Model-Based Signal Processing Techniques for Ultrasonic Flaw Detection: Simulation Studies

Abstract

The ultrasonic signals observed in inspection processes can often be accurately predicted by suitable measurement models. These model predictions can be used to provide important information to guide the development of subsequent signal processing algorithms. Here such a hybrid use of ultrasonic modeling and signal processing is demonstrated in the context of the problem of detecting ultrasonic flaw signals in noise. In particular, we wish to apply this hybrid methodology as an initial approach to solving the problem of detecting hard-alpha inclusions in titanium alloys.The "hard-alpha" inclusions are known to be brittle regions of microstructure caused by oxygen or nitrogen contaminations. During high-stressed manufacturing process or inservice operations, they are likely to initiate cracking, which may subsequently leads to catastrophic failure of aircraft components [1]. Hence, early detection is desired. However, the inherently weak strength of signals from hard-alpha inclusions, complicated further by the presence of high-level correlated grain noise, have long rendered this a particularly serious inspection problem. Furthermore, the lack of appropriate test specimens has made the development and evaluation of detection techniques even more challenging.In our engineering approach toward this problem, we first utilize a measurement model to simulate signals for specific inclusions and superimpose on these noise traces obtained on real samples. We then show that it is feasible to construct matched fIlters to achieve significant signal-to-noise ratio (SNR) enhancement. Based on simulation studies of largescale data sets, we can then assess the matched filter performance and the detectability in terms of the receiver operating characteristics estimates. Examples of detection of different inclusion impedances and sizes are presented in combination with various experimental setups. The results of using split-spectrum technique are also included. A parallel effort of applying neural networks and statistical analysis to the hard-alpha problem can be found in these proceedings [2]. SIGNAL SIMULATIONThe manufacture of test specimens containing seeded hard-alpha inclusions is currently underway in a separate work [3]. Presently, it is believed that these hard-alpha inclusions have similar ultrasonic characteristics as those of weakly scattering inclusions, and can be simulated from models [4]. Here the Thompson-Gray measurement model [5] is employed to simulate the hard-alpha inclusion signals. This model simulates flaw signals through modeling of the entire ultrasound propagation process using theoretical calculations of flaw

Spatial Correlation Coefficient Images for Ultrasonic Detection

“…The efficacy of this approach as applied to ultrasonics is directly related to the quality of the template; in essence, the efficacy is controlled by the operator's ability to predict the signal distortion caused by scattering at an unknown defect. Specifically as related to defect detection in ultrasonics, Xin et al [7], Chiou et al [8], and Srinivasan et al [9] address the difficulties in estimating the template along with the use of model-based approaches for template estimation. Eriksson et al [10] address the problem of correlation-based timeof-flight estimation when the signals have undergone unknown, nonlinear distortions.…”

Creating Noise-Corrupted Flaw Signals for Ultrasonic Nondestructive Evaluation

“…In this paper, an approach is presented for creating "known flaw" noise-corrupted flaw signals which was frrst introduced by the authors in 1985 [8] and which has been used in a number of studies since that time [9][10][11]. Significant inspiration for this approach came from the work of Gray and Thompson in creating noise-free flaw signals [4,5].…”

New ultrasonic signal processing techniques for NDE applications

“…It has been widely used in the area of communications for signal detection in noise. The effectiveness of the matched filter technique for ultrasonic flaw detection was studied by Chiou et al (1993aChiou et al ( , 1993b. Another approach to the detection of signals in noise is Fourier analysis or spec tral estimation.…”

“…Two sets of ultrasonic data were prepared for studying the hard-alpha detection problem (Chiou et a/., 1993a). Important details regarding these data sets are given in Table 3.1.…”