Xuanhe Zhao

⏳ Less than 2 weeks to go! 🚀Join the #AIXSummit to explore how AI is reshaping our world. Don't miss our incredible lineup: 🎙️ Keynotes: @ylecun & @AndrewWLo 🛋️ Fireside Chat: @NobelPrize David MacMillan Secure your spot today 👉aafuturefoundation.org/aix2026/ #AI

I will give a talk @Penn tomorrow (3/26). Welcome if you are around.

🌬️ Flow-induced reconfiguration, made simple A new FE approach couples corotational plates with a simplified drag model. ⚡ Faster than full CFD–FE coupling 🧪 Validated on plates, disks & kirigami sheets 📄 Read: doi.org/10.1016/j.eml.… #FSI #mechanics #EML

🌍 Do slip-weakening law shapes matter? Even with the same fracture energy, the shape of the friction law changes rupture dynamics. ⚡ Different shapes → distinct slip-velocity peaks 🧱 Different responses to barriers 📄 Read: doi.org/10.1016/j.eml.… #earthquake #fracture

@Junsoo_twit Congratulations!

Excited to share our new paper, “Trapping Polymer Entanglements via Prolonged Initiation,” now published in Macromolecules — and honored to be selected as an ACS Editor’s Choice article. doi.org/10.1021/acs.ma…

💡 Photo-mechanical response of azobenzene-filled elastomers UV light triggers trans–cis isomerization → macroscopic deformation or stiffening. 🧪 Experiments on azobenzene–silicone composites 📊 Data supports photo-viscoelastic modeling 📄 Read: doi.org/10.1016/j.eml.…

Join the AASF AIX Summit East 2026 student poster info session THIS FRIDAY, Feb 6 at 7 PM ET. Learn how to submit your AI research poster & what to expect at the Summit. Register now: buff.ly/cHVmHq2

🧩 Designing DMMCs with domain-knowledge enhanced ML 🔧 140k Al–SiC composite architectures 📊 High-throughput FE data 🧠 Deep residual NN + genetic algorithm 🎯 Optimizes modulus & Poisson’s ratio across volume fractions 📄 Read: doi.org/10.1016/j.eml.… #composites #ML #EML

🕷️ How do spiders hear so well? A new study shows that spiders’ hairs tune acoustic impedance, enabling efficient sound energy transfer from air to body. 📐 Hair length & angle matter 🧪 Verified with a bionic hair array 📄 Read: doi.org/10.1016/j.eml.… #biomimetics #acoustics

From all of us at MIT, Happy New Year! ✨

Our long review article is online, at Chemical Reviews, titled, Thermodynamic and Molecular Origins of Crack Resistance in Polymer Networks. @ZheqiChen pubs.acs.org/doi/10.1021/ac…

🧬 Rapid prediction of grain boundary evolution with generative ML A cGAN trained on MD data enables fast, accurate predictions in nanomaterials. ⚙️ Trained on MD simulations 🧠 Captures morphology + atomic details 📄 Read: doi.org/10.1016/j.eml.…

Paula Hammond ’84, PhD ’93, an Institute Professor, MIT’s executive vice provost, and former head of @MITChemE, has been named dean of MIT’s School of Engineering, effective January 16. news.mit.edu/2025/paula-ham…

🛰️ Thermo-mechanical metamaterial for stable microwave antennas Lattice sandwich design gives near-zero thermal warping under non-uniform heating 📊 99.7% less thermal warping vs bilayer 📡 Antenna tests: gain, beam offset & VSWR much more stable 📄 Read: doi.org/10.1016/j.eml.…

I will give two talks at #MRS #F25MRS in #Boston. Mon, Dec 01, 11-11:30, Hynes 109: #Bioadhesive #ultrasound imaging and closeloop therapy. #BAUS. Mon, Dec 01, 3:45-4:15, Hynes 104: Adhesive #nonfibrotic interfaces with extreme #hydrogels. Look forward to meeting many of you!

🧪 The anti-dogbone! Can we design tensile specimens that give richer data for ML-based constitutive modeling? 🔧 Non-uniform shapes 📊 Stress-state diversity metric 🎥 Displacement mapping via CV 🎯 More stress–strain coverage per test 📄 Read: doi.org/10.1016/j.eml.… #ML #EML

I have just asked ChatGTP a question: Who coined the words energy, entropy, enthalpy, exergy? ChatGTP produces an answer, which is the same as what I learned over the years. ChatGPT also gives a table:

🤖 Deep learning for bandgap engineering in lattice metamaterials. 🔧 Parametric lattices → rich design space 📊 FEM-generated dataset 🎯 Target bandgap in → lattice geometry out 📐 Validated in simulation & experiments Read: doi.org/10.1016/j.eml.… #deeplearning #inverseDesign