Wednesday, August 31, 2005

making more ribs

This is a great rainy day project: making ribs for your solar collectors. You'll need quite a few of them once you've decided to scale-up and make a larger working system. Each eight foot collector needs seven of these. Six collectors need 42 and so on. A general rule of thumb I heard for heating a swimming pool was to have one third to one half of the surface area of the pool in the collector, so you can start doing some math on your application.

The ribs shown here are at various stages of completion. The green ones on the bottom have received the two-part marine primer sealer coat and are ready for the top coat. As these ribs are made from wood they need to be sealed from the weather.

All of the ones shown above are one half inch (also sold as 13mm) "baltic plywood" which is a terrific material to use where relative precision and strength are required in a panel material.

The white one with the extra large holes is experimental and has the final resin topcoat in white. The large holes give it more of a space age look but their actual function is to lighten the rib on the bottom of the collector panel to improve balance at the collector tube/pivot. I felt that I could remove material using the large sawtooth forstner bits that I bought to make the through holes for the collector tubes. The baltic ply material is very stiff and punching the holes does not seem to adversely affect the strength of the rib in the direction we need it to be strong (mostly lengthwise).

Making the large holes with a forster bit didn't work out very well with the plywood ribs, since without using clamps for each individual hole, the plywood tended to tear out at the back. Aside from the cosmetics, this created a problem for weathersealing and much patching and sanding was necessary to get a clear looking sample. This was solved but took more work. Each hole (there are nine in each rib) requires individually repositioning and clamping the rib to the drill press table, using two clamps, one on each side of the hole to keep the rib on the table firmly as the forstner bit comes through the rib.

I'm trying some ribs made from from MDF (medium density fibreboard) material which is also a fairly common furniture making material available at many home centers. We'll see shortly how the MDF comes out. You could also make the ribs from a plastic sheet material although I haven't tried that. I suspect that a plastic sheet of acrylic, lexan or UHMW would not have the stiffness of wood, although using plastic would eliminate the coating steps which are significant.

When selecting material for the ribs, the minimum thickness is the half-inch material used here but it could be thicker unless the weight became very much greater. Given that the design needs bolts sunk into the ends of the material with some precision lead me to the half inch material I am using.

Here's my first batch of MDF ribs with the first jig on top of the stack. The funny handles are actually off-cuts from the band sawing rough cut operation but ended up being quite functional as hand grips to guide the jig and rib over the router table for the edge routing operation.

I can get ten ribs out of a 2x4 foot piece of material, since they are about 23 inches at the widest point.

The technique I use is suited to a small shop for hand production of a hundred or so ribs (a dozen collectors takes 84 ribs). I've thought that they also could be molded out of recycled plastic, but here I am individually making each one by cutting it out of a sheet and then routing the edge. The ribs above are sitting on my small routing table (the cheap kind with the hand router hanging upside down underneath).

Here's another detail of the top of the jig on top of an MDF rib which has been edge routed using the jig. The 10cm refers to the focal length of the parabola made from this jig. This means that the copper tube is centered 10cm from the bottom of the reflector to achieve design focus.

The small metal strap is cut from a roll of galvanized strapping and used to brace the glued and screwed attachment of the off-cut handles to the jig (two more small screws come up thru the bootom).

The MDF rib has only a small burr which is removed by a light hand sanding. The edge is very nice with no breakouts. Of course, the inside edge is very much more important than the outside edge, since it is the inside that forms the basis for the curviture of the parabola. The metal rails hold the reflector sheet onto this edge. I estimate that with my jig and technique, I am achieving an accuracy on the parabolic curve of about a millimeter from the ideal.

This is the bottom view of one end of the first jig. There are two of these short pins on the jig, one at each end. The pins are actually small steel nails of about 1mm diameter. They are epoxied into the jig and cut off so that about 3mm protrudes thru the bottom.

The two mating holes drilled into each ribs are actually the first step in making the rib, since they guide the loaction of all subsequent machining operations. The pins lock the jig to the rib.

The jig is handmade from 1/4 inch scrap and patched and filled, especially along the edge, to guide the ball bearing cutter mounted at the top of the router bit. The jig could have been made out of any type of material including metals like aluminum.

There is a second jig for the ribs, and it is shown here on top of the stack, held in place to the top rib with two 1mm dia nails which fall into the locating holes marked with the first jig.

I'd originally thought to just have one jig, with the position of various holes marked in it, but the need for strong handles on the routing jig meant that some of the holes would be underneath the handles. So I simply used the first jig to make a copy out of another piece of 1/4 inch thick scrap and marked and drilled the location of the holes that would be necessary.

Using a small center punch, the second jig is used to transfer the location of all the holes to the top of each rib. Depending upon what type of rib it is (there are several variations), some holes are drilled, while others are omited, but all the holes are marked in this manner in each rib. These small dimples in the surface guide a drill bit's starting location in the subsequent operations.

This is the top of the second jig, showing the highly accurate method of locating the jig using small nails. Remember that the accuracy I'm aiming for here is about a millimeter and this seems to be adequate for the purpose, using the simplest tools and equipment for the job.

The material here is 1/4 masonite which makes a good jig material. The jigs should last a long time, provided that you don't damage them. It's a good idea to make a spare or two, just in case.

Also you can see again the fine slivers of material on the top edge of the MDF. This is nothing to be concerned about. The edge is actually very smooth and precisely follows the jig. The fine sliver comes off easily with a light hand sanding.

It's not a solar day today since it's raining. Time for me to go and make some more ribs.

Thursday, August 25, 2005

frame construction

The hard work of building a solid frame for up to 12 reflectors is pretty much finished now. I'm watching the concrete footings dry in the cool early morning.

I took apart the original structure to build this in it's place. In the last couple of weeks, the night-time temperature has definitely had a chill and without even the two reflector system for heat, the pool temperature has fallen to about 72F. The days are hot, but no-one wants to swim. We haven't used the propane heater at all this year and aren't going to start it up now.

I've built this structure to hold the next set of collectors, up to twelve in all.

It's made of pressure treated 4x4's. They are held together by galvanized metal joining plates from the hardware store using deck screws. The vertical ones are four feet in length and are sunk about 18 inches into concrete in ten inch sonotubes.

The structure forms a couple of straight parallel beams, about four feet off the ground and eight foot six inches apart. The reflectors will hang from this structure, with the pipe spanning the gap. The pipes will be installed with their reflectors at two foot intervals between the beams.

In spite of some difficulties like twisted beams and rocky ground, the final top beams ended up very straight, probably much straighter than they need to be, about a quarter inch over the each 24 foot length and parallel by the same tolerance. Hopefully they will be still be straight and parallel after the frost heaves the ground a few winters!

As I push against the these beams, set in concrete to test their satisfying stiffness I wonder if I haven't gone a bit overboard with the structure? After all, this frame will support the weight of several hundred pounds of water and copper pipe and must also withstand whatever wind load that nature will provide. So I'm happy that this support is adequate for my solar heater.

While it involved ripping out the previous prototype, I've gone to this type of structure primarily for simplicity. The previous method, that of passing the collector pipes through a holes in beams, was far too complicated to line up correctly and required one set of posts for each collector. With the new parallel beams, the collectors will just "drop into" the frame and be screwed into place with two pipe clamps, one at each end, into the beam. Couldn't be stronger or simpler, could it?

It's close to the ground to make it easy to work on and to cut down the wind load. I plan to surround it with a fence of the same, four foot height, to reduce the wind load.

Also considered, but rejected this time around, was to drop the southern facing beam about a foot or so to better match the sun's track, but I decided to make them parallel in the horizontal plane for simplicity of the control mechanism and because a thrust bearing would be required at the lower end. Perhaps I'll tackle that on another iteration.

It's important to store the 4x4 beams for a while in a dry place, like the garage for a month. Turn them over once in a while and check for any beginning warpage. I've been doing a few outdoor structures this year and have of pile of wood in storage in this manner to weed out the troublesome ones that are going to warp. There is no guarantee with wood however and the placement and final orientation to the sun may determine whether a beam is going to warp or not. Incidentally, the warped ones (there were several) were chop sawed into the four foot lengths for the vertical supports. The fact that they were twisted slightly over their length down into a concrete footing didn't make a lot of difference and used up the troublesome timbers.

One final word on the mirrors. We dropped one of the collector pipes into a reflector when removing it and the mirror did break, although not into thousands of little pieces like glass. Here's what it looks like.

Fortunately, the reflector material is easy to change if necessary and with the plastic mirror the cost of each is about $40.

Monday, August 01, 2005

parabolic reflectors

The DIY parabolic heater uses thin sheets of reflective material gripped along the two long edges in a ribbed frame shaped like a parabola. This forms a rigid parabolic trough when the frame is tightened to snug the sheet into the frame.

The reflector ends up looking like a shiny inside-out canoe. The parabola is shaped to bounce sunlight onto the pipe which is at it's focus. The pipe also serves as the support for the collector and it's the pivot to allow the heater to follow the sun. The collector pipe doesn't move.

Each frame holds one two foot by eight foot sheet. What type of sheet you use depends upon your budget and what is available in your local area. As the reflector sheet is one of the main cost elements of the heater, I did a fair bit of research on what was locally available and you will too. In the large metropolitan area outside Toronto, I found that there were plenty of choices and I managed to test a few.

The most expensive, but most impressive, at about $8-12/sq.ft. (German made MIRO IV) is an optically clear anodized aluminum electro-polished material sold by a local distributor in Brampton for use in light fixtures. Curiously, the distributor does not recommend exposure to the elements. We'll see. I've had a sample outside for almost a year now and it seems to be holding out very nicely.

The cost of building only two heaters with MIRO IV is estimated to be about $300 but the cost comes down slightly with larger quantities. MIRO IV is like a metal mirror and the reflection of light in ideal conditions is over 95%. It is an almost perfect mirror that can be bent. One nice feature of MIRO IV is that it is fairly stiff material at about 0.20 inches thickness, so it sits firmly in the frame and is quite strong. My only concern, other than the cost, is the long-term effects of having the aluminum in contact with the galvanized edge of the frame. This will undoubtedly lead to galvanic corrosion at that interface.

In this picture, the reflection at the top is from the MIRO IV material, at the bottom is the clear aluminum sheet roofing material (described below). The one in the middle is another product called Anolux which is only slightly less expensive than MIRO IV.

I also tried galvanized steel sheet which I found at the local building center, conveniently cut into 2x8 sheets at only $15 bucks a sheet (about $0.50/sq.ft.!) and learned about the significance of Herscel's infrared experiment in 1800. The galvanized steel showed me that even with a relatively "dull" material that the infrared portion of the sun's energy contained a lot of heat (about half). In a side-by-side comparison, the galvanized sheet contributed about half of the heat energy as did the Anomet material, making it's cost per BTU a lot lower.

The galvanized sheet when new is quite shiny, in a blotchy sort of way. This is the zinc coating that sacrificially corrodes and thereby protects the steel underneath from rusting. You've probably noticed how galvanized metal turns dull and flakes of white oxide form on its surface. Eventually the zinc disappears and the steel rusts. This process depends on the weather and the amount of acid and salt in the air. Just how long reflectors made of galvanized material would be tolerable I don't know. But they do reflect infrared (the visible part of the spectrum is scattered or absorbed). It's interesting that the galvanized material gets quite warm to the touch which wouldn't be the case if it was doing it's job and reflecting all the sunlight onto the collector tube.

One more thing about the galvanized: it is a low cost way to build a prototype that can be later replaced with something better. To replace the sheet in a frame only takes about 10 minutes, so it is easy to upgrade or replace a corroded sheet.

I tried an aluminum sheet product that is sold to the roofing trade for flashing (for about $1-2sq.ft). It is available in standard widths, usually on a roll, at the home centers and comes in various colors including "clear" or natural aluminum. This material in the clear aluminum form produces heating results similar to the galvanized and is a lot thinner, only about 10 thou thick. This makes it slightly harder to handle as it is quite "flippy". You can distort a sheet by mishandling it and produce a crease in the material that ruins the optical properties in that region of the sheet. Also, as the sheet is not anodized like the Anomet and it will degrade more quickly over time like the galvanized sheet. I decided not to use the aluminum further. Although it may make a great prototype material and seems to be readily available in precise and clean 24 inch widths up to 200 feet long.

Any roll material would need to be cut into 8 foot lengths to fit the collector frame. I found it difficult to get clean cuts using tin snips on any of the sheet metal products and I tried to buy them where ever possible in the 2x8 final form.

The winner so far is an acrylic clear plastic mirror material that I only installed about two weeks ago at a cost of about $1.50/sq.ft. This is a 1/16 inch think cousin material to the plastic mirror used architecturally in public places where they don't want glass mirrors. A local plastics distributor had in stock the 1/4 inch thick acrylic mirror material which I didn't think would bend into the frame as well as the 1/16 inch thick version which they readily ordered in for me.

The acrylic mirror goes into the frames beautifully, although even at 1/16 of an inch it is by far the thickest reflector I've tried. The quality of the reflection and focus is as good, if not better than the Anomet.

The acrylic mirror needs care in handling because it can be broken if bent too far. It becomes noce and strong in the frame but getting it there needs two people.

The long term durability of the plastic is uncertain at this point, but it gives good reflection per dollar and it doesn't get hot in the sun showing that it is doing its job.

I'm not sure about galvanic corrosion where the acryllic mirror (which is probably sputtered aluminum) touches the galvanized frame. We will see with time.