Hello Satish,

        I have done a similar analysis in FEMLAB 3 - much better than Ansys!
:-) - and here are a few comments on your questions:

> > My questions are..
> > 1. Is it right to model the 3D air around the structure to get the
> total
> > capacitance with fringing in all the directions.

        I have done some comparisons between 2D and 3D model.  You will find
that there is an "optimum distance" from the comb structure to the
surrounding volume boundaries which will account for all fringing
fields.

> > 2. Is there any criteria to optimise the size of air model that is
> > modelled around the structure.

        That will depend on the distance between the comb fingers, hence how
much fringing fields will have in your design.

> > 3. what is the significance of applying 0 volt to the air model
> surronding
> > the structure. But even if i don't apply 0 volt to the boundary I
> end up
> > with same results.

        I believe you must define the surrounding volume as an insulator,
instead of a conductor at 0V potential.  I'm not quite sure I
understood your question, though...  I hope that helps.  You can also
post your questions on [mems-modsim], which is the "mems simulation"
group.  Another option is visiting the FEMLAB_Users group at
http://groups.yahoo.com/group/femlab_users/

        - Isabelle