Home Foundry

[GreenTwin]

I built my first furnace back in 2012, and have experimented with oil and propane burners ever since.

I melted aluminum first, and that was pretty easy. You can melt aluminum as easily as Zamak (in my opinion) using the same propane burner and just slightly more heat.
Zamak has more mass to it.

I really wanted to master pouring gray iron, and that took a while to figure out.
If you know what you are doing, casting gray iron is not much more difficult than casting aluminum, although there is a lot more radiant heat that has to be carefully controlled to protect body and eyes.
The slag has to be carefully handled with iron, but that is really no problem.
Gray iron does not have to be degassed, but you do have to let it cool in the mold overnight to prevent hard spots, and add a very small amount of ferrosilicon.

The burner and furnace must be built to withstand iron temperatures, and the crucible needs to be a high-quality clay-graphite unit that is ferrous-metal-rated (I use Morgan clay-graphite "Salamander-Super" exclusively).

I use a siphon nozzle style burner, and diesel for a fuel, with a flow rate of about 2.6 gal/hr.
Variable speed Toro leaf blower for combustion air.
I will be converting to a gear-pump style pressure nozzle soon, which looks just like a siphon-nozzle but does not use compressed air for atomization, but rather a small gear pump instead.

I had to learn pattern making with machining allowances, shrinkage, draft angle, etc., and now I use a lot of 3D printed patterns.

 

[GreenTwin]

There is an art to gating, runners, risers, spruing, pour basins, etc.
I use a variant of some of John Campbell's methods these days, and am getting consistent results with no defects in gray iron.

Gating/risering/spruing/pour basins is a hot topic (no pun intended), and discussing methods in the backyard community is like discussing whether you think a Ford, Chevy, or Chrysler (Hemi) makes the best racing engine.

What I have done is to find a method that works consistently, and stick with that until I find something better.

All of the iron pour events I was aware of were cancelled due to COVID, and so I got very bored, and decided to make a jumbo-belt buckle, or plaque, or whatever you want to call the thing. The intent was to cast something/anything, and the symbolism of the Phoenix was to show that one day the iron pour events would occur again, and the casting community would rise up out of its fear of COVID and start the group casting shows again.

Here is the pattern, which was designed in Solidworks from a hand sketch, and then printed on a Prusa.

.

 

[GreenTwin]

Here is the 3D print, and the front of the print is convex, to get away from just a flat-front look.
I used dual-colors for kicks, to see if that would work, and it did.
I just cut the filament in mid-print, and then fed in another color.

 

[GreenTwin]

Here is the pattern filled with a bit of spackling compound, and then painted.
Filling the pattern was easier than getting Solidworks to insert a fillet around such a complex shape.

 

[GreenTwin]

The pour went without a hitch.
The molds were made using resin-bound sand, which is the cat's meow of molding sand.
The molds were sprayed with ceramic mold coat to give a superb finish.

A simple open-face pour was used, but normally I use a cope/drag closed mold arrangement.

.

 

[GreenTwin]

These are the gray iron castings immediately after I flipped them out of the mold, and before I did any cleanup work on them.
With the ceramic mold coat, there is only very light sand adhesion to the casting, and that can be cleaned off with a few light strokes using a dry paintbrush.

 

[GreenTwin]

Here are the Phoenix gray iron castings after brushing them off with a paintbrush.
It was a very fun day at backyard casting, and a nice relief to the pandemic stuff going on.

With a slight amount of ferrosilicon, machinability is excellent, and there are no hard spots or chills in the iron.

It took me about 8 years to completely figure out the iron process, but I have seen people pick it up in a month when provided with the right information about how to do it. I had to learn mainly by trial and error.

 

[SmithDoor]

I like 3D printing
That will save a lot time.

FYI: The make sheet wax for making cores.

Dave

I built my first furnace back in 2012, and have experimented with oil and propane burners ever since.

I melted aluminum first, and that was pretty easy. You can melt aluminum as easily as Zamak (in my opinion) using the same propane burner and just slightly more heat.
Zamak has more mass to it.

I really wanted to master pouring gray iron, and that took a figure out.
If you know what you are doing, casting gray iron is not much more difficult than casting aluminum, although there is a lot more radiant heat that has to be carefully controlled to protect body and eyes.
The slag has to be carefully handled with iron, but that is really no problem.
Gray iron does not have to be degassed, but you do have to let it cool in the mold overnight to prevent hard spots, and add a very small amount of ferrosilicon.

The burner and furnace must be built to withstand iron temperatures, and the crucible needs to be a high-quality clay-graphite unit that is ferrous-metal-rated (I use Morgan exclusively).

I use a siphon nozzle style burner, and diesel for a fuel, with a flow rate of about 2.6 gal/hr.
Variable speed Toro leaf blower for combustion air.
I will be converting to a gear-pump style pressure nozzle soon, which looks just like a siphon-nozzle but does not use compressed air for atomization, but rather a small gear pump instead.

I had to learn pattern making with machining allowances, shrinkage, draft angle, etc., and now I use a lot of 3D printed patterns.

 

[GreenTwin]

For round cores, I have a variety of sizes of schedule 40 PVC pipe, and I cut a piece to the correct length, using a diameter that is slightly smaller than the rough cast cylinder bore.
I slit the PVC down one side with a thin slitter saw, and make a short slot at the center of the slit.
I fill the PVC with either sodium silicate or resin bound sand, and when the binder sets, I put a screwdriver flat blade in the slot and pry slightly.
A near perfect core drops out.

Most people will cut their PVC in half, which means you have to clamp the two parts back together, and then the corebox halves are prone to getting separated/lost.

Below is an example of cores made using a one-piece PVC corebox.
You can see the coreboxes in the background.
I make multiple cores at the same time.
The cores are vented using an 1/8" or 1/4" wood down in the center, which is removed once the sand sets.
The cores are vented to the top of the cope.

 

[GreenTwin]

Some examples of 3D printed parts that were derived from a 3D model are shown below:

 

[GreenTwin]

This was the first item I cast in gray iron, in 2012, for the green twin oscillator.
I was pretty clueless about how to do iron at the time, and sort of stumbled my way through it.
All things considered, it turned out pretty well, and I used this on the green twin.

Its one of those things, if you don't get in there and try it, then you will never figure it out.

I built my furnace way too heavy, and it took 30 minutes to heat up enough to start melting the iron.
I thought bigger (heavier) was better with furnaces, but later learned the opposite is actually true.

The sand I used is called (tradename) Petrobond, and while it works pretty well with aluminum, it does not work very well with iron.

 

[GreenTwin]

For the next flywheel (this was a group build with two builders, myself and woodguy) I upgraded to OK85 sand and a resin binder that sets the sand into a hard block.

The resin bound sand is the best I have seen for mold making, and is extremely versatile.
The art-iron folks use it a lot.
Resin-bound sand is not easily reusable, unfortunately, but I have not seen a better finish than can be produced with a resin-sand mold, and ceramic mold coat.

 

[GreenTwin]

Here was a test run of the furnace and burner, and I poured some ingots.
Pouring ingots from scrap is ok for aluminum (if you degass), but for iron, it is best to leave the iron in the rough scrap form, and not make ingots.
Since this was just a test of the equipment, I did make ingots, and they were usable, but a lot of time and fuel is wasted if you convert iron to ingots.

I cut one of the ingots, and broke another, and it was good clean iron.

The amount of ferrosilicon that needs to be added to an iron melt is ever so slight.
Any more than a very small amount and you get a lot of shrinkage and bad parts.
The ferrosilicon is not absolutely necessary unless the parts are thinner than about 3/4".

Some folks will insist that you have to use clean scrap iron, but that is not true.
I have seen some people use scrap iron that looked like it had been on the bottom of the ocean for 20 years, and it melted no problem, and created perfect castings.

I use plain gray iron scrap, and scrap that breaks cleanly with a sledge hammer.
I don't use scrap that has phosphorus in it, such as many thin iron castings like radiators, since I have heard this makes for weak engine parts.

 

[GreenTwin]

Here is an oversized Cretor's style flywheel, in 356 aluminum.
I carved one spoke from wood using a dremel tool, cast six of them, added those to a wood rim and hub, and then cast an overall flywheel.
At the time, I was still learning how to do iron consistently, and so I did not attempt this piece in iron, but I will go back and re-pour it in iron one day.

I found a little stainless steel sauce pot on sale at a nearby store for cheap.
The lady asked me "Are you going to do some cooking with this little pot?".
Best I could do was give her a simple "Yes.......cooking." LOL, how do you explain pouring molten metal out of a cook pot?

I now use clay-graphite crucibles exclusively, but in a pinch, for a one-time use thing, you can use a metal crucible, but beware since they sometimes fail in mid-melt.

 

[GreenTwin]

Here is the assembled flywheel pattern, and the mold making setup.
I used sodium-silicate-bound sand for this flywheel, and it turned out ok (this was a proof-of-concept piece), but I prefer the resin-bound sand.
I used a simple naturally aspirating propane burner for this melt, since it is easier to use with small melts.

You can melt iron with propane, but it takes a lot of pressure and the tank tends to get too cold, and the pressure drops too low, so I use diesel exclusively for melting iron.

 

[GreenTwin]

This is how the pattern was assembled.

 

[GreenTwin]

And a comparison between my flywheel and a Cretors No.02 flywheel.
The concept was proven.
You can make a pretty realistic Cretors-style flywheel using just one hand-carved spoke.

 

[GreenTwin]

Here is a straight edge that I cast for a buddy of mine.
He made the pattern.
This piece turned out pretty well.
I was starting to get the process figured out, and was getting consistent pours with my 2nd iron furnace, which is infinitely lighter than my first iron furnace.

I started using snap flasks, and I remove the flask from the sand before I pour the iron.
The two mold halves are adhered together with a ceramic glue.
This prevents any damage to flasks due to heat/spills, etc., and so they last indefinitely.

I don't use hardware on the sides of the flasks for registration, but instead just drill holes with a long drill bit through the cope and into the drag, and then insert rods for cope/drag alignment. The pin method works perfectly.
I normally bend an L-shape on the end of my pins, but these were of hard material and would not bend.
I misplaced my L-shape pins, but later found them.

 

[GreenTwin]

These straight edges turned out pretty well.
I think this was the first time I used ceramic mold coat, and that prevents most of the sand from adhering to the casting.

 

[GreenTwin]

I sent these castings off to my buddy (the one who made the patterns), and he final machined them.
Turned out pretty well for a first-time straight edge attempt I think.

Machined easily with no hard spots.