IMMPACT Lab - Research

Mechanistic Computational Intelligence for Foundation Models

Our lab is building the next generation of separable neural architectures that unifies training, solving, and calibration into one seamless computational framework. Sparked by our foundational HiDeNN research in 2021 and our article in Nature Communications (2025), this new class of AI technology delivers solutions orders of magnitude faster while using dramatically less memory than many existing deep‑learning tools in engineering. For students eager to work at the intersection of AI, simulation, and high‑performance computing, this is where the future is being invented.

Real-time Inverse Control of Additive Manufacturing

Real‑time inverse control has always been one of the biggest hurdles in additive manufacturing, where many different process conditions can lead to the same final material properties. Our breakthrough KHRONOS architecture changes the game by instantly revealing every possible manufacturing pathway—in real-time and without any retraining or calibration. If you're excited about AI‑driven manufacturing and want to work at the cutting edge of engineering innovation, KHRONOS is the kind of technology that will shape your future!

Inverse Design of Microstructures

Design of alloys has been taken to a new dimension by the 3D printing technology. But still a lot of gaps exist in the knowledge required to develop a process-structure-property map for 3D printed alloys. We are developing unique models that can produce target microstructures in real-time based on physics-informed criteria. Our JANUS architecture is distinct from diffusion models or VAE because it is a deterministic network and uses significantly less memory! If you want to be part of this exciting new technology, join us!

Physics of Manufacturing in-space

“The important thing is that we’re going to make it.” — Mark Watney, The Martian (2015)

Our lab explores one of the most fascinating frontiers in modern engineering—the intricate dance between fluid mechanics and microstructure that governs space‑based welding. In microgravity, metals behave in ways that challenge everything we know on Earth, and we’re uncovering the physics behind it. By developing fundamental models that reveal how molten flow shapes material structure, we’re paving the way for reliable manufacturing beyond our planet. For students eager to push boundaries and help build the future of off‑world fabrication, this is where discovery begins.

Materials physics at nanoscale

“There is plenty of room at the bottom” — R. Feynmann

Our lab uses molecular dynamics, density functional theory, cellular automata, and scientific machine learning to understand and transfer knowledge from terrestrial to extra-terrestrial applications. The lab has published several articles over the years in reputed journals that revealed fundamental physical mechanism behind material deformation.

Inverse Design of Aerospace Structures in High-dimensional Space

“The way I see it, you can either run from it or learn from it.” — The Lion King (1994)

Our lab is redefining what’s possible in aerospace engineering by tackling the challenge of inverse design for next‑generation wing structures. Creating novel multi‑fidelity surrogates, we can navigate extremely high‑dimensional design spaces with speed and precision—capturing full‑field physics to create wings that are lighter, stronger, and smarter. For students driven by curiosity and inspired by flight, this is your chance to help shape the future of aircraft design from the ground up.

Additive Friction Stir Deposition for Aerospace Parts

“The only way of discovering the limits of the possible is to venture a little way past them into the impossible.” — Arthur C. Clarke

Additive Friction Stir Deposition (AFSD) is transforming how large‑scale aerospace components are built, offering unmatched speed, strength, and material efficiency. In our lab, we partner closely with industry to develop high‑fidelity digital twins that predict and prevent defects before they ever occur—making AFSD parts not only reliable, but also far more affordable. For students excited about advanced manufacturing, simulation, and real‑world impact, this work sits at the cutting edge of how future aircraft and spacecraft will be made.

Major Recognitions

NIST AM bench 2022

Our group has been awarded in 5 major categories on NIST AM Bench 2022 contest.

MMLDT-CSET CONFERENCE 2021

I have given a talk on mechanistic data science on metal 3D printing of alloys in the first MMLDT-CSET conference 2021.

Research Facilities

SLM 280.2 Laser Powder-bed Fusion Machine

MELD L3 Continuous Feed Additive Friction Stir Deposition Machine

Nanoscale Characterization and Fabrication Laboratory (NCFL)

Virginia Tech Advanced Research Computing Facility (Highlight: TinkerCliffs came online in 2020 and has 346 nodes, 43,776 CPU cores, 117 TB RAM, and 110 NVIDIA A100 GPUs; More CPU and GPU clusters available)

Google Sites

Report abuse