Electricity: measuring and testing – Magnetic – With means to create magnetic field to test material
Reexamination Certificate
2001-01-09
2002-07-23
Lefkowitz, Edward (Department: 2862)
Electricity: measuring and testing
Magnetic
With means to create magnetic field to test material
C324S220000, C324S240000, C702S038000
Reexamination Certificate
active
06424151
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method and an apparatus for evaluation of an eddy current testing (or ECT) signal. More specifically, the invention relates to the method and apparatus useful when applied to flaw detection using an eddy current testing sensor on a multi-coil system, a rotation system, or a two-dimensional scanning system (hereinafter referred to as the multi-coil system or the like).
2. Description of Related Art
Eddy current testing is known as a method for nondestructive testing of a metal. This method involves generating an eddy current in a member to be measured, by a magnetic flux generated by a coil supplied with an exciting current, and obtaining an eddy current testing signal based on a magnetic flux generated by the eddy current as an output signal from the coil. The eddy current testing signal reflects the location, shape, depth, etc. of a flaw existing in the member to be measured. Based on this eddy current testing signal, the member to be measured, which is a metal (magnetic material), can be inspected nondestructively.
To test, for example, a heat exchange tube for a flaw by the eddy current testing method, a coil, which is an eddy current testing sensor, is inserted into the heat exchange tube to obtain an eddy current testing signal. Such a signal representing a flaw changes not only in amplitude, but also in phase. Thus, observation of the eddy current testing signal in one-dimension is not sufficient, and its two-dimensional observation is necessary. Hence, the eddy current testing of the heat exchange tube uses an eddy current testing apparatus which generates a two-dimensional output appearing along an X-axis and a Y-axis. An eddy current testing signal expressed on a voltage plane draws a Lissajous' figure as shown in FIG.
7
.
Such a Lissajous' figure is characterized by its size and its slope relative to the X-axis. That is, the size of the Lissajous' figure is proportional to the volume of the flaw, and its slope relative to the X-axis corresponds to the depth of the flaw. In the eddy current testing of a heat exchange tube, the depth of the flaw is important information. Thus, flaw detection is performed in a predetermined manner based on the phase of the eddy current testing signal. In detail, the phase angle of the eddy current testing signal (a value as a complex number) representing a flaw is measured. The measured phase angle is mapped on a characteristic curve (prepared beforehand) as shown in
FIG. 8
which illustrates the relationship between the phase angle of an eddy current testing signal and the depth of a flaw. Based on a reading taken from the characteristic curve, the depth of the flaw is estimated.
As described above, eddy current testing according to the earlier technologies measures the phase angle of the eddy current testing signal, and maps the measured value on the prepared characteristic curve illustrating the relationship between the phase angle of the eddy current testing signal and the depth of the flaw to estimate the depth of the flaw. However, the phase angle and the depth of the flaw do not necessarily correlate exactly, and the accuracy of flaw detection may be practically insufficient. This is because, given the same depth of the flaw, the phase angle may vary according to various factors, such as the shape of the flaw (e.g., length, width) and the relative positional relationship between the flaw and the coil. Particularly in the case of an internal flaw (a flaw on an inner peripheral surface of a heat exchange tube) showing a low rate of change in the depth of flaw in comparison with the rate of change in the phase angle of the eddy current testing signal, the accuracy of flaw detection may often be problematical.
SUMMARY OF THE INVENTION
The present invention has been accomplished in light of the problems with earlier technologies. An object of the invention is to provide a method and an apparatus for evaluation of an eddy current testing signal, the method and apparatus being capable of improving accuracy in evaluating the depth of a flaw detected by eddy current testing, and increasing accuracy in evaluating the amount of a decrease in the wall thickness of a member to be measured, as well as accuracy in discerning a false signal.
To attain the above object, the invention is characterized by the following aspects:
1) A method for generating a feature amount from an eddy current testing signal, comprising:
generating the feature amount based on the eddy current testing signal obtained by measuring a member to be measured, the feature amount being a numerical expression of not only a phase angle of the eddy current testing signal highly correlated to a depth of a flaw, and an amplitude of the eddy current testing signal, but also a feature highly correlated to a secondary factor which is other than the depth of the flaw and which affects a waveform of the eddy current testing signal.
According to this aspect, signals around a peak of the eddy current testing signal can be incorporated into the feature amount. This makes it possible to quantify an influence which elements affecting the evaluation of an eddy current testing signal, i.e., elements becoming the cause of a noise when the depth of a flaw is evaluated by use of the phase angle and the amplitude alone, exert on the evaluation of the flaw depth. Consequently, this aspect of the invention can provide data for more accurate determination of the depth of a flaw on the basis of an eddy current testing signal.
2) A method for evaluation of an eddy current testing signal, comprising:
generating a feature amount based on a sample eddy current testing signal obtained by measuring a standard specimen as a member to be measured and of a known depth of a flaw, the feature amount being a numerical expression of not only a phase angle of the sample eddy current testing signal highly correlated to the depth of the flaw, and an amplitude of the sample eddy current testing signal, but also a feature highly correlated to a secondary factor which is other than the depth of the flaw and which affects a waveform of the sample eddy current testing signal;
generating an evaluation parameter by learning with use of the feature amount, the evaluation parameter being a parameter for outputting a value with a sufficiently small error relative to known correct answer data as an amount expressing the depth of the flaw of the standard specimen;
generating a feature amount similar to that from the sample eddy current testing signal by use of an actual measurement eddy current testing signal obtained by eddy current testing of a member as an object to be measured; and
generating data representing the depth of the flaw on the basis of the feature amount based on the actual measurement eddy current testing signal, and the evaluation parameter corresponding to the feature amount.
According to this aspect, a feature amount as a numerical expression of features effective for evaluation of the depth of a flaw is statistically processed to generate a single evaluation parameter. The use of such a single evaluation parameter enables the depth of the flaw to be evaluated based on an actual measurement eddy current testing signal. Consequently, this aspect of the invention can markedly improve accuracy of evaluation in comparison with a judgment of the depth of the flaw based only on the phase angle and amplitude of the actual measurement eddy current testing signal.
3) The method for evaluation of an eddy current testing signal as described in 2), comprising:
classifying a type of the flaw, such as an external flaw or an internal flaw, based on the sample eddy current testing signal, and generating a similar feature amount and a similar evaluation parameter according to the classified type; and
classifying a type of the flaw based on the actual measurement eddy current testing signal obtained by eddy current testing of the member to be measured, and performing generation of a feature amount according to the class
Asada Yoshihiro
Kawata Kayoko
Kurokawa Masaaki
Kinder Darrell
Mitsubishi Heavy Industries Ltd.
LandOfFree
Method and apparatus for evaluation of eddy current testing... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for evaluation of eddy current testing..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for evaluation of eddy current testing... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2899374