A processing system has been developed to meet increasing demands for detailed noise measurement of aircraft in wind tunnels. Phased arrays enable spatial and amplitude measurements of acoustic sources, including low signal-to-noise sources not measurable by conventional techniques. The Microphone Array Phased Processing System (MAPPS) provides processing and visualization of acoustic array measurements made in wind tunnels. The system uses networked parallel computers to provide noise maps at selected frequencies in a near real-time testing environment. The system has been successfully used in two subsonic, hard-walled wind tunnels, the NASA Ames 7- by 10-Foot Wind Tunnel and the NASA Ames 12-Foot Wind Tunnel. Low-level airframe noise that cannot be measured with traditional techniques was measured in both tests.
The MAPPS system is an end-to-end system designed to be used by researchers. MAPPS begins at the end of the data acquisition and storage and ends with the processed data visualization. This system is designed to be versatile and robust in its treatment of variable numbers of microphones, number and locations of processors, versatile calibrations, and visualization requirements. This versatility is designed into the system to provide for alternatives if components fail. These component failures will result in degraded results, but the results will still provide the researcher with information to meet their needs. MAPPS provides ease of use for processing and visualization with its system approach and graphical user interface (GUI). The system is designed so that the user may concentrate on research, testing, and data interpretation, instead of data and file manipulation.
An operational design goal for MAPPS was to provide sufficient results in near-real-time to allow the test director and researcher to make future run content decisions. The first operational test of MAPPS was in a recent Flap Edge test using a 100-element microphone array in the NASA Ames 7- by 10-Foot Wind Tunnel. The system had a 9-minute cycle from end of data acquisition to showing results on screen for 166 frequencies with 400 averages and a frequency resolution of 150 Hz. This cycle time was sufficient to obtain results from a small number of points for each run condition and to allow the test director to make model change and run condition decisions for the next run. Another operational design consideration was to have all the data processed and ready for examination by the next day. The ability to batch process multiple data points was also demonstrated at this test.
The MAPPS system (see figure) starts with a raw time history and produces a processed data file and computer visualization of the results. The input raw data file and the output processed data file are both stored in a widely-used self-describing, machine-independent binary file format called Network Common Data Format. Besides the basic data, these data files contain all the information related to the instrumentation, test setup, and test conditions. MAPPS includes many files and computer programs. The user sets all the processing parameters and initiates interactive processing through the Processing Control interface. The Processing Control interface reads the header information in the Raw Time History Data File. Processing parameters may be used from a file or set interactively based upon information derived from the header and displayed in the Interface. All the processing parameters are saved in the Process Settings File. The data processing occurs in four processes that run without user interaction once the Processing Control Interface or a UNIX script for batch processing starts the Control Process. The computer running the Control Process must have access to the Processing Settings File that was written by the Processing Control Interface. The Read Data Process must have access to the Calibration File(s) and the Raw Time History Data File. When processing is complete, the Output Process writes the results and header to the Processed Data File. The header of the Processed Data File contains all the information from the header of the Raw Time History Data File plus all of the parameters used to process the data. Data visualizing is done in a separate process. Visualization requires access to the Processed Data File and Model Projection File.