Structural Software Claims Better Accuracy without Finite Elements

SIMSOLID also does away with having to simplify geometry.

(Image courtesy of SIMSOLID.)

(Image courtesy of SIMSOLID.)

A new simulation company, SIMSOLID, explains that its new platform democratizes the simulation process by eliminating the need for geometry simplification or meshing. If this works, this would result in a tremendous savings of time for design engineers who are doing simulation.

Meshing and geometry simplification take a significant amount of time in the simulation process. Additionally, these processes can be a complicated task, which act as a significant barrier for engineers who are not simulation specialists attempting to simulate early in the design cycle.

This goal of getting simulation early in the development cycle is likely why SIMSOILID has direct integration with Onshape CAD.

“SIMSOLID, with its meshless technology, removes yet another barrier between having to be a specialist and getting a job done. It is structural analysis for everyone,” says Christopher Schaefer, product development engineer at global manufacturer Gardner Denver.

Another complication of traditional simulation is the assessment of large assemblies and parts. However, SIMSOLID reports that they can perform these simulations quickly on a desktop computer.

“All the codes I know are designed for a single part and then they extended the code over time to use assemblies,” said Ken Welch, founder and CEO of SIMSOLID, and a former vice president at MSC Software. “Ours was designed and optimized for assemblies from the start, so it’s much more tolerant to connections than traditional codes. This can be helpful when you are modeling hundreds of parts in an assembly. Our system can perform this simulation in minutes.”

“The ability to analyze large models without having to simplify them first is a definite timesaver,” said Leif Erickson, vice president of engineering at Cenairus Energy Services Ltd.

How Does Simulation without Meshing Work?

SIMSOLID’s solver is based on the “theory of external approximations,” a generalization of the finite element method (FEM) in which arbitrary shapes can be used as finite elements. Additionally, the functions that approximate the physics of interest in the volume are an arbitrary class independent of that volume’s shape.

“With any new technology, people want to understand it. Ours can analyze the geometry directly, so it’s much faster to do design iterations,” said Welch. “Our system then iterates for added accuracy, and it does it for the whole geometry.”

Another difference to traditional FEA is that SIMSOLID doesn’t use point-wise degrees of freedom (DOF). Instead, the DOF are functionals that have geometry support from areas, line clouds, point clouds and volumes. This explains how the platform is able to simulate geometries without simplifying gaps, penetrations and other imperfections.

To learn in detail about how SIMSOLID works, it’s accuracy and how to validate your simulations with it, check out the white papers on their website.

Some other features of SIMSOLID include:

  • Physics: structural statics, modal, thermal and thermal stress
  • Material properties: isotropic, incompressible and rigid
  • Assembly connections: bonded, sliding and bolting
  • Units: SI and IPS
  • Image gallery to save and restore results
  • Monthly and annual subscriptions

How Does SIMSOLID Compare to Traditional FEA?

The next question that every engineer worth his or her salt will ask is, how does this simulation software compare to traditional FEA? After all, the thought of simulation without meshing is likely to cause some skepticism.

To alleviate this concern, Welch said, “We’ve actually been testing it for the last year. We have an extensive validation manual online . . . . Accuracy is a funny thing. You can say, ‘which is more accurate’ or ‘which is closer to reality?’ For traditional FEA, you have to get rid of rounds and holes and you will get stresses that are not accurate to reality.”

Welch also referred to a previous blog on his site that compared results from Autodesk Inventor, NX Nastran (pre- and post-processed in FEMAP and SIMSOLID. The benchmark result was created in NX Nastran using FEMAP to iteratively refine the mesh to 100,000 elements. This was then repeated in Autodesk Inventor.

Simulation comparison of Siemens’ NX NASTRAN (in FEMAP), Autodesk Inventor and SIMSOLID. (Images courtesy of SIMSOLID.)

Simulation comparison of Siemens’ NX NASTRAN (in FEMAP), Autodesk Inventor and SIMSOLID. (Images courtesy of SIMSOLID.)

The simulations were then repeated using the default settings for the previous software and SIMSOLID. These test results saw that, on its first try, NX was 41 percent off from the benchmark. Inventor, even with its adaptive automesher having three passes of refinement, had a 27 percent error on its first try. However, SIMSOLID had an error of only 1.7 percent on its first go.

Now, this is one simulation case from a blog created by SIMSOLID itself. Clearly, more skepticism is healthy. However, it should illustrate that this new simulation technology does show some potential to reduce the amount of work needed to get an accurate simulation result to reality. As a result, it might be worth a look.

Written by

Shawn Wasserman

For over 10 years, Shawn Wasserman has informed, inspired and engaged the engineering community through online content. As a senior writer at WTWH media, he produces branded content to help engineers streamline their operations via new tools, technologies and software. While a senior editor at Engineering.com, Shawn wrote stories about CAE, simulation, PLM, CAD, IoT, AI and more. During his time as the blog manager at Ansys, Shawn produced content featuring stories, tips, tricks and interesting use cases for CAE technologies. Shawn holds a master’s degree in Bioengineering from the University of Guelph and an undergraduate degree in Chemical Engineering from the University of Waterloo.