When used properly and at the earliest stages of design, CAE tools enable engineers to understand, predict, and improve product performance digitally. CAE software also enables the design team to explore more design concepts, cut costs associated with traditional physical prototypes and make faster, more informed decisions as designs progress through the design cycle.
When first introduced, computer-aided analysis (CAE) tools fell into the domain of specialists, analysts tasked with simulating stress or fluid forces on digital models. These results were then passed off to engineers tasked with tweaking, if necessary, the detailed design based on the results of the simulation. Wouldn’t it make more sense—and save valuable design time—for the engineers to analyze their models?
The answer seemed to be “yes,” and vendors began introducing CAE tools targeted specifically for engineers. This new breed of CAE software was integrated with users’ CAD software, and offered improved ease-of-use, reduced technical lingo specific to analysts, and lower prices—all aimed at organizations hoping to improve upon their design process by cutting costs and speeding time to market.
The strategy made sense, so why do engineers still struggle with obtaining meaningful, useful results from CAE? Making the software easier to use encouraged more use by engineers; however, it also led to misuse. Engineers accustomed to getting very specific “yes” or “no” answers from their CAD programs often put too much trust in “red is bad, green is good” analysis results, when often these results are unreliable.
Learning how to use the software doesn’t always translate to obtaining valid results. Let’s look at some way engineers can improve their use of CAE tools.
Modeling matters. The trouble engineers often encounter with CAE starts with the model. According to a report by Autosim, a European consortium focused on the use of simulation in the automotive industry, of the total time needed by engineers to create a simulation, 80% is devoted to generating a model.
To achieve good results, engineers must do more than create a mesh, boundary conditions and loads. They must prepare an analysis model so that it captures a fundamental understanding of the problem they’re trying to solve. Often these problems are ill defined in the software, leading to unreliable results.
Other problems reported include incorrect boundary conditions, inadequately meshed models, results that are not validated and non-applicable material properties. Another common obstacle for engineers using CAE is obtaining good input data.
In some cases, the engineering problems are “multi-physics” in nature, so entry-level CAE software may not be able to solve them. For many of these challenging real-life problems, the expertise of domain specialists may be required to solve them.
Training is essential. In order for engineers to properly use CAE software, they must get the proper training. The focus of the training shouldn’t be on the use of one particular software package, but on practical and theoretical aspects of multi-physics modeling. The training should also teach students how to effectively model real-life problems. Until engineers understand the basics of what the simulation software is doing, learning the specifics of that software won’t help.
Beyond training offered by vendors, another good source of CAE education is the International Association for the Engineering Analysis Community (NAFEMS). This vendor-neutral organization offers training courses throughout Europe and the U.S., as well as e-learning programs and “Hot To” and “Why Do” Benchmark reports. The independent e-courses address all levels of FEA use, from practical introductory basic-level sessions to Non-Linear and Dynamic FE Analysis.
A free option is MIT’s OpenCourseWare, which offers self-guided, online learning. It offers lecture notes, assignments, exams and study materials for courses including “Finite Element Analysis of Solids” and “Fluids I, II, and III.” Yet another option is an MIT OCW video series, “Finite Element Procedures for Solids and Structures,” presented by Dr. Klaus-Jürgen Bath, a pioneer in the development of nonlinear analysis software and a professor at MIT.