thumbnail group

Connect With Us:

ME Channels / Controls & Software

Manufacturing Gets its Day in Simulation

 Bruce Morey








 By Bruce Morey
Contributing Editor

CAE simulation is fast becoming an integral part of design engineering. As computers have grown in power, mathematicians and engineering researchers have devised ever more reliable ways of using CAE to both replace physical tests and move it upstream in development. This has become especially important for big-ticket items like airplanes or earth moving equipment, where building and testing multiple prototypes is prohibitively expensive.

Not to be left out, especially for readers of this column, is the work simulation professionals are doing in advancing new frontiers in manufacturing processes, such as casting, metal forming, joining or welding. A growing infrastructure of support data has mightily helped in this regard, from extensive libraries of materials data to individual companies cataloging test data for calibration and correlation. Ever more advanced software and mathematics helps too.

One example among others is MSC. Recognizing the growing importance of CAE in manufacturing processes, MSC Software in 2015 acquired Simufact. Actually MSC reacquired the core technology, as MSC had actually spun out its core manufacturing simulation technology to Simufact in 2006, according to Arjaan Buijk, business development manager for Simufact. Simufact offers products in three broad categories, metal forming, mechanical joining and welding. The core of these simulations are the same core non-linear finite element and finite volume solvers used in MSC’s design engineering offerings.


Accessibility Important

As many have recognized, simply using an accurate non-linear FEA is not where the real value lies. Accessibility is becoming as important as accuracy. As Buijk explained it to me recently, what the team of developers at Simufact have concentrated their efforts on is making their tools for forming, joining and welding predictions “easily used by In car body construction, the resistance spot welding is used to connect, for example, a B-pillar of a car with a sidewall frame. manufacturing engineers in their process.” The user interface, the software manufacturing engineers actually use, is built in familiar Windows. Computation can be farmed out to a Linux-based server if needed, or performed on the computer hosting the interface. “It is process oriented and easy to use as PowerPoint,” said Buijk. It uses forming technology terminology, with application function sets and templates. Tooling is imported as a CAD model and then converted into a finite element model. “We support all the popular CAD formats as well as STEP, PARASOLID and IGES,” said Buijk. “Most of our customers design their own tooling, so this was an important feature.”

When pressed about the underlying mathematics of mesh elements, Buijk explained that Simufact has already selected the best elements to use in any particular forming operation. For example, Simufact.forming supports simulating processes such as hot or cold forming, forging, rolling, or heat treatment. For each process, the company performed extensive analysis in deciding what type of mesh elements and underlying basis functions are best for that type of process. This means that the end-user need only worry about the process to be simulated, concentrating his or her time on the part rather than the somewhat arcane mathematical rationale in selecting elements and basis function orders. They need not be numerical mathematicians.


Integration Mimics Real Life

Simulating and connecting a number of processes seems to be an especially useful way to employ simulation tools in general. Most parts are manufactured through a number of steps, like rolling, forging, and heat treating. I spoke with an end user whose company supplied monolithic cast aluminum parts to the aerospace industry. He was using simulation to chain 19 different processes, trying to find ways to reduce grains from the casting process. MSC helps in this by providing easy ways to transmit simulation results from one software to the next.

Their other tool, simufact.welding grew out of a joint project with German automotive companies. “Audi in particular had just implemented remote laser welding, and that drastically changed how flexibly they could employ welding,” said Buijk. Simufact.welding simulates beam, arc, pressure, and resistance welding. Their welding tool simulates welding with a moving heat source model.

MSC provides a mechanism for sharing analysis results between its CAE tools. While this facilitates integrating processes with MSC provided tools, there is a catch – it is not easy to use tools provided by others. Buijk described how a major Tier 1 world auto supplier developed a way of transmitting results from some Simufact tools into ANSYS. “They did that on their own and the lack of a sharing standard like STEP or IGES in the CAD world remains a difficulty,” he said.


Accuracy and Calibration

Calibrating the simulation to an end-user’s particular process remains a key part of using simulations, including Simufact. “Simulation always requires a correlation or calibration step to an existing process and part before going into predictive simulations,” explained Buijk. By making sure the material properties and boundary and initial conditions are correct builds confidence in extending the simulations into predictive situations. “We can get forming 95% accurate when compared to measurements of real processes,” he stated. Material databases are available from a variety of sources, either from MSC itself or other partners.

Welding is a bit different, since actual welding processes can vary 20 – 30% from weld to weld. “So, if simulation is within that variability, it still provides a very useful tool for engineers.” For welding in particular, he stresses that a well-conditioned, high-resolution mesh is important for good results. “For comlex assemblies, you should be using a mesh of 500,000 to 1 million hex elements,” he explained. While he noted that it does not simulate a mixed material weld joint, it can predict phase transformations in the heat-affected-zone away from the weld joint. “For example, it can predict formation of martensite in ferrous materials,” he said.

Buijk also related that the company is fully embracing the 3D printed or additive manufacturing world with a new simulation tool, with an emphasis on powder metal additive manufacturing. “We thought we could apply our welding process to this, since on a fundamental level an additive process is basically welding,” he said. Their difficulties prove the point about making individual modules easy to use. “It is actually so different [from welding] than it needs its own environment,” he said. Their additive module calculates laser beam melting thermal stresses and deformations as well. Because of the severe computational load in this application, they developed a computational method of using uniform sized voxel elements – think cubes – with a new mathematical model they call Inherent Strain. “We expect to release this late in 2016,” he predicted.

Published Date : 5/25/2016

Editor's Picks

Advanced Manufacturing Media - SME
U.S. Office  |  One SME Drive, Dearborn, MI 48128  |  Customer Care: 800.733.4763  |  313.425.3000
Canadian Office  |  7100 Woodbine Avenue, Suite 312, Markham, ON, L3R 5J2  888.322.7333
Tooling U  |   3615 Superior Avenue East, Building 44, 6th Floor, Cleveland, OH 44114  |  866.706.8665