Add a new phase time option: t_final_target

[robfalck]

Issue Type

• Bug
• Enhancement
• Docs
• Miscellaneous

Description

We currently have a case where the final time of a phase is used like a parameter within the phase.
We can accommodate this by connecting the last element of timeseries.time back to parameters:{param_name}, but this introduces the need for non-default nonlinear and linear solvers.
We could get around this if we simply allowed another time option: t_final_targets, that would function in the same way that t_initial_targets and t_duration_targets does.

Example

TBD

Environment

dymos 1.5.1-dev