Introducing the eFlaw

Constant problem in NDE training and qualification is the availability and cost of relevant flawed test blocks. The test blocks and, in particular, the defects they contain should be representative to the actual inspection challenges that the inspectors are likely to encounter during field operation. However, manufacturing realistic test blocks and flaws is time consuming and costly. Consequently, the number and quality of test blocks is often limited. In training, where the required number of test blocks is high this leads to reduced representativeness of both samples and flaws they contain. In particular, EDM notches are often used instead of real cracks.

In qualification, the use of representative samples and flaws is considered indispensable. Consequently, the number of flaws limited by the cost of manufacturing. This is becoming increasingly problematic now that there's more demand to get quantitative performance data from qualification. This, in turn, necessitates statistically significant number of cracks and trials. Also, as the number of qualified inspectors and re-qualifications increase, growing number of data-sets are needed.

With the current trend towards automated inspection, there's now a new alternative. In automated inspection, the data gathering and analysis are separated to distinct steps. This separation also allows new possibilities for training and qualification. Since the analysis now operates on pre-recorded data the need for different physical training samples and training data-sets are also separated. The data gathering can be trained and qualified on physical samples while the more demanding data analysis can be completed on separate (possibly unrelated) data set.The needs for the two steps are different. For the data gathering, representative sample is needed, but the costly need for high number of representative flaws primarily concerns data analysis. Consequently, being able to modify the gathered data-sets to include non-existing virtual flaws offers several significant advantages:

  • the number of physical test blocks and flaws can be reduced,
  • the number of flaws in the data can be increased to give statistically significant results and
  • the number of different data-sets available can be increased so that every trainee or qualification candidate receives a fresh data set.

The challenges with virtual flaw's are to:

  • make sure the modified data set is representative and that
  • the modification does not introduce additional disturbance
Thus far, these problems have prevented the use of virtual flaws.

The Trueflaw eFlaw technology overcomes both of these problems. With eFlaw, a mock-up is first scanned in the unflawed condition. Then, the required flaw types are introduced with the acclaimed Trueflaw thermal fatigue process. Then, the sample is re-scanned in flawed condition. With these two data sets, the pristine flaw signal can be extracted for each flaw type. This signal can then be re-introduced to the unflawed data to various locations to produce any number of flawed data sets. The principle is illustrated in Figure 1.

Figure 1. Flaw extraction

The modified data contains real flaw signal acquired with the same process to be used in real inspection, only relocated and replicated to create variety of data sets. The two-scan process makes sure, that no signals unrelated to the flaw are copied. Furthermore, as the flaw signal is added to the new location, the signal gets superimposed to the local noise. Thus, every copy of the same flaw gives a bit different signal due to local sample variation: just as it should.

Contact Iikka Virkkunen (iikka@trueflaw.com, +358 45 6354415) to see how this new technology can help you.

More info

Example data