Some of the most demanding tests that a spacecraft undergoes during its proving cycle are the launch-environment simulations. ESTEC's LEAF, electrodynamic shaker facilities, and future Hydraulic Vibration Facility (HYDRA), are and will continue to be the established tools for simulating this environment. The LEAF simulates the airborne noise produced after engine ignition and during the atmospheric flight phase of the launcher. The launcher's structure-borne vibrations are simulated using the electrodynamic and/or the hydraulic facilities.
The LEAF acoustic facility, which was commissioned in 1990, was designed for the noise levels specified by the launcher authorities at that time for existing and future users. Since then the test levels demanded have decreased below those at which LEAF could produce accurate noise spectra. Deficiencies of up to 10 dBL were observed at frequencies above 2000 Hz, caused by the reduction in levels at lower frequencies giving insufficient noise spill-over to the higher frequencies. To eradicate this deficiency, additional high-frequency noise generators have been installed.
The main features of the LEAF are shown in the cutaway view in Figure 1. The chamber, which is 11 m x 9 m x 16.4 m high, with an internal volume of 1670 m³, has walls made of steel-reinforced concrete. They are 0.5 m thick to withstand the acoustic loads and to attenuate noise levels outside the facility. A thick coating of epoxy resin has been applied to the chamber's inside walls to reduce noise absorption and thus increase the reverberation time.
Figure 1. Cutaway view of the Large European Acoustic Facility (LEAF)
at ESTEC showing, from left to right, the main chamber, horn room, control
room, and gaseous-nitrogen system
The chamber is supported on rubber bearing pads to isolate it from the surrounding vibration-sensitive facilities and residential areas in the environs. A full-height, concrete-filled steel door seals off the chamber during the acoustic test. When open, it permits the insertion of Ariane-5 sized spacecraft.
A gaseous-nitrogen subsystem provides up to 11 kg/s of clean dry gas to drive the noise generators. It consists primarily of:
During operation, the liquid nitrogen is driven from the tank under a pressure of 10 bar and through the heat exchangers. These produce gas at 5 ± 0.5 bar and 20 ± 2°C. It takes 80 sec at maximum flow to achieve this gas pressure and temperature, during which time the unconditioned gas is vented to the atmosphere. Valves in the pipelines to the noise generators control the gas pressure within the 0.2 - 2.0 bar range.
The LEAF was originally specified to achieve an acoustic noise level of 154.5 dBL with a dynamic range of -10 dBL, i.e. a low level of 144.5 dBL. A future extension to 158.5 dBL was also foreseen. Today, the levels being specified are in the range 135 to 147 dBL. Details of the 1/3- octave spectrum required to achieve these levels are shown in Figure 2. Also worthy of note is the lower drop-off of the noise in the high-frequency range compared with the original specification.
Figure 2. Comparison of the new and the original performance
specifications for the LEAF
To achieve the original specification, acoustic powers of 120 kW for the 154.5 dBL and 180 kW for the 158.5 dBL levels were required. To eliminate the danger of contaminating the spacecraft with oil, it was decided to use electrodynamic rather than hydraulic flow modulators. These can also deliver more acoustic power than the hydraulic type. Four 30 kW WAS 3000 noise generators (Fig. 3) supplied by Wyle (Huntsville, USA) were therefore used. These were installed onto horns of 25, 35, 80 and 160 Hz lower cut-off frequencies in the chamber/horn room wall. Additional apertures were foreseen to install two further horns and noise generators to meet the 158.5 dBL requirements. The number of apertures was later increased to eight to allow a level of 162 dBL to be achieved in a reduced chamber volume for the testing of Ariane-5's engines.
Figure 3. The Wyle MU110 modulator, showing the spring, slots and
coil
When operating these noise generator/horn combinations, deficits of up to 10 dBL were observed in the 2 to 4 kHz frequency range (Fig. 4), due to the low noise spectrum in the lower frequencies reducing the noise spill-over to the higher frequencies. It became essential, therefore, to install additional high-frequency noise generators. No suitable systems were commercially available, but a novel solution has been developed, tested and patented by, among others, the staff of the NAE Aeroacoustics Facility (recently renamed IAR) in Ottawa, Canada .
Figure 4. The LEAF noise spectrum without the NAE generators
(cf. Fig. 8)
The NAE high-frequency generator is an aerodynamic device that generates intense tonal noise. A wall jet of gas flows through a rectangular nozzle and impinges on a resonant cavity. The depth of the cavity can be adjusted to give a range of fundamental frequencies for the given generator geometry. By itself, the NAE generator creates tonal noise at the fundamental frequency plus harmonics. The low-frequency noise emitted by the Wyle generators is used to modulate the high-frequency tones produced by the NAE generator. This results in a supplementary broadband contribution of noise that is added in the frequency range above that of the cut-off of the Wyle generator (Fig. 5).
Figure 5. The narrow-band effects of adding a 1/4-inch
generator
Until last year, however, the NAE generators did not have the high-frequency performance now required for the LEAF application. They covered a frequency range of 500 1250 Hz, whereas the LEAF required additional energy from 1000 to 4000 Hz. A conceptual design for a generator to cover this extended frequency range was therefore developed for ESA by Aiolos Engineering Corp. of Canada. A prototype was subsequently tested in the IAR chamber to validate the design. The tests also confirmed that direct control of the sound levels, with sufficient power, could be achieved.
To cover the exact LEAF frequency-range requirements, generators with 2-inch wide and 1/4, 1/2 and 1-inch high cavities were installed in the LEAF's 25 Hz horn (Figs.6 & 7). The results of commissioning tests completed in 1995, illustrated in Figure 8, confirmed that a wide range of spectral shapes can be obtained by various combinations of these modulators and generators.
Figure 6. The NAE high-frequency generator assembly
Figure 7. The NAE generators installed, with their motor
drives, in a section of the 25 Hz horn. They are connected to the
gaseous nitrogen supply via reinforced flexible pipes (centre
left). The Wyle noise generator is the brass-coloured cylinder
in the upper left of the photograph
Figure 8. The LEAF noise spectrum with the NAE generators (cf.
Fig. 4)
The acoustic testing of the Italian national satellite SAX in September 1995 (Fig.9) and of the Structural Model (STM) of ESA's large Polar Platform/Envisat spacecraft in July 1996 have conclusively demonstrated the improved fidelity of the upgraded LEAF facility.
Figure 9. The Italian SAX spacecraft being prepared for
testing in the LEAF
The appropriate acoustic spectrum for a given test is currently set up manually by the facility operator, prior to the introduction of the test specimen into the chamber. This process can take several days and consumes large amounts of liquid nitrogen. In addition, significant differences in the spectrum can occur due to absorption by the spacecraft when it is introduced into the chamber, calling for further manual fine tuning. This has to be done within 10 seconds of starting the test or the latter may have to be aborted to avoid over-/undertesting.
The implementation of an automatic PC-based spectrum control system is therefore under investigation, the advantages of which include:
This will provide the LEAF with the ability to conduct more accurate test simulations of launcher-specific environments more quickly and at reduced cost.
The NAE generators have significantly improved the noise levels in the LEAF in the high-frequency range, and the versatility of the facility in terms of the spectra achievable has been greatly enhanced with the Aiolos upgrade.
Any future work will be directed at further extending the capabilities of the LEAF as and when necessary to keep it in the forefront of environmental testing technology.
The authors would like to thank the staff of Aiolos Engineering Corp., especially G. Elfstrom and B. Clark, and IAR for their cooperation and hard work in this project. They would also like to thank B. Westley, formerly with NAE, for sharing his vast experience with them.