Foundry Work
5 Bears Home

I became interested in casting after constructing the deHavilland Cirrus inline four engine. There were several very nice sand castings which eventually became the crankcase and timing gear cover, among other things. Eventually, I'd like to create my own engines from scratch, and while it is possible to machine complex parts from solid, it is more authentic (and looks better too!) to cast them.

This is the start of a simple pattern which will eventually become an impellor for fuel/air distribution at the rear of the crankcase. The pattern is a plastic impellor from a hand-held vacuum (a "DustBuster"). The penny is shown for scale. The impellor is glued to a piece of 3/4" ply. At the top, a wooden dowl is glued to act as the basin, the area which will be at the bottom of the sprue. Connecting the two is a section of split dowel to act as a gate for the molten metal.
The drag half of the flask is placed over the pattern, and taped in place with masking tape. The inside of the flask half and the pattern is lightly dusted with parting compund. Corn starch or talcum powder works fine.

The pattern and runners must maintain a draft angle of ~ 3 degrees so the pattern can be removed from the sand.

The moulding sand is packed in tight. I am using Petrobond sand, available from Pyramid Products. It is a pre-tempered, oil-based sand with excellent green strength. Very easy to use.

I have found that with small, detailed patterns, I get better results by using my hand and thumb alone to compress the sand around the pattern. The drag is filled, leveled, and flipped. The pattern board is carefully removed.

With the pattern board carefully removed, the mould cavity is revealed. The whitish cast is the parting dust. I use a gentle compressed air blower to remove any loose sand particles from the bottom of the cavity.

A second board contains dowels for a riser (reservoir of molten metal) and a sprue for the cope half of the mould. It too is tightly packed. The sprue dowel is removed, the cope flipped, and the second board is carefully lifted off. The two mould halves are joined.

The furnace and accessories are laid out in a logical fashion. From front to rear in the photo: the furnace itself (Pyramid) and blower. The crucible liftout tongs are next, followed by the pouring shank on 2 firebricks. In the far back is a cast-iron corn-bread pan to recieve excess molten aluminum for re-use.

Not shown is the mould itself and the large propane bottle.

Pour time! Yours truly is using the lift-out tongs to remove the crucible from the furnace. I had already fluxed the melt, followed by a 45 second degassing with dry nitrogen. You don't need to rush the pour, but don't dawdle either.

I have a pyrometer but rarely use it. I find that a slow count to 20 after the aluminum has completely melted will superheat the melt so that I may flux, degas, and pour at the proper temperature. Not scientific to be sure, but results are acceptable.

The degas wand is made from a 2 foot lenth of 1/4" stainless tube, pinched shut at the end and cross-drilled several times with a #60 drill. The nitrogen gas helps remove dissolved hydrogen from the melt which would otherwise show up as interior bubbles in the casting.

Pouring small moulds with relatively larger crucibles can be tricky. It is important that the pouring shank fit the crucible well without wobble. I made a largish pouring dam to direct the flow of molten metal into the sprue hole. Note the moderate level of safety gear... no, I wear no leggings or asbestos suit, but I do have heavy gloves and a face shield, as well as stout leather shoes and long pants.

The molten aluminum is slowly poured into the dam. I am watching the sprue more than anything... stop pouring when the sprue is filled and a reservoir of molten metal rests over the sprue area.

The mould a few minutes after the pour. The aluminum has entered the mold in the foreground in a pouring reservoir, traveled towards the top of the mould, and entered at the sprue. A small brass sheet at the top of the mould is there to prevent overflow.

I have found with small Petrobond moulds that no venting is required (Petrobond gasses less than water sands) and risers can be "blind" (not exposed to the outside). This greatly simplifies mould work.

Initial shake-out. You can see the flow of metal from 12:00 on the photo, to the sprue, into the basin, runner, and the pattern itself. Not visible is the riser at the top of the impellor.

The petrobond chars to black for ~1/4" wherever it is in contact with molten metal. If I had a muller, I could recycle this used sand, but it is easier to simply discard the rather solid, charred chunks. Attempting to break them up and reintoduce them to the sand bin is more trouble than it is worth.

The finished product! Not quite, as it needs to be solution treated and artificially aged. For 356-T6, I soak at 1000 degrees fahrenheit for 12 hours, followed by a cold-water quench. The part is then artificially aged at 375 degrees for a few hours. At this point it is ready to polish and machine.

Sand casting is incredibly flexible and can be used to produce a wide variety of parts. The learning curve is steep, and you will be amazed at the low scrap rate... I'd estimate 80% of my castings are quite useable. It's also a LOT of fun!