Tales from the furnace – a DIY casting blog

To not deviate too much from the scope of my main blog, I decided to put my DIY casting venture into a separate blog. As a brief overview, I am planning to duplicate at least parts of my green stuff / Procreate sculpts in metal. Depending on how successful the initial experiments are, I may decide to exclusively cast them myself or enlist a professional casting company instead. This is not so much a financial decision – in fact, letting professionals cast them for you is far cheaper than I anticipated, not mentioning the likely higher casting quality, they use spin-cast moulds after all). The main reason I am interested in doing it myself is the flexibility it gives me to quickly reproduce intermediate sculpts and projects of less interest to the general public.

My goal for this blog is to document my journey as well as possible. This will make this thread rather technical but will hopefully help other interested people understand decisions I made, avoid problems I encountered and make them feel emboldened to do it themselves as well. Part of this is due to the general lack of information on this topic online. Many of my sources are 20 years or so old and could disappear by any day from the interwebs. 3D printing certainly plays a role in this in more recent times. I think of it as both a blessing (allowing people to create custom miniatures with such ease) and a curse (for traditional miniature companies but also for hand-sculpting and manufacturing supply mainly). Ideally, I would first digitize my sculpts and make them available to the wider public. But scanning technologies so far have not caught up with printing when it comes to affordability…

For the ease of navigation, I decided to provide a table of contents (and also add a placeholder post for images and such because I regretted that I did not do this on my main blog).

List of posts

7 Likes

[Placeholder]

Choices (An introduction)

So the first question: why metal? Short answer: I don’t like resin. Slightly longer answer: Resin also has arguably long curing times and moulds wear out rather quickly – which is expensive in the long run. This would be fine for a few casts but I would be forced to look for alternatives when casting a whole army. Resin also sounds annoying to handle, store and dispose of and is apparently susceptible to environmental factors such as humidity and temperature while curing. I don’t see dedicated equipment such as a pressure / vacuum pot as a big disadvantage as metal casting also requires its own equipment.

Metal miniatures are traditionally cast in either vulcanized rubber or silicone, usually with the aforementioned spin-casting technique. Here, centrifugal motion forces the molten pewter into the mould allowing for more intricate designs. Spin-casting machines however are expensive and I am not yet willing enough to build one myself. The main advantage are the high throughput and more complex designs you can cast. Gravity casting is possible as well, however and Admiral told me that at least one of the companies he worked with use this technique. Knowing the limitations of gravity casting, I designed my miniatures as compact as possible.

I also want to add at this point that you can decide to use either leaded or lead-free pewter. Leaded pewter has arguably better flowing characteristics and a lower melting point which in fact seems so easy that almost everyone is able to do that (keyword: Prince August). Leaded pewter also contains… well… lead (these days usually <15% – just to put this into perspective: many brasses, steels and aluminium alloys also have a low percentage lead content). Interesting history lesson I learned while researching alloys: the US has effectively outlawed lead in miniatures in the early 90s (around 1991-1992) forcing US GW there to adapt much earlier while they continued until late 1997/1998 to use lead in miniatures sold on the European market. They had a short transition phase where they experimented a bit with alloys (my 2nd edition Necrons look different to earlier and later miniatures) before they settled on the white metal alloy they use to this day. I settled on the potentially more difficult lead-free route, more specifically the alloy Sn95Sb4Cu1 (more on alloys in a later post) – if it goes well there is little reason for me to change alloys and if not I can still think about the alternatives. A quick note on metal prices. Since the beginning of the Covid pandemic, raw tin prices have tripled. If this trend continues, resins or thermoplastics may become more attractive for miniature manufacturers. I bought my alloy days before the all time peak in early March (sigh) from a local company specialising in casting materials.

Next, we need to decide on the mould material. A quick internet survey seems to suggest to use two-part silicones, so called room-temperature vulcanizing (RTV) silicones, more specifically variants that are designed for casting metals (Smooth On’s Mold Max 60 for example – conveniently, these silicones are often red but I don’t know if this is convention or due to additives such as iron oxide). Their advantage is that they don’t need a vulcanizer. This is a device that evenly delivers both heat and pressure to cure the moulds. Their disadvantages are that they are susceptible to entrap bubbles (and bubbles in moulds are arguably worse than in the model). So you need to make sure to adequately degas them (or compress the bubbles – I always forget which one is the one you want to use with silicone). If you go the resin route, chances are, you already bought similar equipment for removing bubbles from the pours but for me, that would imply more costly equipment. The industry also does not use them and I suspect this is because of their limited lifespan (i.e. in service but the opened bottles also have a rather limited shelf life) and relatively high cost. Enter high-temperature vulcanizing (HTV) silicones and rubbers. On the first glance, this looks unreasonable. Vulcanizers start in the higher three figure range (Euro that is). On a second glance, vulcanization is not rocket science. Delivering heat in the 150-200 degree C range evenly is something we all need in an ordinary household oven, otherwise our cakes would turn out quite bad. Pressure can be supplied by clamps. This is not a hack I made up but something that has been proven in practise and even HTV silicone manufacturers suggest this for DIY purposes. However, I would not advise to do this in your household oven as silicones can smell quite badly and are generally not food-save.

Unfortunately finding advanced information online can be somewhat difficult. Good starting points are this video by Tom Mason and the rather small but incredibly worthwhile YT channel of Making Miniatures. Beyond that there is not much quality content on YT. (Semi-)Professional jewellers are our best friends here as this is the prime audience these silicones are targeted at and you find most advanced information on forums dedicated to making jewelry. You will quickly find that the gold standard silicone manufacturer seems to be a company called Castaldo – we are interested in their Econosil branded silicone as this can withstand higher temperatures (incidentally the same material Tom Mason uses – I discovered his video after I made my purchase). The first real challenge is to acquire said silicone. If you are outside the US or UK, this may become somewhat frustrating. While I found a local supplier, their prices were a bit high for my standards. Then I noticed that the jeweller’s community in Eastern Europe seems to be way more healthy and alive with the result that I found several Polish and Czech companies selling Econosil. After many emails, partly machine-translated, I found a very friendly Polish dealer willing to send one package (2.25 kg) to Germany, at almost half the cost of the previously found German-based listing, including express shipping! The package arrived within 3 days. So if you live in the EU, highly recommended to check sellers in other EU countries!

As for dedicated equipment: I bought a toaster oven, several tools (which also may come in handy for other DIY projects, so I don’t count them), a solder bath and a laddle made for casting lead bullets I believe. So overall, this seems to be in a similar price range to casting resin.

Having acquired all the raw materials, I will discuss how to make a mould frame in the next post.

5 Likes

Making a mould frame

Vulcanization not only requires even heat but generates a lot of pressure. This is due to the material expanding as it cures. If you think about it, this is a good thing because it allows the silicone will fill out all the nice details you have spent so much time sculpting. But you don’t want it to expand too much because that would make the miniatures noticeably smaller than the green stuff master. As such, silicone manufacturers usually list shrinkage on their website. Econosil, for instance, has a shrinkage of 1.1%. We therefore need to counteract this expansion by using a robust metal frame made out of aluminium or steel. A second advantage of the thick metal is that it more evenly distributes heat (think of the differences between a cheap thin-bottomed cooking pot and a more expensive, heavy one). We should try to keep the material thickness even, however: if one side is more thin than the other, this would imply an uneven vulcanization rate – this can result in more pronounced mould lines when one mould half cures before the other.

If you live in such a forsaken part of the world as I do (for mould making at least), chances are that mould frames are either prohibitively expensive or connected to import from abroad and the high shipping and tax rates that comes with. I was not willing to spend 50 euros or more on a piece of machined aluminium after having spend already so much on raw materials and other parts for this project. It’s just a square hole in a block of 30 mm thick aluminium, right? So enthusiastically, I bought some cheap scrap aluminium at different thicknesses (20, 25 and 30 mm) and started drilling holes into the thickest one (I bought a used drill stand along as well). Well, I soon did not feel too optimistic about my consumer-grade drill surviving this if I were to drill holes around the whole perimeter (it’s 30 mm thick metal after all).

So I settled on a minimum amount of holes and used two hand saws to remove the rest (first a fret saw to connect the holes and then a 300 mm metal saw). This took a few hours in total but I did a bit at a time, so I was finished after a week or so. The result is a rather crude but hopefully functional mould frame. I could not be bothered to make it pretty. I made sure to deburr all the sharp edges such that there is no risk of accidentally cutting myself and that the surfaces where the holes are were planar. I will use C clamps to attach the 8 mm thick aluminium plates on both sides, hopefully creating a more or less sealed enclosure (I think heat transfer is more important than an air-tight seal here).

I also bought some square tubing I intend to test at a later point (I will post a picture at a later point). If the wall thickness is enough (around 5 mm I believe), this would be a far easier way of creating mould frames of different sizes. However, I did not want to introduce more potential failure points right in the beginning which is why I took the more labour-intensive approach described above.

If mould frames are affordable at your place, good for you – then I would suggest you buy them instead. A note on commercial mould frames. Some models feature a small channel along the sides to allow for excess silicone to ooze out. The presence or absence of these channels seemed to be a rather controversial one in the past if they are necessary or not. Quoting a Castaldo representative, the industry generally does not use them but some jewelry makers like them because silicone oozing out of the mould frame looks rather satisfactory. The general consensus seems that there is no clear argument for or against these holes, so they seem to bring little if any benefit. I therefore did not add them for the simple two reasons that (i) you probably don’t want silicone potentially messing up your oven and (ii) I was frankly too lazy to drill a single further hole.

Next, I’m going to do some vulcanization experiments in a toaster oven.

4 Likes

A first foray into mould making

For a first test I figured that my anatomically not quite correct arms may be good objects of study. First, they are relatively planar avoiding advanced mould line engineering. Second, they are made out of green stuff, my putty of choice these days. I want to see how green stuff and its details survive under 165-177 degrees C (320-350 degrees F) – the vulcanization temperature of Econosil. Experiments I did a few years back with green stuff under a FIMO coat showed that it seems to partially degas leaving a crumbly dark green residue. This was several tens of degrees lower, though (I believe you bake FIMO at around 110 degrees C). Third, the test also contains a weapon made out of polystyrene, another material I want to test how it behaves under heat.

If these tests are successful I would move on to a big hat (a hat that is, not the dwarf). This test piece introduces a new challenge and a new material: slight undercuts in the top section where the feathers are – and the material is a mix of green stuff and Procreate, a putty supposedly more stable under vulcanization temperatures. We will put this to the test. Here I need to be more clever about my mould line placement and slightly angle the piece.

But on with the first test. I placed the mould frame onto a 8 mm plate, generously coating everything with baby powder (talcum). This will hopefully prevent the silicone from sticking to the crudely worked mould frame. Then I applied the first layer of Econosil – you can simply cut the strips with a scraper, I used a piece of styrene sheet for a lack of a better tool. This stuff has a consistency comparable to freshly mixed putty, I would say slightly more stiff then Green Stuff – in line with Procreate. I made sure not to get baby powder onto the first layer. Then a second and a half layer to fill the mould frame half-way. I pressed everything down to fill all the gaps. Then I placed the masters in a way that they would fill without leaving cavities (that is the top to bottom direction in the picture).

Perhaps I need to add a vent or two to the weapon’s arm. Vents are there to avoid creating said cavities – the air can escape through the vent instead allowing pewter to fill this section. I then placed several so-called mould locks across the mould half, in my case IKEA-like shelf retainers (these pins you put into Billy-like furniture). They force the mould to lock into a specific state. The only thing worse than pronounced mold lines are shifted mould halves – that’s pretty hard to correct (looking at you 90s GW plastic minis). People also use acorn nuts but these pins were significantly cheaper for me and narrower as well, so I bought them instead – the thing you should note, they remain permanently attached to one mould half. After having added the mould locks, I gave everything a dusting with baby powder and added the remaining two-and-a-half layers of silicone. Once again, try to avoid getting powder in between the layers. Then I pressed everything gently down again such that all gaps where filled. The nice thing about Econosil is that you can simply add a bit more at a location that does not have enough material. When everything was roughly flush with the mould frame, I powdered everything again and placed the second plate onto the frame. Lastly I tightened the clamps.

I had already pre-heated the toaster oven to 170 degrees with air circulation. Castaldo advises to factor in 15 min vulcanization time for each 6 mm, that makes 75 min for 30 mm. It took 20 min to reach the terminal temperature again (the cold metal apparently cooled down the oven quite a bit). Then I left it in there for about two hours after that to account for a small delay in reaching the same temperature inside the mould. A note on oven temperatures – reading online reviews, some models are apparently rather unreliable. I would suggest to buy a (good!) oven thermometer as well. Those are well-spend euros – nothing would be worse than a failure at this point because the toaster oven temperature control decided to wreak havoc. I was fortunate enough that my oven was bang on. Also: Factor in the size of the clamps when you buy a model. After around 30 minutes after putting the mould in, some silicone starting oozing out at the only corner where I had not placed a clamp. I am glad I did not use less clamps as this showed me how much pressure was building up in there. After two hours, the oven rang a satisfactorily sounding bell – I did not immediately take it out but let it cool down by itself. No idea if this is necessary but I figured that preventing thermal shock at this stage cannot be wrong. At the time of this writing, I am still waiting for the mould to cool down (and the kitchen to become less smelly), so I am excited about the outcome.

After 2 hours the mould frame was cold enough to carefully remove it from the oven. The clamps had decided to become permanently attached to the aluminium plate – likely the lacquer melting – so I left them on there for the time being. The rubber easily removed from the frame and after some well-placed surgical cuts with a scalpel I was able to carefully tear open the mould. The result looks for the lack of a better word… amazing. The rubber looks cured and preserved the details as good as I was hoping. The green stuff survived much better than I anticipated, in fact, it looks exactly as before albeit being much darker. I cannot tell any degradation and it is not crumbly. This is great news because that means I can potentially use my green stuff masters more than once! The styrene is a different story. It partially melted / deformed and broke when I lifted it up. Lesson learned: avoid styrene in sculpts for moulds that are going to get vulcanized. It’s not too bad though and I need to make some test casts before I allow myself a final verdict. As for the mould line, the masters submerged more than I was intending into bottom half – a lesson from this could be that I should not press them in as much as I did. Some talcum had collected in spots such as the weapon arm (close to the shoulder and at the tip of the molten axe head)-- fortunately only in uncritical areas but something to look out for – it can potentially create bad mould lines. There are some alternatives to talcum such as mould release spray which I may order and test at some point.

Next, I should probably cut sprues, vents and do some test castings…

8 Likes

I would also recommend videos by Crafsman, he’s done a lot of videos on making molds and casting in various materials that you might find useful.

3 Likes

Wow, amazing stuff and a wealth of information.
This is true dwarven daemonsmith at work.
I’ll be following this with interest!

3 Likes

@cornixt Thanks! A great channel with casting methods across the board! I watched his video on casting pewter into RTV silicone moulds a while ago but completely forgot about it after I decided to use HTV silicone. I can definitely credit him for giving me confidence that lead-free pewters might just work fine. It would be interesting to compare RTV to HTV silicone in terms of ease of use, quality and long term wear at a later point.

A sidenote since I just watched his video on thermoplastic injection moulding with 3D printed moulds: I am somewhat surprised that we have not seen more hobbyists so far attempting hard plastic miniatures. Injection moulding machines are rather simple devices and we don’t really need steel or aluminium moulds for extremely limited runs. After all, if it wears out, we can simply a print a new one.

Another honourable mention with useful information on toaster oven vulcanization and mould making is Craig Dabler’s channel. His content is mostly jewelry-related and for a large part intended for casting wax into HTV silicone moulds (to be later used for lost wax casting). There are some videos on Econosil (and the silicone he sells on his website) and he did cast pewter directly on at least one occasion. On the negative side, I find his content somewhat uncurated: Since the video titles all sound similar I have difficulties finding something specific (or finding useful bits again). On his website he provides detailed information on how to vulcanize moulds in a toaster oven.

@Fuggit_Khan Thank you! :hatoff:

3 Likes

Vents, vents, vents

It’s casting day. So I melted down two bars (250g each) of the tin alloy Sn95Sb4Cu1 (that is 95% tin, 4% antimony, 1% copper) and set the temperature dial to 350 degrees C (that’s the maximum rating of Econosil). Then I pre-heated the mould halves in the oven to 100 degrees C.

For casting I used a different talcum powder I bought along with my metal specifically designed for casting (according to the label it also contains club moss spores). Just to be on the safe side, I wore a dusk mask as talcum can be contaminated with asbestos if it’s not cosmetics-grade. After every second pour, I powdered on new talcum. Quite obviously you should also wear eye protection and some gloves when you handle molten metal.

On top of the molten metal (inside the solder bath) you will see slag collecting – you need to scrape that off with the laddle until the surface is shiny and clean before you start pouring. I used two plastic clamps and a thin MDF sheet on each side to hold the mould halves together. After a few test pours, I realised that air could not escape, so I started adding more and more vents until the metal was able to fill out the mould completely. Apparently, the temperature of the mould plays a role how successful your casts will be: if it is too cold, the metal will get shocked by the temperature difference. So it took a few casts for the mould to become warmer and to achieve better results.


Result of an average cast

Overall, I am very happy with the outcome for a first attempt. The muscular details get reproduced quite nicely across all three arms. After a few attempts, I was able to get the hands on the two individual arms with all fingers cleanly cast – on one hand even with a subtle indication of nails. The axe came out rather nicely – I like it more than the deformed and broken master. Most of the casts feature almost no mould lines and if they do they compare to what you get from commercially available miniatures. However, I find small details (such as the nails and the sharp transitions on the axe head) somewhat washed out – perhaps the reason I don’t get many mould lines: those are details after all. I don’t know if this is due to small air pockets, the temperature differential between mould and pour or something completely different. If it is the former, this may be hard to overcome with gravity casting. The latter could be solved if the mould reaches working temperature. Some people recommend tapping the mould right after pouring but I’ve also read that you should not do that. I’m going to test that more systematically in a future casting session.

It should be said though that these arms are perhaps not the best objects for studying the degree of detail, so I’m going to go ahead with the chaos dwarf hat as initially planned (if it does not work out, I can still send in the master to a commercial caster). Also: I am pleasantly surprised how fast you can cast. If you figure out where to place the vents (like on the arm in the middle), you can churn out quite a few good ones in a short time.


All casts of the day.


Detail on the first arm: most successful casts. After a few attempts, all fingers were cleanly cast, even some indication of nails.


All weapon arm casts: either the arm or the weapon were cast completely. In the beginning I did tap the mould which seems to have resulted in more successful axes but without the arm. Without tapping, the opposite result. But I also added more vents, so this likely contributed to the outcome.


The only complete weapon arm. You see the washed out detail on the fingers


Detail on the third arm: I forgot to add another vent at the triceps here causing miscasts due to an entrapped air bubble

4 Likes