(c)1996 William Beaty

For those who dream of giant solar collectors which can generate temperatures high enough to melt steel, here is a method for building your own Solar Furnace of any size you desire.

These diagrams show a tiny 6 X 6 mirror-array as an example. Also, I used the silicone-glue "hinges" for this version so my array could be reprogrammable. If you only want a fixed-focus furnace, you can use screws to adjust the mirror chips, then glue them permanently into position. Obviously you can increase the size of the array to large as you wish. My dream is to build an array the size of a 4ft X 8ft sheet of plywood. With 1 inch mirror chips, it should create a focused spot which is 4000 times brighter than normal sunlight.

The trick is simple: Make or buy a large number of mirror chips of glass or plastic, each one about 2cm or 3cm in size. 12in glass mirror tiles can be had from the larger hardware stores, and these can be scored and broken into 2cm chips with a glass cutting tool. Note that an "oil-fed" glass cutter from a stained glass supply store works far better than the cheap little diamond-wheel cutters they sell in hardware stores. Or if glass-cutting is too weird for you, you can cut up some acrylic mirror-plastic with a power saw. Search for plastic mirror scrap at window-glass companies, or find larger pieces at plastic sheet suppliers. The smaller your mirrors, the hotter the focus.

[Drill Guide: screw holes in opposite corners of each mirror chip, glue dot at 3rd corner]
Obtain a sheet of plywood and a big bag of screws which are about 1/4in longer than the thickness of the wood (if the wood is 1/2in thick, buy screws that are 3/4in. or slightly longer.) I used 6-32 machine screws because they were cheap, but most any screw will probably work. If your mirrors are larger than 1 in., use longer screws. If your furnace is going to be extremely wide, use thicker plywood to prevent warping. Lay out and drill the holes as shown, choosing a drill bit which is only slightly smaller than the screw diameter. The screws should turn easily with a screw driver. If your furnace is going to be wide, consider making four square regions where the mirrors in each region tilt towards the center. If not done this way, then the toothpick spacer mentioned below might not make the glue-globule tall enough. In that case the aim of some mirrors might be unable to hit the focus. Or, instead of complicated pre-aimed regions, just start aligning your entire array using a focus that is appropriately off-center from the whole furnace! (Idea: make four separate furnaces with the focus in the corner, then arrange them in a square.)

[Detail of one facet]
To make "hinges" for all the mirrors, I put a small glob of RTV silicone caulk at the corner of each mirror. Since this glue blob must stand up and not be smashed flat, I cut a number of round wooden toothpicks into 1/8inch segments, placed a small dollop of silicone sealant at the corner position of each mirror, laid each toothpick sideways into the silicone glob as a spacer, then laid down the mirror on the correct spot on the wood. The silicone lightly attaches the corner of a mirror to the wood, while the toothpick segment will lift up the corner of each mirror; tilting each mirror and giving it a greater adjustment range. The toothpick spacers also keep the mirror from smashing the silicone into a too-thin layer. Carefully perform this operation for all your mirrors. The toothpick in the corner will tilt each mirror one way, and when you later add the screws to the corners, the screws can lift the mirror as desired.
Let the mirrors harden for 24hrs. When the silicone has cured, screw all your screws into their holes just enough to let them touch the mirrors.

Programming the array as a solar furnace

Cut up some Post-it(tm) notes, and stick a square of paper to each mirror to block the light. LEAVE ONE MIRROR NEAR THE CENTER OF THE ARRAY UNCOVERED. This one mirror will act as a reference for adjusting all the others. (As the position of the sun changes, the position of this spot will also change, so you shouldn't need to move the whole device to track the sun while adjusting mirrors.) Take your mirror array out in the sun, and position it a few feet from a convenient target. The distance between the mirrors and this target will become the focal length of the Solar Furnace.

First use your screwdriver to adjust the single bare mirror ship so it is no longer tilted by the silicone glob. Bounce some sunlight from this small mirror towards your target, and move the whole device so the spot of light hits the target. Now remove the Post-it(tm) note from one other mirror chip, and use the screwdriver to adjust this other mirror so the spots of sunlight from the two mirrors combine together. Cover this newly-adjusted mirror, and uncover the next one. Repeat the adjustment process, then cover it and go to the next. Do not of course change the setting of your central "reference" mirror. When each mirror has been adjusted to combine with the "reference" mirror, peel all the paper from the mirrors and see what you've accomplished. All the bright spots should now shine on the same place. Tweak any stragglers to put them in line. If you wish, go through and glue down each mirror permanently and remove all the screws.

A 6X6 array is pretty safe for experimenting. It won't set anything on fire, but in the summer sun it will heat a black garbage bag almost to the melting point. Once you have the whole process learned, try making a 16 X 16 array (256 mirror chips), or even larger. The above process lets you slowly "coat" any flat wooden surface with solar-furnace arrays. With thoughtful planning you could even cover a non-flat surface with a solar furnace array. If you had enough time, you could build one of ANY size and temperature. ***WARNING*** if you build a big one, keep it covered when not in use. If you leave it around the house, moving sunbeams might unexpectedly bounce off it and start a fire!


Once I had a brainstorm, and when I tried it out, it actually worked: send intense light into one end of a fiber-optic cable, and use the other end as a woodburner. I used an expensive glass-fiber cable 1cm in diameter which was about 1/2m long, and I placed one end of the cable at the focus of a 12-in fresnel lens in summer sunlight. The other end of the cable could char a wood surface, but just barely. A bigger solar furnace might have made it impressive. Sign your name as charcoal! Note: if you try this stunt, realize that you probably will damage your fiber optic cable, so don't try it with one that you can't afford to lose! Plastic opto fibers might melt, so use glass if anything.


The smaller your mirror chips, the smaller and hotter the focus. After all, the hotspot is approximately the size of a single mirror. An array of 1in. mirrors a foot across will make 144 beams, but if you use 2in. mirrors for your 1ft furnace, the hotspot only receives 36 beams.

However, if you use small mirror chips and adjust your solar furnace for a very long focal length, you'll find that the hotspot grows larger, fuzzier, and cooler. This occurs because the sun is not a tiny point, instead it is a disk, and the mirror-facets act as the pinholes of a "pinhole camera." Small mirror-chips form an image of the sun, rather than an image of the mirror-chip shapes. Each little square of light will develop a blurry edge, and only the center of each square image will get "full sun."

To compensate for this blurring effect, use larger mirrors. Here's the rule: choose a focal length which is lots shorter than 120 times the width of a mirror-chip. (This 120 comes from 1/tan(.5deg), the sun being about 1/2 degree in angular size.) For example, the 1in. mirrors would give a blurry hotspot if F.L. was longer than a few feet, and at 9.5ft the blurred regions swallow the hot center of the hotspot. This "blur" is an image of the sun. If you want to burn objects from 120ft away, you'll have to build a furnace using mirrors which are wider than 1ft each. The size of the sun-disk is the cause. (If our sun was tiny, but still just as bright, then this blur would be gone, and you could form its light into an intense parallel beam like a laser!)

For a 2-ft solar furnace:
1 in.9 ft.576
2 in.19 ft.144
3 in.28 ft.64
4 in.38 ft.36


Rather than "programming" your adjustable array to form a solar furnace, you could instead use it to form any desired pattern of light spots. For example, make it project your initials. First adjust all the mirrors so they are flat. In the sun, the array should reflect a square grid of light spots. Now choose one-half of the mirrors and adjust them so they form one of your initials. Do the same with the other half, for your second initial. Now you have a "magic mirror" which looks like a bunch of little squares; like an almost-flat mirror, but when exposed to sunlight it creates a giant projection of the initials of your name.

Chinese "magic mirror"

Some science museums have ancient Chinese "magic mirrors" on display. These look just like a flat mirror, but if you bounce the sun off them, you'll see a picture in the projected sunbeam. The surface of these mirrors contain tiny wobbles. We can program the solar furnace to do something similar. First program your solar furnace so it's perfectly flat; so it projects a square array, and so it acts like a big flat mirror when you use it to look at your own face. Then go out in the sun, aim the beam at a far distant wall, then program the mirror chips to make a tiny pattern (such as a number, or your initials). Now use the furnace as a mirror to look at your own face. It acts like a flat mirror, no? The deflections used to produce the pattern are tiny, so the mirror is ALMOST flat, just like the Magic Mirrors.


DeathMeK says to use old CDROMS (or free ones from junkmail) as zero-cost mirrors. Unfortunately the silicone glue used as a "hinge" just makes the thin mirrors bend, so you'll have to attach them some other way. Or perhaps use silicon glue, but only use an incredibly tiny dot of the stuff.



Date: Sat, 27 Nov 1999 12:31:50 -0700
From: rwduncan
Subject: Solar Furnace

Hello Bill,

My name is Robert Duncan and I live in the Phoenix, Arizona metropolitan area. While searching for solar energy-related subjects, I found your website through Infoseek. After reading your 1996 article on the Infinitely Large Solar Furnace, I became interested in your technique and constructed my own. The finished product is a 48 inch by 48 inch solar furnace with 1440 mirrors. While this is hardly infinite, I was able to transform an aluminum beverage can into smoke.

The materials and method of construction differed only slightly from the information posted on your website. I used drywall screws instead of machine screws because they were cheap. I pre-adjusted all the drywall screws for a thirty inch focal length prior to gluing the mirrors in place. I also omitted the toothpick spacers since my final adjusments would be very small. The programming was done by placing an unfrosted lightbulb at the focal point, allowing each mirror to reflect the light onto a grid pattern drawn on a piece of plywood placed along the focal plane. Each mirror was adjusted until its reflection was aligned with it's corresponding grid sqaure on the focal plane.

The result was approximately 1000 watts of solar influx concentrated on an area the size of a silver dollar. Wood ignited with an audible "pop" the instant it entered the focal point. Toast burns instantly. Aluminum melts after 15 seconds. Half inch copper tubing deforms under it's own weight after 20 seconds. Steel glows red in about the same time. The temptation for onlookers to stick thier hand in front of the furnace is strong, so I keep a cardboard box handy to demostrate what will happen to them if they try.

I appreciate the information and I wanted to pass along this success story. All this was accomplished with less than a $100 investment in materials. My next step is to build a heliostat and attenuator to create an apparatus similar to the ones at the National Solar Thermal Test facility in New Mexico.

Thanks again,
Robert Duncan
rwduncan atsign gateway period net

Want books? Try searching

(try "science fair" too)
Created and maintained by Bill Beaty. Mail me at: .
View My Stats