Grad student develops software used by fusion startups to automate stellarator design

Connor Moreno, a graduate student at the University of Wisconsin–Madison, has developed software tools to help leading fusion companies design the first commercially viable fusion power plants. 

A PhD student in the Department of Nuclear Engineering and Engineering Physics (NEEP) working with Professor and NEEP Department Chair Paul Wilson, Moreno developed an interest in fusion during his undergraduate studies when his coursework introduced him to topics like neutron physics and engineering grand challenges, one of which is harnessing energy from fusion. Like many researchers in the field, he was drawn to the appeal of this emerging industry and the promising role fusion plays in addressing climate change.

Early on in his graduate studies, Moreno gravitated towards computational fusion research,  with a focus in parametric modeling, a technique that utilizes design parameters to automate the creation of CAD models. 

Stellarators are challenging to model using traditional methods because of their complex geometry. Recognizing this gap, Moreno developed ParaStell, a software tool designed to automate parametric modeling of stellarators—a type of fusion reactor. As one of the only open-source software packages of its kind, the software has become a valuable resource for private fusion startups that have integrated ParaStell in their design workflows to increase efficiency. 

“I had to build it from the ground up,” he says. “It started as a hodgepodge of different scripts, but now it’s a polished software package.” 

That package also includes support for running neutronics calculations which help optimize stellarators by simulating various performance metrics, including radiation damage to reactor components, allowing researchers to identify the most effective designs. 

Moreno’s work is made possible through funding from a SciDAC partnership called HiFiStell, a multi-institutional research consortium that applies high-fidelity simulations and advanced computing to stellarator design. 

“Our relationship with fusion energy startups that are pursuing the stellarator concept has allowed us to quickly translate the outcomes of this federal funding to influence the design process of fusion power plants,” says Wilson.

ParaStell is currently being utilized by private fusion companies like Gauss Fusion and Type One Energy, a UW–Madison spinoff company co-founded by NEEP Professor Chris Hegna. Moreno worked closely with two NEEP master’s students on Type One Energy project assistantships to integrate ParaStell in their neutronics design workflow for neutron and photon transport modeling. The tool has also been central to the design of their breeding blanket, a device used in stellarators to breed more fusion fuel and shield the magnetic coils from the neutron radiation.

Moreno notes that working with private companies and national laboratories to integrate ParaStell has helped refine the software, making it a more robust tool for the community. Given the complexity of the problem, collaborations between public and private entities are paramount in the effort to commercialize fusion. 

“It’s exciting to see my work having a real impact on these companies,” Moreno says. “It’s not just in this little box that exists only for me; it’s useful for other people.” 

Moreno says his current work in design optimization could be even more impactful. While ParaStell helps users model their reactor designs, his design optimization methodology aims to automate the stellarator design process itself, leveraging ParaStell as its core geometry modeling tool. 

“A lot of other industries use design optimization to design their components, but it’s really underutilized in nuclear,” he says. “It hasn’t been applied so much in the fusion space and definitely not to stellarators.” 

As with modeling, designing stellarators using traditional methods is difficult and time consuming because they are geometrically complex machines. Design optimization is a robust, proven way to automate the process, allowing fusion companies and national laboratories to more efficiently design their reactors. 

Looking ahead, Moreno plans to continue his work in industry after graduation, advancing reactor designs with the goal of providing limitless clean energy for the world. 

“The fusion energy industry is still quite young and needs people like Connor who are trained to help them answer the most challenging aspects of designing fusion power plants,” says Wilson. “This SciDAC project has enabled UW–Madison to serve that need.”