Acoustics Engineering

Silencer DIL & Octave-Band Analysis

Dynamic Insertion Loss methodology for engine exhaust and inlet silencers, including ISO 9613-1 atmospheric absorption and stack directivity corrections.

Hospital Grade DIL

18-61 dB

Per octave band, 31.5 Hz to 8 kHz

Octave Bands

9 Bands

31.5 Hz through 8,000 Hz

A-Weight Range

-39 to +1 dB

IEC 61672-1 corrections

1. Dynamic Insertion Loss (DIL)

Dynamic Insertion Loss is the difference in sound pressure level at a fixed measurement point before and after a silencer is installed in a duct or exhaust system under actual operating conditions with gas flow. Unlike Transmission Loss (TL), which is a laboratory measurement, DIL captures real-world performance including flow-generated noise, breakout noise, and installation effects.

Dynamic Insertion Loss: DIL = Lp_before - Lp_after (dB) Where: Lp_before = sound pressure level without silencer (dB) Lp_after = sound pressure level with silencer installed (dB) DIL is measured per octave band (31.5 Hz to 8 kHz)

DIL vs. Transmission Loss

Transmission Loss (TL) measures the sound power difference across the silencer element in a controlled laboratory setting with anechoic terminations. It does not account for flow noise regeneration, flanking paths, or casing breakout. DIL is always the preferred metric for engineering design because it represents actual field performance.

Testing Standards

Silencer DIL is tested per ASTM E477 (Standard Test Method for Laboratory Measurement of Acoustical and Airflow Performance of Duct Liner Materials and Prefabricated Silencers). Manufacturers provide DIL data at rated flow conditions across octave bands from 63 Hz through 8 kHz, with some providing 31.5 Hz data as well.

2. Octave-Band Analysis

Octave-band analysis divides the audible frequency spectrum into bands, each centered on a frequency that is double the previous band. The standard octave bands used in industrial noise assessment are:

Center Frequency (Hz) 31.5631252505001000200040008000
A-Weighting (dB) -39.4-26.2-16.1-8.6-3.20.0+1.2+1.0-1.1

A-Weighting

The A-weighting filter approximates the frequency response of human hearing. Low frequencies are heavily penalized (31.5 Hz gets -39.4 dB) while mid-frequencies near 1-4 kHz are slightly boosted. The A-weighted sound level in each band is calculated by adding the A-weighting correction to the linear sound pressure level.

A-Weighted Level per Band: L_A,i = L_p,i + A_i (dBA) Overall A-Weighted Level: L_A,total = 10 * log10( SUM( 10^(L_A,i / 10) ) ) (dBA) Where the sum is across all octave bands i = 1 to 9

Why Not Just Use Overall dBA?

Overall dBA is a single-number summary that can mask frequency-specific problems. Consider a scenario where a silencer reduces 1 kHz noise by 40 dB but only reduces 125 Hz noise by 5 dB. The overall dBA might look acceptable, but the 125 Hz component could still cause complaints or regulatory issues, especially at long distances where atmospheric absorption provides little help at low frequencies.

3. ISO 9613-1 Atmospheric Absorption

Sound waves lose energy as they travel through air due to molecular absorption. This effect is strongly frequency-dependent and varies with temperature, humidity, and atmospheric pressure. ISO 9613-1 provides the standard method for calculating atmospheric absorption coefficients.

Atmospheric Absorption Coefficient (ISO 9613-1): alpha = 8.686 * f^2 * [ 1.84e-11 * (Pr/p) * (T/T0)^0.5 + (T/T0)^-2.5 * ( 0.01275 * exp(-2239.1/T) * frO / (frO^2 + f^2) + 0.1068 * exp(-3352.0/T) * frN / (frN^2 + f^2) ) ] Where: alpha = absorption coefficient (dB/m) f = frequency (Hz) T = temperature (K) T0 = 293.15 K (reference) Pr = 101.325 kPa (reference pressure) p = atmospheric pressure (kPa) frO = oxygen relaxation frequency (Hz) frN = nitrogen relaxation frequency (Hz)

Typical Absorption Values

The following table shows typical atmospheric absorption at standard conditions (70°F, 50% RH, 29.92 inHg):

Frequency (Hz) 31.5631252505001k2k4k8k
dB per 1000 ft 0.010.040.130.400.871.483.08.731

At low frequencies, atmospheric absorption is negligible even at long distances. At 8 kHz, the absorption can exceed 50 dB over 1,000 ft, meaning high-frequency noise is naturally attenuated. This is why low-frequency silencer performance is critical for far-field compliance and why simplified broadband absorption estimates can be significantly inaccurate.

Relaxation Frequencies

The oxygen (frO) and nitrogen (frN) relaxation frequencies govern the absorption peaks. These depend on the molar concentration of water vapor, which is why humidity significantly affects sound absorption. Low humidity conditions (common in winter and arid regions) produce lower absorption, meaning sound travels farther.

4. Stack Directivity

A vertical exhaust stack acts as a directional source at high frequencies. The opening radiates sound preferentially in the vertical direction (along the stack axis), reducing the horizontal component that reaches ground-level receptors. This directivity effect increases with frequency because the stack opening becomes acoustically large relative to the wavelength.

Frequency (Hz) 31.5631252505001k2k4k8k
Directivity Reduction (dB) 0000135914
Sound Pressure Level with Directivity: Lp = Lw - L_divergence - DI - alpha*r Where: DI = directivity index (dB, subtracted for vertical stacks) L_divergence = 20*log10(r) + 10*log10(4*pi) for spherical spreading alpha*r = atmospheric absorption over distance r

When to Apply Directivity

Stack directivity corrections apply to vertical exhaust stacks discharging to atmosphere. They do not apply to horizontal inlet systems, side-discharge configurations, or enclosed exhaust manifolds. For inlet silencers, the noise source is typically omnidirectional and no directivity correction is applied.

5. Silencer Types and Grades

Industrial silencers for engine exhaust and inlet applications are classified by their noise reduction capability. The silencer grade determines the DIL performance across octave bands.

Reactive Silencers

Use expansion chambers and resonant cavities to reflect sound energy back toward the source. Most effective at low frequencies. Common in engine exhaust applications where high temperatures preclude absorptive materials.

Absorptive Silencers

Use acoustic packing materials (fiberglass, mineral wool) to convert sound energy to heat. Most effective at mid and high frequencies. Used in inlet applications and HVAC systems where temperatures permit.

Combination Silencers

Combine reactive and absorptive elements for broadband performance. Hospital-grade silencers typically use this approach to achieve high DIL values across all octave bands.

Silencer Grade Comparison

Grade Typical Overall Reduction Application
Industrial (Standard)15-25 dBARural sites with generous setbacks
Residential25-35 dBASuburban areas with moderate setbacks
Critical30-40 dBAResidential areas with tight setbacks
Hospital (Super Critical)40-55 dBAHospital zones, stringent noise ordinances

6. Worked Example

Consider a reciprocating engine exhaust at a compressor station. The unsilenced sound power levels and a hospital-grade silencer DIL are given below. Calculate the silenced noise level at 500 ft.

Parameter 31.5631252505001k2k4k8k
Unsilenced Lw (dB)131130131117115112112114115
Silencer DIL (dB)183648566161575043

Step-by-Step Process

  1. Calculate divergence: For r = 152.4 m (500 ft), L_div = 20*log10(152.4) + 10*log10(4*pi) = 43.7 + 11.0 = 54.7 dB
  2. Apply directivity: Subtract DI values [0, 0, 0, 0, 1, 3, 5, 9, 14] from each band
  3. Calculate atmospheric absorption: Using ISO 9613-1 at standard conditions, multiply alpha by distance for each band
  4. Unsilenced Lp: Lp = Lw - divergence - DI - absorption per band
  5. Subtract DIL: Silenced Lp = Unsilenced Lp - DIL per band
  6. Apply A-weighting: Add A-weight corrections to get silenced dBA per band
  7. Sum logarithmically: Overall silenced dBA = 10*log10(sum of 10^(dBA_i/10))

The result is an overall silenced level of approximately 32 dBA at 500 ft, well below the typical 55 dBA residential limit. Use the Silencer DIL Estimator to perform this calculation with your specific data.

Need more resources?

→ Browse All Fundamentals

Frequently Asked Questions

What is Dynamic Insertion Loss (DIL) and how does it differ from Transmission Loss?

Dynamic Insertion Loss measures the actual noise reduction achieved when a silencer is installed in a duct system under operating conditions with gas flow. Transmission Loss is measured in laboratory conditions without flow. DIL is the more practical metric because it accounts for flow-generated noise, breakout noise, and real installation effects.

Why is octave-band analysis important for silencer selection?

Overall dBA levels can mask frequency-specific problems. A silencer might achieve excellent broadband noise reduction but leave a tonal peak at a specific frequency that causes regulatory non-compliance. Octave-band analysis ensures every frequency range meets the required noise limits.

How does atmospheric absorption affect noise propagation at long distances?

Atmospheric absorption per ISO 9613-1 is strongly frequency-dependent. At 1000 ft, absorption at 8 kHz can be 10-30 dB while absorption at 63 Hz is essentially zero. This means low-frequency noise travels much farther, making low-frequency silencer performance critical for distant receptors.

What is stack directivity and when does it matter?

Vertical exhaust stacks radiate sound preferentially upward at high frequencies due to the directional characteristics of the stack opening. This reduces horizontal sound levels by 5-14 dB at frequencies above 1 kHz. Directivity corrections are important for accurate far-field noise predictions from exhaust stacks but do not apply to horizontal inlet systems.