Nuclear physics and engineering are already challenging disciplines to learn and to put into practice designing and modeling reactors, accelerators, criticality benchmarks, and more. However, navigating the nuances of a myriad of nuclear engineering simulation tools adds another substantial challenge to the field. Many of the most widely used simulation tools, like the Monte Carlo radiation transport code MCNP, were designed independently of each other and thus follow unique input formats and modeling approaches. This often creates a steep-learning curve for a single code, but more importantly, makes communication and data sharing between codes an interesting challenge. This problem is being addressed by taking a metamodel-driven modeling approach.

The metamodel-driven approach focuses on creating a comprehensive model for an individual code which provides a complete characterization of a code’s syntax and semantics. From the metamodel, textual editing support and programmatic functionalities can be derived. By leveraging these two complementary sets of features, a consistent modeling environment is created for a code which provides a vastly improved manual editing experience as well as advanced programmatic capabilities. Currently, this work has focused on MCNP and its integration into editor environments like the NEAMS Workbench and Visual Studio Code as well providing model translation in support of code-agnostic workflows for frameworks like the NRC’s BlueCRAB.