Torsion of a non-circular cross-section with more than one material

[VKulkarni]

I am trying to understand how to tackle torsion of bars that are non-circular and two-material composites. I understand the process of finding the torsion constant for non-circular sections but I am not sure how the warping function changes across a material interface.

For example, if I have a solid square bar made from HDPE and one on surface I bond a thin sheet of Aluminum, how does the torsional behaviour change? Is there a process to calculate the new torsional constant? If this were a bending problem, we could convert the composite cross-section into an equivalent section using the ratios of elastic moduli of the materials and then find the area moment of inertia of the new section. Is there an analogous process for torsion?

 

[SteamKing]

This is to go-to guy for these types of torsion problems:

http://scholar.google.com/citations?user=HsBcGJ0AAAAJ&hl=en

This is one of the papers you should probably lay your hands on:
http://scholar.google.com/citations...J&citation_for_view=HsBcGJ0AAAAJ:UeHWp8X0CEIC

There are several others by S. and his student Mokos which deal with this type of problem. Be warned, know something about BEM and advanced mechanics of materials in order to digest these papers.