Idea of "SMA SunMill"

[Khwartz]

Hello Guys!

I was following the experiments and ideas of Attila Blade with his teeter and sun flower "self-running" on heat energy at very low temperature difference.

https://youtu.be/8BF0UinTAxw

https://youtu.be/4ukwkiUOEs8?list=PLukR_iBSyGXRJNo3KgZIQqJrs-cjWxQ8G


Himself, find his inspiration in a "self-running" moisture wheel invented by one of his friend.

https://youtu.be/GcIeMWNlIGI?list=PLukR_iBSyGXRkd4KOSXuWMJL_fHvr3Y7A


I made some raw calculations and I get an order of magnitude of 10^-5 W for his teeter, for a 10 cm² around area element.

I checked what we could have if implemented of a larger surface but even with 10 m², we would be still in the oom of 1 W.

Indeed, the change of shape are very slow, the mass involved is very little, and same for the length of the displacement.

Then came the idea of using Shape-Memory Alloy to do the same calculations, and what a difference:

- Speed of change from 10 to 100 faster,

- Mass involved from 10 to 100 heavier,

- Length of displacement for 10 to 100 too.

Thus, if we add these windows of improvement, we should expect between 1 000 to 1 000 000 more power for the same size (from 1 kW to 1MW).

But of course, if sun-powered, the power will not exceed the power we will focus on the mill, and for 10 m², if no solar concentrator, the maximum would be in the best conditions as day/night average value 200 W/m², thus 2 kW for 10 m². With wood of else, maybe more.

I do not have Nitinol or other SMA but I wanted to suggest my idea to those who would have some, and if I have messed up in my calculations, don't hesitate to correct me


Here some suggestions:

1. If we use the cylinder sheet technique, with blades going in and out of the cylinder function of the heating, it could works much better the other way! ^_^ I mean when the sun heat, the blades go back in alignment with the average surface of he cylinder. Indeed, we could change the shape of the blade at cold temperature to pull it out and when heated, they just come back at their original location.


2. The width, longer and thickness of the blades should be adjusted to optimise their reactivity to the heat and cold, the simplest system wouldn't have even a liquid side cooling and just cooling by air.


3.  I may be done with just stems for example just plugged in a cylinder wood, for cheap realisations 


4. A solar concentrator, by lenses of reflectors, may be used to increase the power catching a larger area of sun rays reception to concentrate it in the exposed face of the mill.


5. A cover may be use too to only expose a precise width of the perimeter.


6. A more economical design could be to use indeed blades or stems, rods, wires, but to place masses of much cheaper materials at the extremity, to increase the ratio power available versus average price of materials used.


7. The direction way of the blades, rods, wires, may be significant to use the "kick" of the rotating blades, their momentum while they go back in place or when they take their radial direction.


8. A deflector may be used to guide the sun rays so they only come as per an ideal angle compare to the vertical axis of the mill, so that the changes of shapes of each of both configurations, only occur on one side for each. Let's say sun rays coming right, then blade would be only retracted (along the perimeter) on this side and on left side, only in radius direction ("perpendicular" to the perimeter).

It may be even need to expose the bottom surface of he cylinder so that when the cylinder turns, the blade will reach its austenite start and austenite finish temperatures (As and Af) near the bottom of the cylinder.

At the top, the symmetrical and opposite phenomenon would take place with martensite start en finish temperatures (Ms and Mf). (will try to load a new schematic for this configuration)

[Khwartz]

(y, y') is vertical axis

As, austenite start temperature

Ms, martensite start temperature

Green spots symbolise additional masses in cheaper but dense material.

The shift angle "alpha" is to help the rotation.


Some calculation for this example:


1. change of the distance of the centre of gravity of each halve of disc (maybe raw estimate because considered as homogeneous while it is far to be while mass is very in perimeter) :

I have chosen 12 arrays of blades but could be other number.

For here, the length of each blade is 1/12 of the perimeter of course.

Thus, the ratio between the radius difference and the initial radius before extension is:

"Delta r" / r = (perimeter / 12) / r =  ((pi * 2 * r) / 12) / r = (2 * pi) / 12 = pi / 6.

Formula distance of a centre of gravity in a halve of disc:

d = (4 * r) / (3 * pi).

Thus, the ratio of the two distances "d" and "d' " of the two centres of gravity is:

d' / d = ((4 * r') / (3 * pi)) / ((4 * r) / (3 * pi)) = r' / r,

thus,  the ratio of the two distances "d" and "d' " of the two centres of gravity is in the very same proportion than of the difference of radius.


2. Ratio of the torque of each halve:

Let's say

- each halve is of 1 kg.

- "d" is 1 m.

- gravity acceleration is 10 ms^-2.

Torque in "d" is:

1 kg * 10 ms^-2 * 1 m = 10 Nm.

Torque in "d' " is:

1 kg * 10 ms^-2 * (1 m / (pi / 6)) = ((10 * 6) / pi) Nm =~ 19 Nm.

Thus the ratio is:

19 Nm / 10 Nm = 1.9.


PS: I am not at all familiar with these calculations but I think they should demonstrate a potentially in the use of SMA for a SunMill or even any other mill powered by any other source of heating and having its actuators SMA blades or rods, etc.

[Khwartz]

Note / classification:

I would qualify this kind of wheel a partial gravity unbalanced wheel by mean of SMA as actuators activated by average temperature difference; heating could be sun or else, and cooling natural or forced like with water or else.

What do you think? and would you have a try?

[Khwartz]

Set of NiTi wires orderred.

[Khwartz]

Little adding of the initial post (in red), to make it clearer:

"
I made some raw calculations and I get an order of magnitude of 10^-5 W for his teeter, for a 10 cm² around area element.

I checked what we could have if implemented of a larger surface but even with 10 m², we would be still in the oom of 1 W.

Indeed, the change of shape are very slow, the mass involved is very little, and same for the length of the displacement.

Then came the idea of using Shape-Memory Alloy to do the same calculations, and what a difference:

- Speed of change from 10 to 100 faster,

- Mass involved from 10 to 100 heavier,

- Length of displacement for 10 to 100 too.

Thus, if we add these windows of improvement, we should expect between 1 000 to 1 000 000 more power for the same size of 10 m², corresponding to a cylinder of 10 m long and around 2 m of diameter (from 1 kW to 1MW).
"