Fusion Melting Furnace

Project started November 23, 2007
Updated 17 Nov. 2008
© Copyright 2007-2008 Ray and Ruby Brandes. All Rights Reserved
The story begins...

This idea was born from a discussion on Castinghobby regarding the effects of what we do on our invironment. I suggested we look for ways to get 'greener' while melting metal. So I started exploring the idea and formed a general configuration for a green melting furnace. For a power source I chose the nearest fusion reactor, the sun.

At first I thought a tracking parabolic reflector with the crucible at the focal point was a good starting point. However, a heavy crucible filled with metal doesn't lend itself to being positioned out on a boom. Then I discovered heliostats.

A heliostat is a similar system, but it tracks the sun to reflect on a stationary focal point. Now I can set my crucible at ground level.

Contacting Justin at heliotrack.com, he helped me with caculations regarding how much sun to harnes to melt lead. Seems 1 square meter harvested and concentrated to 36 square inches (6" square) has the potential of melting 50 pounds of lead in an hour.
Justin also suggested a 7 x 7 matrix of 6" sq mirrors which works out to just around 1 sq meter.

To design the mirror array I first searched the web for ideas. Below is a link to a home-made system that I found quite interesting.
Tetons Solar Page

Justin's design uses an axial tracking system and the array rotates as it tracks the sun.
The X-Y tracking system seems better suited for my use. Click here to see Leo Gerst's design.

Here are some renderings of my design so far:

This is the mirror sub assembly.
The mirror is glued into an aluminum clip with silicon.
The mounting stud will be bent to aim the mirror.

This mirror bar has seven mirrors mounted on a 1 x 1 aluminum tube. (weight 4 lbs)

This is seven mirror bars framed together. 1.14 sq. meter.

Here is Leo's site. Heliostat.Us

01 Dec 2007
Rough concept of the furnace. Range lead will be placed in the squirell cage. CSL will be focused directly on the bullets. As the lead melts out it will fall into a collection vessel/melting pot below. When most of the lead is out, the cage will be rotated to tumble out any lead trapped in the bullet jackets.
When all the lead is out, the CSL will be re-aimed onto the melting pot and the lead re-melted and then poured into ingots.

05 Dec 2007
Bought 12" x 12" mirrors and made a gage for cutting them into 6" x 6" squares.

06 Dec 2007
Ordered 1" aluminum tubing for the frame.

07 Dec 2007
Ordered sheet metal for clips and had it shipped to my brother who has a brake and shear.

12 Dec 2007
Aluminum tubing delivered.

13 Dec 2007
Brother's tinsmith equipment can only do the end bends. Asked he do that and send the clips for completion here.

14 Dec 2007
Ordered threaded rod, nuts and washers for mounting the mirror clips.
If the bent stud method of adjusting the mirrors doesn't for my 20' focal length, I have two other methods to try. They are more costly so I will try the bent stud method first.
Spherical washer mount
Ball and socket mount

16 Dec 2007
The control system is on the way! I have firmed up the basic design for the mirror.
Heliostat Two

27 Dec 2007
I have reviewed the Basic Stamp stuff while visiting in Dallas. Looks simple enough. I hope to try the basic one mirror heliostat on the next sunny day.
Regarding the expected problem with the sensor seeing many mirrors, I may sink the sensor at the bottom of a tube so only the main (center) mirror will be able to reflect on to it.

28 Dec 2007
Mirror clips from brother Fred get the first mirrors glued in with silicon seal. I have only done enough for one bar.
I need to see if the bent stud method will work for a long focal distance (20+ feet).

30 Dec 2007

Assembled the first mirror bar. Now to test if the bent stud adjustment method will work. I think It will!

31 Dec 2007

Today I got Leo's heliostat up and running. Mostly a test of connections etc. I have to study the software!

3 Jan 2008
I assembled the mirror framework today. Very ridgid and strong. Also played around with the Basic Stamp learning how to run the motors back and fourth. Tomorrow I will play with the photosensors.

4 Jan 2008
Today I machined the arms for the tilt axis. I think I will try plane bearings to start, delrin on aluminum.
No sun today so I didn't study the photo sensors.

5 Jan 2008
Roughed out the plan to balance the thing.
Since I want the master mirror to be at the intersection of the X-Y axii, the CG will be to the rear of the pivot point. A boom with a cable over a pulley to a counter weight should do the balancing act.

10 Jan 2008
Today I got my tracking program worked out. Seems earlier versions expected the processor scan to be faster than it actually is.
Now to gimble up the mirror array and motorize it.

11 Jan 2008
Ordered inch size ball bearings from Enco. Hey! They are on sale!
Spent a little time refining the motor mounting design.

20 Jan 2008
More 1" square tubing is on the way. Today I machined the bearing blocks for the tilt axis.
Playing around with my camcorder and Adobe Elements I compressed the 25 minute job into 5 minutes. Click here to see the video!

5 Feb 2008
Some time has past since I updated these pages, but I have not been idle.
If you watched the YouTube machining flick, here is where that part goes.
It is one of the bearing blocks for the tilt axis.

Then I made a lost foam casting with aluminum for the pan bearing and yoke support.
This was a little tricky to machine, but I went slow and careful and it turned out well.

Once the casting was finished, I needed to make the carriage to tote the helio around with.
Here is the array frame mounted on the carriage.
The mirror you see is the central 6x6, one of 49 that will eventually fill the grid.

Here are two more shots from the back side.

The dark bar is for the counterweight stack. Because the weight is behind the tilt pivot point, the counterweight has to pull up on the back of the array. A pully and cable allow the change of direction. I will make slotted lead disks of about one pound each to stack on the bar as necessary. Without the mirrors, about six pounds balances the rig.

Here is a closeup of the drive screw and nut for the tilt motor.

And for the pan motor.

6 Feb 2008

The counterweight system uses a 1 x 1/4" flat bar with a 2" long cross piece welded at the bottom.

I made a mold for lead weights of about one pound each.

It takes six weights in the stack to balance the rig. The counter balance is easily adjusted by adding or removing weights.

16 Feb 2008

Here is a photo of me after fitting the final mirrors.

Here is the finished heliostat, 49 6"x6" mirrors, about 1.14 sq. meters.

When all the mirrors were syncronized and uncovered, it didn't take long to ignite the target!
Leo's program for the Basic Stamp micro processor took into account clouds etc. Considering the amount of heat I need to melt lead I will only run on clear days, and we have a lot of them here in Florida. Once I understood the system, I modified Leo's program removing things not needed to track the sun. You can see it here.

18 Feb 2008

Here is the fully operational heliostat.

Today I melted lead with it. Just a small amount that could fit in a tuna can. When placed in the hot spot, the bullets began to sweat in just a couple of minutes.

24 Feb 2008

This is a shot of the more perminant setup. I set bricks in the lawn for the helio to set on so when I roll it out it is easliy positioned. The cinder blocks the bottle is setting on will stay in place for now. The air temperature inside the un-insulated bottle hit a high of 265F.

This is a shot looking into the bottle at the squirrel cage with the lead to be melted. You can't tell, but lead was melting and dripping out when I took this photo.

This is how the recovered lead looks. It drips into the collection pot and solidifies immediately.
This is about 5 pounds recovered without rotating the cage.
Parts are on the way to rotate the cage and I expect great results.

Rotating the cage helped a little, but not being able to keep the heat in was the biggest problem. The only window that didn't shatter or melt was mica and that had a dark tint and I think it absorbed a lot of energy.

11 Nov. 2008
I decided to try to reflect the energy up into an igloo type of furnace where at least I could minimize the convection losses from air.
Leo had suggested a water cooled window and I took his idea and applied it to a water cooled mirror.

Two mirrors with copper pipe manifolds top and bottom guled together with silicon seal. Thermo syphon for water circulation. The water in the coffee can got up to 90F.

My design called for an insulated metal can for the furnace. This would sit on a tripod. The curcible would also be supported by the tripod.

This shows the detail of the screws that the furnace can rests on that also hold the bent rods that support the crucible.

An adjustable mirror bracket below would re-direct the CLS up into the furnace and onto the crucible.

Here is the tripod with mirror bracket, mirror in place and furnace can in place.

Here is the rig in operation. The inside of the furnace can is painted with high temp silver to reflect energy back to the crucible.
The crucible is a stainless steel utensil can from Walmart painted with flat black bbq paint. My hand held measured 750F+ on the bottom of the crucible.

The feedstock was bullets I picked up from my backstop. I didn't weigh the ingot, but I am guessing about 3 to 5 pounds.
I am not sure when it was melted to pour, as I had to remove the furnace to look in, but I think it was 30 minutes to one hour.

16 Nov. 2008
A few lessons were learned from the experiment of Nov. 11.
I added a wind skirt around the bottom to try to lessen the drafts from the breeze.

I also added a perminent shelf for the water can.

The new setup is light enough to carry with one hand with no water and the furnace and crucible removed.

Here is a view of the hot spot via the secondary mirror.

I remelted the ingot from the 11th in about 45 minutes, 4lbs. 9oz.
Then some scrap was melted producing an ingot of 7lbs. 3 oz. in about the same amount of time.
A therometer through the insulation at the top of the furnace registered almost 500F.