"Ideal" propeller

[scorman]

Samoa,

Glad you responded because I was looking for an email address in your posts and wanted to ask you a few OT questions.

First let me get back to my thread:

"their baseline rotor is the contemporary Jacobs not the original Jacobs"

Yes, by "Jacobs" I was referring to the contemporary design as contrasted to any of the proposed newly Propid configured designs.

"a two-dimensional projection of span and chord onto a surface. It does not include airfoil thickness."

The term planform as you correctly stated is the "overhead" view of the blade ignoring both the profile and the twist.

As I see the Sandia choices, they split into CP which is the Jacobs planform and the NP which is the Propid free design which resulted in a 18.11" root and 6.2" tip as shown in Table 3.1. for a rotor of 14.5ft with root at 2+ ft station

According to Fig 2.20 comparing the choosen LPT3 taper, the Jacobs is about that same tip cord width of 6"+ but w/o taper or twist.

That said, once you choose the cord widths at each station, and declare the profile used, you have therefore defined the thickness at each station and the corresponding Cl and L/D for a given RE#.

"I believe a lot more goes into an airfoil selection than just Cl and L/D"

yes, but they are still claiming a 10%+ improvement which I IMHO must be the pitch,

since the Cl and L/D don't compute as being better ...still can't figure why they didn't show the model using twist and taper for the USNPS4 profile as a comparison.

Now back to your stuff:

I saw the pic of your three blades:

http://www.otherpower.com/images/scimages/5176/Blades1B.jpg

I would like to get info on the history of your project and where it stands today.

I searched your log and cannot find many details except:

"Yes, that rotor was tested some 20 odd years ago when it was made. Cp=0.42"

AND, how did you determine Cp=0.42??

These are certainly not common shapes

I noted that you chose a TSR of 4.5 ...how/why did you choose such profiles/parameters?

what are the actual dimensions and how did each perform?

did you correlate real performance to modern JavaFoil type modeling that didn't exist 20 years ago?

I really question how much extra you can squeeze out of a fancy design vs NACA4415, etc, especially for low wind speed conditions ..that Sandia report didn't convince me that they had something really special

Stew Corman from sunny Endicott

scorman@stny.rr.com

[scorman]

Murlin,

"The 2004 report was surley designed for a full blown variable pitched rotor."

do you have a link ..been using the 2002 as ref in other parts of this thread, but couldn't google the 2004 you ref

What is the status of your project? any links?

BTW, I modeled the 2002 proform (scaled down as per paradigmdesign thread)in Excel w/ AOA, RE# etc if you want it

thx,

Stew Corman from sunny Endicott

scorman@stny.rr.ocm

[scorman]

Samoa,

As a followup, I used JavaFoil to give me polar plots for a comparison of Cl and L/D.

I choose the NACA4425 that I am using, , NACA4415 which is thinner, USPNS$ which is the Jacobs profile, and the SG6050 which the Sandia project chose.

Some are softer stall than others, some smoother transitions over AOA

three RE# were choosen : 50K, 150K and 300K

At design TSR=4.5, RE#300K is about max for 20mph WS for 4.5 inch cord at tip.

Lower RE# is for stations closer to hub in non tapered case.

Mid RE# is typical for lower WS ie 12mph.

http://www.otherpower.com/images/scimages/7526/LD_ratio.jpg

http://www.otherpower.com/images/scimages/7526/Cl_vs_AOA.jpg

I am trying to get my hands around these analyis tools to see how they give guidance.

If my presumptions are correct ...the L/D dictates the max TSR you will actually get at any WS,  since drag at the tips limits how fast those tips can fly (thinner is better), but the actual lift coeff Cl coupled with the cord width determines how much torque you deliver at that TSR (wider is better).

Two diff rotors at same rpm in same WS, the greater the torque , the better the efficiency.

So, the balancing act is deciding on first a TSR to match the generator rpm, choosing the right blade pitch for AOA so that L/D is best at the 75% station, and then choose the profile and cord width so that the tip is as thin as reasonably possible, while still having enough meat at the root to keep the thing from flying apart. The ideal blade would have the same torque at all stations along the blade at some optimum WS at which point the maximum efficiency will be obtained.

In my prototype , I choose a non-twist, tapered 1:3 so that the RE# is fairly constant,

root is thick for strength, wide for low starting and high torque, but tip is thin to get to correct TSR = 3.5. By mounting on tubular spars, I can reset the pitch angle at will.

I would appreciate comments on the above conclusions.

Stew Corman from sunny Endicott

[finnsawyer]

Why not only one integral, the one for lift?  The drag can then be determined from the resulting blade parameters.  Ideally, we would like all of the available power to go into lift.  Drag always robs some of that power.

[finnsawyer]

It would be nice if you put in the scales.  Do you mean blade angle or angle of attack.  If you really mean blade angle then the AOA must become less with higher TSRs.  For instance, at a TSR of 7 the angle between the apparent wind and the direction of rotation is 8.13 degrees.  For a blade angle of 5 degrees this means the AOA is about 3 degrees.  At a TSR of 14 the AOA would become about a negative 1.  You might redo the calculations keeping both the AOA and the blade angle equal to half of the angle between the apparent wind and the plane of rotation.  

[wdyasq]

"So, the balancing act is deciding on first a TSR to match the generator rpm, choosing the right blade pitch for AOA so that L/D is best at the 75% station, and then choose the profile and cord width so that the tip is as thin as reasonably possible, while still having enough meat at the root to keep the thing from flying apart. The ideal blade would have the same torque at all stations along the blade at some optimum WS at which point the maximum efficiency will be obtained."

I'm not sure you have reached the 'right conclusions'. And, if this is concluded, there is no reason for discussion.

However, at the low Reynolds numbers we are working with I'm not sure tip thickness/thinness is much of a concern. And, I believe it is much more important one grab the power at lower speeds than higher speeds. At higher speeds there is more power than the alternators can handle and the power of the rotor will probably be far greater than the bits of drag we 'worry' with. I THINK this is what the UIUC folks were trying to achieve. At least that is the way I remember the reports when I read them some years back.

As far as airfoils go I believe some of the ones used on the HPVs, Human Powered Vehicles, need to be at least considered. The DAE51 interested me as well as the S8030? and S8050? I'm not sure about those last two numbers as this is from memory and not refreshed over the past few years.

I also don't know if the results UIUC got were from 'rigid vs. VP Jacobs' or 'Rigid vs, Rigid Jacobs'. I may have some time to go back over those reports in the next few weeks. I do know the next 3 weeks are a bit of the female dog for me.

Ron

[finnsawyer]

The power available in an annular ring of width dr at radius R is given by:  

   dP = 2xPixRxpxV^3xdr.

This in turn, taking into account the Betz Limit, is equated to the component of lift force in the direction of the velocity of the blade at that radius times that velocity.  That is,

  FlxVrxdr = 0.59x2xPixRxpxV^3xdr.

This may be a bit confusing as it is not possible to write all of the mathematical operations using a standard computer.  V, of course is the magnitude of the velocity of the incident wind.  The lift force, which must have the units of force per unit length is actually a vector as is the velocity Vr.  So, we can get the left hand side simply by taking the dot product of the two vectors.  The derivation of this is fairly simple.  If you have a problem with this equation, let me know.

Now the velocity of the blade section at radius R can be written as Vr = WxR, where W is the angular velocity of the blade.  So, Vr has the same dependence on R as the power available from the air.  This means that the component of the lift force in the direction of rotation must remain independent of R.  This also means that the torque per unit length on the blade must become less toward the root, as the moment arm becomes shorter.  When I offered the opinion about the blade possibly becoming narrower, I was focusing on the fact that when keeping the proper amount of twist more of the lift force will be in the direction of rotation toward the root due to the twist.  This ignored the effect on lift due to the decreasing value of the apparent velocity.  The bottom line is that when keeping the same value of the AOA and the proper amount of twist the blade will probably get wider, although one can also keep the blade width constant if one allows the AOA to change.  While it may even be possible to design a blade that gets narrower toward the root over a reasonable range of R, it probably would not be very practical.  I apologize if this caused any confusion.