Eparish1/op inf

[eparish1]

PR for op inf branch. The code is primarily orthogonal, and the main design concept is an opinf model that contains the ROM operators as well as a SM model that we interface with. I'll try and highlight changes to existing code blocks:

1. In ics_bcs.jl, you will see the addition of a few "typeof" commands in the blocks for the Schwarz DBCs. These are needed to enable coupling between the FOM and ROM subdomains. If we wanted to avoid these, we'd need to overload the BCs on the model and the coupled domain model.
2. You will observe the addition of side_set_name to SMOverlapSchwarzBC. This is required so that we know what ROM operator corresponds to what sideset.
3. In simulation.jl, you'll notice a slight re-org of how the different classes are constructed. This is done so that integrator and solver can be overloaded based on the model.
4. Subsequently, you will see that HessianMinimizer and Newmark now require a model for construction. They pull the number of dofs from the model class. I haven't done this for all the different solvers and time schemes, but I could.

[eparish1]

@ikalash @lxmota the CI failed, so I'll have this as a draft until I diagnose.

[ikalash]

Thanks for working on this @eparish1 ! I will take a look for sure, but lets have Alejandro make the final call on merging since he may have structural/design suggestions based on his vision for Norma. He will be back from travel on Mon. Jan. 13.

[eparish1]

@ikalash @lxmota the CI failed, so I'll have this as a draft until I diagnose.

CI is fixed, was related to the manifest.toml