The Boundary Element Method does away with that obstacle entirely, with one small caveat which will follow presently. Using this method, SIDT members won't need to use FEM for the mechanics modeling. They'll only need to create surfaces for the other simulation objects within Evolver, and then import the whole geometry into this BEM solver to set up mechanics BCs and calculate the displacement field.
One can also use the Green's Function Library concept to introduce defects, such as voids or internal cracks in the solder, as the "S2" which we could move around freely to examine the impact on stress field. Or use the solder droplets as "S2" and test the effect on an entire package of changing the solder volume or pad geometries. This would merely involve creating new Evolver solutions for the new solder conditions, and importing them in such a way that we show the BEM program what is and isn't being changed, e.g. using the Evolver colors to signal to the BEM program what is S1 and S2 (and S3 and ...).
Evolver seems to be a natural platform for this because it is already used throughout the Solder Interconnect Design Team, and it presents a natural and relatively "simple" method for generating a mesh on a surface. There need to be safeguards to ensure fine enough meshing on the surfaces. But I have ideas on this, which are similar to a 2nd-generation automated Evolver mesh refinement scheme currently being developed by Adam Powell.
And now the caveat. A lot of what people worry about in solder is plastic deformation. This is especially true of reliability modeling, which is entirely based on the extent of plastic strain during cyclic loading. An attempt on this problem will be an important part of the second-generation Green's Function library activity, but will probably involve a consiberable amount of time and effort.