Using Silicone Molds to Make EPS Patterns For Lost Foam

Discussion in 'Pattern making' started by rightbug, Sep 5, 2019.

  1. rightbug

    rightbug Copper

    I saw some folks wondering about the possibility of using silicone molds to create EPS patterns in an older thread. I'm not sure if anyone ever tried it but I've been playing around with the process this last week so I thought I'd share my findings so far.

    My ultimate goal is to cast a brass ball-peen hammer using lost foam casting and it quickly became apparent that with my limited tools and skill set, sculpting a decent foam model wasn't going to be possible. My research into alternative methods led me pretty quickly to EPS beads.

    There are some good tutorials on Youtube for using EPS beads to make duck decoys. Finding a non-industrial source of pre-expanded beads was a little trickier but there is a guy on Ebay who sells them at a good price for making fishing jigs. The beads are smaller than those used for decoys but seem to be working okay. We'll see if the density is an issue.

    Following the advice of decoy makers, I coated the mold with oil and sprinkled in a layer of un-expanded beads to create a smooth outer shell. I pre-expanded the rest of beads halfway as needed in a measuring cup placed inside a pot of boiling water with a steaming rack in the bottom. (Not sure what this little stainless steel rack is called -- I stole it out of my wife's Instapot. It just serves to keep the measuring cup and later the mold off the bottom of the pots.)

    My first thought was to make a hollow core form around the temporary wooden handle I used when making the mold:


    This proved to be a failure as the steam had no way of penetrating the mold. I figured it would be easy enough to drill out the eye for the handle after casting so all of my future experiments did away with the wooden core. The form handle ends (see below) will probably give me a good place to attach gates, risers or vents.

    After the mold is prepared, into the pot it goes for steaming. My pot has a glass lid with a little vent hole for excessive stream. How long to steam for? Decoy makers with their aluminum molds do 60 minutes. After my failed first attempt where the steam didn't penetrate the silicone mold I've been going for an almost certainly excessive 2 hours after which I let it cool down before removing from the mold. Once I have a handful of decent patterns I'll play around with steaming time to dial this in a bit.

    For my second attempt I braced the mold top and bottom to keep it closed as the beads expanded but I ended up with lateral distortion as the beads pushed to the side. Still, the end result was very promising. Addressing this distortion seems to be the primary challenge in using silicone molds for EPS.



    The screen was to keep beads from leaking or expanding of of the hole in the mold. For subsequent attempts I moved it to the end of the hole.

    [​IMG] \

    The picture above shows distortion in the direction I did not brace on that first successful attempt.

    My subsequent attempts having gotten progressively better as I've played around with clamping pressure and bead density. My fifth and six attempts show very little distortion, mostly at the face of the hammer. I'm not sure if this is because I'm still using too many beads or if it's because this is the one direction I haven't braced yet. I'll continue experimenting and let folks know what I find.

    Here are my first four successful attempts. Lateral expansion is getting better with each try but the face distortion is getting worse:


    Here are two views of my sixth attempt:



    That last one is probably close enough to get the job done. I can clean up the issues post casting. That said, I'm going to keep playing around with additional braces and fewer beads.

    Ultimately I think the process shows potential for complex shapes that don't require tight tolerances and can be cleaned up post casting. A welded metal box around the mold would minimize distortion but is beyond my resources currently. We'll see how the density works come casting time.

    I'm very interested in getting advice from folks regarding orientation and placement of sprue, vents and riser for casting. I'm new to the forums - I assume a separate thread in the lost foam forum be the best place for that?
    Last edited: Sep 5, 2019
    BattyZ, Tobho Mott, Jason and 3 others like this.
  2. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Good work there Cliff. Allow me to give you your first like.

    I do a lot of lost foam casting. If you poke around in the lost foam subforum, I have quite a few project and process related posts there. I fabricate all my patterns from extruded polystyrene. Molding EPS as you have is great for reproducing patterns and complex shapes. They look to have a pretty decent finish. Any idea on the density lb/in3? I use 1-1.5lb/ft3. Less dense it typically better if it suits for sufficient pattern strength.

    You could also cast the cavity for the handle. No core required in loose sand lost foam. Vibration of your flask will help packing and mold stability. You'll also want to coat the foam pattern. Drywall mud works fine. I'd probably just sprue and top feed that right in the face of the hammer head with it oriented vertically......or might cock it 45 degree angle. -More on that when you get ready to pour.

  3. FKreider

    FKreider Copper

    Wow that looks awesome!
  4. Jason

    Jason Gold

    You better send me a hammer or I'm telling your wife! My wife made some Beef Bourguignon the other night in ours. It was amazing. Going to be great when it's cold outside.
    Your EPS foam is a slick idea. With some careful cutting of that foam, you won't have to drill the area out for the handle. If you are fighting distortion, stick the original back in the silicon and make a 2 part mother mold out of some plaster of paris for it. Cheap and quick!

    One thing we know how to do around here is make decent molds of crap. You don't need a welder.
  5. Zapins

    Zapins Gold

    How dense is the foam after being formed? It looks like it might have a fairly hard covering on it?

    I thought this idea came up years ago and nobody could get the beads to try it out? Cool that you were able to and that it worked. Next step is seeing if it burns out cleanly.
  6. Peedee

    Peedee Silver

    The beads have a limited shelf life so nobody wan'ts to hold stock of them.
  7. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    A few years ago when I started lost foam casting, I went the same route as rightbug......watched a bunch of duck decoy making videos and joined a couple of those forums. As mentioned above, the materials can be scarce and have shelf life, but for my needs I ultimately decided it was way more work to make molds to make EPS patterns than it was just to machine and assemble the patterns, with perhaps a couple exceptions.

    If you have relatively small parts, a complex shape and readily available master pattern from which to pull a silicone mold (and can stand the dimensional shrink), and/or want to make a number of identical copies, it's a useful hobby method of making lost foam patterns......otherwise I think it's more suited as a production oriented process. It's also better suited for solid shapes or it further complicates mold making.

    After you get used to it, foam is actually quite easy to work with. I think I could hand fabricate a reasonable facsimile of that hammer head in about 15 minutes from XPS.....probably a half dozen copies in an hour. Would they be absolutely identical? No. But close enough for a hammer head.

    Though I doubt rightbug wants to make a lot of them, the hammer head seems to be a good process fit for the most part. If the parts got larger with more surface area, it would become more difficult to do with just a silicon mold.

    I'm happy to see someone using the process, but for me, machining and assembling XPS patterns is a better fit for the nature of my parts.....and it's fast! I think DavidF doinked around a bit with beads and molding them and ultimately concluded the same.

  8. rightbug

    rightbug Copper

    Thanks for the feed back guys. I'll try calculating the density tonight. (Math! Thank god for online calculators and conversion tools!)

    Thanks for the tip -- Will do. How does coating affect the gases? Do I still need to vent or do the gases dissipate through the coating?

    This occurred to me last night. I realized I was over thinking things and straight down would likely work well.

    Shrinkage is obviously one of my biggest concerns. Do you think I should put a riser above the face to feed the hammer as it cools of will the metal in the sprue be sufficient?

    Oh, she knows! She's been putting up with my experiments on the stove for a week now. "So that's where my canola went!" (The basement. It made it's way back to the kitchen eventually.)

    Ah! Brilliant. Thanks, I may well try that.
  9. garyhlucas

    garyhlucas Silver

    Looks like you should cast your mold in a hinged metal box to make sure it keeps its shape.
  10. rightbug

    rightbug Copper

    Houston, we have a problem. Looks like I've dropped from 9.9lb/ft3 to 6.6lb/ft3 between my first and my fifth complete hammer. Time to see if 2 grams is enough beads to fill the mold. (Though it's likely the shell of un-expanded beads is throwing off my calculations a bit so I might not be as far off as I fear.)

    What's my maximum density before I start having significant problems?
    Last edited: Sep 6, 2019
  11. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Check your calcs. I would think its very unlikely they're that dense. I've never used greater density foam than 1.5lb/ft3 so I cant say what the upper limit may be and results would vary depending upon the part.

  12. rightbug

    rightbug Copper

    I know that the outer shell was throwing off my calculations somewhat but not by enough I'm afraid.

    I'm using water displacement to estimate the volume in ml. The first one was 12g / 75ml. The 5th was was 8g which I thought was a huge improvement until I double checked the volume which was closer to 50ml.

    For this most recent attempt, I weighed out 2g of beads, got nervous it wasn't going to be enough and added a smidge more. The weight is 3g but the volume is closer to 20-25ml. That still works out to 7.5 l/bft3 according to this calculator:


    And, of course, the finish is unacceptable.

    I think I'm going to try one with the shell of unexpanded beads and exactly 2g of beads inside. I won't be able to get an accurate read of the density within the shell but if it works and the volume is closer to 50ml,the interior density will be as low as I'm going to get it. From there it will take a test pour (tomorrow?!) to see how much of an issue it is.
  13. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    25ml-75ml is a lot of variation in measurement....but if you were putting 8-12gm of foam beads in, your patterns are quite dense. The amount of blowing agent can be a factor as well as the initial charge and how much you pre-expand the bead stock. Duck decoys probably shoot for higher densities for durability so you may not have enough blowing agent to expand the lower mass of beads enough to get the desired quality of pattern at lower density.....or just need more pre-expansion.

    Opinions vary among hobbyists about how lost foam castings vent, but opinions don't vary in the commercial practice....there it is all through the coating. Coatings slow down the rate at which gases can vent. Higher density foam generates more gas that must escape. Higher temperature metals evaporate foam and generate gas more quickly requiring higher vent rate and more permeable coatings.

    Many attempt to include a vent straw in lost foam castings but IMO it's futile because the vent is not a vent until all the foam obstructing the path is evaporated and then the vent immediately fills with molten no vent.

    If you were going to experiment, you might try no coating (cant vent much better than that) and as a separate attempt, a thin drywall mud coating. Besides vastly improved surface finish, the coatings also prevent sand from imbedding in the casting.

    My comment was more about the small amount of dimensional shrinkage than shrink defects. If you sprue into the most massive area of the part with a sprue slightly smaller than the diameter of the hammer head, I doubt shrink defects will be an issue.

  14. rightbug

    rightbug Copper

    Proof of concept!

    I was hoping to pour last weekend but my crucible gave me a quick lesson in thermal shock (fortunately before I had the brass in) so today was attempt number two. My biggest problem was that I didn't have enough brass. I also initially thought I might have gotten the brass too hot but the end result isn't terrible so maybe not? I think I'll borrow a friend's IR gun for my next pour. If anyone else is casting brass I be curious to know what what a proper pour temp should be.

    Anyway, it looked like such a small amount of brass and the pour was over so quickly that I initially thought it was a total failure. There was a tiny bit of the pour basin, then empty sand, then the outline of the riser (yes, I went with a riser, I might eliminate it for my next pour) and then a lot of sand. I thought that was it but I was curious to see if any metal reached the hammer pattern so I started emptying out the sand and, lo and behold, I found a tiny snub nosed hammer!


    Here it is next to the original for comparison...


    Here's the good news -- It's solid and while there was some shrinkage and not enough brass to fill the head, it doesn't look like foam density was an issue at all.

    While I was for more brass to arrive I'll try cleaning this up and turning it into a tiny hammer. Here's some very rough cleanup:



    Those voids in the face are just on the surface -- A little more grinding and they'll be gone.

    For my next pour I'll get rid of the riser. Is a pour basin still a good idea or should I just pour straight down the sprue?
    Tobho Mott likes this.
  15. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Indeed. Congrats. That's a decent result for first go.

    Thermal shock? What happened?

    Give it a go but don't get your hopes up. IR guns are unreliable in foundry use. There are many things that effect emissivity and the corresponding readings in the IR spectrum.

    Maybe it's just the picture but the casting looks significantly smaller than the pattern. The barrel of the hammer head itself looks incomplete. Did it fill?

    Not sure I'd say foam density wasn't an issue at all.....maybe, maybe not. Depends on what result is acceptable to you but a lost foam pour should perfectly replicate the features of the pattern, shape and surface condition. Did it?

    Could you tell us a little more about:
    • How did you sprue and position this first one? Take a picture for us next time..
    • Did you coat the pattern?
    • Did the pour flame/smoke and/or bubble/belch back at you while you poured? Video of the pour can reveal more details.
    • What did you use for a pouring basin on this one?
    Pouring basins can help maintain a continuous uninterrupted pour and feed of the casting. Sometimes those surface defects are caused by trapped gases and oxides.

  16. rightbug

    rightbug Copper


    I had read varied advice on preheating the crucible. Some people said they never preheated it and it was fine, someone else said put it in the oven, bring it up to 300 degrees and leave it for ten minutes and someone else said he puts it in the foundry before he turns it on and brings it up to heat with the foundry. I went with that last option and the crucible failed.


    This time I preheated it and let it cool down to room temp and then, just to be safe, brought it up to 300 degrees in the oven before using it and it looks brand new.

    The hammer seems solid (if incomplete) so maybe heat wasn't an issue. It was kicking out crazy gases, there was no slag and it left a spotty white residue on the crucible but, again, the metal in the hammer seems great so maybe this is all normal for brass?

    It did not fill! I'm missing two thirds of the head. More brass is on the way.

    For the most part it's a perfect replica, right down to surface imperfections on the pattern. There is some deformation at the bottom of one cheek but that could be a side effect of not enough brass.

    I took a picture! Try to look at it in your peripheral vision so as not to be horror struck by my terrible foam shaping skills. This is obviously upside down so the hammer itself was gravity fed.


    I read your excellent coating thread and followed your advice. I coated the hammer but not the sprue or riser.

    My daughter has an amazing ability to take pictures that cut off the top of people's head. I let her shoot the video and she missed most of the pour itself. But you can kind of see the start and she got the end.

    As seen above, I just used foam with a little lip. I poured into the large area and let it overflow down the sprue. The more I think about it, aside from not having enough brass, I'm extremely happy with the result so maybe I won't change anything apart from any advice you might have. I've cleaned it up some more and it's turning into a perfectly fine hammer with a tiny little head.

    Thanks a ton for all of your advice and feedback. I wasn't expecting success on my first pour but came a lot closer than I expected.
  17. PatJ

    PatJ Silver

    I use Morgan Salamander Super clay graphite crucibles, and I temper them per the manufacturer's instructions before I use them.
    But after I temper them, then I don't bring them up to temperature slowly.

    I charge the cold crucible being careful not to jam the scrap, because the scrap will expand when it heats up.
    I start my burner at full output (about 3 gallons/hour of diesel) and so the crucible sees full output within 5 seconds of me lighting the burner.

    I have never had a Morgan Salamander Super fail using them this way.
    They seem to be pretty tough crucibles as long as you don't jam them tight with metal.

    I know of others who do iron work that use the same crucible brand, and the same full-power burner startup, and they don't have crucible problems either.

  18. PatJ

    PatJ Silver

    Looks like a pretty good pour.

    I consider a pour a "good pour" when I don't drop the crucible, when the crucible does not accidentally fall out of the pouring shank, when I don't trip over something while lifting the crucible, when the mold does not break open in mid pour, when I don't spill metal everywhere and catch half the stuff in the driveway on fire, and when the ingot mold does not go POP and splatter molten metal all over me and the nearby onlookers.

  19. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Wow, that crucible failure is shocking. I'd put that off on the manufacturer's process having more limited curing.......but, all manufacturers do recommend initial tempering and seasoning after long periods without use. Here's Morgan's recommendations.


    Thermal shock is one of the biggest challenges to refractory materials, especially any material that has not yet been fully cured, and that's why manufacturers have recommended cure cycles for crucibles, castable, and other refractories. There's a certain amount of thermal shock that is unavoidable in a hobby crucible furnace, just due to opening the furnace at temperature. Good news is your ceramic fiber furnace is very resilient in this regard.

    The fellas that pour brass will need to weigh in, but if you saw white smoke before the pour, it was probably zinc boiling off the melt. Could be a bit of superheat of the melt, and can be somewhat managed with cover fluxes, but some of that is to be expected.

    I like your pouring well for green sand casting but not as much for lost foam. It's clear you investigated the subject because an offset pouring basin is good practice.......but, in lost foam, I've found that a large portion of casting defects were caused by the initial molten metal contact with foam in open air, so I try to minimize the amount of foam available above the sprue.

    Why? Because upon initial molten metal contact with the foam all hell breaks loose. At that moment, in free air, the foam not only melts, it reacts with oxygen and burns, which produces even more gas than just evaporating foam. Burning foam produces black smoke, evaporated foam in the absence of oxygen is white smoke which is what you typically see when you demold. All this this creates a lot of turbulence in the pouring well which also promotes the formation of metal oxides in the molten metal which lead to casting defects.

    The other issue with the offset pouring well is mold collapse if you don't keep the well 100% full. In bound sand this is a non issue but in lost foam the sand at/near the surface is not contained and will cave in if the metal isn't there to keep it in place. So what to do?

    Use a pouring well made out of a self supporting material.....a tin can or short section of steel tube is popular (often referred to as a Kush cup), then fill the mold with more sand to the top of the cup. This gives you a buffer of metal above the sprue which is helpful because foam patterns don't draw at an even rate. This also allows the melt to develop more head pressure so the pour can settle into a tranquil mode of evaporating the pattern as the gas escapes through the surface into the mold. This is commercial practice.

    Now some will say nonsense, just use big sprues and vent all the gas through the sprue. It's true that even if you do what I say, you will still likely see some gas blow back through the sprue, but that's only because we're amatures and don't have the coatings, patterns/mold design sufficiently developed in the first shot......but it doesn't mean that's not what we should be trying to achieve.

    The other thing that can help reduce defects is adding a bob (don't coat the bob) at the bottom of the pattern. Why? The initial metal that enters the mold is exposed to oxygen and more prone to containing defect causing oxides. That metals flows through the pattern into the uncoated bob which can pass gas to mold more rapidly than the coated pattern. As in commercial practice, I've built flasks that can add vacuum to help further expel gases through the coated surface.

    Is that a lot of trouble? Yup. -Just depends what you're trying to accomplish. OK, had my coffee to face the day.

  20. Tobho Mott

    Tobho Mott Gold Banner Member

    Wow, that crucible... Was it stored someplace damp for a while, maybe?

    Kelly and I use lost foam casting quite differently - he emulates the pros and uses it to make nice car parts and stuff, whereas because it can be done fast and easy, I do quick and dirty one offs that just have to work in some simple way, as a kind of shortcut. Big soup can kush cups and fat sprues and no coatings for me, it's all very much "backyard" style. That said, after a few years of trial and error, I've finally gotten it working for what I need... almost every time.

    But if I was trying to make very nice parts like I try to do when I'm sand casting? I'd certainly copy Kelly's methods.

    One thing to watch out for with the lost foam is the mold trying to trick you into stopping your pour before the mold is full. Your soup can will start to fill up (assuming you can see into it through all the smoke and fire) like the mold can't take anymore metal, so you stop pouring and start preparing to pour an ingot... Then you see the molten metal in the Kush cup suddenly all drain down into the sand and the sprue caves in, and that's when you know you're boned... You want to be ready to resume pouring ASAP if needed, but eventually you'll get a feel for how fast to pour to just keep up with it and not keep running into that problem anymore. Best practice is to do just one long interrupted pour of course, but keeping a nice big reservoir of molten metal above the sand inside the Kush cup is sometimes a nice insurance policy that can buy you that extra half second you need to resume pouring again and keep that sprue open and choked in a pinch.


Share This Page