Fresnel-type solar collector [STLs included]

I was inspired by Erik's post in this thread: [forums.reprap.org] to design a modular, extensible fresnel-type solar energy collector. Before I get into the specifics, here's a render of how the final product is supposed to look.



Note that the array I've illustrated here is much larger than what I would expect any user to build from the start. This particular setup takes 4kg of plastic and is about a meter long per side. However, because of the completely modular nature of the project (each triangular block in the picture is a separate piece, although they are all connected rigidly), an array can be expanded or broken down at will.

I understand that a first-generation machine will almost certainly require some tweaking of this design before it will print successfully. At the very least, this requires a build area of 280mm and the ability to either use support material (see Adrian's prototypes) or span bridges (see Skeinforge). Until anyone actually tried to print this, it is still just a concept. If you do attempt any of the components let me know how it goes - I'm willing to change anything that's necessary to make this work better.

Each mirror is mounted on a ball joint, so each one must be aligned to face the focus point individually (each one is mounted on a ball-and-socket joint), a task that is admittedly extremely tedious. Can anyone think of a better way to solve this problem?

A full 3D model of the concept can be viewed and manipulated at [freewheel.autodesk.com]

The STLs for all of the components are attached. In addition, anyone who has a compatible CAD program and is interested can find a STEP file at [hyperfileshare.com]. If you have the Autodesk DWF viewer (it's a free download, although unfortunately Windows only) you can view the model in full resolution and full color, as well as take measurements from it if you download the "SkyDial5.22.08.dwf" attachment.

For the record, this model was designed in Autodesk Inventor. After exporting the STLs, I opened them in the open-source MeshLab program to "unify duplicated vertices," which greatly reduces file size while preserving file quality.

----------
I've now finished my project, and am looking for something to do with my spare time while I wait for summer to come when I will start work on my RepRap. Anyone have a model that they're dying to test-print that they would like drawn up at no extra charge? I can't guarantee that I'll work on it, but I'm open to suggestions.

Great idea. Also, nice looking design.

The ball and socket concept is good for ease of alignment, but it's also a weak spot. The forces from wind will make the mirrors act like wind-screens that are also a lever jerking on the ball-and-socket assembly. Making the RepRap do all the measurement and alignment for you would be less time-consuming.

I really like the modular nature of the setup.

From the picture, some things are still unclear to me though. I wonder how you would track the sunlight's direction so that the solar rays are focussed onto the tracker during a large part of the day. Which parts will you move, which will rotate or be static? You could move the collector. There is a mechanical advantage to move this one thing instead of hundreds of 'little' mirrors (even if you scale it up).

I made small sketch. I hope you can read it, it's blurry and my handwriting and sketching skills are severely limited!


I'll talk about this picture rotated 90' clockwise.

The square in the middle (in perspective) is the solar array. It can be designed as you have. It should probably not be square at all... but it's just symbolic. There's a tracker, a collector on a tower. This tower is on a rails that goes about 180 degrees around on a rail. The rails should be to the back of the square, so the collector will collect the reflection. The height of the sun determines the height of the collector. This curve varies per season (graph on the bottom). The place on the rail is determined by the time of the day. The orientation of the whole thing is to the south for the northern hemisphere, or north of you're on the southern hemisphere.

If you want to RepRap anything, I think it's the connecting bits that fasten steel rods. Or, if we can RepRap even stronger sparse structures of composite material such as foam filled girder structures. But tell me if you have other ideas...

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

Zowie! I'm glad you're designing solar concentrators, we'll need that to power all our repraps!

Since you so generously offered to donate your spare time to additional projects, I was thinking of a low efficiency version of your tower concentrator. Basically the 1d version of your 2d concentrator.

For the concentrator, have long mirrors running East West, mounted on a hinge or something, that can be turned between north and south by the arduino. Have a fixed, elevated, long absorber running East West, a bit longer than the mirrors, instead of the original point absorber. In the first version the mirrors would be flat, and there would only be three of em, so they could be turned by the standard arduino electronics and gcode. The driver software would be a timer that would send a bunch of gcodes every several minutes. Later versions would turn the mirrors to face down when hail, snow and/or darkness is sensed.

Because the first version would have only three fixed mirrors, the absorber would only get three suns, so the efficiency would be low, but it would be the easiest solar concentrator to build. We can almost certainly build hinges with Darwin. In the version after, you could upgrade the flat mirrors to shallow parabolas, and you could probably get fifteen suns, even with only three motors. Later still versions could keep increasing in sophistication, until you're building your original tower collector.

Edited 1 time(s). Last edit at 06/03/2008 04:03PM by Enrique.

My idea was that the entire contraption would stay stationary relative to itself, but be rotated along two axis to always be pointed directly at the sun, thus eliminating the need to adjust the individual mirrors more than once. If necessary, a soldering iron or other hot object could be used to "fuse" the mirrors in one orientation once they were aligned correctly.

I didn't know the entire contraption would be rotated, I thought you were making a solar concentrating tower.

How about designing a transpiration hot air solar collector?

It is basically a box with many small holes on the south side and an outlet pipe on top. It is wide in the East West direction, thin in the North South direction and tall. It could definitely be fabricated, the only modification for fabrication is that the holes would be teardrop shaped and the top would slope up at about 45 degrees, instead of being flat.

After the collector works, then you could boost the concentration, until just before the plastic melts

I don't understand - how does the collector generate energy?

After you asked, I googled some more and found that the common name is the unglazed perforated plate collector:
[www.canren.gc.ca]

The collector is a dark perforated plate. When facing the sun it gets hot, and air drawn in through the holes is heated by the plate. There is no glazing, the convection losses are instead reduced because the air is continually pulled in.

Perforated collectors are available in two forms. One is a cladding for a building, the other is a stand alone rectangular collector. I suggest making the stand alone version.

A simple test version would be a monolithic box, say ten millimeters deep, forty millimeters wide and ninety millimeters tall with a single teardrop shaped hole five millimeters wide in the center. The top would slope upward at an angle of at least forty five degrees, becoming an outlet pipe about ten millimeters wide at the top.

To get electrical power from this, the hot air could be circulated around the hot side of a heat engine. Heat engines that could run off this medium grade heat include stirling engines, rankine cycle engines and thermocouples. We can make thermocouples for sure, but they have low efficiency. Later we could make the more efficient stirling or rankine engines.

wikipedia has an article under solar mirror [en.wikipedia.org]

there is also a good solar mirror explanation at [www.eskimo.com]

To keep the mirrors in place why not add a hole in the ball joint so you can screw in some 8mm threaded rod to keep it in place?

Or maybe a hole for a self tapping screw to lock it when happy.

Must admit soldering iron and weld is cheap and easy though.

For a solar collector use PC cells but mount them on a heat sink (Fluid cooled could be useful)

Cheers

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

How much would a stirling/heat engine weight?

Suppose we get 1/10 of 1000W = 100W out of this thingy on a sunny day, directed towards the sun with 1 m^2 of area. We would like to have a 100 W stirling engine.

Wonder if one could make an engine out of plastics without being bitten by the Carnot factor :-/

Hi Mimarob,

I think that would be very hard to print because you need at least two conductive surfaces, one for the cold and the other for the hot cylinder of the engine. Thermoplasts are not very conductive. Metals would be better. Also, the regenerator (to get decent efficiency) has to be of porous material with big specific heat but hardly any gas flow resistance. Any resistance anywhere in the system will reduce the efficiency of the engine.

High pressure would improve the engine, so it would be best if the entire engine is strong enough to be able to be highly pressurized. If you'd use a lost wax casting technique with aluminum you'd probably be able to use the RepRap indirectly to print molds for a heat engine.

> Wonder if one could make an engine out of plastics without being bitten by the Carnot factor :-/

You will probably be 'bitten' by the Carnot factor since it's the highest efficiency that is theoretically attainable. I don't think that it's easy to break the laws of thermodynamics. (Unless you can unscramble eggs, of course!)

I do like the combination of concentrated solar and Stirling engines, though! It would be nice to leave out all this semiconductor stuff and do it the old-fashioned way

Erik

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

What about a high temperature thermopile.

With an evacuated glass bulb surounding the hot end the temperature at the focal point should be very high.

Water Cool the cold end with a ground source heatsink (water drawn through loops of pipe buried in the ground.

And have electricity straight out.

I guess if the thermopile was constructed for moderate voltages, A switch mode PSU could convert what came out to something consistent and useable.

Proof of concept could be achieved using those White LED step-up up driver IC's that are cheap/cheerful.

Thoughts for what they are worth.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

As far as I know, thermocouples (or piles of them) require a huge surface area to really produce a decent amount of energy. I don't know how cheaply you could make them, but the ones used for sensors often need specialized amplifier chips (such as the thermocouple board of the RepRap does).

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

Erik - Here is your favorite project and it is just going commercial.

[www.stirlingenergy.com]

Look at the SunCatcher technology item.

I have been following this company for about 10 years. They have the right idea
but its does have its limits.

I am traveling down a different path to achieve the same results.

---

Bob Teeter
"What Box?"

Erik

I agree, I thought very much the same until I read up a little more on thermocouples.

The voltage output is proportional to the temperature difference between the hot and cold junctions.

For each additional pair you add the same voltage on again if the difference between the junctions is maintained (By pair here we are talking 1 each hot junction and cold junction).

In a measuring device you have generaly 1 pair. The Seebeck effect yeilds a voltage on the order of millivolts per Deg C.

The interesting thing is that whilst voltage is directly proportional to temperature (pretty much linearly so, this is why they make good temp sensors) my readings suggest that the current is proportional to how much heat energy the pile is shifting from the hot end to the cold end and can be substantial.

The key factors appear to be

1. The thickness of the wires used in the thermopile construction.
2. The number of junctions in the Pile.
3. The temperature difference maintained between hot and cold junctions.

To quote Wikipedia

"Thermocouples can be connected in series with each other to form a thermopile, where all the hot junctions are exposed to the higher temperature and all the cold junctions to a lower temperature. The voltages of the individual thermocouples add up, allowing for a larger voltage and increased power output, thus increasing the sensitivity of the instrumentation. With the radioactive decay of transuranic elements providing a heat source this arrangement has been used to power spacecraft on missions too far from the Sun to utilize solar power."

[en.wikipedia.org]

I also recently read a fun book "SOE the scientific secrets" about gadgets made by for SOE operations behind enemy lines in the second world war. (ISBN 0-7509-4005-0)

In it there is a photograph of a "Thermal Battery Charger" for charging up the batteries on valve radio transmitters. A quick study of the picture suggests that it is a Thermopile device.

Whilst Stirling Engines look great fun too (I have read a lot on these as well) Thermopiles are a lot simpler for the home constructor/experimenter to kick off with and have no moving parts to wear out.

Given a lathe, time and materials I wold love to construct Stirling Engines there are a lot of engineering challenges to be overcome.

The best commercial unit I have seen so far is the hermetically sealed unit for combined Heat & Power domestic use.

[www.whispergen.com]

As these are fairly new being able to scavenge the stirling units out of them though could be a while and there is not much you could do with it if it was broken (Hermeticaly sealed).

They may sell you them as spares/replacments though but expect a hefty price tag (The company is part of the Worcester Bosch Group).

Cheers

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

... here - [forums.reprap.org] - are some thoughts about fabbing a simple solar thermogenerator with many Seebeck-elements and a temperature-gradient of 500

Viktor

Thanks for that I hadn't seen it.

It looks like we are thinking along similar lines.

I was thinking of a short Batton for the pickup.

I have drawn a quick piccy to illustrate as piccys often work better than words.

Cheers

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

Hi Andy,

... here my idea:


Viktor

Viktor

Generation as I understand it is a function of heat moved from hot end to cold end and is greater for a greater temperature difference.

You may need some form of heat sinking at the rear of the dishes otherwise the cold end will heat up and reduce the temperature difference between the Hot and Cold Junctions.

I think this is where the space PSU worked well as the temperatures in the shade in space are savagely cold.

Other than that sure an array of smaller units should work just like one larger collector and an array of mirrors.

Whether you split the Pile into 5 separate piles or run it as one bundle the performance will be limited only by how much heating/cooling you can bring to bear and how well it is coupled into/focused onto the collector.

It would be fun to build some of each and compare.

Questions re materials ????

Constantan and Iron are popular. What is the performance difference between that pairing and Constantan and copper ??.

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

Using solar power to heat your water is one of the most practical and cost effective ways to harness energy from the sun. Solar thermal systems absorb the sun's heat and use it to heat your residence's own hot water needs.
Here is a link that might be useful: solar heat information

I am currently doing career changes to go into teaching and a heliostat array could be a really good project to do at secondary school level for systems & control.

The great thing is each pupil can make one heliostat each to their own design and they can be tested off and shown functioning as a class project.

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.