Selecting the correct pre- and post-processing software is crucial for fast and accurate analysis. How do you know if your pre- and post-processor measures up? Take into account these considerations:
- Accuracy
- Accessing CAD Data
- FE Model Creation and Idealization
- Solver Support
- UI
- Results Comprehension
- Automation and Customization
- Overall Value and Support
Accuracy
“Does the pre- and postprocessor enable you to control the FE model creation process sufficiently to ensure the creation of efficient FE models without sacrificing accuracy?”
Accuracy of the simulation is not dependent upon the solver, at least not entirely. You do have to pick the correct solver for the problem, but after that it's all about matrix maths. Accuracy of the simulation results is more dependent upon the setup of the model, and that happens during pre-processing.
Meshing, idealization, loads, and boundary conditions must be properly applied to get accurate results. (“Accurate” in this context is defined as results that represent as-tested data or real-world reactions.)
The pre-processor should allow for full control of the mesh – element type, shape, and size. Just as important, and often missing in cad-integrated FEA tools, is the ability to determine convergence of the mesh. Running a convergence analysis in the pre-processor improves the accuracy of the mesh without the time and overhead of having to run repeated solves on the entire FE model. Does your pre-processor have the tools to ensure accuracy without sacrificing speed?
Accessing CAD Data
“Can the pre-processor import and manage geometric data from multiple CAD systems and data formats?”
Geometry can be created from any number of systems in any number of file formats. If your pre-processor only works with its native file type, then speed is sacrificed and errors potentially get introduced as the geometry has to be recreated.
Importing CAD data from the original design source not only saves time, but reduces errors by allowing for a single source of design definition used through the enterprise. But data translation introduces its own potential for errors in geometry or topology. The fewer data translations, the less likely that errors will be introduced into the geometry. Does your pre-processor have the tools to ensure import of native CAD files as well as industry-standard data formats?
FE Model Creation & Idealization
“Does the pre- and postprocessor allow you to idealize certain topologies such as thin-walled models, and create smaller, more accurate finite element models?”
There is a simple FEA workflow that consists of
importing a solid body,
applying a tetrahedron mesh,
and adding loads and boundary conditions.
But, in terms of the mathematics behind finite element analysis, that workflow may be neither the most accurate nor the most efficient technique.
First of all, small curves and sliver surfaces cause excessive numbers of elements which greatly increase solve times, assuming the mesh is even able to generate across those problem areas. Thin-walled structures, like sheet metal parts, are difficult to mesh with the minimum number of elements across the thickness. And structural shapes are best analyzed by beam calculations, not finite element approximations. Reducing the number of elements by properly creating the correct element type reduces the solve time while improving accuracy. Does your pre-processor have the tools to create finite element entities and idealize geometry?
Solver Support
“Does the pre- and postprocessor support the export of solver input files and the import of solver results files? Can it support the parameters needed by industry-leading solvers?”
Each of these analysis types requires unique elements, boundary conditions, and load cases.
The pre-processor is the tool used to prepare the model for solving. Although each FEA software solution usually comes with its own pre-processor that has the elements, boundary conditions, and load cases to ensure a successful solve, using the unique pre-processer requires maintaining proficiency in multiple applications.
Likewise, the results of the solver must be imported into the post-processor. Again, the built-in post processor could be used, but that also would require proficiency in multiple applications. A universal pre- and post-processor can save time and dollars with reduced training and increased proficiency.
Does your pre- and post-processor create input files for any solver and does it import results from the solver? Is your pre and post-processor scalable to handle the advanced functionality of additional solver modules?
User Interface
“Does the user interface promote productivity in the use of the software and expedite learning?”
Ease of use is not defined by the simplification of complex mathematical theories. Ease of use is defined as the interaction between the user and the software.
Regardless of how complex the simulation is, the software should be easy to use in order to quickly and accurately create the idealized model, solve the simulation, and view the results. The user interface should not be cluttered with seldom used features, or features that don't apply to the current analysis type.
The user interface should be intuitive. Help files should be thorough and available on-demand to aid users in finding the right command to guarantee proper application of elements, loads, and boundary conditions. Finally, the interface should allow for customization so each user can tailor their experience for optimum performance. Does your pre-and post-processor have an intuitive user interface to ensure efficient navigation of the application?
Results
Does the post processor allow for complete interrogation of results to provide a true understanding of the simulation? Does it allow for multiple views of the data to extrapolate additional information?
Simulation results are vast. Disregarding the graphics card's ability to project the data onto the screen, it is the post-processor that determines the information to display. And analysts need to be able to interrogate results in any multitude of ways: Von Mises stress, primary stress, strain, internal stresses, moments, reaction forces; those are just a few of the results for linear static analysis. Thermodynamic and fluid flow results add even more.
First, results files are huge, sometimes into the terabytes. The ability to manage results files as a single database or external references are needed when combining simulation with limited IT resources. Second, once the results file(s) are loaded into the post-processor, visualization tools must be intuitive. That means reconstructing an idealized model into something closer to the as-designed geometry via the results. Does your post-processor handle results files of all shapes and sizes and provide the visualization tools to fully interrogate the results?
Automation & Customization
“Does the pre- and postprocessor have built-in toolkits, macro creation capabilities and an application programming interface? Is knowledge capture possible and can workflows be automated and adapted to the organization's processes?”
Like many engineering tasks, finite element analysis is full of repetitive tasks. Automating those tasks saves time on each and every simulation. Likewise, customization of the pre- and post-processor through an application programming interface (API) can be used to enhance the application for more advanced and involved simulations that may be beyond the capabilities of the release version of the software. Does your pre- and post-processor allow simple automation of repetitive tasks and customization of advanced tasks?
Overall Value and Support
“Does the pre- and postprocessor vendor provide sufficient support to ensure maximum productivity? Does the vendor release regular software updates and do they provide useful new capabilities as well as error fixes?”
Even if a pre- and post-processor satisfactorily answers all the above questions, there is still one more consideration. Software changes, and updates to software should be regular and of high quality. Staying current with changes to CAD file formats for easier import of data is necessary. Bug fixes are critical to accurate simulations. Do you know who is behind the software you use?
Conclusion
Defaulting to the built-in pre- and post-processor that came with your solver is an easy decision, but to truly take advantage of finite element analysis, the same rigor in choosing your solver should be used to choose your pre- and post-processor.
FEMAP, from Siemens PLM Software, is a pre- and post-processor that resolves all these questions. Find out if your pre- and post-processor meets requirements by reading the Buyer's Guide and then try FEMAP for free for 45 days.
Additional Resources:
Aberdeen: Cost Saving Strategies for Engineering
Aberdeen: Simulation Driven Benchmark Report
FEA for all Engineers
Top 10 Reasons to Upgrade to FEMAP
Siemens has paid a fee to ENGINEERING.com to promote their simulation solutions. They have not had editorial input to this post. – Scott Wertel