CAESAR II uses point supports, your sea floor is a continuous support. To get the CAESAR II model to "match" the sea floor, you would have to enter many, many point supports. That's not practical. We do a few things in our buried pipe modeler to address this issue:
1) to model soil friction in the axial direction, we can space these friction supports far apart without affecting the axial response due to friction
2) transverse response requires closely-spaced supports to pick up bending - in your case, you would locate more friction supports where bending occurs
3) we assume that the soil supports deadweight, therefor, we eliminate weight on buried sections so that those distant (vertical) supports (handling axial friction) do not creat a deadweight bending problem
4) since we eliminate deadweight on buried sections, the normal loads in the vertical direction are improperly defined for calculating normal loads (N). These normal loads would be used to set friction forces (mu*N; you are suggesting mu=1.2 here). So, instead of friction we use a bi-linear stiffness in the axial direction to model the hold/break-free situation with axial friction.
Does this answer your question? No. But perhaps you can now determine how close your Y supports should be - is there bending? is there weight to be carried?
And the other question - should mu=1.2? I'll leave that to others. Perhaps once you're done with this job, you can add your insights here.
_________________________
Dave Diehl