Scientific AI · Hidden Physics · Transport Closure
Haein Jung
정 해 인

Postdoctoral researcher at GIST.

I develop scientific AI frameworks that go beyond prediction to become tools for physical discovery. My work focuses on inferring hidden states, unresolved mechanisms, and missing closure relations in complex transport systems from sparse observations, governing equations, mechanistic models, and physics-informed learning.

My research background is rooted in phase-change heat transfer and multiphase transport, with additional experience in chaotic advection, transport enhancement, and thermal management systems. I aim to build interpretable, AI-augmented discovery pipelines that reveal the underlying physics rather than merely fitting data.

profile
position Postdoctoral Researcher
School of Mechanical and Robotics Engineering, GIST
education Ph.D., Mechanical Engineering, GIST
M.S., Mechanical Engineering, Korea Aerospace University
focus Scientific AI for inverse problems
Hidden physics & closure discovery
Transport modeling and inference
methods Physics-informed learning
Neural operators · Inverse problems
Agentic scientific computing
domains Phase-change heat transfer
Multiphase transport systems
Thermal transport and mixing
background Experiments · Mechanistic modeling
CFD · Scientific machine learning
§ 01

Research
Direction

I develop scientific AI methods to discover hidden physics and missing closure relations in complex transport systems. Rather than treating AI as a black-box predictor, I focus on frameworks that infer latent mechanisms and governing relations by integrating governing equations, sparse measurements, mechanistic models, neural operators, and physics-informed learning.

unifying theme
AI for Hidden Physics Discovery in Transport Systems

Many transport systems are governed by equations that are only partially known. Unresolved mechanisms, latent variables, and empirical closures create fundamental barriers to predictive understanding. I study how scientific AI can move beyond forward modeling to become a discovery instrument — one that can infer hidden states, identify missing closure relations, and produce interpretable models even when observations are sparse and incomplete.

This direction targets regimes where traditional methods struggle: systems with strong nonlinearity, multiscale interactions, interfacial dynamics, and limited observability.

scientific AI hidden physics closure discovery inverse problems operator learning physics-informed learning agentic scientific computing
domain I
Phase-Change & Multiphase Transport

Phase-change and multiphase systems offer a uniquely challenging testbed for hidden-physics discovery. Key mechanisms — nucleation-site activation, bubble growth, microlayer evaporation, interfacial transport, heat-flux partitioning, and flow instabilities — are often only partially observable even in well-instrumented experiments.

My background includes subcooled flow boiling experiments, density-wave oscillation studies, CHF characterization, and mechanistic modeling in annular and microchannel geometries. I am extending this foundation toward physics-informed inverse modeling, neural operators, and closure discovery for unresolved interfacial transport.

boiling condensation multiphase flow nucleation bubble dynamics interfacial transport flow instability
domain II
Thermal Transport Systems

Thermal transport systems connect heat generation, fluid transport, material response, and performance constraints across multiple scales. These systems often require hidden-state inference because internal heat sources, effective boundary conditions, thermal distortion, and transport coefficients cannot be measured directly.

My recent work has involved high-power laser thermal management, liquid cooling of gain media, heat-load estimation, thermo-optic distortion, and coupled thermal-fluid analysis. I view these systems as emerging applications for scientific AI, inverse heat transfer, and operator-learning surrogates for design and diagnostics.

thermal management laser cooling inverse heat transfer thermo-optic coupling cooling design multiphysics transport
supporting domain
Chaotic Mixing, Flow Structures, and Transport Enhancement

Chaotic advection and laminar mixing provide clean, interpretable settings for studying hidden flow structures and transport enhancement. Stretching and folding dynamics, transport barriers, scalar dispersion, and effective diffusivity can be analyzed through CFD, Poincaré sections, dynamical-systems concepts, and scientific machine learning.

My prior numerical work on groove-embedded and barrier-embedded pipe mixers established a foundation in chaotic transport. I use this background as a supporting application area for hidden-structure discovery and data-driven modeling of nonlinear transport phenomena.

chaotic advection mixing transport enhancement flow structures CFD nonlinear dynamics latent structures
§ 02

Selected
Publications

Selected works tracing the trajectory from chaotic transport and multiphase experiments to mechanistic modeling and physics-informed inverse problems.

2026
Experimental investigation of subcooled density-wave oscillations in concentric annulus
International Journal of Heat and Mass Transfer · H. Jung, S. Lee
2023
Mechanistic model to predict oscillating frequency of flow boiling in large length-to-diameter ratio micro-channel heat sinks
International Journal of Heat and Mass Transfer · H. Jung, S. Lee
2020
Fouling mitigation in crossflow filtration using chaotic advection: A numerical study
AIChE Journal · S.Y. Jung, H.I. Jung, T.G. Kang, K.H. Ahn
2020
Flow and mixing characteristics of a groove-embedded partitioned pipe mixer
Korea-Australia Rheology Journal · H.I. Jung, J.E. Park, S.Y. Jung, T.G. Kang, K.H. Ahn
§ 03

Background
& CV

Research trajectory spanning thermofluid experiments, mechanistic modeling, computational transport, thermal systems, and scientific machine learning.

education & positions
2025.9–
present
Postdoctoral Researcher
School of Mechanical and Robotics Engineering, GIST
~2025.8
Ph.D., Mechanical Engineering
GIST · two-phase heat transfer and flow instability
Advisor: Prof. Seunghyun Lee
~2019.8
M.S., Mechanical Engineering
Korea Aerospace University · chaotic advection and laminar mixing
Advisor: Prof. Tae Gon Kang
research background
2026-
Scientific AI for hidden transport physics
Physics-informed learning · sparse observations · hidden-state inference · neural operators · agentic scientific computing
2024.12-2026.07
Thermal management of high-power laser systems
Direct liquid cooling · thermo-optic distortion · heat-load estimation · multiphysics transport
2019.07-2026
Flow boiling and flow instability
Flow boiling experiments · density-wave oscillation · CHF · mechanistic modeling · thermal transport
2017.05-2019.06
Chaotic advection and transport enhancement
Numerical studies of laminar mixing and crossflow filtration
Get in
touch

I welcome enquiries on hidden physics discovery with AI, closure learning, inverse problems in transport systems, phase-change phenomena, thermal transport, and chaotic mixing.

office School of Mechanical and Robotics Engineering, GIST
Gwangju, Republic of Korea