With the aim of growing the facility’s community of industry users, the ALCF’s Industry Partnerships Program engages with prospective companies of all sizes, from start-ups to Fortune 500 corporations, that could benefit from leadership computing resources and collaborative opportunities with the ALCF and across Argonne.
The ALCF’s leadership computing resources—equipped with advanced simulation, data analysis, and AI capabilities—enable companies to tackle problems that are too computationally demanding for traditional computing clusters.
Access to ALCF systems and expertise allows industry researchers to make predictions with greater accuracy, rapidly analyze massive datasets, and create higher-fidelity models of everything from manufacturing processes to fusion energy devices. The results permit companies to accelerate critical breakthroughs, verify uncertainties, and drastically reduce or eliminate the need to build multiple prototypes.
The ALCF has enhanced its industry outreach program by partnering with other Argonne user facilities and divisions, including the Science and Technology Partnership Outreach (STPO) division. By providing more complete picture of the laboratory’s resources, this collaborative approach has resulted in broader engagements across Argonne with a number of companies.
The ALCF is also actively involved in directing the DOE Exascale Computing Project’s Industry and Agency Council, an advisory group of senior executives from prominent U.S. companies and U.S. government agencies interested in working with Argonne and other DOE laboratories to deploy exascale computing to improve their products and services.
Driving Innovation for Industry
Here are some current examples of how ALCF resources are helping companies to advance their R&D efforts.
3M Reduces Improves Energy Efficiency of Manufacturing Process
As part of DOE’s HPC for Energy Innovation (HPC4EI) program, researchers from 3M are working with Argonne to leverage ALCF supercomputers and AI to improve the energy efficiency of a manufacturing process used to produce melt-blown nonwoven materials. This extremely energy-intensive process is widely used by 3M to produce filters, fabrics, and insulation materials, as well as the N95 mask used for protection during the COVID-19 pandemic. By using ALCF computing resources to pair computational fluid dynamics simulations and machine learning techniques, the Argonne-3M collaboration is working to reduce energy consumption by 20% without compromising material quality.
ComEd Prepares for the Impacts of Climate Change
The ComEd energy company is partnering with Argonne to understand and prepare for the impacts of climate change. The team is using ALCF resources to dynamically downscale global climate models, providing projections and analysis for more localized areas. Their work is providing an understanding of how climate change may affect ComEd’s distribution grid and highlights the need for strategies that adapt to future climate conditions.
Dow Chemical Employs HPC and ML to Optimize Manufacturing Equipment
Dow Chemical is working with Argonne on a HPC4EI project aimed at optimizing the efficiency of gas-liquid turbulent jet mixers used in chemical manufacturing. The team is using machine learning (ML) techniques in conjunction with computational fluid dynamics simulations to speed up and improve design optimization for its advanced mixing equipment. Ultimately, the team’s work will lead to an efficient framework combining ML and HPC for optimizing process equipment, and will provide a demonstration case for ML approaches to enable wider adoption across the chemical industry.
Raytheon Develops ML Models to Improve High-Efficiency Gas Turbines
With support from the HPC4EI program, researchers from Raytheon Technologies are working with Argonne to develop machine learning models for designing and optimizing high-efficiency gas turbines in aircraft. The machine learning models have been trained on computational fluid dynamics simulations of gas turbine film cooling performed on ALCF supercomputers. The team’s framework aims to help the company extend fuel efficiency and durability of aircraft engines while slashing design times and costs.
Solar Turbines Models Carbon Capture Technology for Use in Industrial Gas Turbines
Researchers from Solar Turbines Inc. are partnering with Argonne on an HPC4EI project aimed at modeling cost-effective carbon capture technologies for industrial gas turbines used for power generation, marine propulsion, and oil and gas production. With the goal of reducing CO2 emissions, the team is using high-fidelity large eddy simulation-based modeling to optimize the performance of a novel carbon capture system on Solar Turbines’ industrial gas turbines. Through this project, the company aims to shave months or even years off the product testing and development process, helping to accelerate the time-to-adoption of this promising new technology.
Aramco Americas and Convergent Science Develop Design Process for Hydrogen Engines
In a collaborative project between Aramco Americas, Convergent Science, and Argonne, researchers are using ALCF resources to carry out large-scale computational fluid dynamics (CFD) simulations of ultra-high-efficiency hydrogen propulsion systems. As part of their work, the team has developed a high-fidelity, HPC-enabled analysis-led design process to accelerate the advancement of such systems. Ultimately, their work seeks to help decarbonize the transportation sector by expediting the adoption of clean, highly efficient hydrogen propulsion systems.
Wabtec Advances the Development of Hydrogen-Powered Train Technologies
Wabtec Corp., a leading manufacturer of freight locomotives, has entered Cooperative Research and Development Agreements (CRADAs) with Argonne and Oak Ridge national laboratories to advance the development of combustion technologies to power the next generation of trains with 100 percent hydrogen and low-carbon fuels. As part of the collaboration, Argonne researchers are using ALCF supercomputers to create a modeling framework to study combustion and emission control technologies used in hydrogen combustion engines. The simulation software tool will help predict the behavior of combustion engines and increase understanding the combustion process, helping to drive improvements in energy efficiency and emission reduction.