Report KD 732 about the VIRYA-14 available

[Adriaan Kragten]

Report KD 732 can be copied for free from my website: www.kdwindturbines.nl at the menu KD-reports. The title of this report is: "Calculations executed for the 3-bladed rotor of the VIRYA-14 windmill (lambda design = 6) with the pendulum safety system with a torsion spring connected to the generator type PMG900-I-30KW-100R for grid connection".

The chosen generator is the largest generator of Hefei Top Grand with an inner rotor. The VIRYA-14 is therefore the wind turbine with the largest rotor diameter (D = 14 m) which can be connected direct drive to a PM-generator which can be bought in the market. As development of a big direct drive PM-generator is rather complex and very expensive, this is a large advantage. The maximum power of this wind turbine is about 21 KW at V = 10 m/s. The wind turbine can be used to power the heat pumps of about twenty well isolated houses in a moderate wind regime. KD 732 is only a design report and drawings of the auxiliary shaft, the head and the tower have still to be made. Only professional companies with enough skills and manufacturing capacity should think about the development of such a wind turbine.

[SparWeb]

That would be an impressive machine if it ever became reality.
Have you ever investigated the construction of blades that size (14 meter diameter) following your design?

[Adriaan Kragten]

That would be an impressive machine if it ever became reality.
Have you ever investigated the construction of blades that size (14 meter diameter) following your design?

The blades are almost 7 m long but Lagerwey has made similar almost 9 m long blades for its former Lagerwey 18/80 and that was no problem. This kind of blades are rather simple to manufacture because they have a constant chord, a constant airfoil and no twist. The Gö 711 airfoil is used, which is flat over 97.5 % of the chord and the airfoil is made in the whole blade length. So only the back side of the airfoil has to be machined. A blade is made from massive wood but the beam is made by glueing strips together. The airfoil can be made in only some strokes if you have a rotating cutter with the shape of the airfoil in the cutting tool. Wind tunnel tests have proven that rotors with constant chord blades can have a high maximum Cp if the correct airfoil is used which has a low Cd/Cl ratio for a large alpha range.

[SparWeb]

Wind tunnel tests have proven that rotors with constant chord blades can have a high maximum Cp if the correct airfoil is used which has a low Cd/Cl ratio for a large alpha range.

Indeed, most operation of a WT should be in the range of angle of attack below the peak Cl/Cd point.
This brings to mind the fact an airfoil that achieves very high CL is counterproductive in a WT, and it is best to choose low drag instead.

[SparWeb]

I am not sure if I can find this report again, but if you would like to see, I can look.
In the 1980's the NREL instrumented several sets of WT blades with pressure taps and operated them in the wind, not just a tunnel.  They found that straight blades did not suffer from as large zones of stalled flow as they expected.  It turs out that the induced flow compensates slightly.

[Adriaan Kragten]

Wind tunnel tests have proven that rotors with constant chord blades can have a high maximum Cp if the correct airfoil is used which has a low Cd/Cl ratio for a large alpha range.

Indeed, most operation of a WT should be in the range of angle of attack below the peak Cl/Cd point.
This brings to mind the fact an airfoil that achieves very high CL is counterproductive in a WT, and it is best to choose low drag instead.

Tapered wind turbine blades are normally designed for the lift coefficient for which the Cd/Cl ratio is minimal. This so called optimum lift coefficient is lying a lot lower than the maximum lift coefficient. The optimum lift coefficient is found for the point of contact of a line through the origin and the Cl/Cd curve. If we take the Gö 711 airfoil (see my public report KD 285), it can be seen in figure 4 that the optimum lift coefficient is about 0.95 for the point of contact. The Cd/Cl ratio for this point is about 0.015 which is very low for an airfoil with a flat lower side. I think that this is because the airfoil was measured for a rather high Reynolds value of 4 * 10^5. In figure 3 it can be seen that Cl = 0.95 corresponds to alpa = 3.3°. In figure 4 it can be seen that the maximum lift coefficient is about 1.5 for alpha = 12.5°.

If we draw a line in figure 4 though the origin and the point Cd = 0.025 and Cl =1, all points on the Cl/Cd curve lying left from this line have a Cd/Cl ratio smaller than 0.025. This corresponds to values of Cl larger than 0.57 and smaller than 1.42. In figure 3 it can be seen that this Cl range corresponds to values of alpha larger -1° and smaller than 10°. So this is a very large alpha range for which the Cd/Cl ratio is small enough to expect a high maximum Cp.

Constant chord blades are designed for a low lift coefficient at the blade tip, for the optimum lift coefficient about half way the blade and for the maximum lift coefficient near the blade root. So the average Cd/Cl ratio for a constant chord blade is somewhat higher than for a tapered blade and this results in a somewhat lower maximum Cp. But manufacture of a constant chord blade is much easier especially if the blade is made from massive wood and a much smaller wooden beam is needed.

Another advantage of a low lift coefficient at the blade tip is that the pressure difference in between the front side and the back side of the airfoil is smaller and this may result is lower noice production if a constant chord blade is compared with a tapered blade with the same length and the same design tip speed ratio.

[Adriaan Kragten]

I have separated the Cl-alpha and the Cl-Cd curves from report KD 285 and put them in a file which I have added as an attachment. The line for Cd/Cl = 0.025 was put in the Cl-Cd curve.

The Cl-alpha and Cl-Cd curves can also be used to find the Cd/Cl ratio for a certain value of Cl. Assume Cl = 1.2. In figure 3 it can be seen that this is the case for alpha = 6.6°. Next one goes to the Cl-Cd curve of figure 4 and draws a straight line through the origin and the point on the Cl-Cd curve for Cl = 1.2. All points lying on this line have the same Cd/Cl ratio. This line is intersecting somewhere with the horizontal line for Cl = 1. It can be read that Cd = 0.02 for this point. This value of Cd is the Cd/Cl ratio because 0.02 / 1 = 0.02.

 KD 285 (only Cl-alpha and Cl-Cd curves).pdf (27.04 kB - downloaded 38 times.)

[SparWeb]

It's probably conservative to use Re = 4E5.  Given the rotor diameter is 14 meters, it looks like you could use CL/CD plots for Re over 5 million.
There is a marked difference in the CL/CD ratio, judging by the other airfoils I've worked with in the past (that are similar to the GOE 711).
The discontinuity in the lift curve slope at alpha = -6 degrees goes away completely.
Otherwise, you may be satisfied with under-predicting the performance.  Something to compensate for all the others who exaggerate!!

[Adriaan Kragten]

It's probably conservative to use Re = 4E5.  Given the rotor diameter is 14 meters, it looks like you could use CL/CD plots for Re over 5 million.
There is a marked difference in the CL/CD ratio, judging by the other airfoils I've worked with in the past (that are similar to the GOE 711).
The discontinuity in the lift curve slope at alpha = -6 degrees goes away completely.
Otherwise, you may be satisfied with under-predicting the performance.  Something to compensate for all the others who exaggerate!!

As the Gö 711 airfoil has only been measured for the rather high Reynolds value Re = 4 * 10^5, it should only be used for rather big rotors with a rather high design tip speed ratio. But in the design calculations of the VIRYA-14 rotor it can be seen that for a wind speed of 5 m/s, the Reynolds value is higher than Re = 4 * 10^5 for all stations except the most inner station F (see table 1). The Cd/Cl ratio is given for every station in the last column of table 1 and the average value is about 0.02. If all losses are taken into acccount, this results in a maximum Cp of about 0.44 which is very good for a rotor with a constant chord.

I found the negative bulge in the Cl-alfa curve for negative values of alpha also strange but the measuring points are copied correctly from the original source given in report R443D.