Building Design for Advanced Technology Instruments Sensitive to Acoustic Noise
Michael Gendreau Colin Gordon & Associates
Building Design for Advanced Technology Instruments Sensitive to Acoustical Noise

Presentation Outline

Michael Gendreau, Colin Gordon & Associates
! High technology research and manufacturing instruments respond to internal vibration that can be excited by the external acoustic environment. The degree to which this occurs depends on many factors, but primarily the correspondence between the resonance characteristics of the instrument and the frequency content of the acoustic environment in which it operates. Adverse acoustic environments, such as those often found in operating laboratories, can affect the threshold of resolution achievable by the instrument. ! This presentation will include:
! a review of the basic terminology and criteria used in the acoustical design of advanced technology facilities, ! discussion of the mechanisms by which acoustic noise can interfere with instruments, and ! details of what the building contributes in terms of sources of noise (internal and external) and protection from those sources

I. Acoustics Review (1)

I. Acoustics Terminology Review
There is a lot more to even basic acoustics than can presented in a short time. In this section of the presentation, I have tried to focus on the parameters that seem to be most important in defining room criteria and equipment specifications with respect to acoustic noise.

I. Acoustics Review (2)

Presentation concerned with acoustical phenomena impact via the air medium (as opposed to the same phenomena propagated via structures, which is vibration) Pressure versus power
Noise is pressure fluctuation in air, it is to pressure variations that ears and machine components respond

Decibels (dB)

Logarithmic scale necessary due to large dynamic range (12 orders of magnitude for human hearing) Reference units
! ! Pressure: dB re 20 micropascals Power: dB re 1 picowatt

Types of noise

Tonal versus Broadband
I. Acoustics Review (3) Types of Noise
Tonal versus Broadband Noise
spectrum dominated by broadband noise (500 to 800 Hz) spectrum dominated by tonal noise (2988 Hz)
Sound Pressure Level, dB (re 20 Pa)
<INSERT DRAWINGS HERE>

2500 Frequency, Hz 5000

I. Acoustics Review (4)
! Measurement methodologies

Bandwidth

! Constant (e.g., narrowband) narrowband ! Proportional (e.g., 1/10, 1/3, and 1/1 octave band) ! Overall
Leq, Ln Leq , Ln eq n Linear versus Exponential Averaging Linear versus Exponential Averaging Lmax, Lpeak, etc. Lmax , Lpeak , etc. max peak
Common Noise Level Representations Common Noise Level Representations

! Average ! Average

! Maximum ! Maximum ! Standard ! Standard
Time Averaging Characteristics Time Averaging Characteristics
Slow (1000ms) Slow (1000ms) Fast (125ms) Fast (125ms) Impulse (35ms) Impulse (35ms)
Measurement Position Measurement Position

! Other ! Other

! Stationary (may be best for time domain) ! Stationary (may be best for time domain) ! Space-average (usually best for frequency domain, average or ! Space-average (usually best for frequency domain, average or Space maximum) maximum)
I. Acoustics Review (5) Bandwidth

One-third Octave Band

4000 Frequency, Hz

Narrowband 110

Octave Band 31.5

Frequency, Hz

Overall (0 to 5000 Hz) = 102 dB

I. Acoustics Review (6)

! Constant (e.g., narrowband) narrowband ! Proportional (e.g., 1/10, 1/3, and 1/1 octave band) ! Overall ! Average
Leq (equivalent energy average rms), Ln (statistical centile levels) rms), Linear versus Exponential Averaging Lmax (maximum rms), Lpeak (greatest instantaneous sound pressure), etc. rms),
Common Noise Level Representations
! Maximum ! Standard ! Standard

I. Acoustics Review (7)

Leq, Ln Linear versus Exponential Averaging Lmax, Lpeak, etc.

! Maximum ! Standard

Time Constant Time Averaging Characteristics
Slow (1000ms) Fast (125ms) Impulse (35ms)

! Other

I. Acoustics Review (8) Time Constant
Effect of Time Constant on Measured Amplitude ("Packaging Machine" - Impulsive Noise)

'impulse' setting (35ms)

'slow' setting (1000ms)
40.0 31.500 Frequency, Hz 4000 8000

I. Acoustics Review (9)

Time Averaging Characteristics

Measurement Position

! Stationary (may be best for time domain) ! Space-average (usually best for frequency domain, average or Spacemaximum)
I. Acoustics Review (10) Standing Waves (1)
Noise Level, dB (re 20 micropascals
0m (at Mechanical Room Wall) 2.4m 50 4.8m 7.2m 9.6m 40 12.0m 14.4m 16.8m 30 18.4m (at Opposing Wall) NC-40 NC-35 20
0 31.8000 Octave Band Center Frequency, Hz
I. Acoustics Review (11) Standing Waves (2)
2.4 4.8 7.2 9.14.4 16.8 18.4 Distance from Mechanical Room Wall, m 125 Hz 2000 Hz

I. Acoustics Review (12)

! Indices based on human perception
Frequency and amplitude range
! 20 Hz to 20,000 Hz ! 0 to 120 dB at 1000 Hz (varies with frequency)
Frequency-based criteria: NC, RC, NCB, NR, PNC, etc. FrequencyOverall indices: dBA, dBC, dB
! Indices based on research equipment sensitivity
! Should be tested and specific to the equipment ! Should be tested and specific to the equipment ! Optionally one can identify a standard frequency based ! Optionally one can identify a standard frequency based curve (but not an overall index) that is entirely above a curve (but not an overall index) that is entirely above a tested spectrum but this may be overly conservative at tested spectrum but this may be overly conservative at certain frequencies certain frequencies

I. Acoustics Review (13)

I. Acoustics Review (14) Comparison of Criteria (1)
Comparison of Perception-based Criterion Curves ("Quiet" Labs)

PNC-25

NCB-25

RC-25 Mk II

31.8000 Frequency, Hz
I. Acoustics Review (15) Comparison of Criteria (2)
Comparison of Perception-based Criterion Curves ("Noisy" Labs, e.g., cleanrooms)

PNC-60

NCB-60

RC-60 Mk II (None)

I. Acoustics Review (16)

! Should be tested and specific to the equipment (See Section II) ! Optionally one can identify a standard frequency based curve (but not an overall index) that is entirely above a tested spectrum but this may be overly conservative at certain frequencies
II. Noise Impact to Equipment (1)
II. Noise Impact to Research and Process Equipment
In this section, discussed are some of the mechanisms that make advanced technology equipment sensitive to noise. This is important, as it will help identify what type of tool specifications we should expect to find, and thus help us define our lab noise criteria.
II. Noise Impact to Equipment (2)
! The mechanism by which acoustic noise interferes with equipment
Among other things, the achievable resolution of an optical tool is a function of differential vibration between critical elements in the tool, say, between a lens and the observed target. Vibration of elements within a tool can be stimulated by:
! vibration sources within the tool; ! external vibration sources, transmitting to the tool via its support structure; and ! acoustic noise in the laboratory environment that causes vibration of exposed elements of the tool, which is then passed on to sensitive internal elements via the tool structure. structure.
II. Noise Impact to Equipment (3)
! Acoustic excitation of equipment components (example: casing panels)
Example 1: vibration induced in an 560 x 710 x 0.4 mm thick steel panel due to the presence of five different levels of broadband acoustic noise (white noise)
Example 2: vibration induced in an 210 x 350 x 6 mm thick Example 2: vibration induced in an 210 x 350 x 6 mm thick aluminum panel due to the presence of acoustic noise (white aluminum panel due to the presence of acoustic noise (white noise) noise)
! Demonstration of linear relationship between the sound pressure level impinging on the panel and the vibration level measured on the panel ! Primary impact in 75 to 150 Hz range
! Structures will tend to respond more readily to impinging noise at ! Structures will tend to respond more readily to impinging noise at their modal or natural frequencies their modal or natural frequencies ! This can be especially dramatic in low-damped structures excited ! This can be especially dramatic in low-damped structures excited low at their fundamental, or low order, resonance frequencies, when at their fundamental, or low order, resonance frequencies, when order, these frequencies are high enough that the size of the structure these frequencies are high enough that the size of the structure equals or exceeds the acoustic wavelengths. equals or exceeds the acoustic wavelengths. ! There is a significant amount of vibration at several of the plate ! There is a significant amount of vibration at several of the plate pla modal frequencies (275, 750, 785, 960, and 1370 Hz), but at modal frequencies (275, 750, 785, 960, and 1370 Hz), but at other frequencies relatively little vibration is induced. other frequencies relatively little vibration is induced. ! The amount of vibration induced in a structure is not only a ! The amount of vibration induced in a structure is not only a function of the noise level, but also the frequency. function of the noise level, but also the frequency.

II. Noise Impact to Equipment (4) Noise Induced Vibration (1)
Noise induced vibration in a 560 x 710 x 0.4 mm thick steel panel (Source: white noise at 5 different amplitudes)
RMS Velocity Leve dB (re 1 m/s)

0 -10 -500

II. Noise Impact to Equipment (5)
Example 1: vibration induced in an 560 x 710 x 0.4 mm thick Example 1: vibration induced in an 560 x 710 x 0.4 mm thick steel panel due to the presence of five different levels of steel panel due to the presence of five different levels of broadband acoustic noise (white noise) broadband acoustic noise (white noise)
Example 2: vibration induced in an 210 x 350 x 6 mm thick aluminum panel due to the presence of acoustic noise (white noise)
! Demonstration of linear relationship between the sound pressure ! Demonstration of linear relationship between the sound pressure level impinging on the panel and the vibration level measured on level impinging on the panel and the vibration level measured on the panel the panel ! Primary impact in 75 to 150 Hz range ! Primary impact in 75 to 150 Hz range
! Structures will tend to respond more readily to impinging noise at their modal or natural frequencies ! This can be especially dramatic in low-damped structures excited lowat their fundamental, or low order, resonance frequencies, when order, these frequencies are high enough that the size of the structure equals or exceeds the acoustic wavelengths. ! There is a significant amount of vibration at several of the plate plate modal frequencies (275, 750, 785, 960, and 1370 Hz), but at other frequencies relatively little vibration is induced. ! The amount of vibration induced in a structure is not only a function of the noise level, but also the frequency.
II. Noise Impact to Equipment (6) Noise Induced Vibration (2)
Noise induced vibration in a 210 x 350 x 6 mm thick aluminum plate (Source: white noise)
RMS Velocity Level, dB (re 1m/s)

0 -10 -1800 2000

II. Noise Impact to Equipment (7)
! Assess likelihood of impact by dividing noise into three general frequency regions
low frequency range (low impact probability)
! acoustic wavelength is significantly longer than the dimensions of the tool structures, coupling between the two is relatively inefficient. ! Exceptionally, very low frequency pressure fluctuations may interfere with tools with open beams (some interferometers, atomic force microscopes, etc.).
mid frequency range (higher impact probability)

! In the mid to high frequency range, especially at the coincidence coincidence frequency (where the acoustic and structural bending wave speeds are equal) and above, the structure is more likely to be excited by acoustic energy. ! As with the aluminum plate example, the middle frequency middle range might also contain easily excitable low order resonance frequencies.
high frequency range (low impact probability)
! acoustic excitation of structures is often less of a concern due to fact that there is usually less acoustical energy available with increasing frequency (see Section III), among other reasons.
II. Noise Impact to Equipment (8)
! Should be tested and specific to the equipment ! Optionally one can identify a standard frequency based curve (but not an overall index) that is entirely above a tested spectrum but this may be overly conservative at certain frequencies
II. Noise Impact to Equipment (9)
Research equipment sensitivity is not well represented by overall noise indices such as dBC. 1. An optical tool that probably functions better in one 70 dBC laboratory (Lab A) than in another 70 dBC laboratory (Lab B).
RMS Sound Pressure Level, dB (re 20 micropasca
Measured Sensitivity Curve for Optical Tool X
Lab A (70 dBC) Lab B (70 dBC)

30 31.8000

Octave Band Center Frequency, Hz
II. Noise Impact to Equipment (10)
Research equipment sensitivity is not well represented by overall noise indices such as dBC. 1. Four laboratories that meet the NC-60 HVAC design criterion, with an 11 dB range in equivalent dBC performance.
RMS Sound Pressure Level, dB (re micropascals)

31.8000

non-clean lab 38 (79 dBC) clean fab 14 (76 dBC) clean fab 25 (73 dBC) non-clean lab 35 (68 dBC) NC-60
III. Lab Noise Sources and Control (1)
III. Laboratory Noise Sources and Control
The final section of this presentation regards the types of noise sources that will be encountered in the facility, typical noise spectra, and an outline of noise control methods.
III. Lab Noise Sources and Control (2) Michael Gendreau, Colin Gordon & Associates
Noise sources and typical spectra
HVAC noise sources Typical operating cleanroom and lab noise levels and source frequency ranges
! HVAC (fans) broadband and sometimes tonal noise, usually predominant in the 63 to 250 Hz octave bands (occasional motor or other tones to 1000 or 2000 Hz) Equipment and support equipment broadband and tonal, 250 to 2000 Hz range (occasional higher frequencies, especially from air flow or air activated components in tools) HVAC (diffusers / flow noise) broadband and occasionally tonal 1000 to 8000 Hz range Fluid flow, valve, and air leakage noise broadband 1000 Hz and above

III. Lab Noise Sources and Control (3) Michael Gendreau, Colin Gordon & Associates

Lab Noise Levels

Statistical distribution of measured operational non-clean laboratory noise levels - 20 data records (each data record is a space-averaged 20s Leq, "slow" time constant, no frequency weighting)

20 31.8000

III. Lab Noise Sources and Control (4) Michael Gendreau, Colin Gordon & Associates

Cleanroom Noise Levels

Statistical distribution of measured operational class 1 cleanroom noise levels - 28 data records (each data record is a space-averaged 20s Leq, "slow" time constant, no frequency weighting)
III. Lab Noise Sources and Control (5) Michael Gendreau, Colin Gordon & Associates
Research equipment (tools) and tool support equipment (dry pumps, compressors, mini-environment fans, air miniand fluid flow noise, mechanical actuators and solenoids, etc.) People noise (voices, telephones, transportation activity, etc.)
! ! More likely to be controllable by user More likely to be controllable by user
External noise (Air and ground traffic, mechanical equipment, etc.)
! ! Primarily a low frequency problem, as a function of the Primarily a low frequency problem, as a function of the typical insertion loss of building facades. typical insertion loss of building facades.
III. Lab Noise Sources and Control (6) Michael Gendreau, Colin Gordon & Associates
Lab Equipment Noise Sources
Sound Power Levels of Various Advanced Technology Tool Support Equipment
Dry Pumps 1 Minienvironment Chiller (Compressor) 2 Chiller (Compressor) 4
Dry Pumps 2 Chiller (Compressor) 1 Chiller (Compressor) 3 Chiller (Compressor) 5
Sound Power Level, dB (re 1 picowatt)
III. Lab Noise Sources and Control (7) Michael Gendreau, Colin Gordon & Associates
! More likely to be controllable by user
III. Lab Noise Sources and Control (8) Michael Gendreau, Colin Gordon & Associates
! Primarily a low frequency problem, as a function of the typical insertion loss of building facades.

III. Lab Noise Sources and Control (9) Michael Gendreau, Colin Gordon & Associates

External Noise Sources

Sound Pressure Levels of Contemporary Commercial Jet Aircraft (noisy urban environment - 1 km from airport in flight path)
Large Twin Engine Commercial Jet (777) Take-off, Lmax (71 dBA) Small Twin Engine Commercial Jet (e.g., 737) Take-off, Lmax (79 dBA) Large Twin Engine Commercial Jet Take-off, Lmax (63 dBA) Large Twin Engine Commercial Jet (777) Take-off, Lmax (70 dBA) Small Twin Engine Commercial Jet (e.g., 737) Take-off, Lmax (81 dBA) Small Twin Engine Commercial Jet (e.g., 737) Take-off, Lmax (77 dBA) Ambient, Lmax (58 dBA)
20 31.500 Frequency, Hz 4000 8000
III. Lab Noise Sources and Control (10) Michael Gendreau, Colin Gordon & Associates
Sound Pressure Levels of Automobile Traffic (noisy urban environment - 300m from roadway at building facade)
Ground Traffic Accelerating, Lmax
Ground Traffic, Lmax Ground Traffic, Lmax Ambient, Lmax (58 dBA)
III. Lab Noise Sources and Control (11) Michael Gendreau, Colin Gordon & Associates

Noise control

! at the source
Fans and air handling units
Fan selection / specification Dump walls Spliiters (internal silencers) Devices producing laminar inlet flow
Terminal elements (diffusers, duct elements, air terminals, etc.)
selection / specification

in the path

distance (long duct paths) duct lining (where possible) silencers (lined and unlined)
Tool support equipment noise
equipment selection (depends on having good noise data from manufacturer) improved casework
pump chases, sub-fabs, etc. enclosures
III. Lab Noise Sources and Control (12) Michael Gendreau, Colin Gordon & Associates
Noise control (continued)

Tool noise

! ! ! ! (probably designed to work with itself originally primary concern is add-ons, such as mini-environments) addminiLocal enclosures for sensitive components Surface treatment, clean class, and outgassing partition design, basic considerations
single vs double partitions leaks (penetrations, doors, HVAC services, etc.)
Reverberation control External noise control
III. Lab Noise Sources and Control (13) Michael Gendreau, Colin Gordon & Associates
! (probably designed to work with itself originally only concern is add-ons, such as mini-environments) addminiSurface treatment, clean class, and outgassing Building faade and partition design, basic considerations
single versus double partitions leaks (penetrations, door and door seals, HVAC services, etc.)

Reverberation control

External noise control