Platform-Independent Integration of SysML with 1-D Simulation

This presentation was made at the NAFEMS Americas Seminar "Model-Based Engineering: What is it & How Will It Impact Engineering Simulation" held on the 1st of October 2019 in Columbus Ohio

Resource Abstract

Model-based systems engineering uses representations of system requirements, behaviors, and architectures to communicate complex system specifications among engineering teams and their stakeholders. Systems modeling languages capture these representations to help organize a wide variety of engineering activities, many of whom develop discipline-specific simulation models. Many simulation models are built with equation-based languages, containing detailed behavior specifications and analytics to predict system dynamics. They are often built separately from systems models, leading to additional work to resolve inconsistencies and lower quality due to fewer multi-disciplinary simulation concerns reflected in the overall systems design. To address these problems, a widely used systems modeling language (SysML) has been extended to represent physical interactions and signal flows (also known as 1-D or lumped parameter models) that can be translated to simulation tools for execution. All translated simulation structures can be traced to their corresponding systems structure, along with any simulation errors related to the systems model.



This presentation describes the SysML Extension for Physical Interactions and Signal Flow Simulation (SysPhS), which enables automated translation of SysML models to executable simulation models. The extension fills modeling gaps between SysML and simulation tools and languages. These gaps (simulation concepts missing from SysML) were identified by a comparison of constructs and semantics in SysML to constructs and semantics that simulation tools and languages have in common.



This presentation also gives an overview of best practices for using the extension to develop and translate SysML models. These guidelines for modeling and debugging systems models ensure smooth translation to and execution of engineering simulation models. They emphasize transformation and transmission of conserved physical substances and numeric information within system structures, with equations chosen to describe these processes. Debugging procedures identify defects in system models that cause simulation execution failure or incorrect simulation results. Finally, an example SysML model built with the extension is translated and executed on different simulation platforms, showing consistency with the same results.