Yuan Chiang

Welcome to my cyber nook! I am Yuan Chiang (江元, pronounced as ‘You-and John’). I also go by the name Cyrus. Currently I am a PhD student in Materials Science and Engineering at UC Berkeley and Lawrence Berkeley National Laboratory, under the guidance of Prof. Mark Asta. I leverage quantum mechanical calculation and develop machine learning on supercomputers to drive our understanding and innovations in physical sciences. I recently work on the devlopement and benchmark of foundation machine-learning interatomic potential (MLIP) and use them as a probe to understand and design materials under extreme environments in Gen IV molten salt fission reactors (MSRs).

My research interests broadly lie in computational materials physics and chemistry at the atomic and molecular levels, with an emphasis on the theory and application of ab-initio calculations, molecular dynamics, and machine learning to tackle challenges across energy, materials, pharmaceutics, devices, and computing.

Previously I earned my BS in Civil Engineering and MS in Structural Engineering at National Taiwan University, under the supervision of Prof. Shu-Wei Chang. I was a structural engineering intern at T.Y. LIN in 2018 and was a full-stack intern at Ucarer in 2021.

Yes! I had an unorthodox transition from “civil” to “materials science” (they are at the opposite ends of spectrum in terms of the length scale if you will) after years of working with Prof. Chang in civil engineering department because he is too nice and I was too numb not seeing what I really want to pursue, thanks to the spoon-feeding education system in Taiwan and my self-deception from the uninformed, vague childhood interest in civil structure 🙃. Fortunately, I had Prof. Chang’s unwavering guidance that enlightened me to use my gift and interest for fundamental science and technology.

News

Oct 23, 2023	We release a universal machine lenaring interatomic potential on Huggingface... Read more
Nov 28, 2022	I am giving two talks at MRS Fall 2022 in Boston. Check out the abstracts in the link for details!... Read more
Jul 20, 2022	Our paper “Encoding protein dynamic information in graph representation for functional residue identification” is published in *Cell Reports Physical Science*.
Apr 12, 2022	Migrate to this new site!

Latest Posts

Jul 23, 2023	Solve spinodal and binodal curves using Julia
Jul 10, 2023	Use LaTeX with matplotlib on HPCs
May 14, 2023	Embedding jupyter notebooks as jekyll blog posts

Selected Publications

2022

Encoding protein dynamic information in graph representation for functional residue identification

Yuan Chiang, Wei-Han Hui, and Shu-Wei Chang

Cell Reports Physical Science 2022

arXiv Bib HTML

@article{chiang2022encoding,
  title = {Encoding protein dynamic information in graph representation for functional residue identification},
  author = {Chiang, Yuan and Hui, Wei-Han and Chang, Shu-Wei},
  journal = {Cell Reports Physical Science},
  volume = {3},
  number = {7},
  pages = {100975},
  year = {2022},
  publisher = {Elsevier},
  url = {https://www.cell.com/cell-reports-physical-science/fulltext/S2666-3864(22)00261-2},
}

ImageMech: From Image to Particle Spring Network for Mechanical Characterization

Yuan Chiang, Ting-Wai Chiu, and Shu-Wei Chang

Frontiers in Materials 2022

Abs arXiv doi HTML

The emerging demand for advanced structural and biological materials calls for novel modeling tools that can rapidly yield high-fidelity estimation on materials properties in design cycles. Lattice spring model , a coarse-grained particle spring network, has gained attention in recent years for predicting the mechanical properties and giving insights into the fracture mechanism with high reproducibility and generalizability. However, to simulate the materials in sufficient detail for guaranteed numerical stability and convergence, most of the time a large number of particles are needed, greatly diminishing the potential for high-throughput computation and therewith data generation for machine learning frameworks. Here, we implement CuLSM, a GPU-accelerated compute unified device architecture C++ code realizing parallelism over the spring list instead of the commonly used spatial decomposition, which requires intermittent updates on the particle neighbor list. Along with the image-to-particle conversion tool Img2Particle, our toolkit offers a fast and flexible platform to characterize the elastic and fracture behaviors of materials, expediting the design process between additive manufacturing and computer-aided design. With the growing demand for new lightweight, adaptable, and multi-functional materials and structures, such tailored and optimized modeling platform has profound impacts, enabling faster exploration in design spaces, better quality control for 3D printing by digital twin techniques, and larger data generation pipelines for image-based generative machine learning models.

2021

Bridging the gap between NMR measured mean silicate chain length and nano-scale silicate polymorphism of calcium silicate hydrates

Yuan Chiang, and Shu-Wei Chang

Cement and Concrete Research 2021

Bib doi Code

@article{chiang2021bridging,
  title = {Bridging the gap between NMR measured mean silicate chain length and nano-scale silicate polymorphism of calcium silicate hydrates},
  author = {Chiang, Yuan and Chang, Shu-Wei},
  journal = {Cement and Concrete Research},
  volume = {140},
  pages = {106268},
  year = {2021},
  publisher = {Elsevier},
  doi = {https://doi.org/10.1016/j.cemconres.2020.106268},
}