Coreform received a DOE grant to study nuclear reactors.
Nuclear energy is more prevalent than we think. According to the Office of Nuclear Energy, U.S. nuclear power plants produced 778 billion kilowatt-hours of nuclear energy in 2021—more than anywhere else in the world. This accounts for around 20 percent of the U.S. energy supply and 50 percent of the country’s clean energy. The agency went on to say that nuclear fuel is extremely dense, so much so that all the waste from the last 60 years would fit on a football field with a depth of fewer than 10 yards. Though impressive, for those who can envision a football field, this amount of waste is not the flex the agency might be hoping for.
Perhaps that is one reason that Coreform is working to use simulation to optimize reactors. Specifically, the goal is to speed up reactor simulation throughput and make the process of designing and developing reactors even faster. The Orem, Utah simulation startup says that preparing CAD geometry for meshing is the most time and resource-intensive part of the simulation process. So, it offers a few ways to make the process work faster.
Recent wins in this area for Coreform, and its Cubit software, include a partnership with Radiant Nuclear and a $1.15 million Department of Energy (DOE) Small Business Innovation Research grant for a project titled “Enabling solution verification and efficient, high-accuracy simulations through spline-based adaptivity in MOOSE.”
There’s a strong feeling of partnership at Coreform, as the MOOSE (Multiphysics Object-Oriented Simulation Environment) project is developed with the Idaho National Laboratory (INL) and the company also licenses some of its software from the Sandia National Laboratories.
So, after that introduction, let’s dig into Coreform’s nuclear simulation offerings.
Nuclear Reactor Simulation—How Hard Can It Be?
On some level, every engineer thinks that their job is difficult—not just any schlub could do what we do—especially when it comes to simulation. Sometimes there’s a structural study, or a computation fluid dynamics problem, or a heat transfer test that someone else has run and, after looking through the results, we think that it makes sense and could probably reproduce something similar with the right tools. This isn’t always the case for nuclear reactor simulations. These engineers might be able to claim a few extra degrees of difficulty when it comes to their simulations.
For instance, it requires a heavy knowledge of particle physics to truly simulate nuclear reactors. How do neutrons react against different energies and materials in a reactor, and after the neutrons scatter, what happens to that energy?
There is a good article on MIT News about Amelia Trainer working on these deep questions and finding ways to model neutron scattering. The International Atomic Energy Agency (IAEA) is also working to create better simulations of nuclear energy by developing the Nuclear Fuel Cycle Simulation System (NFCSS), a tool that’s used to look at fuel cycle assessments and develop long-term plans for planners and policymakers. The IAEA offers a full slate of simulation tools that focus on education and the parameters of nuclear processes. Most of these examples are not what we think of as traditional simulation.
So, after all this research, what can a traditional simulation company like Coreform do? The answer lies in the software’s abilities to run studies fast and give users a high degree of confidence in the results.
Where Does Coreform Cubit Come into Play?
To that point, Coreform says that model preparation is a big bottleneck in the nuclear simulation process. If a company wants faster results, then getting the study from CAD models to mesh faster is one way to go. Coreform Cubit has tools for modifying geometry, auto-healing functions for fixing errors caused by import operations and a smart defeaturing tool to remove details that will slow down a simulation without raising the fidelity of the results. The meshing tools also give users complete control over the process—or the ability to use auto-meshing tools.
Spline-based simulation is a big product differentiator for Coreform. Splines are used in programming to approximate polynomials with math models. They help provide a good fit that approximates the CAD geometry with something the simulation tool can understand—even if the geometry doesn’t quite fit the traditional lines or arcs.
I often tell my students that creating splines in CAD is difficult but that in the late ’80s I used a French curve to approximate curves on drawings and am convinced that no digital frustration can ever match that experience.
Coreform was a pioneer in taking the idea of spline-fitting math curves and using that on meshing. The company says that its spline-fitting method is “more accurate per degree of freedom than FEA, more computationally efficient at every level of accuracy and more robust.”
For a use case of the tool, Radiant Nuclear, a company founded by former SpaceX engineers, wanted a micro-reactor that could be transported around the world for events where high levels of energy were needed on nonpermanent time frames. One of the big pushes for Radiant was developing its reactors quickly, with an eye on a two-year development cycle instead of the traditional five-to-eight-year one. This meant that the design phase needed to happen much faster than normal. By using the Coreform tools, the company was able to shrink the timeline. Radiant noted that the simulation company’s spline-fitting methods, automated meshing functions, and hex-element mesh tools were all reasons to enter the partnership.
Beyond Nuclear and the Big Picture
Even though the focus of the DOE grant is nuclear energy, this technology can be used for a wide range of applications. For instance, the MOOSE tool was first used in a groundwater analysis in Australia and the team sees possible applications everywhere.
The Coreform tools discussed here are designed for a specific phase of the simulation process, but the company has a full suite of tools that can help engineers run studies from the CAD phase to the results phase. There’s even a free edition of the software called Cubit Learn, which can be licensed without charge for noncommercial use.
If nuclear energy continues to be a part of our renewable energy future, then developing new reactors faster will be a huge technological leap. Researchers are working to build complex mathematical models that will help us understand neutron reactions, while simulation companies are working to make sure that simulations can run faster and provide us with more confidence in the results. The number of partnerships at work here is amazing to see, with companies working together with regulatory agencies and government labs to move toward a more energy secure future. It will likely require many breakthroughs on many different levels to solve all the energy problems on the horizon, so it’s promising that simulation will help us develop the next generations of energy production faster.
