Postdoctoral Appointee-Developing Exascale MuPhFASa; Multi Phase Flow Adaptive Simulator

Postdoctoral Appointee

The Argonne Leadership Computing Facility's mission is to accelerate major scientific discoveries and engineering breakthroughs for humanity by designing and providing world-leading computing facilities in partnership with the computational science community. We help researchers solve some of the world's largest and most complex problems with our unique combination of supercomputing resources and computational science expertise.

The Computational Science Division focuses on solving the most challenging scientific problems through advanced modeling and simulation on the most capable computers. The division aims to build lab-wide cross-cutting simulation application capabilities integrating with mathematics, computer science, domain science, and advanced computing architectures and facilities. Additionally, the CPS provides an interdisciplinary home for spawning simulation programs and projects, often in collaboration with the ALCF.

The ALCF and CPS division are seeking a postdoctoral appointee to develop computational fluid dynamic (CFD) tools that make exascale computing accessible to a broader set of users. The successful candidate will develop a massively parallel solver, capable of running simulations on the Aurora supercomputer, using AMReX and the lattice Boltzmann method (LBM). The candidate will develop flow/geometry-aware refinement strategies that go beyond traditional error estimators, creating physics-based adaptation algorithms that intelligently predict where refinement will be most beneficial for smarter, more efficient simulations.

We seek someone with a strong CFD background and LBM expertise who can develop methodologies broad enough to tackle a diverse set of problems, appealing to a wide range of computational science users. Example problems are: bubbly flow, emulsions, sedimentation, wetting, gas-mixing, red-blood cell flow in the human vasculature. We are seeking a candidate who is intellectually curious and enthusiastic about computational research.

They are intrinsically driven, goal-oriented, and can work collaboratively with others.

Working closely with the CPS division, the postdoc will leverage AMReX and the LBM to develop an integrated framework to explore advanced workloads including simulations with in-situ visualization and, possibly, machine learning integration. This work will inform future ALCF platform procurement decisions.

Position Requirements

• Recent or soon-to-be-completed Ph.D. (typically completed in the last 5 years) in mechanical/aerospace/chemical engineering, applied mathematics or a related discipline.
• Experience in numerical methods and CFD development using mesh-based scientific codes.
• Expertise in the lattice Boltzmann method (LBM) as evidenced by their publications
• High performance computing (HPC) experience in code development with parallel programming techniques using the message passing interface (MPI) library
• Proficiency in writing code with C, C++ and/or Python
• Ability to demonstrate strong written and oral communication skills
• Ability to model Argonne's Core Values:
  Impact, Safety, Respect, Integrity, and Teamwork.

Desired

Skills:

• Experience with two-phase/multi-phase flows as evidenced by their publications
• Experience programming GPUs with CUDA, SYCL, HIP or OpenMP
• Experience using and developing code with AMReX
• Experience in performance engineering to improve code scalability and reduce time-to-solution