1st September, 2012
Mr. Gustavo Pinto Pires, Equipment Engineer at the CENPES, kindly accepted to answer our questions about CIVA.
CENPES – Petrobras' Research and Development Center, Brazil
What is the role of CENPES?
Provide technological solutions to the Petrobras’ system, acting in three key axes: expanding the limits, adding value and diversifying products, and guaranteeing sustainability.
What is your role in CENPES
I work as an equipment engineer in the inspection group of a department concerned with the integrity of Petrobras’ industrial installations. In our group, we adapt and develop original non-destructive techniques to address specific demands for which there are no available commercial solutions. These demands come from onshore and offshore units, both in emerged or submerged environments.
Since when is your department using CIVA?
Which kind of applications do you use it for?
To help us plan challenging inspection problems, and to evaluate the effectiveness of different approaches in terms of coverage and detectability. This can be accomplished with the aid of the raytracing tool for simple scenarios, and both the field calculation and the echo response modules for more accurate tasks.
To help better understand real collected signals by comparing them to simulated results or simply by checking the trajectories and arrival times using the raytracing tool.
And to generate synthetic signals to test post processing algorithms.
What did it change for you to use simulation?
Well, from time to time we face difficult inspection scenarios that demand sophisticated tools to handle them. The instinctive action is to seek for computational resources to help accomplishing these tasks. CIVA is one of these resources. But it is necessary to understand the limitations of any modelling/simulation tool, in order to be able to use them reliably in the working environment. Another important issue to consider is the amount of effort needed to learn and to use these tools. Are these required resources (human, computational and time) available and compatible with the demands in order to accomplish the inspection goals? Certainly, not in all situations. But there are some in which these tools play an important role.
In your opinion, what are the advantages of this software?
In a few words, it is meant to have inspectors as its main users. I read it, and I believe it is true, that most part of the code is devoted to the development of a user friendly interface. Besides, it is kept updated with technologies developed by CEA and partners (Universities and Technological Institutes) spread around the world. This combination seems to be a clever strategy to maintain its reputation as a reliable and easy to use tool for the NDT/NDE industry. But, at the same time, the research centers could benefit more from the software if it were conceived in a more opened way. I understand the commercial issues involved, but I think the strategy could be improved by allowing researchers or advanced users to have access to data that has been generated but is not available to users. Still in the same context, since a considerable amount of researchers use MATLAB in their environment, it would be very interesting if CIVA could provide MATLAB APIs to those wanting to go deeper in the potential of the tool, making it even more powerful and usable for scientific and technological purposes. For instance, CIVA could be used inside an external loop controlled by MATLAB, in which both inputs and outputs could be exchanged accordingly to the .m algorithm, in order to perform a specific task; for example, changing configuration parameter values in an optimization loop. It seems to me that there would be a considerable increase in the industrial usage and in the number of publications using CIVA after these improvements take place. Depending on the commercial issues involved, this might be implemented as an Advanced License, for example.
How did it help the CENPES?
It helped, and is still helping, in different tasks:
- Planning the configuration of phased array weld inspection of various components;
- Visualizing the viability of non-conventional approaches to inspections of components with complex geometries;
- Improving the layout of conventional transducers in inspection tools developed with specific purposes;
- Providing the geometries for mock ups and the distribution of intended defects inside them;
- Confirming hypothesis of possible trajectories of rays inside components and in the coupling medium.
You told me that you are going to study at the University of Cincinnati next year, are you going to keep on focusing on NDT Knowledge?
Yes. I will work with imaging techniques for ultrasonic signals. But unfortunately CIVA is not capable of providing the tools we will need because of the semi-analytical approximations, even if used solely in the propagation region.
In September, EXTENDE and the CEA will release the new CIVA-ATHENA 2D module. It is a coupling between CIVA and a Finite Element code. Such a combination of the two calculation methods keeps calculations fast. This module will take into account all the physical phenomena that will appear in a box defined by the user around the flaw(s). But the beam propagation out of the box will still be calculated with CIVA semi-analytical method. After the computation, you will be able to visualize the beam propagation interacting with the defect. Have you heard of it? If so, what do you think of it?
Yes, since 2008. I think it is very interesting. It will open a range of possibilities in terms of more realistic field modelling and echo response calculations, especially for non-canonical flaws and specimen models, releasing the inspector from the task of choosing in advance the interaction algorithm (Kirchoff or GTD) to use for each defect. It is a pity that it delayed so much from the scientific article mentioning this hybrid approach to the commercial version. I once saw a version of the CIVA software installed in a computer of a CEA (now Extende) staff member that had the link to the Athena module in 2009. Now, there are other research institutes in England and in India that also have developed similar hybrid algorithms, according to what I saw at the 2011 QNDE conference. But at least (and at last) we will have it commercially available now in CIVA! Good to hear that.