HyperMesh and Custom Export Template Streamline CFD Analysis in Research Projects at Arizona State University

Arizona State University is an institution committed to excellence, access, and impact. It pursues research that contributes to the public good and assumes major responsibility for the economic, social, and cultural vitality of the communities that surround it. The School for Engineering of Matter Transport and Energy (SEMTE) at ASU encompasses mechanical, aerospace, materials, and chemical engineering. The Integrative Simulations & Computational Fluids Lab research group at SEMTE is strongly focused on developing and utilizing tools to investigate complex engineering and physical systems on massively parallel machines. They perform their research with the open-source computational fluid dynamics (CFD) solver Nek5000 which is used in a broad range of applications including thermal hydraulics of reactor cores, transition in vascular flows, atmospheric and ocean modeling, and combustion.

The Integrative Simulations & Computational Fluids Lab researchers at the School for Engineering of Matter Transport and Energy (SEMTE) at Arizona State University (ASU) were faced with the challenge of using the commercial code HyperMesh as a general preprocessor to mesh complex geometries for use with the spectral element CFD code Nek5000. The Nek5000 code requires 3D hexahedral elements, which posed a difficulty as most CFD tools use tetrahedral meshes that are easier to generate for conventional geometries. The researchers wanted to benefit from the rich functionality of advanced meshing tools like HyperMesh, capable of producing high-quality hexahedral meshes, while using the Nek5000 solver code. Before the project started, the researchers had no general process for meshing in place. Most of the meshing was handled with custom-made tools that were developed 15-20 years ago and have seen minimal updates since that time. Other users created their own meshing tool for specific problems in software such as MatLab.

To overcome this challenge, the researchers at SEMTE set up a project to develop a converter tool that could export a mesh from the commercial code HyperMesh into a form that the Nek5000 code can use. They chose HyperMesh because of its solver-neutral functionality, outstanding documentation, and open architecture. The researchers started with very small problems that did not include more than five to ten elements and gradually worked through the development and debugging process to test it on larger domains. After weeks of intensive research and coding, they finalized the export template, creating a more user-friendly and less error-prone process. The converter tool starts by translating the nodal coordinates for each element from the native format within HyperMesh to the Nek5000 data structure. The users also supply the boundary conditions through HyperMesh and this data is converted to the Nek5000 format after the geometric conversion takes place. All of this data is then written to a file that can be fed into the Nek5000 solver. The engineers also added mid-side node support for the meshes to increase geometric flexibility and the converter automatically implements this information based off the element type used in HyperMesh.

• The development of the converter tool and the custom export template has provided researchers and students at SEMTE with a universal approach to evaluate virtually any application and its complex geometry with the spectral element CFD code Nek5000. The new process is extremely user-friendly and generally applicable, paving the way for an advanced and highly integrated engineering workflow. It has increased the ability to collaborate with other researchers who use different tools and has sped up the learning curve of all students working on these kinds of projects. The new process has also improved the precision of results for all analyzed tasks and has allowed for evaluating flows within complex geometries. There are currently plans to enhance the converter and rework sections of it to optimize its functionality and make it work even faster.