The words unsteady and periodic have different meaning depending on the context they are used in. My answer is mainly with regards to axial or radial turbomachines.
Lets say our machine is running at a specific operating point. The rotation of the blades (the shaft) does not change. The massflow-rate through the system is constant.
Even though this operating-point is steady the flow inside of the machine does not have to be steady.
For practical purposes one usually designs every blade row in its specific (relative) frame of reference. This means the designer 'sits' on the air-foil.
Now we assume further: steady flow- and boundary conditions. Doing this makes the aero and mechanical design feasible. When designing the blades and vanes for a turbomachine one will not design 'every' blade but assume that the flow around the blades will be the same for all blades of one blade-row. (Even though this assumption is usually not valid at the boundaries of the operating range it is a very practical approach to get the design-process started.) This means the flow is believed to be steady and periodic.
However! There is one thing called the 'Unsteadiness Paradox' 1. It basically boils down to the fact that a turbomachine has to be unsteady in order to work. But the unsteadiness in this case refers to the rotation of the blade-row.
The figure shows a sketch from Greitzer. If one assumes the turbo-machine to be 'black box' which just adds or extracts work than it is not possible to describe the turbo-machines analytically. It only works when adding a stationary and rotating black-box.
Without any simplifications of the flow and the geometry it would not be possible to design a turbomachine and predict its operating characteristics.
Two things have to be simplified:
In case of the aerodynamics the unsteadiness (eddies and turbulence) in the flow is usually averaged out. In case of the geometry one assumes periodicity (see the figure below from Oxford University):
For the version (a) only one passage was simulated and simply copied (all passages look the same). In version (b) all passages were simulated showing slight differences under certain operating conditions.
The second simplification to the geometry is the introduction of mixing planes so the rotating and stationary part of the turbomachine can be calculated separately (see figure below from Cambridge University):
The unsteadiness which is introduced by the relative motion of blades and vanes (right) is averaged-(smeared)-out (left).
These simplifications are necessary in order to have fast optimisation and development of the turbomachine and are later corrected and checked by experiments or high fidelity simulations if necessary.
Dean, R.C., "On the Necessity of Unsteady Flow in Fluid Machines", ASME J. Basic Eng. March 1959, pp 24-28