Dale A. Schick Cassandra L. Dutt Acoustic Research Systems, Shepherdstown, WV
William C. Raynor Jeffery R. Pope Matthew D. Dela Cruz US Naval Research Laboratory, WashingtonDC

ABSTRACT

Several years ago, various studies were performed in the Reverberant Acoustic Chamber facility at the US Naval Research Laboratory (NRL) to examine the feasibility of augmenting the nitrogen-driven exciters with electrodynamic options. Following the positive results of the latest study, NRL decided to fully refurbish the chamber with a combination of commercial- off-the-shelf and bespoke acoustic devices by Acoustic ResearchSystems, making this retrofit the first completely electrodynamic reverberant chamber capable of launch-vehicle acoustic testing. This paper describes the details of the background studies which led to the chamber retrofit project including the challenges of implementation and post-retrofit chamber capabilities.

KEYWORDS: reverberant chamber, acoustic testing, spacecraft

INTRODUCTION

Reverberant Acoustic Testing Facilities [1] (RATF) have been traditionally used to qualify spaceflight hardware for launch by exposing the unitunder test (UUT) to similar input spectra as recorded or assumed by the launch vehicle manufacturer. As launch vehicles evolved, the specifications for payloads have also changed to include more high-frequency energy or to envelope several rockets when the payloadmanufacturer is unsure of which will be used. RATF chambers typically provided excitation between 30 and 3000 Hz, on average, but were unable to control the energy around the bandpass of the chamber, where many of the test specifications required input. Lower frequencies (between 20 – 80Hz) were also difficult to control in RATF chambers depending on their size and horn configuration due to characteristic room-modes. It has been beneficial, in these cases, for chamber operators to explore supplemental drivers to augment the existing excitation systems.

NRL CHAMBER

In the mid-2010’s, after almost 40 years of reliable service in providing high quality launch environment simulation for the flight qualification and certification of spaceflight hardware, the Vibro-Acoustic Reverberant Chamber Facility at the NRL in Washington D.C. was ready for a full refurbishment of its systems due to equipment end-of-life issues and parts obsolescence. The legacy facility designed in 1973 and commissioned in 1975, provided a state-of-the-art capability to simulate the vibration and high-intensity acoustic noise environment experienced by spaceflight hardware during the launch vehicle ascent, see Figure 1. The facility consisted of a 10,000 cubic feet (285 m³) (21.5 feet deep, 17 feet wide, and 27.4 feet high) fully reverberant type acoustic test cell, a LING A340 30,000 lb-force electrodynamic vibration shaker, machinery room, network of computers and amplifiers, 26,000-gal liquid nitrogen (LN2) storage facility, and an assortment of handling and test fixtures.

Figure 1 – Naval Research Laboratory RATF, circa 1989

Noise generation was provided by a gaseous nitrogen (GN2) distribution and control system consisting of a single low frequency horn fed by fourelectropneumatic transducers (EPT) and two small high-frequency floor-mounted EPT’s, capable of achieving an overall sound pressure level (OASPL) environment of 153 dB between 32 Hz to 10 kHz; controllable from approximately 32 Hz to 4 kHz, and with uncontrolled reverberant roll-off out to 10kHz, see Figure 2. The 30,000-lb force (2 in. stroke) shaker was mounted in the center of the chamber floor to provide random mechanical vibration excitation in the 20 Hz to 2000 Hz bandwidth in addition to acoustic excitation of test specimens. Additionally, the shaker could provide sine vibration from 1 Hz to 2000 Hz.

Instrumentation and Control of the chamber sound pressure level (SPL) was provided through a Spectral Dynamics 1500 acoustic controller connected to up to 12 microphones suspended within the chamber. For shaker vibration, a Spectral Dynamics 2550 provided control and limiting of up to 32 channels of accelerometer response. The facility had the capability to perform digital data acquisition of up to 300 channels using a HPVXI E1432 digitizer with I- DEAS® postprocessing.

To minimize loss of acoustic power due to absorption and to minimize transmission of acoustic energy to the surrounding area, the test cell enclosure was designed to remain rigid to an acoustic environment of 170 dB OASPL; and provide a minimum of 50 dB attenuation through the chamber walls.

To deliver the GN2, a complex system of evaporators utilized propane gas to evaporate liquid nitrogen (LN2) rapidly. This system of delivery required extensive maintenance and special safety precautions to keep personnel safe around the chamber before, during, and after use.

 As with many RATF’s, the post-test environment in the chamber is nitrogen rich and poses an asphyxiation hazard until proper ventilation can restore oxygen levels.

Figure 2 – Diagram of the Naval Research Laboratory RATF

BACKGROUND TEST – IN-CHAMBER AUGMENTATION

Previously published efforts to reinforce reverberant chambers with electrodynamic transducers focused around using a quantity of loudspeaker enclosures placed inside a chamber to excite the room alongside the existing GN2 driven horns [2], as shown in Figure 3. The expectation was to use the electrodynamic loudspeakers while controlling with a narrowband MIMO control strategy to enhance the frequency response of the existing electropneumatic horns and 1/3oct-band acoustic control system inside the chamber.

Figure 3 – Test Configuration for In-Chamber Speaker Augmentation Experiment

 While this approach did offer improvement to frequency response by providing control for certain frequency bands, especially in the frequencies below 40hz, one of the primary issues was that the large wooden loudspeaker enclosures made a significant impact on the reverberation time (RT) of the reverberant chamber. This was exacerbated by the need for the DFAN system to utilize multiple different types of enclosures (Mid-High, Sub-bass, etc.) to address the entire active frequency range of the chamber. As a result, this study concluded that this configuration would only be plausible for tests which were below 135 dB OASPL.

The results of utilizing an electrodynamic system to drive the response of the chamber showed promise, but the downsides of this approach –namely the reduced reverb-time and challenges of operating in the chamber with minimal floor-space – outweighed the benefits at the time of this study and the augmentation of the chamber with electrodynamic loudspeaker enclosures was not explored further.

BACKGROUND TEST – ON-CHAMBER AUGMENTATION

If an electrodynamic system were to realize the proposal of replacing the nitrogen driven system, the system’s concentration of energy would needto be sufficient that it would require minimal surface area. It was thusly proposed to conduct a study that placed the devices on- chamber rather than in-chamber.

Figure 4 – Test Configuration for On-Chamber Augmentation

In 2021, such a study was conducted jointly by Acoustic Research Systems (ARS), m+p International (MPI), and the U.S. Naval ResearchLaboratory (NRL). In this study, six ARS

©Neutron acoustic devices were used in two stacks of three and were placed in the opened chamber doors, as shown in Figure 4. A plywood barrier was constructed and installed in-place around the acoustic devices and foam was added to the rear of the barrier to aid in personnel safety. As in the prior in-chamber study, the on-chamber study was performed with the original LING A340 shaker present. The panel (not shown) was instrumented and suspended from an overhead crane positioned directly above the shaker.

The results of this test offered many conclusions, some of which led to a revised design and proposal to not simply augment the existing nitrogen system, but to completely replace it.

Firstly, the m+p Acoustic Control System (1/3 Octave control based on ANSI S1.11 [3]) demonstrated an ability to control the acousticenvironment even when the microphones were not directly in the path of the acoustic devices, offering “DFAN-like” control [4] from 25 Hz to 10 kHz in an RFAN environment. Furthermore, Multiple Input and Single Output (MISO) control in this environment produced an acceptably diffuse field, unlike MISO in DFAN applications. [4]

Secondly, the 120V supply power for this test was not adequate to provide the current needed to power the devices, reducing the overall power output.

Thirdly, the wooden frame and structure of the acoustic devices, even though they were not completely in the chamber, hindered the reverberant time of the room somewhat.

Figure 5 – 1/3 Octave Results of On-Chamber Background Test

Ultimately, the test was able to achieve ~145 dB OASPL evenly distributed in the chamber, as shown in Figure 5. This achievement, a full 10 dB over the previous in-chamber configuration study, was used as the basis for the revised design proposal.

REVISED PROPOSAL

In 2021, ARS proposed to NRL a chamber modification method that would allow for complete replacement of the nitrogen-driven system which was, at the time, nearly inoperable. The proposal consisted of ten full-range ©Neutron devices in two stacks of five in the corner that housed thelow-frequency horn and three stacks of three custom high-frequency (HF) devices, one stack in each of the remaining corners, see Figure 6. The HF sections would contribute somewhat to a Direct Field Acoustic Noise (DFAN) testing environment inside the chamber.

 The full-range acoustic devices in the proposal were modified such that any major exposed structural component in the horn-section was clad inaluminum, increasing the reflectivity and lessening the effect of the device on the chamber’s reverberant characteristics. The exposed horn section on the HF sections were also clad in aluminum.

Figure 6 – Proposed Chamber Layout

The proposal also included recommended modifications to the chamber, most of which were necessary to complete the installation and were as follows:

• Remove low-frequency horn
• Remove high-frequency generators
• Cut large hole to accommodate the 10 ©Neutron Devices
• Pour level concrete pad to support custom frames for the acoustic devices
• Resurface and repaint chamber
• Installation of 415V Transformer

 The proposal was accepted by NRL in 2021 for execution beginning FY2022 as an internal capital improvement project (CIP).

INSTALLATION AND COMMISSIONING

Prior to the installation, a significant amount of floor-space in the laboratory was recovered by removing the LN2 evaporation system, which consisted of large indoor evaporator unit fed by a sizable propane storage facility located outdoors. The removal of the low-frequency EPT, which extended well beyond the chamber as shown in the overhead view of Figure 2, opened significant space behind the chamber. In the controlroom, several large cabinets were removed and replaced by a dual-computer workstation which only required a single-cable connection to the amplifier units below.

With the hardware from the previous chamber excitation method removed and all recommended modifications made, the replacement hardware was installed, see Figures 7 and (Additional considerations for the installation are outlined in detail in subsequent sections.)

Figure 7 – Chamber Installation HF 3 & ©Neutrons – Rear Wall View

Figure 8 – HF Sections 2 & 3 Installation – Payload Access Door View

To confirm that the installation did not affect the reverberation time of the room significantly, an RT60 measurement was performed using the ©Neutrons as sources. Measurements were taken at eight locations each one third space off the walls of the chamber using a single mic andRational Acoustics’ Smaart software as both the source output and data acquisition system. The measurements were made solely for ARS to reference internally and were not ISO 3382 compliant. The impulse was a short burst of pink noise at 60dB above the measured noise floor while fully sealed. The results at each 1/3 octave band are presented in Figure 9 below.

The average reverberation time from 20 to 8,000 Hz was found to be 2.8 seconds, whichh will likely improve with the installation of a reflective aluminum closeout section between the acoustic devices and the chamber wall. At the time of testing, several layers of soundproofing in this area were exposed, likely significantly affecting the RT measurements.

Figure 9 – RT60 Measurement of RATF Post-Modification

The primary focus of the chamber commissioning was to determine the optimal control strategy at various levels of testing. Though no test specification is the same with respect to the requirements of each individual octave band and the Overall Sound Pressure Level (OASPL), it wasdetermined to run the highest-level test ever required by this chamber. This specification happened to be roughly 148 dB OASPL and was requiredby a piece of hardware developed at NRL for a past program.

Many configurations were tested to characterize the chamber, but the two most significant are reported here for simplicity and clarity: A four-drive and an 8-drive configuration. Since the acoustic devices used in this installation do not rely on constructive interference to operate, these two drive configurations represent an optimally incoherent option (8 drive) for the best uniformity of field and a more coherent option (4 drive) when extra power is required. The representation of the ten ©Neutrons and the three high-frequency corner installations are represented in Figure 10 below.

Figure 10 – Four-Drive vs. Eight-Drive Configuration

In the four-drive configuration, the level was achieved using less power and exhibited stable control, reaching the 0 dB portion of the test within roughly 30 seconds, which is a vast improvement over the duration required to build a stable pressure environment in a typical RATF. The distribution over the sixteen control microphones was like the original commissioning of the chamber (not included in this report,) with the bulk of the 1/3 Octave dispersion contained within a +/- 5 dB boundary, as shown in Figure 13. The Octave and PSD data, shown in Figure 11 and Figure 12, respectively, produced a similar result to the original chamber commissioning.

Figure 11 – 148 dB OASPL Octave Data – 4 Drive Configuration

Figure 12 – 148 dB OASPL PSD Data – 4 Drive Configuration

Figure 13 – Dispersion (in dB) – 4 Drive Configuration

In contrast, the eight-drive configuration produced a tighter grouping and was more like the results of a typical DFAN test. The grouping of theOctave and PSD data, shown in Figure 14 and Figure 15, respectively, are much improved from the original chamber commissioning. Similarly,the dispersion plot in Figure 16 contains the bulk of the energy in a +/- 3 dB spread. It should be noted that the placement of the microphones in thecommissioning phase was much more disparate when compared to the typical placement around the test article. The reasoning behind this placement method was to determine the overall usable space in the chamber.

Figure 14 – 148 dB OASPL Octave Data – 8 Drive Configuration

Figure 15 – 148 dB OASPL PSD Data – 8 Drive Configuration

Figure 16 – Dispersion (in dB) – 8 Drive Configuration

Since 290 m² were available for testing in this chamber, the space the test article occupied had to be considered for overall uniformity of field, as inany RATF. This is true for both historical testing in the chamber and post-modification of the chamber.

An observation made by the NRL personnel during the commissioning was that the operation of the chamber, when compared to that of the nitrogen-driven excitation methods, was drastically simplified in both complexity and time. Firstly, the safety factors involved in the operation of a nitrogen-rich environment were completely mitigated. Secondly, the time required to setup a test was reduced to minutes rather than hours or daysof preparation. Thirdly, the time required to step from lower levels to the ultimate test level are eliminated. Since the electrodynamic drivers are actively controlled and donot rely on chamber pressure to function at-level, the user may drive the chamber from ambient to full (0 dB) level in as few as 20 seconds, depending on the start level.

In contrast to the above, the ultimate chamber level of 153 dB OASPL has been reduced to roughly 149 dB OASPL, depending on the profile shape and drive configuration. It’s worth noting that very few modern launch vehicles require a test specification for anything above 148 dB, even for Protoflight requirements, unless the test article is located outside of a faring.

INSTALLATION CHALLENGES

All reverberant acoustic chambers are unique and therefore present unique challenges to consider before deciding to retrofit. The items listed here were of special significance in the installation phase of this project and may be useful to those considering such a retrofit in the future.

Typical RATF Concrete Construction

Traditionally, reverberant acoustic chambers are constructed using reinforced cast-in-place concrete. [5] Some cases of reverberant chambers constructed using fully grouted reinforced concrete masonry units (CMU) have been noted, though this condition appears to be atypical.

The thickness of the concrete walls could vary based upon several factors, such as the geometry of the chamber and the design maximum OASPL.Reinforcing bars should be expected in both the horizontal and vertical directions on each face of the concrete wall. Depth to reinforcing bars may vary based upon the best practices at the time of chamber construction. Variances in concrete reinforcing placement and depth should be expected.

Cutting Existing Concrete Walls

The size and location of the new opening in the concrete wall was dictated by chamber geometry and accessibility. Drawings were provided by ARS to NRL to define the required minimum opening size of 8’-9” W x 19’-6” H. Ultimately, the opening was widened to 9’-0” due to the location of existing rebar at the edge of the marked opening as seen in the rightmost photo of Figure 17.

Figure 17 – Process of Removing Concrete in Chamber Wall

A contractor was hired to create the opening in the existing concrete chamber wall. Due to the size of the opening, concrete was removed in approximately 3-foot square chunks starting at the top of the opening. Using a track mounted wet saw, all horizontal cuts were made. The track was then rotated and/or moved to accommodate the vertical cuts. A lifting plate was mounted to the segment of concrete being cut and connected to a temporary overhead hoist. Once the cuts were complete, the hoist was used to gently lower the concrete chunk to the ground. This process was repeated until the opening was complete.

To limit the spread of dust, the concrete was cut with a wet saw. Wet cutting concrete turns the dust into a concrete slurry which must be cleaned, contained, and disposed of appropriately which can present a challenge when working indoors or near cleanroom facilities.

In this case, it was determined by facility engineers that no reinforcing or structural header was required to accommodate the new opening. Note that, based upon existing site conditions and required opening size, structural headers or jambs may be required. For anyone considering a similar chamber modification, facility engineers should be consulted during the design phase of the project to verify requirements.

Removal of Existing Concrete Horn

The existing low frequency horn was removed as a part of the process of cutting the required wall opening. The horn was framed with ½”± thickstructural steel angles with studs embedded in the concrete on all sides, which appears consistent with traditional horn construction. As the bottom framing angle of the horn was cast into the floor slab, it could not be fully removed without removing a section of floor slab deep enough to remove the embedded anchors. The angle on the right side of the horn when viewed from the chamber was cut as flush to the inside of chamber wall as possible. However, a truly flush cut was not achieved which created complications in the placement and mounting of the acoustic insulation and close out cover.

 Where possible, it is recommended that the cut to remove the horn be oversized to remove the entirety of the steel frame and embedded stud. Note that there is likely to be a greater concentration of reinforcing bars around the horn framing which can present challenges in cutting or drilling.

Anchoring in Existing Concrete

The ©Neutron 1016i are walled mounted units. To provide flexibility in mounting, a series of vertical Unistrut channels were installed. Use of the Unistrut also limited the number of concrete anchors that had to be installed to support the units. This was a highly effective solution in the corner that was not adjacent to the chamber door.

It was quickly determined that the frame for the chamber door was cast into the concrete and extended more than 8” beyond the inside face of thedoor on each side. While technicians were able to drill into the concrete, the drill bits hit steel when the hole was approximately ¾” to 1¼” deep.The required depth for the specified concrete fasteners was greater than the depth of hole that could be drilled. This made mounting the Unistrut adjacent to the door challenging.

A series of aluminum plates were manufactured onsite to provide a structural connection from the Unistrut to the wall away from the existing doorjamb. Three (3) aluminum plates with (2) eccentric fastener holes were mounted to the backside of the Unistrut (top, middle, bottom). As thefastener holes extended several inches beyond the centerline of the Unistrut (away from the door jamb), technicians were able to drill and place concrete anchors to the manufacturer’s specified depth without hitting the embedded door framing.

In future efforts, ground penetrating radar (GPR) is a methodology that can be used to identify the location of existing reinforcing bars and otherembedded steel within the concrete walls. In many cases, however, this is unnecessary. In cases where structural drawings indicate high concentrations of reinforcing bars or structural steel elements may be cast into concrete, GPR could be utilized to determine the most appropriate mounting locations.

CONCLUSIONS

Primarily, this paper has shown that not only is the replacement of nitrogen-driven exciters in RATF’s possible, it may also be desirable in somecases. Due to the proximity requirement for DFAN tests and similar hardware used in such situations, there may be a limitation to the size of the chamber that such an installation is possible, but this requires further investigation. In this case, the control of the chamber was somewhat more uniform and could be more easily shaped to the specification between 31.5 and 10,000 Hz.

Additionally, the maintenance and overhead cost of an electrodynamically-driven chamber is drastically reduced when compared to a nitrogen-driven chamber.

Installation of such a system, as it requires drastic structural changes, is subsequently challenging. The cost of comparative repairs aimed at reviving legacy systems is objectively higher, overall, than the retrofit described in this paper.

Perhaps most importantly, the risk of test personnel is nearly eliminated due to the lack of use of Nitrogen. Immediately following a test, the NRL personnel can inspect the test article and proceed to subsequent tests with minimal time in between.

REFERENCES

Q. Z. S. W. J. Z. Xiaojun QIU, “A case study on the new reverberation room built in University of Technology Sydney,” Proceedings of the 23rd International Congress on Acoustic, 2019.

D. Hayes, “Reverberant Chamber Enhancement,” Journal of the IEST, vol. 58, no. No. 1, pp. 12, 13, 2015.

American National Standard, ANSI S1.11 Specification for Octave-band and Fractional- Octave-Band Analog and Digital Filters, Melville: Acoustical Society of America, 2004.

Schick, Hoffmann, Friedlander and Kinsella, “Verifying a Novel Approach to Acoustic Testing: Acoustic Devices & Control Systems,” in 32nd Aerospace Testing Seminar, Virtual, 2021.

P. C. Fagerman, A. W. Maine III, R. A. Noreen and R. M. Slone, “An Overview of Three New Acoustic Facilities,” in 3rd ESTEC Environmental Testing for Space Programmes, Noordwijk, 1997.

BIOGRAPHIES

Dale Schick is Director of Applications for Acoustic Research Systems. He has experience in environmental testing as applied to Aerospace andover 18 years of practical experience in the industry related to vibration, acoustic, shock, and thermal testing technologies.

Cassandra Dutt is a Staff Project Engineer with Acoustic Research Systems, providing structural and technical oversight for test campaigns andsystem installations. Cassandra has a Bachelor of Science from Lehigh University. A practicing structural engineer with 12 years of professional experience, she has worked on the design and construction of state-of-the-art entertainment venues, commercial and residential developments, and specialty structures for marine and aerospace applications. Since 2021, she has been involved in the design and analysis of structural rigging systems used in Direct Field Acoustic Testing and has acted as the structural engineering lead for several large scale DFAN test campaigns.

William Raynor is the Associate Superintendent of the Spacecraft Engineering Division at the US Naval Research Laboratory’s Naval Center for Space Technology. The division is responsible for conducting basic and applied research and development for spacecraft science and technology primarily focused on supporting US national security missions. Prior to becoming Associate Superintendent, from 1996-2021, Bill served as Section Head of the Spacecraft Processing & Test Section, responsible for planning and executing the mechanical fabrication, assembly, integration, functional & environmental testing, and launch processing of spaceflight systems. He has been employed at NRL since receiving his BSME from the University of Maryland in 1991. Bill is a member of the Aerospace Testing Seminar (ATS) advisory board (1998), a US delegate and working group convener to the International Standards Organization (ISO) for Space System Interfaces, Integration, and Test (2000), and AIAA Associate Fellow (2014).

 Jeff Pope is a Systems Test Engineer within the Spacecraft Engineering Division at the Naval Research Laboratory (NRL). He attended VirginiaTech and received a Bachelor of Science in Mechanical Engineering in 2015. Jeff has over 9 years of industry experience in vibration, shock, modal, and acoustic testing and has acted as the lead dynamic testing engineer for various programs at NRL.

Matt Dela Cruz is a Senior Engineering Technician for the Systems Test Section at the Naval Research Laboratory, Washington, D.C. Hegraduated from The Navy Nuclear Power School and is a 12 year veteran of the U.S. Navy’s Submarine Force. He has over 20 years of practical experience in environmental testing with a focus in Dynamics and Acoustics.