Extruded or Cast

Today I spent a good part of the day trying to track down a distributor of the Acrylic double-wall panels.  This was not as easy as I had hoped.  I contacted a number of the so-called distributors listed on the Deglass website, but the result was not very helpful.  Most of them did not know what I was taking about, or it was not in stock and needed to be special ordered.  This would not necessarily be a problem, except for the manufacturer's requirement of a 20 sheet minimum order.  I ended up on the manufacturers website, signed up for a user account, and logged into their technical support section, which was a question/answer forum.  In addition to leaving a rather long question that I am doubtful I will get a response, I searched the site for questions regarding "water".  One of the first answers to come up was interesting:

We do NOT recommend using ACRYLITE FF sheet for any applications containing water...The molecular weight of ACRYLITE GP sheet is about ten (10) times that of ACRYLITE FF sheet. When either material is subjected to constant water immersion and absorption, it will absorb up to about 2% water. The effect of water in the material is analogous to lowering its molecular weight. In the case of ACRYLITE FF sheet, this will cause a significant reduction in its mechanical properties and craze resistance. On the other hand, because the molecular weight of ACRYLITE GP sheet is so much higher to start with, the effect of water absorption on its mechanical and craze resistance properties is small.  In some instances, ACRYLITE FF sheet has been used successfully in applications involving continuous contact with water. These applications involve minimal mechanical loads and minimal stresses. In general, using ACRYLITE FF sheet in such applications is very risky since some stress or loading is usually present. Therefore, we recommend ACRYLITE GP sheet for ALL water containing applications.

It turns out there are two primary types of acrylic, extruded and cell cast.  Extruded is inexpensive to manufacture, but has a lower molecular weight.  Cell Cast is more expensive (substantially), but its molecular weight is much higher.  So far as I can gather, the extruded needs to be a low molecular weight so it can be pressed through the extrusion slots (or whatever they are called).  A good quick review of the types of acrylic can be found here.  

Well, two steps forward and one step back.  If I go with the extruded, which would be wonderful because it is sooooooo much cheaper.  Based on the response of the tech at Evonik Industries, I am guessing that the Acrylite FF is extruded and the Acrylite GP is cast.  If I build the panels out of extruded, I run a gamble on long-term reliability.  If I use extruded I will need to design the panels so they have very little stress, which is going to be almost impossible with the double-wall design.  To lower the stress the panels will require almost no pressure drop from inlet to outlet, and to achieve this the "flow chamber" for the particles need to be large.  Of course, I am leaving out other issues, but my conclusion is that making the panels out of double-wall extruded panels could be problematic.

So what now?  The bottom-line is that these panel need to last from 5-10 years and my goal is a solar thermal panel that is 2/3 less expensive then what is available currently.  The payback for the panels should occur in the first two years, and the lost cost barrier to entry means many more people will be able to afford green energy.  The most important property of the panels is low cost with comparable efficiencies.  If I have to use expensive plastic then that's the end-game.  I have a plan.


I have designed a particle panel out of single-sheet (extruded) acrylic and have ordered the parts.  The design will minimize internal pressure because there will only be one large flow chamber.  It should be easy to fabricate.  I anticipate that the particle-water slurry inside the flow chamber will cause constant damage to the inside of the panel.  Although the particles are very fine, I am predicting that they will "craze" the inside of the panel over time, and I'll bet this will happen with the cast type as well as the extruded.  After reading about crazing on Wikipedia, I discovered an interesting fact:

The initial energy absorption per square meter in a craze region has been found to be up to several hundred times that of the uncrazed region....Crazing occurs mostly in amorphous, brittle polymers like PS, PMMA and polycarbonate; it is typified by a whitening of the crazed region.

The first part would indicated that crazing will improve the panels ability to resist cracking.  The second part is not so good.  White is bad because it means the light is reflected away.

I still have no data on this, so I just need to build something and test it.  To simulate heavy aging, I will build another panel and use fine sandpaper (the silicon carbonate particles I have are actually used to make sandpaper) to roughen up the inside and compare the two efficiencies.  If there is a huge drop in power efficiency then I have a problem.  If the drop is minimal, then I am one step closer.






 

 

Acrylic to the Rescue

After the disappointment with the "clear" Coroplast not actually being clear, I have taken this morning to look in a few more directions and I have made what I think is a significant discovery.  Let me review the problem first.

For the particle panel concept to work I need to find a plastic that is highly transparent that will serve as the container for the particle "absorption flow chambers".  I built the first prototype out of polycarbonate, as it was available at a local distributor and its "double wall" configuration makes it easy to construct the particle flow chambers.  The problem with polycarbonate, as I learned later, is that it degrades in the presence of hot water.  Since water will be pumped through it, this is obviously not good.  I went looking for another plastic and thought I found it with Coroplast, which is actually polypropylene.  Polypropylene is highly hydrophobic and consequently highly resistance to water, so this appeared to solve the problem.  Since I am trying to build a particle panel out of off-the-self components, it helps tremendously if I can get it pre-formed into channels or tubes.

When the shipment of "clear" Coroplast arrived, it was not actually clear.  I have not bothered to measure exactly how much light is passing, but just by looking at it I would guess at least 15-20% of the visible light is reflected.  The result is that the panel looks white.  Unless I can find polypropylene in a double-wall configuration, this is another dead end.

Ideally I need a double-wall plastic sheet that is both transparent and can withstand long exposure to water.  I think I may have found it, and wouldn't you know, this stuff seems to have been around for awhile. 

I started by checking out Tap Plastics, to see what sort of plastics they had available.  I was looking for a solid sheet of acrylic that I could bend to form a channel, having not had any luck searching for "double wall" acrylic.  Acrylic is used for aquariums, so my thought it that its OK around water.  I actually just visited the Monterrey Bay Aquarium with my girl friend after a business trip and while she was marveling at the sharks swimming by, I was marveling at the weld between the two 6 inch think panels and how clear the Acrylic was despite its thickness.  I have later learned that acrylic is more transparent then glass.  I have not yet found much information about long term-exposure to hot/warm water.  I seriously hope this OK.  The price on Tap Plastics for a solid sheet was not all that great, so I searched for what I could get in bulk.  I noticed the brand Acrylite came up multiple times, so I thought I would check out their website.  I took a look at their product listing, and came across a "double skinned" configuration.  Well wouldn't you know...all this time I was searching for "double wall acrylic" and got nothing.  On the product specs, they claim "increased light transmission of 91%, compared to 72% for 16mm polycarbonate".  Well that, my friends, is exactly what I was looking for!

Looking at the product specs I was pleasantly surprised to find that the channel widths are actually much wider then the polycarbonate panels.  This will have a wonderful effect of increasing efficiency of the particle panel because it will lessen the internal resistance to flow, which will allow the water to flow faster (increases absorption efficiency), which also lets me use larger particles (cheaper) and larger wire mesh (cheaper).  They also have a "NO DRIP" coating that is applied to the outside and inside of the panels.   I don't yet know much about this coating, but it sounds nice because it lets the panels "self-clean" after a rain which would help to keep the panels clear over time.  As far as the coating on the inside, i think the particles would wear it down after while.

The panels are marketed under the brand Deglass.  A look at their product listing indicates that the "DEGLASS IMPACT" product is the best bet for the absorption channel and possibly the glazing as well. They state it is "the most optically clear multi-skinned sheet available".  Its rib spacing is 2.5'', which will substantially lower the resistance to flow.  They make a 4-wall (oops, I mean quadruple-skinned) sheet that has an IR reflective "Heatstop" coating.  This answers my question about if a transparent IR reflective coating existed.

If I glazed the panel with the same material as the absorption channel and joined them with acrylic resin, I could get exception transmittance onto the particles.  Most of the light is reflected at the air/plastic boundary.  Acrylic has an index of refraction of 1.48 and air is about 1.  Whenever two substance differ in their index of refraction, light will be reflected at their interface.  Check out Fresnels equations to learn all about this.  The inside of the absorption channel is water, not air.  Using Fresnel's equations at a normal incidence, I get 3.61% reflectance at an air/water boundary and .28% at the acrylic/water boundary.  This means that the absorption plate should actually reflect less then 3.89% of the light (accounting for internal reflections).  With a design where I join the panels with resin, I can eliminate the reflectance at the air/acrylic boundary and create a structurally sound solar panel by simply laminating panels together. I'll bet I can get the transmittance up to ~92%.  Combine this with the inherent efficiency of the particle panel absorption and we just might have the worlds cheapest and most efficient flat solar thermal panel...and it can be built from off the shelf components with cheap tools. Well, that remains to be seen, but I have a plan...and so far, so good!

I still need to look at the cost of these panels, but from what I have read PMMA (another name for acrylic) is cheaper then polycarbonate.

Now I just need to contact a distributor of Deglass acrylic "double skinned" panels and order a few sheets and get my workspace in order.  I should also enquirer about the reflective IR coating.  If I could get that on a double-wall panel, that would be the icing on the cake.

Plastic Welding, Coroplast and VC's

Well, it sure has taken awhile.  About three weeks ago I ordered some clear sheets of Coroplast.  This is a brand name of polypropylene corrugated sheet, most often seen in on lawns for political signs.  Based on some user feedback I was worried that "clear" Coroplast was not actually clear.  Unfortunately, this is true.  I ordered the 2mm panel, the thinnest I could find.  Its obvious right out of the box that this stuff is not going to cut it.  They will simply reflect too much light back out of the panel.

The Coroplast was not a total loss, as I was able to test out my new plastic welder, and I have a back-up project planed for the material should they not work out for particle panels.  The plastic welder is a $200 tool which is nothing more then a glorified heat gun with some nozzle attachments and the ability to set the temperature in 10F increments.  It takes a little getting use to, but the principle is similar to metal welding: you raise the temperature of two surfaces you want to weld and melt a rod of like material into the joint.  The result is a solid joint.  Of course, its not as easy as it sounds.  If you go too fast or too slow, if the heat gun is too hot or too cold, or if the materials being joined differ in thickness, things may not work out just right.  This last condition is going to be the challenging one when welding the manifold onto the end of the panel.  Since the panel is thin (~1/64-1/32), it will not take much heat to melt and deform it.  If the manifold material it thick, or the welding rod, by the time those components start to melt the panel will have collapsed.  This will indeed be challenging.  Of course, that's assuming I can find the right plastic.

It appears I must now take a closer look at polypropylene.  The wikipedia article indicated that it can be made "crystal clear" when it is biaxially oriented.  I have no idea what that entails, so i'll have to look into it.  Another option would be acrylic, since that could hold up to the water exposure. However, I cannot find it in corrugated double-wall sheets and I suppose this is because it would be rather brittle.  My troubles with finding the proper plastic in a corrugated double-wall form has led me to think up some other designs that I would prototype with solid sheets or tubes.

On a non-technical note I met with a gentlemen representing an early seed capital firm.  It was a good meeting, but mostly entailed me getting him up to speed on solar thermal and the potential market for very cheap solar thermal.  This subject is deserving of its own post, so I'll hold off on that for a later time.

SolSwitch

Its been a few weeks since the last post, mainly because I have been on a business trip and a vacation.  I am back, and there are some developments!  The most significant development is that I have discovered a competing idea.

A couple weeks ago I was browsing this article.  On page 4 I came across this picture:

 On further googling I found a published paper on the idea.  I cant display it here or reproduce the images for fear of copyright infringement, but I'll summarize.  You can glean almost all the information you need from the above picture.  After some more googling, I found a website for the idea.

The concept is wonderful.  When the panel is off, a chamber between the plastic "saw tooth" glazing and the black absorber plate is filled with air, or perhaps more accurately, it is drained of water.  If the sun light enters the panel, the difference in the index of refraction of the plastic (~1.4) and air (~1.0) will cause total internal reflection, a phenomena that all high school physics student learn about.  A real-world example is if you are in a swimming pool just under the surface.  If you look straight up you can see outside, but if you look out at an angle, the water is acts like a mirror and you just see the bottom of the pool.  However, if the chamber is filled with water, the index of refraction increases to 1.33 and...the total internal reflection goes away!  Thus, when the water is not being pumped the panel is reflective/non-absorbing.  When the water is flowing it is absorbing. The idea is that you use the property of total internal reflection to create a "light switch". 

The first problem I thought of is that for this idea to work the sun must be normal to the panel.  However, the suns position in the sky changes.  In the winter (when we really need the heat), the sun is low on the horizon.  In the summer it is high in the sky.  If we are going to use these panels to heat my house (that's all I really care about at the moment), I will want to angle the panels so they are normal to the sun in the winter to maximize the energy I can capture.  In the summer the sun will be almost vertical in the sky, like this:

 The panels will work in the winter, but they fail in the summer from my thinking.  Interested readers  should check out this website, which includes a nice little interactive application.

After reading the published paper, it turns out that a double-layer configuration does work for all seasons: 

The above image is from the sol-switch website set up by the inventor.

Problem solved?  Well, sort of.  On first blush, there is no obvious problems.  The panels can turn on and off by pumping water, and that's really a solution to the overheating problem.  However, when I started to think about the design as it will look like in a completed solar panel, including the absorption plate, insulation and glazing, the problems start to appear.  When I compare SolSwitch to ParticlePanels, I appreciate how simple the particle panel concept actually is.

SolSwitch is a mechanism for controlably reflecting away or transmitting light onto an absorption plate.  Its an added layer, between the absorption plate and the glazing.  ParticlePanels solve the overheating problem by controlling the absorption plate directly.  There is no added layer, because the mechanism is both the absorption plate and the light switch.  


For SolSwitch to be economical, the design must use a small amount of plastic.  As the size of the grating is reduced (reducing the plastic), the channel for the water is also reduced, and at some point pumping water into it is a problem.  A read of the paper, however, indicates they are thinking about using a phase-transition of the liquid.  In other words, if water is in the channels, if the temperature gets above boiling then the water goes to a gas and the panel shuts off.  This seems very problematic to me.  Along with the transition to a gas, there will be a huge pressure spike.  This is going to cause the plastic to deform, which in turn will cause the angles that the light hits to all change, and this could mess everything up.  

I would be very interested in comments from readers.  On the one hand I think this idea is remarkably clever.  One the other hand, when I think about reducing it to a workable solar thermal panel, a whole lot of issues start to be raised.  Can these problems be overcome?  That's what innovation is all about!  If you dont have a problem, you cant innovate.  Hey, we all have problems, and Particle Panels are no exception.  The only difference is that I am telling the world about mine.  I emailed the inventor to see if I could get any information about the idea and possible commercialization.  He was very cagey, only saying that "they are working on optimizing the structure".  Given the seemingly simplicity of the idea, and the fact that they have stated "the idea is ready for commercialization", I find it curious that they are still optimizing the structure and not working with a company to make some panels.  Perhaps they are, but I just don't know.  Perhaps they just cant figure out a problem.  If they opened up the idea to the world and let everybody think about it, I wonder how much more efficiently the problem would get solved?

In any case, its good to have a competing idea.  It helps to highlight design advantages and weaknesses and focuses the development effort.  Competition is always a good thing.