Midsurface modeling 1

[kotawsu]

what is the advantage of juss modeling the top face of a sandwich panel over modeling the midsurface of the panel? any insight. or can a sandiwch panel be idealised by midsurface modeling?

 

[GBor]

Modeling sandwich stuctures is generally a little different than modeling general plate structures. Sandwich structures are often built (analytically) from the top surface down so that layer #1 is the top surface, layer 2 is next, etc. When the calculation engine (the processor) calculates stresses, it will generally bend at the mid-plane of the core layer...kinda' like a beam bending about it's neutral axis, sandwich structures bend about a "user defined neutral axis"...the core mid-plane. Since the processor needs a reference point as it builds the sandwich, you mesh the "top" surface, but this is not true in ALL processors...make sure you understand what yours does. Some do want you to model at the mid-plane of the core.

For post-processing, you have to know which layer is on top and which is on bottom. If you are entering an unsymmetric laminate, you have to know which layer you are looking at when you evaluate the stress in each layer, so you need to know that layer 1 is on top or on bottom. The direction of the build is generally defined either by the position of the mesh to the origin (i.e. if the mesh is above the origin, you may actually be meshing the bottom of the plate...if the mesh is to the right, it will radiate outward to the right), or by some pressure vector. This is one reason why many packages actually do want you to mesh at the core mid-plane...cores usually line up at joints.

You also have to make sure that orient everything properly. You have to define a primary fiber direction (generally considered to be 0 degrees in your lay-up) relative to the global axis system and then define each layer relative to the primary fiber direction.

It's a lot of bookkeeping...

Garland

Garland E. Borowski, PE