The Shop > Metal Stuff

Small Core Boxes


Just thought I'd start a thread about small core boxes. They can be a challenge to make for a number of reasons, Round "stick" shaped cores are the easiest to do - generally you can make a mold for a stick core from split pipe or tubing if the needed core is a convenient size or if not, fabricate a split wooden core box shaping the core mold with a drill or by boring on the lathe.

But for more complicated core shapes, for instance the core pattern for a 2 cylinder Westinghouse style steam engine block (see photo) it gets harder. There are two routes: an assembled and glued core made from individual pieces, or a complete single molded core.

I'm mainly interested in talking here about single piece cores -- mainly because they present more difficulties. But they are simpler and faster to mould, so definitely also an advantage, if possible.

I also should mention here at the beginning that, yes I do know that cores themselves can be made of sodium silicate, sand, and hardened with CO2, or made from sand and epoxy as a binder. My main interest, as always, and as a dyed-in-the-wool MadModder is in using readily available low tech, low cost materials by preference to modern specialty product and commercial practice. It's not through ignorance that I'm continuing on in my way. And also pretty likely, I already do know about "the right way" to do things. It's just pride of prejudice on my part, a preference for the old fashioned, cheapness and stubbornness, among all my other negative character qualities, in general. And why I'm here in this small forum gritting my teeth in the 21st century.

That is not to say, I wouldn't read, enjoy, and learn from somebody else's thread  about their methods and their own "higher tech" materials. Just that this thread isn't that.  :zap:

Okay here's a pattern for a small single somewhat complicated core:

Okay so this core was intended to create the initial rough bore for two cylinders and also create the crankshaft housing space in the block. That bottom block casting shape is trapezoidal (a trapezium to you guys, we don't know what that is here, and to the north  :wack:). The dowels are 3/4" dia. (19mm) to give you scale. And I made this pattern about a decade ago.

To make a core box for it, I tried casting it in plaster of Paris as a split mold. I did the usual of forming a box first, larger than the pattern, fastening the pattern in place, the box half way with wet plaster, waiting for cure, and then forming the upper half the same way.

That worked, but molding a core afterward from it didn't. At the time I was using baked sand with molasses water and wheat paste as binder for cores. I had not yet seen Ironman's Portland cement core, or plaster casting technique for a core box, made by sinking into a greensand false cope. Both neat innovations.

The biggest problems in moulding cores of this size and complexity are:
1.) that they want to break when trying to release them
2.) adding wires and reinforcement is difficult
3.) ramming is difficult
4.) connecting the two halves is difficult if trying to make a 1 piece core, or
5.) cementing the two halves together as a 2 piece is even harder to reinforce, and the core halves even more fragile

So here are some ideas, observations to improve:

1.) If making a plaster core box, I like the greensand "false cope" idea. It avoids the likelihood of a fragile parting edge in the resulting plaster mold because of the meniscus formed in the first half.

2.) In a split plaster mold ALL flat vertical surfaces of the pattern should be drafted to the parting line, which is the split line of the mold. This means a double taper, either way on vertical surfaces that cross the parting line. This is much more important on a core box than on actual metal casting, where you can get away with rapping an un-drafted vertical plane. You can't rap a core, or a pattern cast into a plaster mold.  :poke: My pattern did not have appropriate draft.

3.) In making the outer wooden box for pouring the plaster, you should shape it irregularly (non-rectangularly?) so that there is a flat face adjacent to the end of any cavity. Let me explain......what you want is to be able to ram the sand from any necessary direction while the mold is closed up.

If you look at my pattern, and if the two dowels were right at the edge of the assembled plaster pouring box, once the plaster was cast, it would leave two open holes where you could ram sand down while the mold was closed up.

All well and good. But also you need to ram sand into the trapezoidal cavity - which means ramming from right angles to the dowels. Well if you build your box with one of those trapezoidal ends tight to the box face, it will also leave a cavity that you can ram into.

Well you could make a complicated box to do all of that. But, simpler would be to make a rectangular box, but pad out the sides with blocks of wood to meet the pattern faces where needed, which would --- well, act like prints to leave spaces where you want them. Loose prints for the core pattern, I guess they could be called.

to be continued.....


--- Quote from: vtsteam on March 04, 2022, 11:31:11 AM ---cheapness and stubbornness

--- End quote ---

Definitely positive characteristics for a madmodder. Keep trying until it (whatever it is) works.


Hey Russell!  :beer: I was just thinking about when we and shipto built engines together -- this engine core reminded me of that time. That was a lot of fun here!  :dremel:

Thinking more about this core box stuff tonight, you only need one hole for ramming sand into a particular shape, but two is better -- if you need to add wire reinforcing. You could insert wire blind, but I find I tend not to do that straight on the first try, and come out too near the side of the bore. Usually I get it straighter, second try. You wouldn't be able to see how the wire lays if one end is blind.

Wire reinforcing is needed for small cores like the one above, there's little chance those 3/4" dia. cylinder bore cores would survive before iron molding even starts, without a wire running through them. In fact it's hard to unmold them from the corebox in one piece without reinforcing. For cylindrical cores of 2" diameter or more, reinforcing isn't needed.

Design considerations:
When designing a part like the Westinghouse steam engine crankcase, you can design to make coring easier. Cylindrical cores are easier to mold than those for faceted solids -- like the trapezoidal shape above. That's because cylinders have automatic draft, if split lengthwise.

If the lower crankcase of my engine had been designed as a cylindrical shape (like those for model airplane engines), there would have been no need for adding draft to the corebox pattern, and the core would have been easier to ram.

Or, if the outer trapezoidal crankcase shape was an aesthetic need, the interior space could still have been cylindrical, allowing a cylindrical core. It would only have been necessary to make the core diameter greater then the crank arm swing.

A slight disadvantage in casting such a shape is that it would yield thicker and thinner sections in the thinnest part of the crankcase, and in general one should try to keep section thicknesses consistent. That's not an absolute requirement, but it is often helpful in avoiding shrink problems, chill, etc.


[0] Message Index

Go to full version