The enormous complexity of today’s vehicles is well recognized. Not only are automakers turning to simulation as a cost-effective means to enhance physical test, it is becoming a fundamental tool in the design process. This is true in sometimes surprising ways, as computers get faster and engineers smarter.
“In just the last three years, an important trend for us is the increasing use of 3D simulations,” said Curtis Collie, Principal Technical Specialist for Gamma Technologies, in an interview with Automotive Engineering. This is significant for a company that established itself in 0D and 1D engine simulations with its GT POWER software. Now, with its expansion into GT SUITE and software that simulates and optimizes entire vehicles, expect the mantra of simply “simulating more” to continue in targeted ways.
Collie sees the future as “increasing levels of optimization to match increasing levels of complexity.” He noted that from simple models, engineers "can now simulate more components connected through more interfaces with complex interactions. Optimizing those complex interactions are what is needed and what the industry will continue to grow towards.”
Are you part of an EMDO?
Just as importantly, the organizations that design them are also growing in complexity. Companies need a varied engineering staff to meet a complex welter of worldwide regulations while satisfying ever more specialized customer niches. This has led to automakers employing thousands of expert engineers, often in a network of engineering centers worldwide.
That is certainly true for Ford, according to Yan Fun, Manager and Technical Leader, Product Development and Strategy Analytics. She made some key points in a presentation at a symposium titled Engineering Analysis & Simulation in the Automotive Industry on April 28. The event was sponsored by NAFEMS, the International Association for the Engineering Modelling, Analysis and Simulation Community.
In her talk, Fun described how Ford is developing an Enterprise Multi-disciplinary Design Optimization, or EMDO, system.
“One of our objectives was to understand how we can help Ford utilize our tens of thousands of engineers worldwide,” she explained. Most engineers are experts in their field, from brake design to vehicle NVH, so integrating that knowledge to produce an optimized vehicle is an ever greater challenge.
Equally important is the need to build a system that helps train new engineers and capture the knowledge of experienced ones. Developing a system to combine that knowledge is just as important, Fun said, as vehicles require a balance of attributes to perform well.
There are emerging tools that provide opportunity to meet these goals. Mobile computing such as smartphones connected by the internet is a key enabler in Ford’s EMDO. High performance computing, or HPC, combined with advanced databasing, and 24/7 web based services enable a flexible, worldwide optimization process.
Another key enabler is using advanced commercially available multi-disciplinary optimization tools, in their case ModeFrontier and SOMO from Esteco. ModeFrontier, as a desktop solution, is used by individual engineers and small groups to streamline an engineering process. SOMO, on the other hand, is a web-based platform for integrating multidisciplinary design projects across multiple departments and geographically distributed organizations such as Ford.
While presenting the results of a pilot study that used a prototype of this system to balance NVH and Safety attributes, Fun also pointed out the emergence of a new type of engineer – the MDO expert. This is an engineer who combines individual attribute models created by specific experts into a multi-disciplinary optimization run. The emergence of such a discipline makes sense, since there are a wide variety of MDO algorithms to choose from. These include algorithms that are gradient based, deterministic, or heuristic, so choosing the right one might require expert knowledge.
Connections and standards - FMI
Simulation tools also vary greatly based on their underlying principles. A CAE model for NVH is quite different from one used to analyze drivetrain performance or vehicle aerodynamics. A system or vehicle optimization method must both be able to accept such models and combine their individual results. While MDO tools can be customized to support model exchange, an important emerging standard is the Functional Mock-up Interface, or FMI.
“The reason FMI is such a hot topic in the industry right now is that it addresses critical needs for users to exchange, integrate, and deploy models in a tool neutral way,” said Ed Ladzinski, manager for Modelon. “Once the output from a modeling tool is put into the form of a Functional Mock-up Unit (FMU), it typically frees the simulation model from requiring further access to the authoring tool.”
Also, output as FMUs from a range of modeling formalisms, including 3D finite element stress analysis, a controls model created in Simulink, and computational fluid dynamics, can all be combined into a system model providing more realistic simulation of the actual real word performance.
As an emerging standard, it requires the CAE and systems modeling companies to produce a standard output that allows for collaboration and interoperability. The good news is that there is a growing list of these vendors that see the advantages – and the market demand.
“As FMI continues to mature in the next few years, expect to see even larger and more integrated full vehicle simulations and optimizations across a range of disciplines, modeling formalisms, and tools,” Ladzinski noted.
Continue reading »