Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >
Skip to content

Validation of CIVA software (September 2005)

   
Charles R A Schneider and Davide Kleiner

TWI, Cambridge CB1 6AL, UK

Extended abstract of paper given at the BINDT conference September 2005.

1. Introduction

Theoretical models can be used to support ultrasonic inspection programmes by:

  • allowing the design of the inspections to be optimised;
  • allowing inspections to be qualified more quickly and cost-effectively by avoiding the need for large numbers of test specimens;
  • providing physical insight into inspection data.

It is, however, necessary to validate such models and establish any limits to their validity, especially if the models are to be used to underpin the inspections of safety-critical plant.

2. Approach

This paper describes progress during the first year of a two-year project, which ultimately aims to validate the CIVA models for ultrasonic testing (UT) for clad pipe/plate. The project has approached verification of the CIVA software incrementally, by considering increasingly complex modelling problems. In this way, it is hoped that the generic benefits to different industry sectors will be maximised. The three stages of the two-year project are as follows:

Stage 1
Determine limits of model validity for unfocussed UT of smooth planar flaws in unclad material, by comparing the predicted signal amplitudes with those predicted by simpler models that have already been extensively validated against experiment.

Stage 2
Determine limits of model validity for focussed and/or phased array UT of smooth planar flaws in unclad material, by comparing the predicted signal amplitudes with those from experimental trials and with exact solutions (where available).

Stage 3
Review the data available from experimental trials of focussed and/or phased array UT of clad plate/pipe. Where appropriate, supplement these data by acquiring or manufacturing new representative test pieces containing realistic planar flaws (lack of fusion, lack of penetration) and collect inspection data from them using TWI's phased array system. Compare the experimental signal amplitudes with modelling predictions.

This paper describes progress to date on Stages 1 and 2 of the project in verifying the CIVA model for UT of unclad material. The remaining work (including Stage 3) will be reported in a future publication.

3. Results and Discussion

Stage 1 of the work (unfocussed UT of unclad material) has quantitatively compared various predictions from the CIVA model with:

  1. other semi-analytical models that have already been validated against experiment,
  2. some of the corresponding experimental results,
  3. TWI's 2D finite element model (in just two cases), and
  4. an exact solution for a simple on-axis beam profile [1] .

The CIVA model (which uses Kirchhoff theory to model the scattering from a flaw) was found to generally exhibits a similar level of accuracy to that of previously validated models based on Kirchhoff theory, whenever the latter models lie within their stated regime of validity. The results suggest that, at least in some cases, CIVA can provide useful results outside this regime. However, more extensive comparisons are needed before TWI can provide definitive guidance on CIVA's regime of validity.

Stage 2 of the work (focussed UT of unclad material) is still in progress. Initial results, from very simple beam profiles in water, show some promise, and are broadly consistent with the Stage 1 work on unfocussed beams.

4. Conclusions

  1. TWI has gained confidence in the validity of the CIVA software for modelling ultrasonic testing of unclad material with unfocused beams. We therefore regard the CIVA software as a strong candidate for the modelling of ultrasonic testing of clad pipe/plate with focused beams.
  2. The CIVA model (which uses Kirchhoff theory to model the scattering from a flaw) appears to exhibit a similar level of agreement with experiment as previously validated models based on Kirchhoff theory, whenever the latter models lie within their stated regime of validity.
  3. The CIVA model has a wide-ranging functionality, which few other codes can match, for modelling the ultrasonic NDT of clad components using focused beams.

Acknowledgements

We wish to acknowledge the support received from CEA staff in overcoming the initial difficulties we had in using the CIVA software.

References

  1. Krautkrämer J and Krautkrämer H: 'Ultrasonic testing of materials'. Springer-Verlag, New York, 1990. ISBN 0-387-51231-4.

For more information please email:


contactus@twi.co.uk