What am I doing wrong with Lost PLA investment casting (aluminum)?

I've tried several times to get lost PLA investment casting to work, but it never turns out great.  In the photos you see an attempt at casting some 3D printed pulleys.  Interestingly enough, the one furthest from the sprue turned out the best, but there are still pieces missing and the details aren't as good as they could be.

I used 6061 scraps for the melt.  I've tried adding a bit of boraxo/boraxo+salt.  I've tried without it.

The mold was made by using 325 mesh silica and plaster of paris about 45 to 55% respectively.  This produced far better details than with play sand, but still not great.

After casting the mold I let it sit for 12 hours or so, and then baked it for 2 hours at 350 F, and then 2 hours at 500 F (upside-down), and then put it in the propane BBQ grill for 2 or 3 hours at max heat.  I then blew out the ash using a can compressed air.  In this photo I even kept the mold heated until just before pouring, and then I put it into the furnace until the top level of aluminum re-melted.

You can see that I have several vents (2 for each pulley) and a very large sprue.  The sprue fills quickly before anything comes out of the vents.

Once the aluminum melts I set a timer for 3 minutes and allow it to continue to heat before pouring.

Any advice is welcome from those with experience.  Also any photos or videos of how you've made it work would be great!

The pulley's I've printed are from http://www.thingiverse.com/thing:343343.

Picture of What am I doing wrong with Lost PLA investment casting (aluminum)?
sort by: active | newest | oldest
Toga_Dan9 months ago

Iirc, when jewelers cast rings and such, the molds r spun in a centrifuge. Otherwise u just don't get much detail.

In very small things that;s probably true, because there are going to be surface tension effects. In larger sand castings, I have seen chalk writing on the pattern reproduced in the metal !

whoa chalk?? I wonder if that was the thickness of the chalk or something about the way it cooled the metal.

You'd be amazed at the detail you can see.

als_liahona (author)  Toga_Dan9 months ago

I'll have to decide if it is worth the effort to build a centrifuge.


Actually you don't need a centrifuge, You can do this (AT YOUR OWN RISK) by swinging the casting round on a rope or chain. BUT your plaster needs to be very hot to keep the metal molten.

Needless to say this is dangerous

Downunder35m9 months ago

I tired PLA a few times but never really liked the results in the form.
As you noticed sometimes there is residue that just won't get out, especially if it is flaky and stuck in some corner.
What I noticed is that it highly depends on the type of PLA in use.
For example black is not really good, most orange colors are fine but the best is natural.
Noone really wants it for decent prints so it is often used as a bridging or filling material.
But since there are no pigments it hs the best chance to burn off nice and clean.

rickharris9 months ago

A number of possibilities strike me.

1. I have only done lost wax casting but:

2. Your plaster needs to be almost as hot as your aluminium so that it will flow to fill all of the space.

3. Your definition looks poor which suggests the aluminium is cooling too fast.

4. You need to pour as fast as it will allow so the aluminium needs to be very fluid ie. hot.

5. Possibly steam or air are spoiling the casts. Although you say you burn out thoroughly.

6 are your casts from sand just the same? You could try a 2 part sand cast so you can remove the pattern.

7. Has anyone produced good casts from this pattern?

als_liahona (author)  rickharris9 months ago


2. Thanks, I can try that. It's tricky without a second foundry going - which I can do if necessary.

3 and 4. That I agree with. It may not have helped that temperatures yesterday were below freezing.

6. I've only tried sand (greensand) casting a few times. It's a lot of work frankly -- probably because I built too big of a flask and it takes a full 5 gallon bucket of greensand to fill it, and then its too heavy to move easily. I made another 3D printed flask, but have not tried it yet. I also made a two-part plaster mold for a large part, and that worked okay, but you get a seam line. For most things that is okay. It wouldn't work too well with the pulleys.

7. IDK. I'm not really even in need of pulleys -- obviously I can just buy some if I need them. I'm just trying to learn how to make it work.

LOTS of youtube video showing lost way casting. However watching a movie and doing are different I appreciate that.

As your casting something with a flat bottom you could do this as an open mold, ie cast the pattern upside down up to the base then remove and pour direct to the mold.

The attached image is a casting done at school by an 11 year old. The material is pewter and the mold was made from MDF CNC engraved with the design hence the slightly rough edges.

However the casting process in a 2 part mold was just the same. All first year students cast their hose badge as a project. The only difference is the letter and the colour. The colour is added after casting with cold enamel - a kind of transparent resin.

There is a pin cast into the back to attach to the uniform.

A reasonable amount of detail is achieved considering this is MDF. We had much better results with a speciallt made material for making casting molds using CNC but the material was too costly.

als_liahona (author)  rickharris9 months ago

That is nice. When I made the casting in the image below, I also had, as explained somewhere earlier, created a second mold, with the base of the 3D print pressed against the bottom of the plaster mold. When I burned out the material I was able to see what was left and still blocking molten metal flow (or what might be). Since I had plenty of aluminum, I did try casting into that mold too, but without any vents, it only went 1/4" down and did not form the part of the pulley with the details. It pooled up on top but did not get pulled in. It may have been that I'd already let the aluminum cool too much in pouring the first casting.

Toga_Dan9 months ago

an option that I tried without success: after pouring, put a damp rag over the casting, and press it down with a board. Steam presses the metal into fine detail. I read about it and tried it with brass.

als_liahona (author)  Toga_Dan9 months ago

Didn't work though huh? That would be easier than building a centrifuge. :)

This does work. I suggest wet cardboard though as it makes a better seal. The top of the casting vessel needs to be flat too.

We had a lot of success using tuna tines and small tin cans as casting vessels.


There is some practical advice here on that subject.

als_liahona (author)  rickharris9 months ago

Thanks, I will give it a try. If it doesn't work I'll have to decide if it is worth the effort of building a centrifuge, because I will not be using a rope and swinging it around!

If I can build a foundry, and everything else I've built lately, I can build a centrifuge. :)

I've also read about swinging the mold over your head by a rope or wire to centrifuge it. That idea made me nervous.

Toga_Dan9 months ago

http://www.bing.com/images/shop/search?view=detailV2&ccid=RZC0KPLG&pccid=z5OXLfzK&id=2F4A6FA78D7435EF0A872098E717B5D6954FBCD7&pmid=8DB34143D4B498511ABD6D3CC46A8E99830786C4&q=6+tooth+belt+pulley&qpvt=6+tooth+belt+pulley&psimid=608015766659534107&iss=VSP&selectedIndex=1&count=35&PC=VZW. Just thinking out of the box. I used to make these in a machine shop.

als_liahona (author)  Toga_Dan9 months ago

Thanks, yes, I have a lathe and mill and could machine things like, but the goal was to learn to do more complex things that can't be done as easily that way.

Toga_Dan9 months ago

how big are those? About 1/2 inch dia

als_liahona (author)  Toga_Dan9 months ago

The widest part is about 20 mm.

Downunder35m9 months ago

Did you use the "reservoir" that is to the left of the model?
Can see if you cut it off or just not made it.
Without the amulinium won't hold enough heat and won't flow properly.
The parts seem to be very small too so I have my doubt if the little vents are sufficient.
To me it looks like you tried to pour through one vent hole without using the massive reservoir to the left.
And I somehow doubt you will get enough detail for an aluminium cast of a little pulley...

als_liahona (author)  Downunder35m9 months ago
Yes, the square to the left in the model was the sprue. That is where I poured. It is 25mm square. The vents have a 2mm radius. I took photos after I cut off the sprue and most of the vent from the middle pulley. It was filled solid to the top.

I've been making the sprue larger and larger every time I try and adding vents. For the few things I've tried casting in greensand I use a large round sprue and (smaller [than the sprue]) round vents so that after I cut them off I can put them in the lathe and turn them into usable bar stock. So far that is the only really useful thing I've gotten from casting.

The PLA looked like it all burned out clean, but... Just out of curiosity, I printed the exact same model with vents and sprue, and made another mold (this time with play sand and plaster because I was out of 325 mesh silica) and I put the base of the print on the bottom of the mold so that when it burned out I could see what was left. It was "baked" about the same amount of time as the other mold. When I used compressed air to blow it out I could see pieces of ashy plastic that still seemed to be stuck. It took a lot of blowing directly on that area to get it out. I'm considering trying MoldLay or the MachinableWax.com filament, but they are expensive - so I'd rather only resort to that if it will really help and I can't get PLA to work.

2mm vents. That looks like the bottleneck. Increase it to full dia of the part. 8mm?

If u increase the Depth of the sprue to 1 foot (30cm?) you'll hav about 1.3 psi at the part. This may help with detail.

als_liahona (author)  Toga_Dan9 months ago

Thanks, I will try that.

Do you think that the feeder is too big too, if the reservoir isn't filled ?

als_liahona (author)  steveastrouk9 months ago
The sprue (feeder) was filled to overflowing. Nothing but steam came out of the vents so poured what was left in the crucible on top of them after the sprue was completely full and slightly overflowing. The aluminum just pooled on top.

Since that is becoming a common question, I'll attach a photo of the cut off sprue. It looks terrible because I tried to pour what was left into the vents, but keep in mind that this was only after nothing came up out of them as I'd expected. I've tried before without doing that and the results were similar.

The image shows the underside of the sprue and overflow. The dimples are where the vents were.