Mechanics, Modeling and Machine Learning (M3) Lab
Welcome to the M3 Lab at the Indian Institute of Technology (IIT) Madras! We work at the intersection of mechanics, computational science and artificial intelligence to design, simulate and optimize next-generation systems and devices. In particular, we focus on developing physics-based and data-driven models to understand and predict the mechanical behavior of engineering systems across multiple length scales, from macroscale to micro and nano regimes.
Complex static and dynamic phenomena in solid and fluid mechanics are typically governed by non-linear partial differential equations (PDEs). Our lab develops analytical techniques, numerical models (including Finite Difference, Finite Element, Finite Volume models) and state-of-the-art machine learning methodologies (such as neural operators, foundation models) for accurate and efficient solution of these PDEs. We use these models to support downstream tasks of practical relevance such as design optimization, uncertainty quantification and inverse problems, with applications in diverse domains including micro/nanofluidics, hydrodynamic and elasto-hydrodynamic lubrication, data storage (e.g., hard disk drives), 3D Integrated Circuits (ICs) and micro-electro-mechanical systems (MEMS).
Fun fact about IIT Madras: in addition to being the top ranked institute in India, it is also well-known for its beautiful campus, which is home to diverse flora and fauna. The header image on our website is a real photograph of a couple of deer spotted on the campus!
Explore M3 Lab
Research
Key areas of interest include: (1) Scientific Machine Learning for Engineering Simulations, (2) Structural Mechanics and Dynamics, with a focus on System Identification, Design Optimization and Uncertainty Quantification, (3) Fluid Mechanics and Fluid-Structure Interactions, particularly at the micro- and nanoscale.
People
Prof. Siddhesh Sakhalkar is the Principal Investigator of the M3 Lab. We are actively looking for motivated undergraduate, graduate (Masters/PhD) and post-doctoral candidates to join our team! If you would like to join us, please contact Prof. Sakhalkar (siddheshsakhalkar1992@gmail.com) and include your CV with research interests.
Publications
Please review our published papers and patents to learn more about our prior research. In particular, we encourage you to explore our recent work that combines statistical analysis of measured population Frequency Response Function data with Finite Element Method simulations to predict device-to-device variations during high-volume manufacturing.
Collaborations and Parternships
Our core expertise lies in developing theoretical frameworks, numerical models and machine learning methodologies to study static and dynamic problems in solid and fluid mechanics. If you are interested in collaborating with us on a research project, please feel free to contact us.