Chandrasekhar Annavarapu Department of Civil Engineering Areas of interest: Extended Finite element method, Computational fracture mechanics, Geomechanics, Multi-physics simulations

Dr. Chandrasekhar Annavarapu is an Assistant Professor in the Department of Civil Engineering at IIT Madras. Dr. Annavarapu earned his B.Tech. in Civil Engineering from IIT Madras in 2007. Subsequently, he pursued his graduate education at Duke University in Durham, NC, USA and earned a doctoral degree in 2013. On graduation, he worked as a scientist at the Lawrence Livermore National Laboratory (LLNL) – a Department of Energy research facility in Livermore, CA, USA until May 2018. Later, he worked at ExxonMobil Upstream Research Company in Houston, TX, USA as a computational scientist until October 2019, before joining IIT Madras as a faculty member.

What is the area of your research?

My research expertise is in computational mechanics – a subfield of engineering mechanics focused on developing computational methodologies and tools used by practicing engineers to analyze complex physical systems.

In particular, my team and I focus on addressing the limitations of finite element methods for complex and evolving interface problems in mechanics. Research in our group advances a special class of finite element methods known as the eXtended finite element methods (X-FEM) -- these methods alleviate meshing constraints by allowing interfaces such as crack surfaces and material boundaries to be embedded in a background finite element mesh. In addition, our group is also developing energetic approaches to model material failure.

Research conducted at our lab is interdisciplinary and involves a good mixture of engineering, mathematics, and computing. Most projects involve a healthy combination of developing new numerical algorithms, and their application to problems in civil, and mechanical engineering, geosciences, and defense.

How was your PhD and PostDoc experience with the computing facilities?

As the name suggests, computational mechanics has a heavy emphasis on computing. Essentially, computational mechanics provides a virtual laboratory to investigate novel theories in mechanics and their applications in engineering. With growing computational power, High Performance Computing (HPC) is increasingly becoming indispensible to active areas of research in computational mechanics.

In the past, I have been fortunate to have access to some of the most powerful computational facilities in the world. At LLNL, we had access to several machines that featured among the top 10 fastest super-computers in the world. Further, even at ExxonMobil, the HPC cluster was one of the fastest in the world. In fact, the other oil and gas super-majors, Royal Dutch Shell, BP, Chevron, and Total all have dedicated facilities with some of the most-powerful computers outside of government institutions.

It is not surprising, however, considering that the earth’s subsurface can be challenging to model. To generate realistic models, one has to consider several length-scales, time-scales, as well as the inherent geometric complexity present in the earth’s subsurface. Personally, I have benefitted immensely from exposure to such an environment where HPC was heavily leveraged to answer pressing concerns by industry and national labs. I am happy to join an active and growing HPC community at IIT Madras. My hope is that I can translate the lessons from my past experiences and set-up a competitive research lab here that leverages HPC for both fundamental and applied research.

What do you suggest to PostDocs who wish to join academia in your field?

To any doctoral, and post-doctoral research scholars who wish to join academia in my field, first, I would like to emphasize the importance of visible research output. It is important that the scholars present their ideas in various national, and international forums. In addition to getting an objective perspective on their research, they will also open up avenues of collaborations. Secondly, publications are, obviously, the currency for any Ph. D. student and postdoctoral scholar. I would encourage them to publish as often as possible without compromising on scientific quality.

Finally, I strongly believe that the best way to make meaningful contributions in the field is to collaborate with teams that bring complementary skills. One such way to collaborate is through contributions to open-source software platforms in the field. This has two benefits: (a) it fast- tracks the scholar’s progress in their doctoral/post-doctoral research by avoiding redundant work, and (b) when the scholar joins as a faculty member, the open-source platform will still be available to them, at no cost, to continue their own research as a PI.