Is MDX the Future of Engineering Simulation?

Print Friendly, PDF & Email

In this special guest feature, Gemma Church from Scientific Computing World investigates how CD-adapco is helping engineers to collaborate and move complex simulations into the design phase.

SCWaug15CDAdapcoWant to make an engineer nervous? Just utter the words “untested design” and step back.

Within the engineering world, testing during the design process is a vital yet difficult task. In the past, with limited computer resources, and simulation tools that lacked complex physics models, engineers were sometimes forced to make bigger simplifications than they were comfortable with. The simulation process was also expensive, in terms of hardware and licensing costs, and often deployed so late in the design process that only a handful of worst-case simulations were possible.

One proposed resolution to these simulation issues is a process called MDX (Multidisciplinary Design eXploration), and it is a process that engineering simulation company CD-adapco is pioneering.

What is MDX?

As product designs become increasingly complex, engineers are now required to simulate entire systems to produce the quality products their customers require.

David L Vaughn, global vice president of marketing at CD-adapco, told Scientific Computing World: “An uncomfortable truth about modern engineering is that there really are no easy problems left to solve. In order to meet the demands of industry, it is no longer good enough to do “a bit of CFD” or “some stress analysis.” Complex industrial problems require solutions that span a multitude of physical phenomena, which often can only be solved using simulation techniques that cross several engineering disciplines.”

By adopting an MDX philosophy, engineers are able to test designs automatically from the early concept stages and against all of the physical factors that might influence a system’s performance. It assesses which set of design parameters will break a system, and which will improve it. This pushes back the simulation process to force engineers to question every assumption they have made within a design, and optimise it appropriately by assessing a simulation with multiple operating scenarios.

In essence, MDX allows users to understand a product’s performance throughout the design space, and CD-adapco expects this will accelerate the decision-making processes as a product is designed.

Vaughn added: “Engineering simulation software has gone from being strictly a troubleshooting R&D tool to a standard part of the product design process. Deployed earlier and earlier in the process, it is increasingly being used in a design exploration mode to investigate hundreds of possible design variants and predicting how they will perform under real-world operating conditions.”

The MDX landscape

MDX often involves the coupling of various simulation software tools; each one is specialised in addressing a specific part of the problem, which allows an engineer to model an entire system instead of its individual components.

This can be quite a difficult task in terms of the computation power and the complexity required to run such simulations, so the MDX process has several important characteristics.

In order to capture the influence of various physical phenomena that might vary substantially over length or time-scales, it is often necessary to adopt a ‘multi-fidelity’ approach, resolving some physics in more detail than others. MDX also allows engineers to automate every step of the simulation workflow to minimise user intervention and speed up the simulation process.

CD-adapco offers a clever licensing model, which allows users to harness as many processors as they have available for a single fixed cost to accommodate any spikes in usage. This licensing scheme gives a level of affordability and flexibility not often seen in the simulation space.

The software

CD-adapco is also one of the few remaining privately owned engineering simulation software providers, giving it the ability to satisfy the needs of its customers instead of its shareholders. Vaughn said: ‘Most of our profits are invested into the company and the development of our software. Compared with other companies, we have a much bigger software development team. This is reflected in the pace of our software development (three releases a year) and the quality of our software products.’

CD-adapco’s suite of products has been optimised for MDX. Its STAR-CCM+ simulation software allows users to affect CAD clients remotely to explore the design space as the CAD design changes while running on a remote cluster.

To reflect the distributed nature of the modern project team, STAR-CCM+ is also written using a client-server architecture. The server performs the numerical calculation while the client (a workstation) is used to control and visualize it. This client-server architecture allows for visualising results as they are generated remotely on an HPC cluster or a remote cloud-based computing facility.

Communication matters

With such a multidisciplinary approach, it’s not just the software that needs optimising – so do the humans at the other end. Vaughn adds: “Historically engineers have tended to align themselves strictly along disciplinary lines: the fluids engineers do CFD (Computational Fluid Dynamics), the stress engineers do FEA (Finite Element Analysis), the chemical engineers do all sorts of other stuff that no one else understands. Getting individual engineers to talk to each other was often as much of a challenge as interfacing the individual software tools.”

This cultural shift for many engineers may be difficult to manage, but it is a necessary move as Vaughn predicts that in the near future MDX will be omnipotent in the engineering simulation space. He added: “Engineers will work with software tools that seamlessly allow them to solve problems using whatever physics is necessary to properly predict the behaviour of a product or design under real-world operating conditions. They will also routinely use design exploration to characterise the entire design space, and find the optimal design parameters for each operating condition.”

This ability to model and simulate such complex systems from the design phase should, as Vaughn succinctly puts it, enable engineers to “discover better designs, faster.”

This story appears here as part of a cross-publishing agreement with Scientific Computing World.

Sign up for our insideHPC Newsletter