Ben Wallander
Ansys Inc.
Pittsburgh, Pa.
Assembly analysis is one of the most complex techniques for predicting how multicomponent structures will perform. But thanks to advancements in automation, engineers can now benefit from an easier assembly analysis with proven methods. As with any tool, users must understand its capabilities, limitations, and potential hazards to make the most of it. Armed with this knowledge, they’ll soon discover that all analysis methods aren’t created equal.
The common-mesh technique is the most basic assembly-analysis method because it simulates what’s called perfectly bonded geometry — the structure is one continuous finite-element mesh. The technique lets engineers assign different material properties to different parts. It works best on assemblies that are intended to perform as though they are perfectly bonded. Examples include parts welded together without regard for weld strength, or bonded materials such as metal plating. While the technique gives reasonable answers for such situations, users must be aware of its limitations.
For example, actual assemblies typically combine welded and moving parts, with surfaces coming into contact. The common-mesh method, however, assumes all components are rigidly fixed to each other. There is no way to account for interaction among parts. This also means there is no way to account for an individual part’s nonlinear behavior. The common-mesh technique typically indicates areas of stress or deformation that are inaccurate due to individual part movement not being taken into consideration. In short, unless the actual assembly is perfectly bonded, results can be misleading if not blatantly wrong. While excellent programs are available for conducting true assembly analysis, beware of programs that only provide for common-mesh assembly analysis.
The Node-to-node assembly analysis technique is an improvement over the common-mesh method. The primary difference between the two is node-to-node’s consideration of interaction between parts. The node-to-node method simulates contact between parts by using contact elements. A “coordinated mesh” places nodes across the interface between two independent parts but still lines them up with one another. Contact elements are placed between nodes that separate one material property of an assembly from another. The technique puts strict requirements on mesh generation for an entire assembly, often at the expense of meshing robustness and accuracy for a given part. The technique assumes the accuracy required by the whole assembly is greater than the needed accuracy of individual parts.
However, although the method allows some part movement, serious limitations to this technique go beyond just the assembly mesh issue. For example, contact locations must be known if node-to-node contact elements, or gap elements, are going to be used. This can be nearly impossible in assemblies with moving parts. To make matters worse, contact problems usually handle relatively small sliding motion between contact surfaces of parts, even in cases of geometric nonlinearities. So, even if users can accurately determine which areas may contact each other, they are still limited to models in which nodes of individual parts line up perfectly, relative part sliding is negligible, and part deflections or rotations are small.
Node-to-node assembly analysis can be frustrating, because result convergence is typically cumbersome at best, and individual part accuracy is questionable.
The “node-to-surface” technique performs true assembly analysis using several advanced techniques. Unlike node-to-node analysis, this method accounts for large deformations and large relative part sliding without needing to know the exact locations of contact. The node-to-surface technique also eliminates the need for contacting parts to have compatible meshes. Node-to-surface elements typically model point-to-surface contact, such as those found with snap-fits.
The surface-to-surface technique is even more recent and handles real-world interactions between parts in an assembly. The technique uses either rigid-to-flexible or flexible-to-flexible contact elements. These elements have a target surface and a contact surface to form a contact pair. Once formed, FEA software can simulate real-world interactions among the parts.