How to properly bond 2 body's - erratic frequency analysis results

[FBastiaansen]

Hello,


I am modelling a coated cantilever beam in ansys (an AFM probe with a thin gold coating on the back side). Constructed a bimetal beam, used "add frozen" for the second layer, and then "form part" to finish it.

Now the simulation: I want to analyze the eigenfrequencies and mdoes. The frequency finder results show that the coating is resonating as if it were not bonded to the substrate material at all! I should point out that the dimensions for this simulation are pretty small (cantilever beam thickness ~1um, with a 50nm coating). I included one of the results

Does anyone have an idea/solution/hint on how to properly bond the two together?


Thanks a lot,
Ferry

 

[FBastiaansen]

Additional info: I work with Ansys 10 Workbench, geometry was created in mm (smallest allowed setting)

 

[hendersdc]

I have not worked with a bimetal beam but one thing that I have heard is that the deformation results in a modal analysis are basically meaningless. I would suggest taking the output of that and puting it into a random vibration analysis and see what happens.

 

[vibration24]

There is an option to bond the two layers with different material properties, if you want to try in Mechanical APDL (ANSYS).
Postpro.> modelling> Operate> booleans> glue
Select volume, are or line accordingly.

 

[FBastiaansen]

Hi and thanks for the replies sofar. I have found a way to keep the materials bonded, with the "contact tollerance settings". A proper solution would still be nice though.

In the normal ansys i would indeed use the boolean/glue. With workbench this option is not available. How do i access the Mechanical APDL via the workbench environment?