The sequence of codes Serpent/DYN3D has been developed by the Helmholtz-Zentrum Dresden-Rossendorf and successfully applied to core static and transient analyses of sodium-cooled fast reactors (SFRs). The successful application of the sequence to SFRs was made possible thanks to several recent development activities. The activities included modifications of DYN3D to account for in-core thermal expansions effects and development of a Serpent-based methodology for the generation of multi-group cross-sections (XSs). Despite the recent developments, the Serpent/DYN3D sequence presents still some limitations. That is, the domain of the sequence analyses is restricted to core-level and considerable computational efforts are required both to generate XS libraries for transient applications and perform related calculations. This thesis aims at overcoming such limitations through: The extension of the modeling domain of the sequence from core to whole SFR systems, the validation of the newly extended computational tool, and the optimization of the accuracy of solutions against computational efforts. The extension of the analysis domain to SFR systems is achieved by coupling the sequence with the thermal hydraulics code ATHLET capable of modeling the liquid sodium flow. Included in the extension is the development of ATHLET-based modeling methodologies to account for the thermal expansions of out-of-core structures that may strongly affect the reactor neutronics. Verification and validation activities inherent to the extension of the coupled tool are performed against numerical and experimental benchmarks based on the Phénix and Superphénix reactor data. Options focused on the simplification of XS libraries are considered to optimize computational times while preserving the accuracy of solutions. The options include the selection of optimal condensed energy groups structure for SFR analyses and the representation of parametrized XSs via first-order derivatives. The research activities conducted in this doctoral thesis led to the successful development and validation of a new computational platform for steady-state and transient analyses of SFR systems. The platform is currently applicable to the analyses of scenarios that do not involve sodium boiling or core damage. The options for the simplification of XS libraries and acceleration of neutronics calculations were implemented and tested revealing wide margins for significant speedup of analyses with the introduction of practically negligible errors.