Recycling and filling/extending silicone

There’s nothing better than silicone rubber if one wants to make reliable and long-lasting moulds. It is really second-to-none for reproducing detail, in addition to having the ideal combination of flexibility and strength. But silicone rubber isn’t cheap, averaging about £22 per kg and unlike many other materials there are no ‘cut-price’ versions. It’s rare to find one for under £20 per kg (in the UK www.tomps.com may offer the best deals) and even the reduction one would usually expect when buying in bulk isn’t really that significant in this case. Moreover, unless one definitely needs to use a substantial amount all at once, buying silicone in bulk can often lead to wastage because its shelf life is comparatively short.

The common practice of building up a silicone ‘skin’ or layer over a form (rather than pouring a block) helps to reduce the amount and cost. See the series of posts Making a hollow 2-piece cast in fibreglass from August 2012 for an account of this method. But there are additional, perhaps lesser known, ways of reducing the cost which at the same time modify or extend the capabilities of the material. Once cured, silicone cannot be simply re-melted (as vinyl can) but hardened leftovers and moulds which are no longer needed can be put to good use. In short, if these can be granulated they can be added as a filler to newly-mixed silicone.

In the photo below I am using a traditional-style kitchen mincer to turn silicone leftovers into granules. I’d heard about this being done before but didn’t really make much note of it until I came across this YouTube video below. I suppose I just hadn’t believed before that silicone could be granulated so easily! The commentary is in Portuguese, but it is fairly self-evident.

http://www.youtube.com/watch?v=-t6T92DFya4

I then discovered that the practice is featured in a book that I’ve actually got! .. good old Thurston James’ The Prop Builder’s Molding & Casting Handbook 1989. I must confess that I read this book thoroughly a long time ago and probably still owe a lot to it without being conscious of that anymore. But I remember that after first reading I stowed it away with a little frustration, purely because many of the materials he features are either a little out-of-date now or only available in the US. I’m now planning to read it again with a bit more respect!

I bought my mincer ( a KitchenCraft brand ‘cast iron mincer’) at John Lewis in London for £18 last year, but Robert Dyas also has them and there are many online for between £17-£25.

I’ve found that it works best if the silicone is cut (with knife or scissors) into strips rather than chunks and that the receiving funnel is fairly continuously supplied while turning the handle. At times this is tough to turn, but ‘back pedalling’ a little before continuing will usually make it easier. These cast iron mincers usually come with a few extrusion plates with differently sized holes. I assume these will affect the size of the granules but I haven’t tested them yet. According to the YouTube demonstration the granulated silicone can be put through the mincer again to make it finer if needed. Another point to make here is that, although I cleaned the metal parts of the mincer as best I could before using it the action of the mincer appears to darken the silicone granules a little. I think it’s something in the metal itself but, as far as I can tell, it doesn’t affect the bonding of the silicone.

What I found surprising was that this doesn’t just work with the standard softer types (Shore A 20-25) but equally with a harder one. As long as it’s cut into manageable strips, as below, even a tougher Shore A 55 silicone (this is RTV-101 from Tiranti) will granulate without too much effort. Perhaps this is because the harder ones often have a lower tear resistance.

The sample to the left (above) includes a few different brands of silicone which have the same or similar Shore A value, while the pile to the right is just the one harder type. So far I’ve found that granules of different brands or even different hardnesses can be mixed into fresh silicone and the only thing I’m assuming will not work (and haven’t bothered to try) is mixing condensation cure silicones with addition cure (see ‘Vocabulary’). I didn’t wash or wipe the old moulds or leftovers first. Some might have been a little dusty and I don’t think that would make much difference in the end, but as a rule it’s probably better at least to wipe them first.

But to find out more about what could be done I made the following test pieces.

For some of these small blocks I catalyzed an amount of new silicone first (roughly half the volume to be filled)  and poured or brushed a little of this as an initial layer in the Lego containment before mixing the remainder with its equal weight in granules, then adding it on top. The silicone should always be catalyzed before any other additions to it because otherwise these could affect the even spread of the catalyst. If a proper mould were being made this first ‘pure’ layer or detail coat would be essential, to keep any granules at a little distance from the mould surface. With some of the other blocks I waited until this first ‘pure’ layer was cured (i.e. added the recycling mix the next day) just to see if this might make a difference to the bonding of the layers.

Above are the test pieces made mixing ‘like for like’ granules and new silicone together (i.e. recycled and new silicone of the same type; Lukasil 429 Shore A 20), the one poured in two stages (with a day between) and the other all at the same time. There was no difference in feel between the cured blocks and certainly no delamination (i.e. separation of the layers) or cracks appearing when strongly flexed.

But the recycled granules do seem to have quite an impact in terms of flexibility! I made a test block of the same size using the Lukasil 429 pure to compare this, and the first photo below shows how this can be flexed with ease compared to exerting the same amount of pressure on the recycled block below it. Although the recycled granules had the same Shore A value the block was much tougher.

The samples were made on a perfectly smooth ceramic tile surface and the blocks match this without a blemish when relaxed, but I did notice that the underlying granular structure is noticeable when flexed (if this comes across at all in the photo below). I would assume however that this wouldn’t affect the surface during casting.

Below are similar flexing tests on samples made using granules of the harder silicone (Tiranti RTV-101 Shore A 55) in which the layers can be better seen. Although I expected more problems since the silicones are not ‘like for like’, again there was no fissuring or delamination even under strong flexing. Also as expected the block was tougher than the previous one but, interestingly, not by that much considering the RTV-101 is so much harder.

While I was about it I thought I might as well see what happened mixing other materials with the silicone instead of the recycled granules. It provided an interesting comparison and led to a surprise discovery. The sample block below contains cork granules and, as expected, the silicone doesn’t bond with them. The block started to fracture when flexed.

But then I tried Fillite and the results were different. Fillite is an industrially produced ash commonly used as a filler for resins and plasters. I use it a great deal for extending polyurethane resin.

As before I made one block laying a thick detail coat of pure silicone first which was left to cure and for the other I just poured in the mix. I mixed almost the same weight of Fillite to silicone (actually 10g of Fillite to 12g of silicone because it already seemed quite a lot by volume). It mixed in very well and the mixture was still pourable though very thick. As with all the other tests, there was no effect on the normal curing time.

On curing the silicone appeared to have bonded firmly with the Fillite, including the Fillite-filled layer with the pure silicone layer, with no sign of fracturing when flexed. In the process the composite had become substantially tougher, changing the Shore A 20 silicone into, at a guess, a Shore A 70 or more. The block composed wholly of silicone/Fillite mix above was very resilient, very slightly flexible but quite difficult to bend.

The Fillite and silicone had fused so well in fact that I suspected that some form of bonding was taking place at a molecular level i.e. not just ‘mechanical gripping’ but rather that the silicone polymer chains were linking directly with the Fillite. I have to say that before this I had been largely ignorant of what silicone rubber (i.e. what we buy off the shelf) actually is! I know at least a little more now. For example .. yes, ‘silicone’ is related to ‘silica’. Silicone rubber is derived from quartz sand and one reliable scientific source I looked at defined silicone rubber as ‘made from crosslinked silicon-based polymer reinforced with filler’. So it makes perfect sense that the long chain molecules of silicone rubber .. the flexible version of silica .. will attach themselves naturally to any less flexible version of silica, but because they are long will still keep much of their flexibility. Incidentally, this is also the reason why silicone rubber, which owes its popularity to the fact that it will not stick to most things, will stick fast to clean glass! Fillite is, as I now know, alumino-silicate, and the silicone rubber we buy off the shelf contains similar fillers as standard. Cabosil for example, which Thurston James also mentions as something which can be used to extend or thicken silicone, is so-called ‘fumed silica’. The silicone rubber manufacturers I looked at list fumed silica, precipitated silica, aluminum silicate or alumina trihydrate as their standard fillers.

What I’m suggesting this means is that one can modify any standard silicone rubber oneself, without fear of ‘weird’ consequences, whether you want to economise by extending the volume or whether you just want to increase the hardness. It’s what the manufacturers do anyway!

To sum up

My tests were not 100% scientific i.e. not rigorously controlled and I only tried a few variations, but at least they gave me some indications of possibilities and drawbacks. For me they definitely suggested that;

– addition of silicone rubber granules will toughen the resultant composite even if both the recycled and the fresh silicones used are relatively soft and of the same type. I would assume that when a fresh silicone with a low Shore A value (soft) is being used this will be toughened up in proportion to the amount of recycled silicone added, although I haven’t tested the range specifically.

– recycled silicone and fresh silicone don’t necessarily need to be ‘like for like’ or the same hardness to bond together

– the old silicone moulds or leftovers don’t need to be rigorously cleaned if they’re going through the mincer anyway.

– there was no noticeable effect on the normal cure time of the fresh silicone. There was no obvious effect on the general strength of the cured composite but the effect on tear strength in practice and in the long term would have to be further examined.

– if other fillers are used they should be ‘silicate’ and there is a range of these available. Fillite is one of the cheapest, will bond firmly with fresh silicone and, dependant on the ratio, can be used to adjust the flexibility/hardness of the composite from just a little tougher to almost plastic hardness. I’m assuming that fillers will always alter the flexibility of the cured silicone in some measure.

– according to all my tests so far ‘silicone will bond with silicone’, whether the first layer is still fresh or tacky or whether it’s been left to fully cure (as long as the surface has been kept clean and free from dust!) and this applies equally between different silicone brands, different hardness, filled or unfilled. It is however unlikely to apply between condensation cure and addition cure silicones!