Adaptive Nonlinear Modeling for Ultrasonic Signal Processing

The purpose of this presentation is to introduce to the NDE community an empirical modeling technique which has been under development during the past thirteen years and applied particularly to many problem areas in the last five. The application of interest to today's audience is that of classification of flaw geometry from ultrasonic signals. Disciplines Materials Science and Engineering | Structures and Materials This 3. flaw characterization is available at Iowa State University Digital Repository: http://lib.dr.iastate.edu/ cnde_yellowjackets_1974/4 ADAPTIVE NONLINEAR MODELING FOR ULTRASONIC SIGNAL PROCESSING Anthony N. Mucciardi Adaptronics, Inc. Mclean, Virginia The purpose of this presentation is to introduce to the NDE community an empirical modeling technique which has been under development during the past thirteen years and applied particularly to many problem areas in the last five. The application of interest to today 1 s audience is that of classification of flaw geometry from ultrasonic signals. The project in which we will be applying this methodology will first be described and then I will spend the remaining portion of the talk giving a brief summary of the technique itself. This project is an eighteen-month program with the NDE branch of the Air Force Materials Laboratory. It has just begun six weeks ago. The main program objectives are to evaluate the efficacy of this particular signal processing methodology for characterization of material flaw descriptors. Our program consists of two tasks. Task 1 is to demonstrate the capability of classifying flatbottom holes in the range of 0 to 8/64ths (in steps of l/64th) from ultrasonic signatures. The second task is to infer fatigue crack length over a range of 0 to 250 mils. Other objectives of the program involve assessing the information content of ultrasonic NDT signals. The returned ultrasonic signal will be parameterized using different types of parameter domains that I will describe shortly. We will be interested in identifying those parameters that contain the most discriminatory information relative to distinguishing between different hole sizes or estimating different crack lengths.