Choosing the Analysis Procedure: Solid or Shell

Numerical simulation uses discrete approximations of continuum mechanics. Good numerical methods seek to balance accuracy, robustness, and performance.

The solid analysis procedure uses the finite volume method for Fill, Pack, and Cool analyses. This is a discrete approximation that applies equally to the three spatial dimensions. Results are more accurate when you use a fine mesh. However, you might decide to use a coarser mesh to achieve a faster, though less accurate solution.

The shell analysis procedure uses further numerical approximations to model the cross-stream flow and the heat transfer without using a mesh that spans the cross-stream direction. The shell analysis procedure applies primarily to thin parts with essentially uniform thickness.

These approximations assume that the flow is always parallel to the solid boundaries, which means that the shell analysis procedure cannot properly model cross-stream flow. The shell analysis procedure is further limited to injection processes that use a single polymer material.

Hele-Shaw Flow for Shell Analysis	Hagen-Poiseuille Flow for Shell Analysis
Laminar flow between two parallel flat plates separated by a small gap	Laminar flow in a long cylindrical tube

The solid procedure typically requires a minimum of five mesh elements to span a flow passage. To mesh a thin part, you typically need at least five to ten times more solid mesh elements than shell elements. Judicious use of the shell procedure saves computation time.

The shell procedure will be less accurate wherever the basic assumption—that the flow is always parallel to the solid boundaries—is not valid. This can happen, for example, when the flow impinges on the cavity wall opposite a valve gate, when jetting occurs, or when a cavity flow encounters a protuberance or step change in thickness.

However, such inaccuracies might be local in nature and would have only small effects on the overall flow results. In addition, even though the solid procedure might be able to capture some or all of these effects, it also requires a fine enough mesh resolution to do so. Without sufficient mesh refinement, the solid procedure might not be significantly more accurate than the shell.

Recommended Analysis Procedure Based on Part Geometry

Part Geometry	Example	Shell Procedure	Solid Procedure	Comments
Thin parts with few or no thickness variations		Yes	Yes	The shell procedure is most efficient for this type of part.
Thin parts		Yes	Yes	The shell procedure can handle small variations in thickness and some holes in the part.
Thick parts		No	Yes
Parts with large thickness variations		No	Yes
Perforated plates		No	Yes	The shell mesh procedure is suitable only in cases where the perforations are fewer or spaced far apart.
Grids		No	Yes
Rings or frames		No	Yes