I’d probably recommend using something like PyMol or Chimera. You should be able to take the duplicated protein, move it around relative to the other chain, and then save both chains to the same file.  (How exactly to do this depends on the program you’re using, so look to that program’s documentation for details.)

I’d recommend this because it allows you to position the two proteins in approximately the location and orientation where they’ll end up in the final docked form. Rosetta docking does better if the two chains are close to where they’re supposed to be. It won’t do very well if the initial chains are superimposed, which is what you’ll get if you just do a naive text concatenation of the files.

The other thing you’ll want to do as you’re adjusting the proteins is to change the chain letter designation of one of them. The Rosetta docking protocols often identify which partner is which by the chain letters, so it’s helpful to have the two partners be different chains.

Another option, given that you’re probably looking for a symmetric homodimer, is to look at the SymDock protocol: https://www.rosettacommons.org/docs/latest/application_documentation/docking/sym-dock  — For that protocol you don’t need to manually duplicate the chains, but rather use a “symmdef” file which specifies the symmetry (probably C2 in your case). Rosetta can then build the symmetric partner from the asymmetric unit based on the information in the symmetry file.

Sorry, it looks like there was a spurious non-printing character on the link try:  https://www.rosettacommons.org/docs/latest/application_documentation/docking/sym-dock