Location

Snowmass Village, CO

Start Date

1-1-1995 12:00 AM

Description

In this paper, we assess the feasibility of using an adaptive spatial averaging technique for general ultrasonic flaw detection, and in particular, for the detection of hard-alpha inclusions in aircraft titanium alloys. Fig. 1 depicts the different components and processes in a ultrasonic sequential detection procedure. In order to achieve the best overall performance of this procedure, it is necessary to optimize each of the processes. Toward solving the hard-alpha problem under the Engine Titanium Consortium [1], the implementation of a multi-zone system [2] represents the current attempt to optimize the processes in the data acquisition stage. For the optimization of the other processes in the succeeding stages, a signal and image processing subtask has been identified and established. Within this subtask, the research team at Iowa State University is undertaking a series of investigations which involve various approaches including signal processing [3–5], statistical analyses [6–7], and artificial intelligence [6–7]. Earlier work in signal processing was focused on the traditional processing of individual A-scan waveforms (top route in Fig. 1). Presently, the effort is extended by incorporating scan plan relationship into signal filtering to further enhance the flaw detectability. Conceptually, this extension is closely related to the statistical K-sample inference method described in [7].

Book Title

Review of Progress in Quantitative Nondestructive Evaluation

Volume

14A

Chapter

Chapter 3: Interpretive Signal Processing and Image Analysis

Section

Signal Processing

Pages

727-732

DOI

10.1007/978-1-4615-1987-4_90

Language

en

File Format

application/pdf

Share

COinS
 
Jan 1st, 12:00 AM

Ultrasonic Signal-To-Noise Ratio Enhancement Using Adaptive Filtering Technique

Snowmass Village, CO

In this paper, we assess the feasibility of using an adaptive spatial averaging technique for general ultrasonic flaw detection, and in particular, for the detection of hard-alpha inclusions in aircraft titanium alloys. Fig. 1 depicts the different components and processes in a ultrasonic sequential detection procedure. In order to achieve the best overall performance of this procedure, it is necessary to optimize each of the processes. Toward solving the hard-alpha problem under the Engine Titanium Consortium [1], the implementation of a multi-zone system [2] represents the current attempt to optimize the processes in the data acquisition stage. For the optimization of the other processes in the succeeding stages, a signal and image processing subtask has been identified and established. Within this subtask, the research team at Iowa State University is undertaking a series of investigations which involve various approaches including signal processing [3–5], statistical analyses [6–7], and artificial intelligence [6–7]. Earlier work in signal processing was focused on the traditional processing of individual A-scan waveforms (top route in Fig. 1). Presently, the effort is extended by incorporating scan plan relationship into signal filtering to further enhance the flaw detectability. Conceptually, this extension is closely related to the statistical K-sample inference method described in [7].