UNIVERSITY PARK, Pa. – A team of researchers from Penn State and NASA’s Jet Propulsion Laboratory (JPL) developed software that’s being used by engineers and scientists around the world to make advances in materials modeling, space technology development and data science.
Their software package, PyCalphad, placed second in the 2019 NASA Software of the Year competition. The contest highlights software that is vital to NASA’s role in developing aeronautics and space technologies and transferring them to government and industry, according to the space agency.
“We are thrilled and delighted to receive this recognition, which follows several years of hard work for our team,” said Richard Otis, a materials technologist at JPL, who began work on the project after joining Penn State in 2012 as a graduate student. “When looking at the list of past recipients, some of which include major NASA missions, it is very humbling to see our software join that list.”
PyCalphad is a free and open-source Python computer language library for computational thermodynamics modeling that uses the Calculation of Phase Diagrams (CALPHAD) method.
“I see this as a foundational piece of work,” said Zi-Kui Liu, distinguished professor of materials science and engineering at Penn State, who has served as a technical adviser on the software development. “Thermodynamics is the foundation for all materials in the universe, including in human life. PyCalphad will improve our ability to model these systems.”
CALPHAD started as a tool to model thermodynamic properties of metallic alloys and is now widely used in materials design and manufacturing. It is a key tool in Integrated Computational Materials Engineering (ICME) as noted in reports by the National Research Council and NASA Vision 2040. CALPHAD databases can help researchers understand how materials will react under various conditions like temperature and pressure, enabling scientists and engineers to tailor chemistry and manufacturing parameters to design new materials and improve the performance of existing materials, Liu said.
“For example, in the nickel-based super alloy world, jet engine turbines are a big application,” said Brandon Bocklund, a graduate student in materials science and engineering who works on the PyCalphad project. “What CALPHAD can do is tell you in this high-dimensional composition space, this group of elements from the periodic table, what combinations will give you stronger nickel-based super alloys that operate at higher temperatures with better fuel efficiency.”
PyCalphad advances the technology by providing a flexible, powerful interface for manipulating CALPHAD data and models, the researchers said. The Python code gives users an intuitive interface and the freedom to customize models beyond other available commercial and open-source software.
“That’s the killer feature of PyCalphad,” said Bocklund, who is also a NASA Space Technology Research Fellow, said. “Researchers are not only able to do these calculations, but now they can write new models and do new, interesting research.”
Liu said the software has applications for a broad swath of the manufacturing industry and could serve as a backbone for an approaching industrial revolution centered around digitalization.
From relatively humble beginnings, the software is now widely used by government agencies, academics and Fortune 500 manufacturing companies, the researchers said.
Otis began working on the project on his own time while a graduate student working with Liu and later received funding for the research after writing a proposal and receiving a NASA Fellowship. The project was the focus of his dissertation research at Penn State.
He continued developing the software after graduating in 2016 and joining JPL. Bocklund joined the project in 2016 when starting his graduate studies at Penn State with Liu.
“For our small team of Richard, Dr. Liu and me to win second place is a real honor,” Bocklund said. “It’s a testament to the hard work Richard started and the quality of work he did, and the vision of Dr. Liu to carry this forward and help find gaps for us to fill and ways to provide value to the community through this software.”
In addition to PyCalphad, Bocklund has been working on developing ESPEI, software for uncertainty quantification in CALPHAD modeling using PyCalphad as the computational engine. Both PyCalphad and ESPEI are developed as open-source software on Github at http://github.com/PhasesResearchLab. The project websites can be found at https://pycalphad.org/ and https://espei.org.