Lost Foam Casting

[vtsteam]

It's been a really long time since I did a lost foam casting. As many of you know I'm pretty much a dyed-in-the-wool conventional greensand caster. And I actually like making wooden patterns -- maybe as much as casting metal itself.

Nevertheless, there are some things you can do with lost foam that are difficult or impossible with conventional casting. Complex undercuts, deep draftless and narrow dimensioned parts, and big changes in dimension with a resistance to forming shrink cavities.

The drawbacks to lost foam are foam dust, foam melting fumes, the need to make up a new pattern for every part cast, more rapid contamination of sprue and supporting metal for remelting, and very different orientations and methods for getting sound castings. I don't find lost foam casting pleasant. But occasionally it serves a purpose, at least for me.

So this time, I wanted to give it a try again for a water-cooled hot air engine block casting in aluminum. This is an adjunct for my No. 83 engine experiments. If I build a much larger engine in the future, lost foam casting may become important to reducing the thickness and weight of aluminum needed, compared to a model sized engine.

For the present, I'm still working with an average 1" bore size engine, as a test of lost foam methodology. I drew up the shape I wanted in the old free Google SketchUp and then through a long series of graphic transformations got that shape into a cutting file for my old foam wing cutting software, JediCut. This runs on my CNC hot wire foam cutter.

I won't go into all the work I had to do just to get that shape to cut, but it was NOT the most fun part of this project! I could have made 10 wooden patterns by hand in the time I spent eventually getting one CNC foam pattern cut

Anyway, here it is, finally....with some flaws. There's a piece of excess at the top, a bad lower left hand corner (defect in the scrap foam piece I was cutting from) and two gaps where the wire had to cut through to get to interior hollows. Those will all be plugged with glued foam bits.

 

[vtsteam]

This pattern actually needs a back, so I used a piece of 1/4" fanfold foam for that, and glued it in place with rubber cement.

 

Then trimmed close to the pattern:

 

Then sanded. I hate the sanding dust that foam makes, so I try to minimize this by first cutting VERY close to the line with a used hacksaw blade. The same tool I used to make my full sized rowing boat -- and many model airplanes over many years. Rubber cement for foam is also a long used glue for me when assembling foam planes.

 

[vtsteam]

Here I've glued in the foam bits I needed to correct the pattern. I used offcuts of fanfold foam, which has a somewhat different texture and density than the regular Dow foam I made the rest of the pattern from. This may show up in the finished part. Probably will....

 

Then sanded again...

 

 

Then I added a sprue that I just shaped by hand out of a piece of foam. Again, more foam dust, only this time I moved outside on the porch to do it. Nevertheless, static electricity tends to stick it to you and everything else nearby. even in a breeze, as there was. The sprue was attached with hot melt glue, since the joint was irregular shape and hot melt is good at filling in small gaps -- unlike rubber cement, which produces a very thin glue line.

 

[vtsteam]

Finally I covered the whole thing in refractory by dipping and dripping it on. And then hung the pattern up to solidify.

[tom osselton]

Doesn’t it need a vent hole? Another guy I know had his burp while doing his pour maybe pouring to fast?

[vtsteam]

Hi Tom, I've both vented and not vented in the past very few lost foam castings I've done -- just needle pricks. Ideally you shouldn't need to (as far as I know) if the wash is porous enough. The surrounding sand is dry and much more porous than greensand.

Also ideally the foam should not burn, just vaporize and be driven through the porosity of the coating. I am not an expert on lost foam, but those are my understandings. I have not vented this coating, and plan to just pour it.

[vtsteam]

Well the part is cast:

First fill -- I just used a section of steel tubing as a pouring cup. I mainly used it to try to keep sand inclusion in the sprue to a minimum so I can re-melt it later without a lot of junk in it.

 

[vtsteam]

Here's the first pour out of the sand into another bucket. I was glad to see the mold looked like it filled:

 

And flipping it over also looked good:

 

[vtsteam]

After flaking off the coating, this is what the casting looked like:

Top:

 

and bottom:

 

and here's a detail of the corner where the little bit of foam was glued in. Notice the seams and also the difference in textures of the wire cut foam, and the sanded edge fanfold foam. Amazing detail preservation.

 

[vtsteam]

The large hole shape in the casting would have been bored out for a press fit on a cast iron cylinder. A facing plate would have sealed off the chamber which would contain water for cooling.

While the casting turned out great, I've already changed my mind about the design specifics, so this one probably will be remelted to cast a somewhat different configuration. It was mainly cast as a test piece to prove out the foam construction method, and the adhesive (Borden's Rubber cement) which I've used many times in building model airplanes.

I just have to draw the new design up, and try again -- with the recent experience of this one under my belt. I'm very glad it came out right the first time, and there's very little I'd change about the method used.

This part also could have been cast from a conventional wooden pattern in greensand with a core piece for the bore hole. The sides would have needed a little draft and therefore a machining allowance. but for something this small, that doesn't really amount to much metal loss. There are no undercuts so it wasn't essential to do it in lost foam.