Interview - Mariana Burrowes Guimarães

20th September 2022


Today, we have the sincere pleasure of interviewing Mariana Burrowes Guimarães, researcher at the Federal University of Rio de Janeiro.


Mariana Burrowes Guimarães


You are a researcher at the Federal University of Rio de Janeiro (Brazil). Can you explain what the main activities of your lab are, especially in NDT?

In 2009, our lab, LNDC (Laboratory of Non-Destructive Testing Corrosion and Welding), was built from a partnership between Petrobras, a Brazilian oil & gas company, and the Federal University of Rio de Janeiro, aiming to deal with the challenge of producing oil & gas from the pre-salt area. Some of these challenges are the depth of the oil reservoir, nearly 7.000 meters below sea level, and the presence of H2S and CO2 in the oil. It has almost 8.000 m² of built area which links the triad: welding, corrosion, and NDT. It has today about 60 employees (administrative, engineers, industrial chemists and technicians) and over 20 students (under graduation, masters and Ph.D. levels). LNDC has a symbiotic relationship with the industry, as it receives technical demands and at the same time, foments industry with state-of-the-art scientific input to upscale. Specifically, the NDT group comprises of 4 researchers, two professors and two technicians. We provide services regarding standard conventional NDT, but our main expertise is to develop new features regarding new materials through non-conventional techniques such as thermography, magnetic characterization, ECT Array, SLOFEC, ACFM, CT, Phased Array and others. Besides all experimental framework, our team has various research lines dedicated to post-processing analysis and data analysis.


Could you describe your role and your main research topic in the lab?

My line of work is mainly theoretical. I have dedicated the last ten years of my research to understanding reliability and its importance to structural integrity maintenance. To do so, I have analysed purely experimental industrial data, purely simulated data and hybrid data. We also have the expertise of applying parametric and non-parametric reliability quantification methods. Lately I have been dedicating my time to understanding how to apply a Bayesian approach on reliability studies using simulation tools to provide the "prior" information. Besides that, as the two main topics of what I think is worth working on are understanding the concept of probability of rejection instead of (or in addition to) probability of detection, and going deep on how to properly describe sizing flaw accurately. The last few years, I have been counting on international partnerships that not only brought those important matters to my attention, but have been working with me to better understand them. I can mention people from the EXTENDE team such as Fabrice Foucher, and my associate and friend Ed Ginzel who is THE expert on UT and everything that relates to it.


How does CIVA help you in your different activities? How often do you use it?

I use CIVA on daily basis. Our NDT group uses simulation as a tool for scientific state-of-the-art research, robust R&D projects, but also to simply simulate characterization and development of new probes, for example. We work with almost all CIVA modules, but the ones we use the most are the UT modules (inspection and beam simulation), parametric and POD studies. The SHM module is a relatively recent CIVA module that shows how the software is focused on keeping up with technical improvements. At the beginning of my life as a reliability researcher, I had experimental data that I could use to validate CIVA simulations and POD curves. From that moment on, CIVA has become a trustworthy tool for all my studies. So, in that way, I used CIVA to understand how small changes on experimental setup and statistical changes on uncertain parameters definition would affect POD curves. Regarding POR analysis, CIVA could provide estimated sizing values, since it takes advantage of CIVA Script and metamodeling to do so, and it would also help understand how small changes on flaws size would affect POD and POR.


Do you think that simulation will take a larger place in the NDT community in the future, and why?

I think it is already taking a huge place in the NDT community. The more simulations are validated with experimental and mathematical grounds, the more NDT fellow researchers and the industry itself are not only believing, but requesting simulations results to back up decision making. It is a curious thing, because a few years ago, not that many years, we had to convince our researchers, partners and clients to consider NDT simulations as valuable results. Especially regarding reliability simulations, because the very regular experimental was already repulsed for its cost and specificity. I remember including simulated POD curves for free many times to help spread the word of simulated reliability studies. Nowadays, I observe a very different scenario when big, respectful oil & gas industry clients come to us requesting and paying for simulated results. And I must add that they request CIVA simulated results!


Do you also use other modelling tools? Could you highlight the main advantages of CIVA? Which main improvements would you expect from the software?

Our team uses others simulation software and modelling tools, but no other is semi-analytical like CIVA. Being based on physical NDT phenomena is an advantage regarding computational costs. Besides that, CIVA is an extremely user-friendly software. But of course, as any other simulation software, it requires a minimum of knowledge of NDT techniques not only to properly set up the case study, but to understand what the software provides as a result.


As part of a university, I believe that you also have new students every year, young people discovering "the world of NDT". Do you think that simulator tools such as TraiNDE UT and RT can be of interest to train them and attract more "masterminds"?

Yes, without any doubt, TraiNDE would help young student understand the physical principles of UT and RX. But we must keep in mind that a small amount of people in a full classroom is really willing to proceed towards NDE-related careers. Here at the university, the undergrad NDT discipline is mandatory, so it is not really an option for the student. I think the main application for TraiNDE would be in preparatory courses for UT and RT qualification or professionalizing courses. Here in Brazil, we have an institute called ABENDI (Brazilian Association of Non-Destructive Testing and Inspection) that certifies professionals to work performing nondestructive inspections. There are many courses that prepare those professionals to be evaluated by ABENDI, and these students would have a great advantage being prepared with such realistic tools. Besides qualification for certification purposes, there are many courses that aim to instruct young students to be kind of immediately absorbed by the industry. In that way, every realistic tool would also be a great help for their training.