CFD Online Discussion Forums - CFD analysis on wind turbine rotor

Hi everybody!
Now get this straight and at once: I am not mechanical engineer nor Physicist, so I do not know jargon, and I know only that CFD is technique for simulation of turbines (and whole WPS farms), since I am Systems Engineer and usually we get formulas prepared by someone else and are given examples of results to be able to check if program we made is working correctly or not.
Now I am in an awkward position, since I invented new kind of turbine for extracting the kinetic energy of wind, which is so different from all others that some formulas cannot apply, and I do not know exactly which ones I can use.
My turbine is different from others in one main aspect, and this is direction of wind flow. In all other turbines that I have studied, wind continue to move on same vector, usually slowed down after spending some of its kinetic energy on the blades/paddles or wings. in case of my turbine air exit on side, at 90 deg. angle to original wind flow. That mean that in original flow direction its speed is zero, and for such case Betz formula is not working/is giving wrong results.
Now before You all start telling me that I am wrong, I shall remind You why Betz Limit actually exist. Usually wind is slowed down on blades and such slowed down air becomes obstacle to next batch of wind that would pass trough has to spend good portion of its own energy to push this slower air downwind. Since in my turbine such air that passed turbine has exited on the side, then wind that has not entered the turbine but is flowing parallel to it is spending its energy to push such air downwind.
That means, in regard to wind that has entered turbine it would not count as unproductive spending of energy, so next batch is entering just as first.
Even if turbine would have 100% efficiency in extracting kinetic energy from wind and air on exit from turbine would have zero speed, it would present no problem since exit rotates and leaving zero speed air at one place, it moves and have new position to leave more zero speed air in path of original wind that bypass turbine, right?
But, my turbine is little more complicated than that.....
It is not slowing down air speed inside it, it is making it flow faster!
First part of turbine (cylindrical in shape) after entrance is funnel that act as air speeder like Venturi or Konfuzor, and with slopes at 30 degrees or less, drag should be non consequential, at least according to what I have read.
Wind is entering parallel to axle of rotation, and after passing the funnel has increased speed and reducted volume (therefore total energy remains the same) and from there it enters set of symetricaly positioned air chambers.
Simples turbine has just two air chambers, so turbine looks like an Savonius from top wiew, except in Savonius air enter from one side and other side causes drag, going againest the wind, where my turbine is round and rotates paralel to wind flow so there is no drag.
Those air chambers tapper in radial direction and have exit on their narrowest part which act as nozzle, expelling air almost tangential to turbine surface.
Effect is like it would be possible to channel all wind trough blades of standard turbine, to be left out at very tips in counter clockwise direction.
Air chambers also compress and speed up air which gain density as well, and in addition to this, I managed to harness also Centrifugal force which also speed up air in radial direction.
Air expelled trough nozzles produce reaction force and this turns turbine, so what is loss at other turbines (wind that pass trough turbine without contact with blades) is gain in my turbine.
You guess by now that such trurbine would present challenge even to CFD expert, which I am not, so I delegated CFD analyse to engineer that is making the prototypes since year 1996.
Unfortunately, he really has problem with getting results (using CFDdesign 3D) as he reported first that 90cm diameter turbine is making 2000 RPM at 200N on 1m stick as load, and thus producing 44 KW of power, with 5Km/hr or 1.39m/sec wind speed. Since I made small program that calculate energy input from turbine diameter and wind speed, I calculated that only 2.2 W has entered in form of kinetic energy so results were 20 000 times greater than possible. I concluded that mistake must be in decimal point of diameter, because in my country it is written as coma. If program ignored coma then diameter would be 100 times greater, and 100 squared is 10 000, right?
But he claims that diameter is in milimeters and that he input it correctly.
While I can explain energy gain of 200% if error was in size of turbine diameter, I know it is not possible to get that much greater result, it would be not even a miracle but totally impossible. Then he told me that somehow it was working with noncompresible flow, and that it has caused such results.
But now turbine act like it is using noncompressible flow as air do not want to enter it and at best in nozzles there is original wind speed achieved........
I know this is so because my small hand made model almost do not turn if put under the water flow, where it is rotating very fast on the wind of moderate speed.
Now with new parameter turbine turns 8-12.5 turns per minute, which is very slow for wind of 10m/sec. (36Km/hr), where in previous simulation it was turning as expected (for 90m diameter, that is).
So I am asking for help from all of you CFD experts, can You help me and my engineer coleague?
I would post some pictures from both simulations, but it is not possible here as page expect them to be at some WEB address and they are on my HD.....

Regards from Croatia, the Homeland of one of greatest inventors from 19th and 20th Centuries, Engineer Nikola Tesla!