Damping Materials & Techniques

Every mechanical structure vibrates. When an optical table rings after a door slam, when a mirror mount oscillates after a positioning stage indexes, or when a laser cavity drifts during a building's HVAC cycle — the root cause is energy stored in structural modes that takes too long to dissipate. Damping is the mechanism that removes this energy.

In formal terms, damping is the irreversible conversion of mechanical vibrational energy into thermal energy within a structure or material. Without damping, a structure excited at one of its natural frequencies would oscillate indefinitely at ever-increasing amplitude. With sufficient damping, the same resonance decays to negligible levels within milliseconds.

1.1Damping vs. Isolation

Two distinct strategies address vibration in photonics systems, and confusing them leads to poor design decisions.

Vibration isolation decouples a payload from external disturbance sources by introducing a compliant boundary — a spring, an air bearing, or a flexure — between the vibration source and the sensitive system. Isolation reduces the transmission of energy across that boundary. It does nothing about energy already present within the isolated system.

Vibration damping dissipates energy that is already stored in structural modes. A damped optical table does not prevent floor vibrations from reaching the table surface (that is the isolator's job); it prevents vibrations generated on the table — by motorized stages, shutters, or vacuum lines — from persisting and amplifying at resonant frequencies.

Effective vibration control requires both: isolation to block external disturbances, and damping to suppress internal resonances. This chapter addresses damping. Isolation principles, isolator hardware, and optical table design are covered in companion topics within the Vibration Isolation category.

1.2Why Damping Matters in Photonics

Precision optical systems operate at dimensional tolerances where even nanometer-scale relative motion between components degrades performance. A Michelson interferometer with λ/10 path stability at 632.8 nm requires relative mirror displacement below 63 nm. A scanning confocal microscope resolving 200 nm features cannot tolerate stage vibration above ~20 nm peak-to-peak during acquisition. These tolerances are easily violated by undamped structural resonances.

The problem is particularly acute because the metals used in optical tables and mounts — stainless steel and aluminum — are excellent conductors of vibration but terrible at dissipating it. A bare stainless steel plate has a material loss factor on the order of η ≈ 0.0001–0.0006 [1, 2]. Without added damping, a table resonance excited at 200 Hz can ring for seconds, corrupting data and frustrating alignment.

1.3Chapter Roadmap

This chapter builds from fundamental damping parameters (Section 2) through the physics of viscoelastic materials (Section 3) and a survey of damping materials used in photonics (Section 4). Sections 5 and 6 cover the two principal passive surface treatments — free-layer and constrained-layer damping. Section 7 introduces tuned mass dampers, including the distributed systems used in high-performance optical tables. Section 8 examines how these techniques combine in modern optical table systems. Section 9 covers measurement and characterization methods, and Section 10 provides a practical decision workflow for selecting the right damping strategy.

Damping in a vibrating system can be quantified by several interrelated parameters. Each originates from a different measurement context or analytical framework, but all describe the same underlying phenomenon: the rate at which vibrational energy is converted to heat. Understanding these parameters and their interrelationships is essential for interpreting vendor data, comparing materials, and designing damping treatments.

2.1Loss Factor (η)

The loss factor is the most widely used parameter for characterizing material damping and is the standard metric in viscoelastic material datasheets. It is defined as the ratio of energy dissipated per cycle to the maximum strain energy stored during that cycle [2]:

Where: ΔW = energy dissipated per cycle (J), W_max = maximum strain energy stored per cycle (J).

For viscoelastic materials characterized by a complex modulus, the loss factor emerges directly as the ratio of the loss modulus to the storage modulus [2, 11]:

Where: E′ = storage modulus (Pa) — the elastic, energy-storing component, E″ = loss modulus (Pa) — the viscous, energy-dissipating component.

The loss factor is dimensionless, frequency-dependent for viscoelastic materials, and ranges from ~0.0001 for structural metals to ~0.5–1.0 for optimized damping elastomers [2, 3].

2.2Viscous Damping Ratio (ζ)

The viscous damping ratio is the standard parameter in vibration dynamics textbooks. It expresses the system's damping as a fraction of critical damping — the minimum damping required to prevent oscillatory free response [1]:

Where: c = viscous damping coefficient (N·s/m), k = stiffness (N/m), m = mass (kg), c_cr = critical damping coefficient = 2√(km).

For light damping (ζ < 0.1), the loss factor and damping ratio are related by [1, 2]:

The exact relationship, valid for all damping levels, is:

For most photonics applications, ζ < 0.05 and the approximate form introduces less than 0.1% error.

2.3Quality Factor (Q)

The quality factor is the reciprocal of the loss factor and is commonly used in resonator and filter contexts [1, 3]:

A high Q means low damping — a sharp, tall resonance peak. A low Q means high damping — a broad, attenuated peak. An undamped stainless steel table mode might exhibit Q > 1000; a well-damped table targets Q < 30.

2.4Logarithmic Decrement (δ)

The logarithmic decrement is measured from free-decay time-domain data — the natural ring-down of a structure after an impulse [1]:

Where: x_n, x_n+1 = amplitudes of successive peaks in free decay.

For light damping [1, 2]:

2.5Half-Power Bandwidth

The half-power bandwidth method extracts damping from the frequency response function (FRF) of a driven system. At the resonance peak, the response amplitude is maximum. The two frequencies where the response drops to 1/√2 of the peak (−3 dB) define the half-power bandwidthΔω [1, 2]:

Where: Δf = bandwidth between −3 dB points (Hz), f_n = natural frequency at the resonance peak (Hz).

This method is the most practical for extracting damping from experimental FRF data of optical tables and other structures.

2.6Interrelationship Summary

The damping materials used in photonics vibration control are overwhelmingly viscoelastic — they exhibit both elastic energy storage and viscous energy dissipation simultaneously. Understanding viscoelastic behavior is essential for selecting the right material and predicting its performance across the temperature and frequency ranges relevant to a given application.

3.1Elastic vs. Viscous vs. Viscoelastic

A purely elastic material (an ideal spring) stores all input energy and returns it completely upon unloading. A purely viscous material (an ideal dashpot) dissipates all input energy as heat, with stress proportional to strain rate rather than strain. Real damping materials fall between these extremes.

A viscoelastic material stores some energy elastically and dissipates the remainder through internal molecular friction. When subjected to cyclic loading, the stress response lags behind the applied strain by a phase angle δ_p (not to be confused with the logarithmic decrement). This phase lag is directly related to the loss factor [2, 11]:

The term tan δ (read “tan delta”) appears frequently in polymer datasheets and is identical to the loss factor η.

3.2Complex Modulus

The complete mechanical behavior of a viscoelastic material under harmonic loading is captured by the complex modulus [2, 11]:

Where: E* = complex Young's modulus (Pa), E′ = storage modulus (Pa) — in-phase component (stiffness), E″ = loss modulus (Pa) — quadrature component (damping), i = √(−1).

An analogous complex shear modulus G* = G′ + iG″ = G′(1 + iη) applies for shear deformation, which is the relevant mode in constrained-layer damping treatments.

The storage modulus determines the material's stiffness contribution to the composite structure. The loss modulus determines its damping contribution. Both vary with frequency and temperature.

3.3Frequency Dependence

Viscoelastic damping is inherently frequency-dependent. At very low frequencies, polymer chains have time to relax fully during each loading cycle — the material behaves more like a viscous liquid (high η but low E′). At very high frequencies, the chains cannot respond — the material behaves more like a glass (high E′ but low η). The peak loss factor occurs in a transition region between these extremes [2, 11].

For practical damping design, this means a material optimized for 100 Hz may perform poorly at 1000 Hz and vice versa. Vendor datasheets should always be consulted at the actual operating frequency, not just at the test frequency reported in headline specifications.

3.4Temperature Dependence and Glass Transition

Temperature affects viscoelastic behavior as profoundly as frequency. Every amorphous polymer has a glass transition temperature T_g at which it transitions from a hard, glassy state (high E′, low η) to a soft, rubbery state (low E′, low η). The loss factor peaks sharply at T_g, where the balance between elastic storage and viscous dissipation is optimal [2, 11].

This peak defines the material's useful damping temperature range. A material with T_g = −20°C will perform well in a 20°C laboratory (it is in its rubbery plateau above T_g), but its loss factor may be only a fraction of its peak value. A material with T_g = 25°C will deliver peak damping at room temperature but stiffen dramatically if the laboratory is cooled.

3.5Time-Temperature Superposition

Frequency and temperature effects on viscoelastic properties are not independent. The time-temperature superposition (TTS) principle states that increasing temperature has the same effect on the complex modulus as decreasing frequency — both allow more molecular relaxation per cycle [2, 11]. This allows construction of master curves: a single plot of E′, E″, andη spanning many decades of frequency, assembled from measurements at multiple temperatures using a shift factor a_T.

The Williams-Landel-Ferry (WLF) equation provides the shift factor for temperatures near T_g:

Where: a_T = horizontal shift factor, C₁, C₂ = material constants (typically C₁ ≈ 17.4, C₂ ≈ 51.6 K for many polymers when T_ref = T_g), T = temperature (K), T_ref = reference temperature (K).

Master curves are essential for selecting damping materials: they reveal whether the material's peak η coincides with the target frequency at the expected operating temperature.

3.6Dynamic Mechanical Analysis (DMA)

Dynamic mechanical analysis is the standard laboratory technique for characterizing viscoelastic materials. A small sample is subjected to controlled sinusoidal strain (or stress) at a range of frequencies and temperatures. The instrument measures the in-phase (storage) and out-of-phase (loss) components of the resulting stress (or strain), yielding E′, E″, and η as functions of frequency and temperature [10, 11].

DMA data is the foundation for all quantitative damping design. When selecting a damping material, the DMA-generated master curve — or at minimum, the η vs. temperature plot at the target frequency — is the critical piece of information.

Selecting a damping material for a photonics application requires matching the material's peak loss factor to the target frequency and operating temperature while respecting constraints on mass, outgassing, chemical compatibility, and physical space. This section surveys the principal material classes used in optical and precision systems.

4.1Elastomers

Elastomers are the workhorse damping materials in photonics laboratories. Their long-chain polymer structures provide high loss factors through internal molecular friction during cyclic deformation.

Sorbothane is a proprietary polyether-based polyurethane developed specifically for vibration damping and shock absorption. It offers one of the highest loss factors of any commercially available material (η ≈ 0.5 at room temperature), combined with good memory and long fatigue life [9]. It is widely used as vibration-isolating feet for breadboard-mounted equipment and as pads between sensitive instruments and table surfaces. Sorbothane is available in multiple durometers (Shore 00-30 to 00-70) and standard sheet thicknesses. Its primary limitation is load sensitivity: exceeding the recommended load per unit area shifts the material out of its optimal damping range and into amplification [8, 9].

Neoprene (polychloroprene) provides moderate damping (η ≈ 0.02–0.10) with excellent chemical resistance and a wide operating temperature range (−40°C to +100°C). It is commonly used in gaskets, mounting pads, and vibration-isolating washers. Its damping is lower than Sorbothane's but its environmental robustness is superior.

Butyl rubber (polyisobutylene) offers high damping (η ≈ 0.2–0.4) with very low gas permeability, making it suitable for vacuum-adjacent applications. Its glass transition temperature near −70°C means it operates well above T_g at room temperature, in the rubbery plateau where η is moderate to high.

Silicone rubber provides the widest operating temperature range of any common elastomer (−60°C to +230°C) but at the cost of relatively low damping (η ≈ 0.01–0.05). It is preferred when thermal stability is paramount — for example, damping treatments near laser sources or in thermal vacuum chambers.

Nitrile rubber (NBR) combines moderate damping (η ≈ 0.03–0.08) with excellent oil and solvent resistance. It appears in O-rings and gaskets used in optomechanical assemblies where fluid exposure is a concern.

4.2Viscoelastic Polymer Sheets

Specialty viscoelastic sheets are formulated specifically for constrained-layer and free-layer damping treatments. These products are designed to deliver high η at specific frequency and temperature ranges.

3M ISD-112 is an acrylic-based viscoelastic polymer sheet widely used as the core layer in constrained-layer damping treatments. It delivers η > 1.0 at its peak (near 25°C at ~200 Hz) and is available in pressure-sensitive adhesive-backed sheets for direct application to structural surfaces [2].

EAR C-1002 (now marketed under various names following corporate acquisitions) is a specialty isodamp material used in aerospace and precision instrument applications. It provides controlled damping with well-characterized DMA data across a broad frequency-temperature space.

These sheet products are the standard choice for retrofitting damping to existing structures — breadboards, enclosures, and equipment housings — without modifying the base design.

4.3Polymer Foams

Open-cell and closed-cell polymer foams provide combined damping and acoustic absorption. They are less effective than solid elastomers for structural vibration damping (the cellular structure reduces shear stiffness) but valuable for suppressing acoustic excitation of thin panels. Polyurethane foam and melamine foam are common choices for lining enclosures around noisy equipment.

4.4Bitumen and Asphalt-Based Sheets

Bitumen-based damping sheets (sometimes called “anti-drumming” compounds) are inexpensive, high-density damping treatments applied as free layers to thin metal panels. They are widely used in automotive and industrial noise control but occasionally appear in photonics contexts for damping sheet-metal enclosures or equipment housings. Their primary disadvantage is mass: effective bitumen treatments can add 2–5 kg/m² to a panel.

4.5Material Selection Considerations

Beyond the loss factor, several practical factors influence material selection for photonics applications:

Outgassing: In vacuum systems or near sensitive optical coatings, the material must meet outgassing requirements. Silicone rubbers and some polyurethanes can release volatile compounds that deposit on optics. NASA-qualified low-outgassing grades are available for critical applications.

Creep and static deflection: Viscoelastic materials under sustained static load will creep over time. For isolation feet supporting heavy equipment, static deflection and long-term creep must be accounted for in the design.

Frequency range: The material's peak η must align with the target vibration frequencies. A material optimized for 10 Hz building sway is useless for a 200 Hz optical table resonance.

Temperature stability: Laboratory temperature variations of ±2–5°C can shift η significantly for materials with narrow transition peaks. Choose materials with broad damping plateaus for environments without tight temperature control.

Free-layer damping, also called extensional or unconstrained-layer damping, is the simplest passive treatment for adding damping to a vibrating structure. A layer of viscoelastic material is bonded directly to the surface of the structure with no additional constraining layer on top.

5.1Mechanism

When the base structure undergoes bending vibration, the outer fibers experience alternating tension and compression. The viscoelastic layer, bonded to the surface, is forced to undergo the same cyclic extensional (tensile/compressive) strain. The material's internal viscosity converts a portion of this strain energy into heat on each cycle, reducing the amplitude of vibration [2, 5].

The damping effectiveness depends on the viscoelastic material's loss factor η₂, its thickness relative to the base structure h₂/h₁, and the ratio of their moduli E₂/E₁. Because the viscoelastic layer is typically much softer than the metal substrate (E₂/E₁ ≈ 10⁻⁴ to 10⁻³ for an elastomer on steel), the damping layer must be relatively thick to contribute meaningfully to the composite structure's energy dissipation.

5.2Ross-Kerwin-Ungar (RKU) Equation for Free-Layer Damping

The composite loss factor of a base structure with a free viscoelastic layer was derived by Ross, Kerwin, and Ungar and remains the standard design equation [5]:

Where: η_c = composite system loss factor, η₂ = loss factor of the viscoelastic layer, e = modulus ratio = E₂/E₁, n = thickness ratio = h₂/h₁, E₁ = Young's modulus of base structure (Pa), E₂ = Young's modulus of viscoelastic layer (Pa), h₁ = thickness of base structure (m), h₂ = thickness of viscoelastic layer (m).

The quantity en(3 + 6n + 4n²) is the stiffness-weighted strain energy fraction of the viscoelastic layer. When this product is small (which it almost always is for elastomers on metals), the expression simplifies to:

This reveals the fundamental limitation of free-layer damping: the composite loss factor scales linearly with the modulus ratio e, which is very small for soft elastomers on stiff metals. Achieving significant damping requires either a very thick layer (large n) or a stiffer damping material (larger e, but this typically means lower η₂).

5.3Design Guidelines

Thickness ratio: A common rule of thumb for free-layer treatments is h₂/h₁ ≥ 2 (the damping layer should be at least twice the base thickness) to achieve meaningful damping on metal substrates [2, 3].

Placement: The damping layer should be applied to the surface of maximum strain — the outer surface farthest from the neutral axis. For a simply supported beam or plate, this is either face. For asymmetric structures, strain analysis identifies the optimal location.

Mass penalty: Free-layer treatments add mass proportional to the productρ₂h₂. For a 6 mm damping layer on a 3 mm steel plate (density ratio ~0.5), the mass increase is approximately 100% — often prohibitive.

Advantages: Simple application (bond and forget), no constraining layer needed, effective over a broad frequency band if the material's η is frequency-stable.

Limitations: Poor weight efficiency on stiff substrates, thickness requirements often impractical for thin panels, damping decreases at higher-order modes where strain distributes more uniformly.

Constrained-layer damping overcomes the weight-efficiency limitation of free-layer treatments by forcing the viscoelastic material into shear rather than extension. This fundamentally changes the energy dissipation mechanism and produces significantly more damping per unit added mass.

6.1CLD Configuration

A CLD treatment consists of three layers [4, 5]:

1. Base layer — the vibrating structure (e.g., an optical table skin, breadboard panel, or enclosure wall)

2. Viscoelastic core — a thin layer of high-loss-factor material (the damping element)

3. Constraining layer — a stiff sheet (typically metal) bonded to the outside of the core

When the base structure bends, the base and constraining layers deflect together but, because they are separated by the compliant core, they undergo different extensional strains. This differential strain forces the viscoelastic core into shear deformation. Because shear strain in the core can be much larger than the extensional strain that a free layer would experience at the same surface location, the energy dissipated per cycle is substantially greater [4, 5].

6.2Kerwin's Shear Parameter

The performance of a CLD treatment depends on the shear parameter g, which characterizes the ratio of shear stiffness in the viscoelastic core to the extensional stiffness of the constraining layer [4]:

Where: G₂* = complex shear modulus of the viscoelastic core (Pa), E₃ = Young's modulus of the constraining layer (Pa), h₂ = thickness of the viscoelastic core (m), h₃ = thickness of the constraining layer (m),λ = wavelength of the bending vibration in the base structure (m).

The composite system loss factor peaks at an optimal value of g that depends on the layer geometry and stiffness ratios. Design consists of selecting core thickness h₂ and core shear modulus G₂ to place g near its optimum for the target mode's bending wavelength [4, 5].

6.3CLD Design Optimization

The full Kerwin CLD model yields the composite loss factor as a function of the shear parameter and layer properties. Key design guidelines [2, 4, 5]:

Constraining layer: A stiffer, thicker constraining layer forces more shear into the core, increasing damping — up to a point. Beyond the optimum, the constraining layer dominates the composite stiffness and the core's shear contribution saturates. For practical designs on metal substrates, a constraining layer thickness of h₃/h₁ ≈ 0.1–0.5 is typical.

Core thickness: Thinner cores produce higher shear strain for a given differential displacement, but very thin cores limit the total energy dissipated. Typical CLD core thicknesses range from 0.25 mm to 2 mm.

Core material: The core should have the highest possible loss factor η₂ at the target frequency and temperature. Acrylic-based viscoelastic polymers (e.g., 3M ISD-112 with η > 1.0 at peak) are preferred over general-purpose elastomers.

Edge effects: Shear strain in the CLD core is not uniform — it peaks near the edges of the treatment and drops toward the center for short treatment lengths. Full-coverage treatments eliminate this concern; partial-coverage patches should extend well beyond the region of maximum curvature [4].

Coverage area: CLD patches are most effective when placed at locations of maximum bending curvature (maximum strain). For a plate vibrating in its fundamental mode, this is typically the center; for higher modes, the locations of maximum curvature shift.

6.4CLD vs. Free-Layer Comparison

For the same added mass, CLD consistently outperforms free-layer treatments by a factor of 5–50×, depending on geometry and frequency [2, 5]. The constraining layer adds minimal mass (it is typically thin metal) but redirects the deformation mode of the viscoelastic material from low-efficiency extension to high-efficiency shear.

The tradeoff is complexity: CLD requires a third bonded layer and more careful design (the shear parameter must be optimized), whereas a free layer is simply bonded to the surface.

Free-layer and constrained-layer treatments add distributed damping to a structure. An alternative approach targets specific resonant frequencies using a discrete mechanical device: the tuned mass damper (TMD). TMDs are lightweight, require no modification to the base structure's material, and can deliver dramatic attenuation of individual resonances — properties that make them the dominant damping technology inside high-performance optical tables [6, 7].

7.1Single-DOF TMD Theory

A TMD is a secondary mass-spring-damper system attached to the primary structure. When tuned so that its natural frequency matches a resonance of the primary structure, the TMD vibrates out of phase with the structure at that frequency, generating an inertial force that opposes the structural motion and dissipates energy through its internal damper [6].

The key parameters are the mass ratio μ, the frequency ratio (tuning), and the TMD's internal damping ratio:

Where: m_d = TMD mass (kg), M = modal mass of the targeted structural mode (kg).

For undamped primary structures (a reasonable approximation for bare metal optical tables), Den Hartog's classical optimization yields the tuning and damping that minimize the peak response [6]:

Where: ω_d = natural frequency of the TMD (rad/s), ω_n = natural frequency of the targeted structural mode (rad/s).

This last expression shows the fundamental tradeoff: a heavier TMD (larger μ) produces a lower peak response, but at the cost of added mass. For optical tables, mass ratios of 1–5% per mode are typical, yielding peak amplitude reductions of 60–90% relative to the undamped structure [7].

7.2TMD Behavior at Resonance

A properly tuned TMD splits the original single resonance peak into two smaller peaks — one above and one below the original frequency. The amplitude of these split peaks is determined by the TMD's internal damping. At Den Hartog's optimal damping, the two peaks are equal in height and represent the minimum achievable maximum response [6].

If the TMD damping is too low, the two split peaks become tall and sharp — the system now has two problematic resonances instead of one. If the damping is too high, the TMD mass effectively locks to the primary structure and provides no benefit. The optimum lies between these extremes.

7.3Narrowband TMD in Optical Tables

Inside a typical optical table, discrete TMDs are placed at the corners where bending and torsional mode shapes have maximum displacement. Each TMD consists of a steel bar (the spring) with a mass attached to its free end, tuned to a specific table resonance. Viscous or elastomeric damping is added to the spring element [7, 8].

A conventional discrete TMD system targeting two modes (first bending and first torsion) uses 2–4 individual dampers. Each damper is tuned to one mode. While effective at the targeted frequencies, discrete TMDs introduce new degrees of freedom, creating secondary resonance peaks at adjacent frequencies [7].

7.4Distributed Tuned Mass Dampers (DTMD)

The distributed tuned mass damper concept, pioneered for optical tables in the 1990s, overcomes the secondary-peak limitation of discrete TMDs [7]. A DTMD system consists of hundreds of small individual TMDs, each tuned to a slightly different frequency within a design band that spans the table's resonances of interest.

Because the individual damper frequencies are spread across the band, the collective DTMD behaves as a broadband absorber within the design band — suppressing both bending and torsional modes without introducing the spurious secondary peaks characteristic of discrete TMDs. The total DTMD mass is comparable to a discrete system (1–5% of the table mass), but the vibration energy is distributed across many small absorbers rather than concentrated in a few large ones [7].

Commercial implementations embed DTMD elements within the honeycomb core of the optical table during manufacture. The dampers are shaped pieces of inhomogeneous elastomeric material that act as though they contain a spectrum of masses at varying distances — effectively a continuous distribution of natural frequencies in a single physical element [8].

Modern optical tables integrate multiple damping strategies to achieve the quietest possible work surface. Understanding how these techniques combine — and what performance metrics to evaluate — is essential for selecting the right table for a given application.

8.1Honeycomb Core Damping

The stainless steel honeycomb core that gives optical tables their exceptional stiffness-to-weight ratio also provides a modest degree of inherent material damping. The many cell walls, joints, and interfaces within the core dissipate energy through micro-friction at bonded surfaces and internal hysteresis in the adhesive layers. However, this baseline damping is insufficient for demanding applications — typical honeycomb core loss factors remain belowη ≈ 0.005 without additional treatment [7, 8].

8.2Broadband Damping

Broadband damping techniques add viscoelastic materials or polymer foams within the table core to absorb energy across a wide frequency range without targeting specific modes. The damping is frequency-indiscriminate: it reduces all resonances by a moderate, roughly uniform amount.

Broadband damping is effective for general-purpose tables used in relatively quiet environments. It provides a baseline level of vibration reduction — typically 5–10 dB at resonance — that suffices for spectroscopy, velocimetry, and other applications with moderate vibration sensitivity [7].

8.3Tuned and Distributed Damping

For higher-performance requirements (interferometry, nanopositioning, holography), broadband damping alone is insufficient. Tuned mass dampers — either discrete or distributed — target the table's lowest and most problematic resonances, typically the first bending and torsion modes between 150 and 300 Hz.

As discussed in Section 7, discrete TMDs can reduce specific resonance peaks by 20–30 dB but may introduce secondary peaks. DTMD systems provide comparable peak reduction while suppressing secondary peaks across the design band. The highest-performance commercial tables combine broadband core damping with DTMD systems for comprehensive vibration control [7, 8].

8.4Active Damping

Active damping systems use vibration sensors (accelerometers or velocity transducers) mounted on the table surface to detect residual vibration, paired with force actuators (voice coils or piezoelectric stacks) that apply counteracting forces. A feedback controller drives the actuators to cancel detected vibrations in real time.

Active systems excel at suppressing low-frequency resonances (below 100 Hz) where passive techniques lose effectiveness, and at rejecting transient disturbances that might excite multiple modes simultaneously. Their limitation is cost, complexity, and the potential for instability if the control loop is poorly tuned [7].

8.5Hybrid Damping

The highest-performance optical tables combine passive DTMD systems with active feedback damping — a hybrid approach. The passive system handles the predictable, well-characterized structural resonances, while the active system addresses residual vibration, environmental transients, and any modes not covered by the passive design. This hybrid strategy achieves compliance values below 10⁻⁹ m/N across the critical 50–500 Hz band [7, 8].

8.6Compliance Curves as Performance Metrics

The dynamic performance of an optical table is quantified by its compliance curve: the ratio of surface displacement to applied force as a function of frequency [7, 8]:

The compliance curve is typically measured at the corner of the table (the location of maximum modal displacement for the lowest modes) and plotted on a log-log scale. Lower compliance means a quieter table. Key features to evaluate:

Resonance peaks: The height and sharpness of compliance peaks directly reflect the damping at those frequencies. A well-damped table shows broad, low peaks; an undamped table shows tall, sharp spikes.

Broadband floor: Between resonances, the compliance is determined by the table's static stiffness. Thicker, stiffer tables have a lower broadband compliance floor.

Frequency of first resonance: Higher is better — it moves the first problematic resonance further from the low-frequency vibration spectrum of most laboratory environments.

When comparing tables from different manufacturers, always compare compliance curves measured under the same conditions (same excitation point, same measurement point, same boundary conditions). A 6 dB difference in peak compliance corresponds to a 2× difference in displacement amplitude at resonance.

Quantifying damping requires careful experimental technique. The measurement method must match the question being asked: material-level properties for selecting damping compounds, or system-level performance for evaluating an assembled table or structure.

9.1Oberst Bar Method (ASTM E756)

The Oberst bar is the standard method for measuring the damping properties of viscoelastic materials applied to structures [10]. A uniform cantilever beam (typically steel) is treated with the damping material — as a free layer, a constrained layer, or a sandwich — and its frequency response is measured under controlled conditions.

The beam is excited by a non-contact method (electromagnetic driver or loudspeaker) and its response is measured by a lightweight accelerometer or laser vibrometer. The FRF yields resonance frequencies and half-power bandwidths, from which the composite loss factor is extracted at each mode. By comparing the damped beam's response to that of the bare (undamped) beam, the contribution of the damping treatment is isolated.

ASTM E756 specifies beam dimensions, support conditions, excitation methods, and data reduction procedures. It provides results at discrete frequencies (the beam's resonant modes), so characterizing a material across a broad frequency range requires beams of different lengths to shift the modal frequencies.

9.2Half-Power Bandwidth from FRF Data

For assembled structures (optical tables, breadboards, mounted equipment), the half-power bandwidth method is the most practical route to system-level damping. The structure is excited at a point (by impact hammer or shaker), the acceleration or displacement response is measured at another point, and the FRF is computed.

At each resonance peak, the loss factor is extracted as η = Δf/f_n, whereΔf is the frequency width at −3 dB from the peak. This method requires clean, well-resolved FRF data — frequency resolution must be at least 10× finer than the expected bandwidth Δf to avoid systematic overestimation of damping [1, 3].

9.3Logarithmic Decrement from Free Decay

When FRF measurement equipment is unavailable, free-decay measurement provides a simpler alternative. The structure is given an impulsive excitation (tap, pluck, or sudden release of a static deflection), and the resulting time-domain response is recorded. The logarithmic decrement is computed from successive peak amplitudes.

For structures with closely spaced modes, band-pass filtering of the time signal may be necessary to isolate individual mode decays. The method is less accurate than FRF-based techniques but requires minimal instrumentation — an accelerometer and a data logger suffice.

9.4Dynamic Mechanical Analysis (DMA)

DMA, introduced in Section 3.6, measures material-level viscoelastic properties (E′, E″,η as functions of frequency and temperature) on small coupon samples. While DMA does not directly measure the damping of an assembled structure, it provides the fundamental material data needed for analytical design of free-layer, CLD, and other treatments.

DMA results are typically presented as master curves (using time-temperature superposition) or as nomograms — two-dimensional plots that allow reading E′ and η at any combination of frequency and temperature. These nomograms are the essential design tool for selecting constrained-layer core materials.

9.5Compliance Curve Measurement

Optical table compliance curves are measured by applying a known sinusoidal (or broadband) force to the table corner using an instrumented impact hammer or electrodynamic shaker. The resulting surface velocity or displacement is measured at the same point (driving-point compliance) or at a remote point (transfer compliance) using an accelerometer or laser vibrometer [7, 8].

The driving-point compliance at the table corner is the standard metric reported by table manufacturers. It captures the worst-case modal response because the corner is an antinode for both bending and torsional modes. Transfer compliance between diagonally opposite corners captures the torsional mode specifically.

With multiple damping techniques available — each with different weight, cost, complexity, and frequency-range characteristics — selecting the right strategy requires systematic analysis of the vibration problem.

10.1Identify the Problem

The first step is characterizing what needs to be damped:

Resonance amplification: A specific structural mode is being excited, producing a tall, narrow peak in the response spectrum. This is the most common problem in optical systems — an optical table mode, a breadboard resonance, or a mount resonance amplifying floor vibration or on-table disturbances.

Broadband vibration: The vibration energy is spread across a wide frequency band with no single dominant peak. This occurs when multiple modes contribute, when the excitation itself is broadband (traffic, HVAC), or when the structure has been partially damped but residual broadband response remains.

Acoustic excitation: Airborne sound waves excite thin panels (enclosure walls, equipment housings). The treatment must target the panel modes excited by the acoustic field.

10.2Match Technique to Problem

10.3Mass Budget Considerations

Every damping treatment adds mass, and mass affects the system's natural frequencies and load capacity:

TMDs: 1–5% of the modal mass per targeted mode. Minimal impact on static load capacity.

CLD: 10–40% mass increase on the treated panel, depending on core and constraining layer thickness. Acceptable for optical table skins (where the mass is designed in) but potentially problematic for retrofit applications.

Free-layer: 50–200% mass increase on the treated panel for meaningful damping. Often impractical for precision applications.

Sorbothane pads: Negligible mass addition to the system; the pads add compliance that must be accounted for in the support design.

10.4Rules of Thumb

For optical tables: Demand compliance curves from the manufacturer. Compare curves, not marketing grades. If the application involves interferometry or nanopositioning, specify a table with tuned or distributed damping, not just broadband.

For breadboards: If a breadboard resonance is problematic, CLD patches applied to the bottom surface can reduce the resonance significantly. A 1 mm viscoelastic core plus a 0.5 mm aluminum constraining layer is a standard retrofit.

For equipment isolation: Sorbothane pads solve most equipment-to-table coupling problems. Size the pads for 50–70% of rated load to stay in the optimal damping zone.

For enclosures: Line thin-walled equipment enclosures with CLD tiles. The mass increase is acceptable for stationary lab equipment, and the noise reduction is immediately noticeable.

10.5Common Mistakes

Confusing isolation and damping. Placing an optical table on pneumatic isolators does not damp its internal resonances. If on-table sources excite table modes, isolators cannot help — damping is required.

Overloading Sorbothane pads. Exceeding the rated load per pad shifts the material out of its optimal viscoelastic range and can cause amplification at the natural frequency instead of isolation.

Applying free-layer treatment to stiff substrates. Free-layer damping is ineffective on thick steel or aluminum panels unless the layer is impractically thick. Use CLD instead.

Ignoring temperature sensitivity. A damping material specified at 25°C may lose half its loss factor at 30°C if its transition peak is narrow. Verify performance across the expected laboratory temperature range.

Targeting the wrong mode. Before designing a damping treatment, measure the vibration spectrum to identify which modes are actually contributing to the problem. Adding damping to a mode that is not being excited wastes mass and budget.