Drowning in Simulation Data? One Company Has a Lifeline
Roopinder Tara posted on June 15, 2018 |

There may be more engineers using simulation than ever before. Thanks to its democratization, simulation is ever so easy to use—even rank-and-file product engineers or designers can operate what was once the domain of the few and privileged. 

3D NVH simulation results shown in VCollab. (Image courtesy of VCollab.)
3D NVH simulation results shown in VCollab. (Image courtesy of VCollab.)

But all the work has been going into preprocessors, says Prasad Mandav, CEO and cofounder of VCollab, who showed his product at CAASE 2018. No one has made it any easier to look at results. Meanwhile the mountain of data has been growing. It doesn’t help that the increasing complexity and size of models has also expanded.

Multiphysics can make the simulation process so much more complicated. Fluid flow, nonlinear analysis adds to cycles and storage, and if multiphysics tools are created by different vendors, their format will be different, too. Automatic meshing can make the model size immense. When engineers can push a button to discretize CAD geometry, as they can with most CAD programs, the result can be millions of 3D elements. Digital twins can be sending torrents of bits to each other.

All of these factors combine to make a crazy quantity of mismatched data.

At CAASE18, two companies were trying to make sense of it all. VCollab has created a proprietary CAX format that all manner of simulation data can be poured into. Ceetron also plans to unify simulation data with a cloud-based application in July.

Industry-standard formats, like STEP, have been working on it, according to Prasad. “But we’re ahead,” he said. STEP handles many things, mostly those related to geometry, and has been late to turn its attention to simulation. Meanwhile, VCollab has been unifying simulation data for around 18 years. The company is headquartered in Troy, Mich., near its automotive customers.

Big auto companies seem to have embraced VCollab. The company’s customer list is practically an automotive Who’s Who. GM, Chrysler … Ford is missing, however.

GM employs different solvers—about 200, said Prasad. Of that total, 60 of have been developed in-house.

VCollab reads data from Abaqus, MSC Nastran, MSC Marc, ANSYS, LS-DYNA, Fluent and many others. Not only can VCollab combine various formats into one, but it can also greatly reduce the size of the dataset, using “smart extraction and storage of key simulation information,” according to information VCollab provided. File sizes are reduced as much as 99 percent, per the company.

VCollab reads in simulation data from most—if not all—of the popular simulation programs. It can display the results on a basic free viewer, but two versions of an enhance viewer, with the most expensive one able to create geometry in the CAX format. VCollab viewers are based on HTML5, so they work on any device and OS without requiring installation or plug-ins.

It’s usually a one-way street, though. VCollab does not write data back to the same native format that it imports from, nor does it convert data to another format. It does export to standard formats, however, like STEP, 3D PDF, JT STL and VRML.

“We haven’t heard much of a need to go back to the native format,” said Prasad. 

Recommended For You