The project was decomposed into three technical workpackages (WP1: Innovative acoustic liner; WP2: Modal detection solution; WP3: Validations and demonstration at system level) and one workpackage dedicated to Management, Dissemination and Exploitation (WP4). In the first workpackage, several concepts of liners with low-frequency- and middle frequency-attenuation capabilities (hereafter denoted as LF and MF liners, respectively) have been designed by ONERA. POLY-SHAPE and ATECA performed an analysis of manufacturing solutions with regards to the industrial constraints given by Liebherr. The costs, manufacturing robustness, temperature resistance, mass issues and associated risks were addressed. Consequently, the technology chosen for the manufacturing process started at a TRL 2 rather than the TRL3 level initially planned. Indeed, a small Formlabs’ stereolitography machine had to be purchased by POLY-SHAPE, so that a brand new resin adapted to hot temperature could be used for 3D printing the LF parts. For the MF part, composite materials were combined with the additive manufacturing technique. Samples of the liner concepts were manufactured by POLY-SHAPE and ATECA, and then tested by ONERA in its laboratory facilities to assess their acoustic impedance, especially under flow conditions close to the industrial application. At the end of this experimental work, LF and MF liner concepts have been selected, both for a SDOF (Single Degree Of Freedom) and a DDOF (Double Degree of Freedom) architecture. In parallel, activities on in-duct modal detection were performed in WP2. ONERA numerical tools have been improved to take into account a shear flow profile in the modal detection process. Several designs of microphone rings were investigated. The best design was selected with regards to constraints on the number of microphone available for Liebherr and with regards on accuracy and robustness of the modal detection in the frequency range addressed in this project. A modal detection ring associated to this design was manufactured and used during an experimental campaign conducted in the Liebherr’s acoustic facility to assess the noise source of a realistic air pump. These results, as well as the liner impedance values measured in WP1, have then been used as inputs for numerical simulations of the acoustic attenuation that could be expected from the selected innovative liner concepts. As a result of this phase, and also taking into account the manufacturing risk assessment, it was decided to select the DDOF geometry for the manufacture of a full-scale prototype by ATECA and POLY-SHAPE. ATECA manufactured the MF parts while POLY-SHAPE manufactured the LFL parts by assembling several small modules that were 3D printed through stereolithography using the Formlab machine. At the end, ATECA assembled both MF and LF parts. The manufacturing of the full-scale prototype appeared to be very challenging, mainly because of issues in the assembly phase of a large number (almost 100) of small-scale parts. The stereolithography process with the hot-temperature-resistant innovative resin was not mature enough to obtain mechanical parts with an acceptable quality for an efficient industrial assembly process. New 3D-printing machines that were delivered on the market mid-2019 would have solved this problem by making possible the manufacturing of the complete prototype in only two or three parts. ATECA and POLY-SHAPE managed anyhow to manufacture the prototype. During the second experimental campaign in the Liebherr’s facility, the prototype was installed downstream the air pump. Comparisons with the results obtained during the first test campaign allowed the assessment of the acoustic attenuation provided by the liner, all over the frequency range and for each acoustic mode. A significant noise reduction was achieved above 1kHz. However, spurious noise appeared at low frequency, which prevented the noise reduction from being as high as expected over the frequency range 200-5000Hz. This spurious noise seems to be due to both a discrepancy between the specified and actual geometry of the prototype, and to an unwanted noise source created by imperfect junctions within the prototype. These issues could be solved by a more mature manufacturing process, as explained previously.
