:

A. 0.25

B. 0.707

C. 1.0

D. 1.414

Correct Answer: D. 1.414

Question 2

Increasing viscous damping in an isolator generally has which combined effect?

A. It increases the resonant peak and improves all high-frequency isolation

B. It reduces the resonant peak but may increase high-frequency force transmission

C. It eliminates the natural frequency

D. It has no effect on transmissibility

Correct Answer: B. It reduces the resonant peak but may increase high-frequency force transmission

Question 3

The primary function of a tuned mass absorber is to:

A. Increase the static load capacity of the primary structure

B. Transfer vibratory energy to an auxiliary oscillator near a selected frequency

C. Raise every natural frequency of the structure equally

D. Remove the need for structural damping

Correct Answer: B. Transfer vibratory energy to an auxiliary oscillator near a selected frequency

Question 4

If the static deflection of a vertical isolator is increased, its natural frequency will generally:

A. Increase

B. Remain unchanged

C. Decrease

D. Become equal to the excitation frequency

Correct Answer: C. Decrease

Question 5

A major disadvantage of an undamped tuned absorber is its:

A. Inability to respond at its tuned frequency

B. Sensitivity to mistuning and nearby resonance peaks

C. Requirement for a very high damping ratio

D. Independence from absorber mass

Correct Answer: B. Sensitivity to mistuning and nearby resonance peaks

Question 6

Which condition most directly creates a parallel stiffness that bypasses a machine isolator?

A. A flexible electrical cable

B. A rigidly connected pipe attached to the isolated machine

C. A calibrated accelerometer

D. A correctly tuned absorber

Correct Answer: B. A rigidly connected pipe attached to the isolated machine

Question 7

A semi-active vibration-control device differs from a fully active device primarily because it:

A. Cannot use sensors

B. Changes its properties but generally does not inject substantial mechanical energy

C. Always operates without a power supply

D. Has no stability or saturation constraints

Correct Answer: B. Changes its properties but generally does not inject substantial mechanical energy

Question 8

In active feedback control, the closed-loop state matrix for u = -Kx is:

A. A + BK

B. A - BK

C. B - AC

D. C - DK

Correct Answer: B. A - BK

Question 9

In the illustrative grinding-machine case, why was flexible piping necessary?

A. To increase the machine's operating speed

B. To provide additional rigid support

C. To prevent the piping from bypassing the isolators through a rigid transmission path

D. To tune the spindle absorber

Correct Answer: C. To prevent the piping from bypassing the isolators through a rigid transmission path

Question 10

Which design choice most directly improves absorber robustness to frequency variation?

A. Eliminating all absorber damping

B. Reducing absorber stroke to zero

C. Adding suitable damping or using an adjustable absorber

D. Attaching the absorber at a modal node

Correct Answer: C. Adding suitable damping or using an adjustable absorber