Pre/Post Processor FEMAP Puts Users First

FEMAP Director Talks of GPU Compatibility, Model Merge and Global Local Tools.

Mark Sherman, FEMAP Product Development Director explains a customer based improvement to FEMAP

 Last week at the Femap Symposium Mark Sherman, Product Development Director, described the direction of the pre and post processing solution. The overwhelming answer is that it’s up to the users. User feedback led to a variety of updates in 11.1, including enhancements to mid-surface extraction, the addition of model merge, a new global-local modeling workflow and continued expansion of Femap’s graphics architecture. And it is user feedback that pushed for output data organization and further graphic updates in 11.2, due this October.

The moral of the story: If you have an issue, let Sherman know. You might see the solution in the next release.

As Femap is only a Pre/Post processor, users must purchase a solver and CAD service which will be compatible with the system. However, Femap is compatible with most solvers and CAD systems out there.

Knowing which NX Nastran features to support is easy for Femap since both work closely together under Siemens. However, Sherman described supporting other solvers in a different manner. With ANSYS for instance, the Femap team relies on their customers to indicate which release features require Femap support. Additionally, he was unsure how the sale of the NEi Nastran solver to Autodesk would affect the future. Although, he was quick to provide assurance to the users that compatibility between the products is paramount and that the Femap team is constantly listening to their user base to implement important features and correct any issues. Unfortunately, this game of catch up between users’ input and 3rd party solvers can be seen as a lack of direction.

Customer First Program Updates

Sherman has been working at Femap since 1985. Back then FEA experts were programming for the needs of FEA experts, and this drive has not left the company. The same core team works to make sure users are taken care of in each release of Femap.

Even during his presentation a user stopped Sherman about a usability issue, to which Sherman replied, “That is an easy fix, I’ll get that up by Monday and have a maintenance release.” It was Thursday at the start of a conference. He added, “Please call us when you run into an issue or performance bottleneck,” as he tells a story of helping another user reduce a 12-hour job to 30 seconds! 

“Tell us soon if you know of any error or bugs,” says Sherman, “11.1.2 will be coming up fairly shortly and we have an opportunity to fix it.”

If Femap is unable to make a fix for you, then I suggest you make your own using the API programming tool. Many users noted the ease of making their own fixes and tools thanks to the help from Femap documentation and support. However, these are stories for another time.

Midsurface Extraction Tool

First on the list of updates is the Midsurface Extraction Tool used to turn thin solids into shells. “The Parasolid geometry team had written a new mid surface technology which takes solid models, that are kind of thick shells anyway, and converts them into a thin surface representation. We had written our own years ago and it worked great on a lot of parts but the Parasolid team are basically geometry experts and we saw it was a lot better and faster,” explains Sherman.

Sherman remembered that while working at Boeing, the practice was to do this when a shape was one tenth as thick as it was long and wide. Fortunately, this would include many aerospace parts, such as fittings, skins or frames. Naturally, a CAD expert would model the part as a solid; however, this isn’t efficient for FEA analysis. Despite the large advancements in PC’s and Laptops, it would take a long time to run a system with so many solid meshes.

Model Merge and Global-Local Tools

Model Merge allows for model entities to be copied between models safely.

Sherman notes that “one of the biggest things in version 11.1 is Model Merge. A lot of our customers were interested in this.”

He explains that Model Merge allows users to reuse a mesh model, or sub-component, in future models. Essentially, “If you model one frame of an aircraft you can reuse it numerous times in different positions and model merge will make sure that nothing clashes. This includes ensuring model ID numbers, groupings, and coordinates will not cause issues. That is unless you want to over write something. You will have complete control,” says Sherman.

Sherman notes that many of his customers are dealing with a multi-vendor environment. For instance, the keynote speaker Mark McGinnis, the James Webb Telescope Lead from SGT Inc., noted that they used models from both the European and Canadian Space Agency. Model Merge would have allowed SGT to seamlessly incorporate these additional models into their own model.

“I think it’s great. The changes to Femap will really help the pre and post processing I do for the James Webb Space Telescope,” said Dr. Emmanuel Cofie, ISM Lead Thermal Distortion Analyst at SGT Inc. “The work you do in pre and post processing can be very monotonous, anything to help make the work faster is useful. I learned a lot at the conference about API programming. I will use the new features like Model Merge and the API to help my work. I will also suggest to my colleagues at SGT Inc. and NASA/GFC to use them as well.”

Conversely, Femap has included a new function called Global-Local to do almost the opposite procedure. “We implemented a Global to Local modeling tool to automate the tedious process of using global model results to drive detailed local models to recover accurate stresses. You take the loads from the big model and transfer them intelligently to a more detailed local model,” said Sherman. The Global-Local functions also preclude overwrites and clashes like model merge.

“Global-local modeling is going to make life a lot easier for analysing breakout models,” said Victoria Harris, Project Engineer at ATA Engineering Inc. She adds, “I also like the way they are constructing RBE elements. This new way to construct the elements in Femap will be useful.”

Graphics Architecture to OpenGL 4.2

Sherman recounts that “over the last few releases we have been transitioning to OpenGL 4.2 graphics. This will allow you to leverage all the GPU processors in the graphics card.”

This is impressive as some GPU cards can have thousands of graphics processors. Sherman explained that an order of magnitude improvement can be seen over older OpenGL graphics. “It makes working with model with millions of nodes that much more interactive. This will allow users to use the graphical interface of Femap in a more reasonable time frame. It makes the analyst’s life that much easier. Its real cutting edge technology,” expressed Sherman.

Results Files Pointers

Traditionally, Femap files would read in the results data from a solver to interpret the data and input it within a Femap database. This would essentially make a second copy of your results and a really large Femap file.

“What we are doing now,” Sherman explains, “is keeping track of what’s in the results file, and where, using pointers. This way we can access it dynamically without internalizing all the data. This allows you to postprocess huge files that were cumbersome and time consuming to handle before.”

NX Nastran and Other Solver Support

“Most of our customers run the Siemens solver NX Nastran, I would guess about 70%. The NX Nastran guys update their product continuously, and as a pre and postprocessor, we need to keep up to date. So every release we have to see what’s new and decide how we will support it. For instance, there is a new pyramid element type we now support,” marks Sherman.

Femap now also supports Restarts directly in the user interface. Previously, users could access this functionality in Nastran by using type commands. Now that it is incorporated into the GUI, it’s much easier to use. “This allows for users to use modal analysis results downstream in a subsequent dynamic analysis without needing to resolve the modes. This will help a user to leverage existing runs without starting over,” clarifies Sherman.

Other NX Nastran functions now supported by Femap include:

  • Design optimization, which helps users optimize a design and avoid natural frequencies using mode frequencies and eigenvalues as design constraints
  • Dynamic response output includes RMS Von Mises stress output for random analysis, and tracking results on individual lamina in composites
  • LOAD and constraint set combination display for load combinations and SPCADD/MPCADD combinations

However, Sherman did admit that they do not scan through ANSYS releases as thoroughly as they do for NX NASTRAN. Therefore, if you have any compatibility issues with new ANSYS updates make sure to let a Femap representative know. As previously mentioned, they listen to their users.

As for NEi’s sale to Autodesk, Sherman notes that “NEi was a long term cooperative relationship. We are here to provide ongoing support to the Femap customers. We’ve worked with mutual customers and will continue to do so.”

What’s Coming Up in 11.2

On-the-fly Post-Processing of Complex Data with a few clicks.

 “The organization of the output interface will be the biggest feature of 11.2,” said Sherman. “Results from Nastran, and all FEA solvers, can be ridiculously voluminous, with thousands of thousands of time steps and output sets. The new capability will keep this nice and organized and then quickly postprocess using the organized data. So you can organize a bunch of max/min diagrams by clicking one button. It is a streamlining of post processing of large data sets and envelopes which can include hundreds of data sets.” Essentially, this allows you to create max/min or data probe the outputs on-the-fly with a simple right click. Data is calculated within memory immediately and can be displayed right away.

The new graphics expansion into OpenGL 4.2 will continue in 11.2. It will include bringing constraints, masses, rods, bars and beams (curved and parabolic) into the new graphics set up. All element types will eventually be included into OpenGL 4.2.

Thickness data will also no longer be included in the properties. It will be included in the elements themselves allowing only the material to be located as property. “This is just a cleaner book keeping of the model,” Sherman explains.

“People always ask me, ‘What’s the long term road map for Femap?’” notes Sherman. “I always tell them we are always listening to users. I’m going to respond in the next version of Femap with what helps my users the most; years from now I’ll do it again. So I never really know what’s coming in a year from now. Things might change. We always want to make it faster and more responsive but small details are hard to identify downstream.”

Images courtesy of Siemens PLM Software.

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, 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.