NAFEMS Americas and Digital Engineering (DE) teamed up (once again) to present CAASE, the (now Virtual) Conference on Advancing Analysis & Simulation in Engineering, on June 16-18, 2020!

CAASE20 brought together the leading visionaries, developers, and practitioners of CAE-related technologies in an open forum, unlike any other, to share experiences, discuss relevant trends, discover common themes, and explore future issues, including:
-What is the future for engineering analysis and simulation?
-Where will it lead us in the next decade?
-How can designers and engineers realize its full potential?
What are the business, technological, and human enablers that will take past successful developments to new levels in the next ten years?

Resource Abstract

Space-time simulation (STS) is useful across the product lifecycle: to support iterative design processes, evaluate possible manufacturing processes and account for them in designs, and predict service life-times from sensor data recorded in digital twins. STS is cost-efficient and time-saving compared to physical prototype testing. Furthermore, STS can predict costs of material design choices, evaluate whether designs meet requirements, and determine whether safety regulatory margin constraints are satisfied. However, STS faces at least two challenges:

1. Model complexity limits broader usage across manufacturing organizations (simulation democratization)

2. Lack of reliable STS results put decisions at risk if not properly managed (simulation governance).

Interoperability is key to improving simulation democratization and governance by helping determine when models from multiple vendors are equivalent based on a simplifying abstraction of simulation (interchange parts for simulation). Such an abstraction would reduce complexity of usage (democratization) and facilitate communication about simulation across organizations (governance). Mathematical equations are a simplifying abstraction, but are not as accessible to humans and machines as graph-based formalisms, e.g., as used in time-only dependent models such as power-coupled Kirchhoff network (e.g. Modelica, Simscape), signal-coupled network (e.g., Simulink), bond-graphs (e.g., 20sim). There is currently no formal and widely adopted model abstraction for STS.

In this work, we apply machine-readable physics (MRP) from previous work, an approach that represents physical laws as computational graphs instead of equations, making them readable by humans as well as machines. Our methodology using MRP captures decisions in layers of models that apply physical laws to engineering problems and prepare them for numerical simulation. The objective of this model-based approach is to

1. Democratize simulation by representing these decisions in a simple and traceable way, documenting the simulation development process and making it reusable on multiple engineering projects. It also enables the same models to be translated to multiple software tools, we well as creation of domain-dependent library models by linking them to existing STS numerical solvers.

2. Aid simulation governance by enabling simulation models to be assessed for applicability to specific engineering problems (currently unreliable due to proprietary model descriptions), especially when multiple software packages are used, e.g, in OEM-suppliers relationships.

We will review MRP and the methodology around it, explaining how it simplifies simulation development and management as compared to current practice. We will also describe an ongoing implementation of storing and accessing these models, and integrating them with domain-dependent solvers.