Reverberant Chamber Acoustic Testing of Spaceflight

Hardware – Test Article Absorption Effects

Aron D. Hozman and William O. Hughes, NASA Glenn Research Center, Cleveland, Ohio

The exposure of a customer’s aerospace test article to a simu-

lated acoustic launch environment is typically performed in a
reverberant acoustic test chamber. The acoustic pre-test runs that
will ensure that the sound pressure levels of this environment can
indeed be met by a test facility are normally performed without a
test article dynamic simulator of representative acoustic absorp-
tion and size. If an acoustic test facility’s available acoustic power
capability becomes maximized with the test article installed dur-
ing the actual test, the customer’s environment requirement may
become compromised. To understand the risk of not achieving the
customer’s in-tolerance spectrum requirement with the test article
installed, an acoustic power margin evaluation as a function of
frequency may be performed by the test facility. The method for
this evaluation of acoustic power is discussed in this article. This
method was applied at the NASA Glenn Research Center Plum
Brook Station’s Reverberant Acoustic Test Facility for the SpaceX
Falcon 9 Payload Fairing acoustic test program.
Figure 1. ISO354 Interrupted Noise Method of determining rever-
Exposing aerospace test articles (spacecraft and payload fair- berant room decay rates.
ings) to simulated, launch-event, high-intensity acoustics is ac-
complished by one of two methods. specified at the one-third octave-band (OTOB) center frequencies.
Aerospace test articles may be exposed to a direct acoustic field To address the effects of the test article’s absorption on quan-
created by a shroud of dynamic speakers that are in close proximity tifying the maximum achievable SPL versus OTOB frequencies
to the test article. This test method is commonly referred as the of a company’s reverberant acoustic test chamber, the following
direct-field acoustic test (DFAT) method. parameters are needed:
Reverberant acoustic testing is the other method. With this more • The test chamber’s linear dimensions (width, depth, and height)
established method, a reverberant acoustic field is created within a and geometry. (If the chamber’s shape is not rectangular in each
large chamber and excites the aerospace test article installed in the plane, then chamber drawings should be provided. It’s also advis-
chamber. One advantage of using reverberant acoustic testing, over able to have drawings that include the layout of the chamber’s
the DFAT, is when a customer’s acoustic spectrum requirements horns and vents to determine those surface areas.)
exceed a 148 dB overall sound pressure level (OASPL). Such high • The reverberation time (RT60) values per OTOB for the empty
pressure levels are beyond the current capability of DFAT. chamber condition (no test article inside the chamber).
For reverberant acoustic testing, it is recommended1 that the • The reverberation time (RT60) values per OTOB for the chamber
volume of the test article is no greater than 10% of the empty with the test article installed.
chamber’s volume to ensure that the presence of the test article The RT60 values are determined from the measured decay rates
does not interfere with the spatial uniformity of the reverberant in the chamber and are equal to the time in seconds for 60 dB of
field. Many of the (approximately 25) active reverberation chambers decay to occur after the sound excitation source has been turned
throughout the world are capable of meeting the 10% test article/ off. If the RT60 values are not available, they may be calculated
chamber ratio volume guideline for many aerospace test articles. from a recording of the time-history decays of sound inside the
These test facilities all have their own unique shape, volume, wall chamber. The Interrupted-Noise Method4,5 is one approach often
surface absorption, and compliment of noise sources/coupled used to record and measure RT values. Because of the noise source
horns. Less unique is the sound medium in the chamber during excitation levels and the microphones’ dynamic range limits due
test, being either that of gaseous nitrogen (GN2) with its correspond- to their sensitivity, it is often necessary to extrapolate to a RT60
ing low oxygen level (typically < 10% oxygen), or low-humidity value from a lesser decay, such as a RT20 value.
compressed air. GN2 is the preferred gas due to its cleanliness and As an example with this Interrupted-Noise Method, an RT20
its lower absorptive characteristics above 4 kHz. calculation may be performed with a minimum of 35 dB of dynamic
The Reverberant Acoustic Test Facility (RATF) at the NASA range from the excitation level to the measurable background
Glenn Research Center Plum Brook Station in Sandusky, OH, is noise level. This 35 dB dynamic range is made up of: the required
an example of a large-volume (101,000 ft3), very powerful (163 dB 5 dB of early roll-off decay when the excitation noise source is
OASPL) reverberant acoustic test chamber2,3 that uses GN2. The turned off; a 20 dB of actual linear decay; and an allowance of a
front cover of this S&V issue shows a photograph of the SpaceX 10 dB margin above the background noise floor. This RT20 value is
Falcon 9 Payload Fairing in this NASA test chamber. The volume simply extrapolated to the RT60 value. (A direct measurement of
of the Falcon 9 test article was 10% of the NASA RATF’s empty RT60 would require a dynamic range of 75 dB from the excitation
chamber volume. level to the background noise level, which is often not attainable
This article describes a method to mitigate the test risk of achiev- in testing.) Figure 1 illustrates the Interrupted-Noise Method for
ing a customer’s target sound pressure level (SPL) spectrum during decay measurements.
reverberant acoustic testing due to the absorption inherent to the Before the delivery of the customer’s test article to the acoustic
test article. Each particular test article will have its own absorption test facility, it is common practice to perform initial empty-chamber
values that can be estimated as a function of frequency. Typically, testing to the customer’s specified acoustic target SPL spectrum.
both the target SPL spectrum and the test article’s absorption are This empty-chamber testing determines what complement of horns
and modulators are best suited to meet the spectral levels. This
Based on a paper presented at Noise-Con 2014, Fort Lauderdale, FL, Sep- target-level testing would be sufficient only if the test article’s
tember, 2014. absorption is assumed to be insignificant. If the test article’s

6 SOUND & VIBRATION/MARCH 2017 www.SandV.com

absorption is expected to be significant, then further attention is The reverberation time is obtained from the measured decay
required to address the additional acoustic power needed within rate, d in dB/sec (Equation 6):
the chamber to overcome this acoustic power sink.
60 (6)
Therefore, a method is needed to estimate these absorption ef- RT60 =
fects when the test article’s absorption is significant (i.e., accounts d
for the additional acoustic power required). The decay rate of the chamber with the test article will be larger
relative to the decay rate for the empty chamber test. The RT60 of
Method to Determine Additional Acoustic Power the chamber with the test article will be shorter than the RT60 for
The following steps are the approach taken to determine if the the empty chamber test. This change in RT60 indicates that ad-
empty-chamber test was performed with adequate sound power to ditional acoustic power will be necessary to meet the same target
ensure that the target SPL spectrum can be reached when testing spectrum with the test article in the chamber.
with the test article installed in the chamber. Step B – Obtain the Reverberation Times of the Empty Chamber.
Step A – Overview of Fundamental Acoustic Equations. Re- Perform tests to determine the reverberation time of the empty
view the fundamental acoustic equations6 related to a reverberant chamber condition (RT60EC) for each OTOB frequency and the
acoustic chamber test. subsequent calculation of the acoustic absorption (in units of
The SPL in the reverberant chamber is calculated by a power metric Sabines).
balance between the chamber’s input power and the power dissi- These RT60 decay measurements may be obtained by using the
pated due to absorption and other energy loss mechanisms within test facility’s existing noise sources (modulators/horns) via the
the chamber. The spatial averaged mean-square sound pressure, Interrupted Noise Method4,5 as previously discussed. Alterna-
<p2>, in the chamber is given by the Room Equation (Equation 1): tively, the interrupted noise method can be applied to obtain the
decay measurements by using conventional dynamic speakers as
4rc (1) the noise source. The dynamic speaker excitation source would
p2 = Wc
Rc be less costly from both a facility operations and labor resources
where Wc is the acoustic power input to the chamber in watts, Rc perspective, and it would also be easier and quicker to implement.
is the room constant in m2, r is the density of the gas within the Note that using dynamic speakers as the sound source instead of
chamber in kg/m3, and c is the speed of sound of the gas in m/s. the chamber’s noise generators may affect the relative decay rates
The room constant Rc defined in Equation 2, is determined from for frequencies above 1 kHz (where the gas dominates the total
the absorption of the chamber surfaces and from the gas absorption absorption), especially if GN2 is used as the sound medium for
and has units of metric Sabines (m2): the actual test article acoustic test.
Step C – Estimate the Test Article Absorption. As mentioned ear-
Rc = Â Sia i + 4mV (2) lier, it is necessary to include the effects of the test article’s acoustic
i
absorption when performing the empty chamber tests to ensure that
where Si is the area of each ith surface of the room, ai is the diffuse adequate acoustic power will be available when the time comes to
field surface absorption coefficient of each ith surface, V is the vol- perform the test with the test article inside the chamber.
ume of the room in m3, and the “m” is the energy absorption coef- Based on experience, evaluate the customer’s test article con-
ficient of the gas with units of inverse meters (m–1). For an empty figuration (shape, size, construction, materials, cavities, surface
chamber, the summation of the individual surface area absorptions area) and its expected acoustic absorption to estimate a delta dB
in metric Sabines accounts for all walls, floor, and ceiling surfaces, to be added to each OTOB of the customer’s target spectrum. To
plus the chamber’s horn and vent openings. account for the test article absorption, for example, +4 dB could
The (metric) Sabine equation (Equation 3) relates the room be added to the customer’s acoustic test SPL at each OTOB when
constant to the room’s reverberation time, RT60: performing the empty chamber test. Note that it is not necessary
that the estimated delta dB be uniform as applied to all OTOBs.
V 60
Rc = (3) Step D – Perform Empty-Chamber Test using Bolstered Test
1.086c RT60 Levels. Perform the empty-chamber test with a bolstered acoustic
The RT60 values equal the time in seconds in each OTOB for 60 test spectrum that includes the delta dB estimate from the previous
dB of decay occurring after the sound source has been turned off. step. Confidence to proceed with the actual test article’s full-level
The room equation (Equation 1) combined with Sabine’s equa- test is achieved only if the bolstered SPLs can be reached success-
tion (Equation 3), results in an expression of the mean-square fully in this empty-chamber test.
acoustic pressure as a function of the reverberation time, as pro- If the bolstered SPLs were not achieved for any OTOB, then
vided in Equation 4: other modulator/horn combinations, drive-signal filter bandwidth
settings, and modulator pressure flows may be adjusted to reach
4.343 rc 2
p2 = Wc RT60 (4) the bolstered SPLs. If none of these adjustments work, then it is
60 V necessary to discuss the situation, maximum achievable SPLs and
where, Wc is the acoustic power input to the chamber in watts, assumptions (of Step C) with the customer. This test should be
RT60 is the reverberation time of the chamber in seconds, V is performed months before the arrival of the test article at the test fa-
the chamber volume in m3, r is the density of the gas within the cility to mitigate risk and allow time to achieve the bolstered SPLs.
chamber in kg/m3, and c is the speed of sound of the gas in m/s. Step E – Perform Empty-Chamber Test using Bolstered Test
The compact relationship between the SPL and the acoustic Levels, Again. Assuming the previous step was successful, it is
power level (PWL) in the chamber can be expressed in decibels prudent to repeat this bolstered SPL empty-chamber test shortly
(dB) by Equation 5. This equation is valid for metric units and when before the test article is placed inside the acoustic chamber. This
GN2 is the gaseous sound medium in the chamber: retest provides confidence that the chamber’s acoustic character-
istics have not changed over time. This also provides a timely
SPL = PWL + 10 log (RT60 ) - 10 log (V ) + 14.17 dB (5) check-out of all the necessary acoustic control and noise generation
equipment and systems.
where, SPL is the measured sound pressure level in the chamber Step F – Obtain the Reverberation Times of Chamber with Test
in dB relative to 20 micro-Pascals, PWL is the measured acoustic Article Installed. Assuming success for Step E above, perform a
power level in the chamber, in dB relative to 1 picowatt, RT60 is low-level (–12 dB from the customer’s full-level target spectrum)
the reverberation time for sound to decay 60 dB, in seconds, and with the test article installed in the chamber. At the end of this
V is the volume of the test chamber in m3. –12 dB exposure, continue to record the microphones’ decay rates
The introduction of the test article affects the above equations to derive the RT60 in each OTOB for the chamber with the test
by decreasing the chamber’s volume (V), and increasing the total article (RT60Cw/TA).
surface absorption due to the Siai contribution of the test article. Note that Step F can alternatively be performed using conven-

www.SandV.com SOUND & VIBRATION/MARCH 2017 7

tional dynamic speakers in lieu of the chamber’s noise modulators
and horns. Using the dynamic speakers for this step should result
in a savings of effort, schedule and cost.
Step G – Calculate Actual Delta dB. Knowing the reverberation
times from Steps B and F, one can now calculate the actual delta
SPL values needed to overcome the test article’s absorption. Rear-
ranging Equation 5 to solve for PWL for the two tests conditions
of: (1) empty chamber (EC), and (2) chamber with test article (C
w/TA), results in Equations 7 and 8.
PWLEC = SPL EC - 10 log (RT60 EC ) +
(7)
10 log (VEC ) - 14.17 dB

PWLCw /TA = SPLCw /TA - 10 log (RT60Cw /TA ) +

(8)
10 log (VCw /TA ) - 14.17 dB
Combining Equations 7 and 8 by setting PWLEC = PWLCw/TA
(modulators produce the same amount of power for both cases),
one can solve for the change in SPL, from hereafter known as “delta
dB.” The calculated delta dB is provided in Equation 9.

SPL EC - SPLCw /TA = -10 log (RT60Cw /TA ) + Figure 2. Illustration of delta SPL dB for two notional test articles.
(9)
10 log (VCw /TA ) + 10 log (RT60 EC ) - 10 log (VEC ) articles’ absorption during their empty-chamber tests. A notional
representation of the two test articles’ calculated delta dB (from
Defining the volume of the chamber with the test article as: their low-level test measurements) is also shown in Figure 2. The
calculated delta dB for Test Article A is below the applied predicted
VCw /TA = (VEC - VTA ) delta SPL dB, and therefore the full-level testing can proceed
(10) with confidence, with minimal risk relative to the test chamber’s
where VTA = actual volume of test-article
available acoustic power. However, the situation is different for
and substituting Equation 10 into Equation 9 results in: Test Article B where the calculated delta dB exceeds the applied
predicted delta dB in two of the OTOBs (at the 160 Hz and 200
SPL EC - SPLCw /TA = 10 log (RT60 EC ) - 10 log (RT60Cw /TA ) + Hz OTOBs). In this case, options would be discussed with the
(11)
10 log (VEC - VTA ) - 10 log (VEC ) customer on how to proceed; primary to these discussions would
be the amount of the exceedance (0.5 dB would be more tolerable
Equation 11 may be reduced to Equation 12 as follows: than say 2 dB).

SPL EC - SPLCw /TA = 10 log (RT60 EC / RT60Cw /TA ) + Summary

(12) It is important to realize that some test articles may have sig-
10 log ((VEC - VTA ) / (VEC ))
nificant sound absorption that could challenge the acoustic power
With the definition of “delta dB” = SPLEC – SPLCw/TA, this may capabilities of a test facility. Therefore, to mitigate this risk of not
now be reduced to Equation 13: being able to meet the customer’s target spectrum, it is prudent to
demonstrate early on an increased acoustic power capability that
delta dB = 10 log [(RT60 EC / R60Cw /TA ) ¥ compensates for this test article absorption. This article describes
(13)
(1 - (VTA / VEC ))] a concise method to reduce this risk when testing aerospace test
articles that have significant absorption. This method was success-
For the special case when the volume of the test article (VTA) is fully applied during the SpaceX Falcon 9 payload fairing acoustic
less than 10% of the volume of the empty chamber (VEC), Equation test program at the NASA Glenn Research Center Plum Brook
13 simplifies to the approximation given in Equation 14. Station’s reverberant acoustic test facility.
delta dB ª 10 log [ RT60 EC / RT60Cw /TA ] (14)
Acknowledgements
By using the derived Equation 13, the actual delta dB is com- The authors would like to thank Dr. Jerome Manning and Dr.
puted. If this actual delta dB does not exceed the estimated delta Patricia Manning from Cambridge Collaborative Inc. for their tech-
dB (from Step C) that was applied in the empty-chamber testing nical contributions to this article. Appreciation is also given to Mr.
(Steps D/E), the risk of having potential deficient acoustic power Jesse Oliver and his SpaceX Falcon 9 payload fairing team for their
due to the test article absorption has been successfully mitigated. cooperation and assistance when implementing this methodology
In this scenario, proceed with the customer’s test plan to eventually during the test program.
expose the test article to the full-level target spectrum.
However, if this actual delta dB (from Step G, Equation 13) References
exceeds the estimated delta dB (from Step C) that was applied in 1. High Intensity Acoustics Testing, Institute of Environmental Sciences and
Technology, IEST-RP-DTE040.1, January 2003.
the empty-chamber testing (Steps D/E) in an OTOB, then options
2. R. N. Sorge, “Space Power Facility – Capabilities for Space Environmental
need to be discussed with the customer for this scenario. Note that Testing Within a Single Facility,” NASA/TM-2013-217816, November
having an exceedance is not a statement of the facility’s limitation 2012.
to reach these higher SPLs, but rather that its capability to do so 3. W. O. Hughes, M. E. McNelis, A. D. Hozman, and A. M. McNelis, “The
Development of the Acoustic Design of NASA Glenn Research Center’s
has not yet been demonstrated. As such, one option would be
New Reverberant Acoustic Test Facility,” NASA/TM-2011-217000, July
to subtract this exceedance from the customer’s target spectrum 2011.
(lowering the target SPL for the affected OTOBs). This would result 4. Acoustics – Measurement of Sound Absorption in a Reverberation Room,
in testing closer to the lower test tolerance value of the customer’s 2nd Edition, ISO 354:2003(E), May 15, 2003.
5. Acoustics – Measurement of the Reverberation Time of Rooms with Refer-
test target spectrum. Another option, although programmatically
ence to Other Acoustical Parameters, ISO 3382:1997(E), 1997.
undesirable, would be to remove the test article from the chamber 6. J. E. Manning, “Risk Factors Associated with the Prediction of Large Re-
and repeat the empty chamber measurements (Step B) with the verberant Chamber Acoustic Levels Part 1: Chamber Absorption Values,”
now-known delta dB. Cambridge Collaborative Inc. Report CC-13964-4, Rev 5.1, 2010.
Examples of these two scenarios are illustrated in Figure 2,
where a +4 dB of delta dB was predicted/used for both of the test The author can be reached at: aron.d.hozman@nasa.gov.

8 SOUND & VIBRATION/MARCH 2017 www.SandV.com