This paper presents the results of the analysis of different 3D designs for the first stator and the rotor of a 1.5-stage turbine test rig. A tangential endwall contouring for the hub and the shroud, a bowed profile stacking, and a combination of those have been designed for the first stator. In addition, a tangential endwall contouring has been designed for the hub of the unshrouded rotor. Part I of this two-part paper deals with the design process and the numerical analysis of the results. All designs have been optimized using the stage efficiency as target function. For the design of the 3D stator vanes, the optimization led to an unexpected result: The secondary flow vortex strength increased. However, the secondary flow pattern is rearranged by the 3D-designing, leading to a smoother radial exit flow angle distribution. A subsequent reduction of the rotor losses overcompensates the higher stator losses. In order to understand how the 3D vanes affect the stator secondary flow pattern, a detailed analysis of vortex stretching and vortex dissipation is presented in this paper. With this approach, the various impacts of the 3D designs on the secondary flow vortices' strength can be quantified. In addition, the potential theory effect of the self-induced velocity is introduced here in order to explain the effects of a tangential endwall contouring on the trajectory of the pressure side leg of the horseshoe vortex (HVps). To the best of our knowledge, both approaches are new for the analysis of turbine secondary flows. The impact of the stronger but rearranged stator secondary flow on the rotor secondary loss development is analyzed by means of unsteady simulations. The results show that the rotor secondary flow can be effectively reduced through a proper stator secondary flow pattern. In Part II of this paper, the analysis of extensive experimental results validates and supplements the numerical analysis.