Aviation Stack Exchange is a question and answer site for aircraft pilots, mechanics, and enthusiasts. It only takes a minute to sign up.
Sign up to join this community
The ideal propeller has one blade. For obvious reasons, that would be impractical. However, adding more blades causes each blade to operate in each others slipstream/downwash.
How to quantitatively estimate this interference?In other words, what is the real angle of attack and airflow speed seen by each blade
For a simple 2D approximation, you can do it in a similar way to how you would analyze the effect of wake on a tailplane or any other body, i.e. by solving the flow field around both at the same time. Since the movement of the blades is cyclical and the flow is subsonic, each blade affects both the ones ahead and the ones behind it. Therefore, you would would need to solve for an infinite number of blades simultaneously so that the one in the middle would have a realistic airflow, but since the effect of the N-th blade will tend to nill, a good approximation can be had with just a few blades on either side of the studied one.
The process would look like this:
Choose the spanwise station of the blade you are interested in and the operating point (rpm and forward velocity) you want to study. These will give you the effective angle of attack for the blade station.
Knowing the angular blade spacing around the hub, the spanwise station you selected, the rpm and the forward velocity, find out where the first airfoil will be relative to the second one in a 2D representation.
Solve the flow field for an effectively infinite number of airfoils spaced this way. In practice solving for a relatively small number of blade instances will provide a good approximation of the real flow on the central blades in the series, as the ones on the edges are only used to provide boundary conditions resembling the real world case: 
It's possible to calculate the losses for a given configuration, but it's not an easy task. Meanwhile, there is a hundred years worth of testing and experience with propellers, so if you have a classic prop design and want to know typical trends, you can find most of the results in the literature.
One of the best classic sources on this topic is NACA-TR-640 (1938!)
I'll quote one graph from it that is directly relevant to the question:
As can be seen, adding an otherwise identical blade causes a power loss of maximum 6-8% per blade (in static conditions, at zero speed), and very little (1-4%) in normal flight conditions.
Comparison of the respective propulsive efficiency curves (Fig.35 in the work) is even more interesting: the difference is even smaller, and in some regions (when the blades stall) the multi-blade props can be more efficient than two-blade ones.
In practice, when you need to absorb extra power from the engine and are constrained by the diameter, you have to increase solidity of the prop. You can do it either by adding blades or by increasing the blade chord. The paper demonstrates that adding blades is better in terms of efficiency. Thus the premise "The ideal propeller has one blade" is not necessarily true, given the realistic constraints.
Power loss due to swirl is up to 10% of the applied power, depending on the applied thrust. From Prouty, Helicopter Performance, Stability and Control.
