Casting shrinkage voids

[kelvin2164]

I do my own castings for my 5" gauge locomotive and most things work ok. But I'm having a lot of trouble with the cylinders. Quite a large lump of bronze. 4" long with a 1 3/4" bore. I used a 1 1/4" core for the bore. The first one turned out great, but the second one had a large depression inside the bore, pulling up into a big section of bronze that forms the base for the steam chest. Too deep to machine out.
I used a piece of 3/4" steel bar inside the core to help the bore of the casting to cool first. Seemed to work on the first cylinder, why not the second one. Am I pouring too hot? I have a thermocouple that will read to 1,000C if required.
I've attached a pic of the cylinder sawn in half.

 

[ddmckee54]

Are you sure that is a shrinkage void? Maybe your steel core chilled it too fast and the part just didn't completely fill. I would think that if it was shrinkage then the edges would be smooth while I can see rough edges on that void.

Don

 

[Mechanicboy]

Maybe too much moisture in sand in the core who boiled inside the cylinder before the metal is cooled. Sand in core must be dry as possible without to distort core or loose up by shaking from molten metal.

 

[kelvin2164]

Mechanicboy's comment may be on the right track. The first cylinder, that worked out fine, I cooked that core in the oven for an hour to make it more easily handled. The core for the second cylinder was not cooked at all.

 

[kelvin2164]

Also, as I have a K type thermocouple with a digital readout ($18 on Ebay from China) is the pouring temperature critical at all.

 

[abby]

You do not say what bronze you are casting but mention a thermocouple that will read to 1000° C , the pouring temperature of most bronzes will be 1100°C or higher.
Bronzes need good feeders to limit shrinkage , agitating the feeder with a piece of rod will keep it liquid for a little longer whilst the casting solidifies.
Treat the inside of the mould with mouldcoat or similar and burn off with propane torch before pouring , or use petrobond around the pattern surfaces.

 

[kelvin2164]

The alloy is a bit of a home brew. I use a scrap bronze which the supplier says is 95% Cu and 5% Sn. I've added 5% Zn to try to get "gunmetal".
I'm a bit confused on feeders. I understand that they act as a resorvoir to supply metal as the casting cools. But this is a big casting, 4" x 3" x 3", and the feeder would freeze before the casting and pull metal out of the part being cast. The feeder would need to be bigger than the casting in order to stay molten. Please correct me. I find that a lot of 1/8" vents works best for me.
What is Mouldcoat and what does it do.
I'm not very good at this, but learning quickly.

 

[ThomasSK]

The feeders and risers needs to be so large as to contain a reservoir of molten metal that is feed into the item molded, as it cools and shrinks.
You could try a "Bronze casters flask" that molds the part standing up, that way, the shrinkage would be at the end, instead of in the middle of the part. A picture is here: http://www.backyardmetalcasting.com/casting/alumflask01_verticalflask.JPG

Did you make sure that the core you use for the part can vent gasses? I find its easy to make cores that does not allow for wenting of the gasses when the metal is poured. Wenting the core with a pice of wire helps.

 

[Mechanicboy]

The alloy is a bit of a home brew. I use a scrap bronze which the supplier says is 95% Cu and 5% Sn. I've added 5% Zn to try to get "gunmetal".

Be careful to add more metal in copper who can create bronze harder than copper alone. To much metal alloy in copper --> difficult or impossible to machine the hard bronze. (Source: Machine work by Peder Lobben)

The archeological period where bronze was the hardest metal in widespread use is known as the Bronze Age.

 

[kelvin2164]

I had another go at the cylinder, taking onboard the advice I received and got a casting with no huge internal void, as before, but now I have cracks that run through into the bore. Again this didnt happen on the first cylinder.
Attached is pic.

 

[10K Pete]

Wow, that's sure a shrinkage problem. I suspect that a big riser is needed
real close to there to feed that web area between the core and outside. It
looks to me as though the shrink flaw has moved from the inside to the
outside...

Pete