Savonius Rotor Gap Ratio

[FranklinsAce]

I've been doing research over the past several months and am finding conflicting information regarding the best gap ratio when positioning the Savonius impeller.  I wanted to know if any of you have some real world advice regarding the gap ratio or links to somewhere it has been discussed previously?

Below is a brief summary of my key findings:

The article, "An experimental study on improvement of Savonius rotor performance" (S111001681200049X) -> implies that no gap provides the most power but indicates elsewhere that a ratio of 0.15 is best.

In one of my books, "Wind and Windspinners", Michael Hackleman has the following illustration which shows a gap of almost 0.35.


Now to be clear, I'm not planning on using oil drums.  My next prototype iteration is using some 5 gallon buckets cut in half

Thanks in advance for all the great advice I receive here.  Someday I hope to know enough to answer some questions as well as ask them.

[Adriaan Kragten]

In my public report KD 599: Measurements of Savonious rotors available on the Internet" from 2009, I give references to eight reports about Savonious rotors. The summary of the report: Increase in the Savonious Rotors efficiency via a parametric ----" given at chapter 2.6 of KD 599 gives an answer to your question. However, the measured Cq values can be doubted because of tunnel blockage (see chapter 2.8 ). KD 599 can be copied for free from my website: www.kdwindturbines.nl at the menu KD-reports.

[SparWeb]

I can't say much to the size of the gap between the buckets, but I'll point out that the diameter of the central shaft blocks a large part of the gap. 
So it seems more fair to subtract the shaft from the spacing between buckets, before calculating the gap size ratio.

[FranklinsAce]

Thanks @Adriaan.  I read through the report, thank you for all that research and putting it into such a usable format.

@SparWeb - yes, you are correct about subtracting the shaft diameter.  In my case that is 1/4" for a solid shaft so that won't impact the calculations too much.

I think based on what I've read and the advice here is I need to run a small set of experiments and see what gap ratio works best for me.  It will be several weeks but I'll report my results here 

[Adriaan Kragten]

If you want to do some tests, may be you can test the following idea.

I big disadvantage of a so called 1-phase Savonious rotor is that the starting torque fluctuates very much depending on the position of the buckets with respect to the wind direction. This fluctuation is normally flattened by using a 2-phase Savonious rotor which is a combination of two Savonious rotors which make an angle of 90° with respect to each other. However, this makes the construction more complicated as you need three in stead of two horizontal disks.

It is proven that a Savonious rotor with three buckets is much worse than one with two buckets because there is no fluent flow pattern through the rotor even if the buckets don't touch each other in the center. But may be that a Savonious rotor with four buckets works better. If every bucket is just half the bucket of a Savonious rotor with two buckets (so a bucket covers 90° in stead of 180°) you don't need more material and there is a rather big free space in the center for the internal flow. The torque fluctuation for four buckets will certainly be much lower than for two buckets.