Lost foam casting question, sort of.

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

lee webster

Well-Known Member
Oct 4, 2019
Reaction score
Cornwall UK
I have seen other craft sites where a two part expanding foam liquid is mixed and then poured. The foam expand quickly and fills the mould it was poured into. would this foam be suitable for producing a foam pattern for lost foam metal casting? I have a feeling it's the wrong sort of foam and is probably toxic.
That is likely to be PU foam, no idea how it burns out though maybe one of the home foundry type scan shed some light on it as PU in the form of insulation boards would be quite a convenient way to get the material and it is also quite a fine grain which would be easier to work than expanded polystyrene
I was thinking of something like PU foam, and now I've read a bit about it, I don't think it will burn out without producing toxic gasses.
I would think all the foams have some form of toxic gas after all you are burning plastic. Some may be worse than others.

Real questions would be does it burn out easily, does it leave too much ash, does it need a separate burn before pouring or will it melt out as the metal goes in without problems?

If it did burn OK it would be an interesting material if you had a batch to cast. One printed mould to expand as many patterns in as you wanted and then use lost foam casting. Would need a good release agent as it is sticky stuff!
I watched a video about pouring PU foam into a single sided mould. They did two pours, one "open" with nothing to stop the excess foam spilling out, and one "closed", a sheet of plastic with a few holes in was placed on top of the mould to let the excess foam escape, but keep the foam under a bit of pressure. The closed foam method produced a much better finish. I can's answer the questions you pose as I have never tried lost foam casting. It is an interesting subject though!
Urethane gives off all sorts of cyanate based toxins when burnt out.

That being said I have used it with a little water to roll cast a plaster mold with great success.

I've also, in the long past, experimented with putting a dribble in a suitable mold and placing it under vacuum, which was maintained until the PU hardened. Under vacuum it foamed up very well. Excellent details with a very very low density, far lower then the density of ambient cast stuff.

I was using casting PU not an off the shelf product.

If you played a bit with a vacuum chamber you might get something low enough in density that burns it out cleanly with minimal toxic fumes.

I never experimented with PU and sand casting, Styrofoam always did the trick for me.

An old msds sheet will give you an idea of the fumes you are dealing with.

Mixing in a little water helped the reaction speed and volume.

Good luck.
Very useful information Mr. Troll. Did you carve the styrofoam to shape? It's a pity styrofoam can't be poured like PU.
I would be very cautious of the fumes generated by burning foam.

You probably can't undo the damage that those fumes can cause, even in small amounts of exposure.

I read the MSDS (material safety data sheet) for everything I use, including sand, and I take the necessary precautions with commerical respirators and even powered respirators with a remote air source.

The industrial lost foam process uses expanded polystyrene beads (expanded with steam/heat), with a permeable outer coating that allows the gasses generated by the melting foam to pass out of the mold cavity.

I have seen a lot of recent success with the pink lost foam method used in hobby applications, and it can't be discounted, even though it is not a true expanded bead application such as used in industry.

I can't answer the question asked, ie: "Is the two part foam suitable for lost foam casting applications?".
The expanding foam thing is much like the expanding plystyrene beads application, as far as foam expanding to fill a mold.

It seems like some backyard guys can get just about anything to work in any situation, but again, I always say "Show me sections of the casting, and put it under stress, to see if it has cold joints that will fail".
Just because someone can make something work, does not necessarily mean it is the most efficient, safe, or cost effective method for others.

For non-structural, non-airtight castings, pretty much anything goes with lost foam.

Below is a commercial application of the lost foam process, to make tractor castings/parts.

For rapid prototyping, the printed resin-sand mold method is very cost effective, since there is no mold required to expand foam into, and changes can be in the 3D program, and then castings made rapidly after that with a printed sand mold.
Once the prototype design concept is proven, then permanent molds can be made for expanding the beads into.
Last edited:
Resin-bound and sodium-silicate-bound mold parts can be glued up into a complex assembly, often with cores.

I have the adhesive for this application, and I used this method several times with good success.

I have joined molds in a twin fashion, with both molds made from a single pattern, and both molded poured as a single unit with a single sprue.

Troll, when you say you burnt out the PU was that a separate operation or did you use the poured metal to burn it as it was poured?

I wonder if you could make the biodegradable foams that you see used in some packaging, that should expand into a printed mould and unlikely to give off toxic fumes. It could save on 3D printer time and filament as you could print a fairly thin wall negative mould with little fill and then expand the foam into that rather than have to use a lot of filament to print a more durable pattern that could be sand cast. The foam would also suit the slurry method.

Would have saved some filament for this cylinder pattern printed over the weekend that I did the CAD for, it is 13" long and as well as the other half there are 8 prints to form the required cores.


  • 54f1886a-60f8-4e33-bb75-81e77784af15.jpg
    94.2 KB · Views: 0
  • d4ca6473-3b31-4ce6-b3aa-03ecad495bdc.jpg
    111.2 KB · Views: 0
Last edited:
I have sent a request to East Coast Fibreglass Supplies asking for advice. In the meantime I am trying to get my head around designing the patterns/moulds required to produce a foam part. Whem making a flywheel pattern for example, I print the two halves of the pattern, then sand and fill for a good surface. But making a pattern for foam would mean making the inverse(?) of the pattern to take a foam pattern from. If using an FDM printer, which I prefer to use for pattern making, then it would be harder to get a goos surface finish. The layer lines would be internal, not external where they are accessible. Badly explained, but I hope you get my drift!

Reply from East Coast just now, Hi Unfortunately we mainly sell this as a general void filler for example buoyancy chambers on boats. We do not have any knowledge of this being used for moulding. Kind rgds Martin
I suppose it depends on the part, take that yellow one above, that pattern has both concave and convex surfaces so some are easy to access others less so. Would be the same with a negative some easy to sand other s not so easy. If you are going to be using it as a way to produce multiple foam patterns then a bit of work put into one mould would be better than having to prep multiple prints of the same item. Also as you only need the one print you could set a smaller layer height as the extra time spent printing that one would be worth it for the overall time saving

The designing is easy just draw the patten, draw a "box" and use a boolene subtract to cut the pattern out of the box leaving a negative cavity. This is how I do Core boxes, as an example the cylinder I showe dabove has cores for the bore and also the water jacket.

Draw the water jacket

Draw your box and subtract the core. In practice you draw the box as two items one either side of your split line

ws core.jpg

Troll, when you say you burnt out the PU was that a separate operation or did you use the poured metal to burn it as it was poured?
Sorry, I said I never tried burning out PU. My PU work was to make PU objects.

I DID burn out white styrofoam in green sand and unbonded dry play sand. Lots of venting was key. It did work for simple shapes. Even the un-bonded sand made AL objects with minimal clean up required. (Talking crazy simple geometric shapes here)

It was at least 20 years ago and I can't remember stuff like pour rate sorry.