Casting of Violin Parts Using Silicone Rubber

Transcript of a presentation given at the First Australian Violin Makers' Conference, Guildford, Victoria in March 2003

Introduction

This afternoon I am going to show you how to make a silicone rubber mould of one side of a violin scroll, and use it to produce a plaster cast. You will be able to use this technique to make permanent copies of parts of instruments and bows for your own reference and study. But also, and more importantly, we are hoping that you might be able to make up reference material to share with your colleagues.

I will just begin by making sure you are clear on the terminology. By the way, there is a transcript of this talk so you should not need to take notes.

Terminology

The first stable rubber compounds were made by adding sulphur and heating, and this process was called VULCANISING (as in volcanos). This chemical reaction was discovered in 1839 by a man called Charles Goodyear. Dow Corning started to manufacture silicone compounds in 1943, and the SILICONE RUBBERS (such as SILASTIC which is a Dow Corning trade name) were known as RTV (or room temperature vulcanising) rubbers. This is a rather paradoxical term as it really means "cool heating".

The Silastic rubbers work like any of these 2 part synthetic compounds - you mix some of the BASE with a certain proportion of a catalyst or CURING AGENT and this starts a reaction to set the rubber. There are two types of reaction - ADDITION and CONDENSATION (don't ask me what the difference is - I haven't managed to get a clear idea myself). It is possible that the presence of a third chemical in any of the materials you are using can interfere with this reaction, and this interfering compound is called an inhibitor and the problem is called INHIBITION. The addition reaction rubbers are much more susceptible to these problems than the condensation type. One of the possible inhibitors is sulphur, which is ironic as the original vulcanising process required the presence of sulphur!

I should also point out that the term MOULD is the form made from a shape (so it is a sort of negative) whereas a CAST is the positive shape which is formed by filling the mould, so it should obviously be a copy of the original. Also note that the American spelling for mould is MOLD - so keep these terms (and the alternate spellings) in mind if you are searching on the internet for information.

Instrument Safety

The usual precautions apply in handling and working with fragile and valuable instruments. You should work within the limit of your expertise and experience. You should have some skills in assessing varnish stability and recognising potential problem areas such as loose, flaking, cracking or unstable varnish (both original and retouching) or open cracks, and so on. You will also hopefully have some knowledge of the varnishes of various makers - for example, I would be very concerned about doing this on an A.E. Smith violin, because his varnish can often be extremely fragile.

As in all things, the ONLY approach is to experiment on a worthless instrument until you feel completely comfortable with the technique. However, feeling comfortable with the technique will not cover all the potential problems you may have with different varnishes. Practising on a worthless instrument will not tell you how it will work on 1700's Italian varnish.

I have tried to work out a technique that gives good results and is fairly simple. I have not tried it out a lot of different instruments, and certainly on nothing valuable. I have also only worked with a few Silastic products and there may well be a better, safer alternative.

I think it would be very advisable, as a general precaution, to use an isolating layer between the Silastic and the varnish and for this purpose I have been using a microcrystalline wax product called Renaissance Wax. It is available through Alex Grant Violins in Melbourne, and there is an information sheet available. You should also be aware of the limitations and precautions to be observed in the use of these products.

Also it may be a good idea to mix a small amount of Silastic and place a test spot on an inconspicuous area (such as under the chin of the scroll) to test the effects. Use a hand lens to examine the area and evaluate the varnish.

Remember - this in not essential repair work, like fixing a soundpost crack - it is an optional procedure that is useful but not absolutely necessary. So if you are in any doubt, don't try it!

Finally, it is of course important to get the owner's permission before taking a mould, and you should carefully advise them of exactly what you will be doing, what the risks are and why you consider it to be safe.

Equipment and Materials

I have tried to use materials that are readily available and avoid the use of any expensive equipment. However there is one item that is very useful and since it has many other uses around the workshop I think it is well worth the investment. It is a small electronic kitchen scale like this one from ARLEC. You can also use it to weigh bows, violin plates, varnish materials and so on and it is very fast and accurate enough for this sort of work. Check out the specifications before buying one though, as different brands vary. This one measures to the nearest gram and has an accuracy of +/- 0.5%. Apart from that all other materials and equipment are cheap and easy to obtain - except, that is, for the Silastic.

Properties of Silastic

There is whole range of Silastics made by Dow Corning, and I have only worked with three of the possibilities. You can see a complete list and get all the data sheets from their web site - www.dowcorning.com. If you search on Silastic you will get about 80 products that are available in Australia. They all have different properties (such as viscosity, hardness, elasticity, etc) that have been tailored for different applications. The ones I used were very elastic, with a high resistance to tearing. This makes them ideal for moulding intricate shapes, and they can even handle fairly severe undercuts. They all have excellent release properties.

The first product I tried was Silastic 3481- it is one of the most widely used for general artistic mould-making applications. You can buy it in a 1.05 kg pack which contains 1 kg of the base and 50g of the curing agent (which is called 81). The cost is around AUS$45 + GST for the pack. It is a condensation cure product, so it is very reliable and I had no problems with it not setting.

It has got a use-by-date and this is stamped on the box. Unopened it has a shelf life of 9 months. The shelf life after opening is supposed to be 1-2 months. However I mixed up some old material (opened and 18 months past its use-by-date) that was given to me and this was the result. It was thicker than normal when it was mixed but, as you can see, it still sets. Once it is mixed, you have a working time of 90 - 120 minutes. Under normal conditions the Silastic will be fully set in 24 hours and will achieve full strength in 7 days.

It is possible to speed up the reaction time by using more than the required amount of catalyst, but of course this means that you will end up with some of the base left over - I haven't experimented with this at all. Also you can get a fast catalyst (81-F) which reduces the setting time to 5 hours, and even a very fast (81-VF) for a 90 minute setting time. The shrinkage rate is up to 0.5%, but it can increase with age and use of the mould, possibly up to 2%. In general, the addition cure products have a better dimensional stability and a lower shrinkage than the condensation type.

The 3481 was very easy to use, and gave great results. HOWEVER, I did find that after doing 4 or 5 moulds it was having a noticeable effect on the varnish of my test violin. The MSDS sheets for the curing agent state that ethanol can be formed when it contacts moisture, so it is quite likely that this was causing a problem.

I then tried Silastic E which is the one that Gregg Alf used, and is discussed in his article (which is a transcript of a talk he gave to the VSA). The article can be found on his web site, and there is a link to it on my site. This cures by way of an addition reaction and so there are more limits to the materials that can be used in contact with the rubber. For example, Gregg Alf talks about the need for an inhibitor-free clay and a special non-inhibiting tape for masking off the f-holes and endpin hole. So just be aware that if you are using any of the addition cure Silastics you will need to be much more careful about testing your materials, and also very consistent with your technique. As an example, I was using commercially available play dough and I found that the blue was OK, but the red play dough inhibited the setting - that's how unpredictable it is!

For some reason, this product is supplied in Australia in a 1.1 lb pack (approx 500g), and these cost AUS$37, or $41 with GST - so it works out to be more expensive than the 3481. The working time is 90 - 120 minutes - the same as 3481, and it cures in 18 to 24 hours. It has a slightly longer life with a use-by date of 12 months. The shrinkage rate is better than the 3481 at 0.1%. The main drawback is the very high viscosity - 55,000 mPa.s compared to 21,000 for the 3481 - which I found made it more difficult to use. Also it is harder (37 Shore A) so there is a bit more potential for damage in de-moulding.

I then tried another addition cure product called Silastic S which had properties much closer to the 3481. The viscosity is 13,500 mPa.s and it has similar mechanical properties. But a big bonus was the much faster curing time - only 6 to 8 hours which means that the Silastic is in contact with the varnish for a much shorter time. The working time is 40 - 60 minutes. It was available in a 1.1 kg pack which cost AUS$56 or $61 with GST.

There is a list of Dow Corning suppliers in the handouts, and also the product information and MSDS sheets for each of these three products. The technique I am going to show you is basically the same whatever you are using, so I will use the 3481 for the demonstration. But before we go on to the practical part, does anyone have any questions or comments about what I have covered so far.

Making a mould support

The first step is to make up some sort of support for the Silastic and it only needs to be a simple wooden box like this, with a cutout for the neck. I have glued and nailed the sides, and I will tape on the bottom with duct tape so it can be taken off to get the mould out. The actual dimensions are not that important - but it will of course determine how much Silastic is needed for each run. You need a bit of thickness around the sides to give some stability - Dow Corning suggest at least a 1/2 inch.

By the way, I have made the cutout for the neck off-centre. I realised afterwards that this means that this box is only good for doing one side of a scroll. It would be better to make the cut-out in the centre, then you can just tilt the head one way or the other.

You can work out roughly the amount of Silastic you will need by setting up the violin in the mould box and filling it to the required height with a granular material. I used sand, but you could use rice, etc. The internal dimensions of my box are 80 mm x 130 mm x 50 mm deep and the volume I needed came to 375 ml. You get 1 kg of Silastic 3481 in a pot and the volume of this works out to about 800 ml. This means I could comfortably get two half scrolls from one pack. The box probably doesn't need to be quite this big.

Make the box fairly accurately - this is useful for doing the plaster casts later as the mould can be reversed in the box to block the neck cutout. I just slip a piece of plastic down the side to separate the plaster from the wood. The construction material isn't important - I have used pine and MDF. Tape the bottom on with a good solid tape, and then wipe the inside surface with petroleum jelly.

Preparing the instrument

As I said before, I am still quite nervous about the effects of any of these products on varnishes. I think it is generally advisable to put a good layer of a protective coating/release agent onto the area that will be in contact with the Silastic. And of course you should test whatever you are using.

The pegbox and pegholes also need to be filled and for this I used play dough. It is water-based and cleans up very easily. You can use the commercially available variety, but it is very easy to make and I have included a recipe in the notes.

I use a piece of cling film to encase a small roll of play dough which is packed into the pegbox. This just helps to make the clean up job a bit easier. Pack a bit more on top but keep it below the level of the pegbox, so that the shape of the pegbox walls will be defined and you can get an idea of the wall thickness, etc. Fill the inside first and then add a bit more to each peghole on the appropriate side. Use your finger to make a slight depression so that the pegholes will just show up nicely.

Now, work out how to support the violin with the neck set into the mould box. You will need to pack up the mould box to get it to just the right level. Now we are all ready to go.

Mixing and Pouring the Silastic

Good preparation is the key to success. It will make things much easier if you have everything ready to hand and the setup is completely organised BEFORE you mix the Silastic.

The main problem you will encounter is air bubbles, and there are two types - small bubbles and large air pockets. The mixing process will introduce air bubbles into the mix. The higher the viscosity the more pronounced they will be, so in the Silastic E mould I passed around you may have noticed they were quite large. The recommended procedure is to remove them under vacuum. For those of you with your own personal vacuum chambers (and I am guessing there are not that many) you will need to use a large container because it will expand 3-5 times in volume before it collapses again. You should vacuum for a further 1-2 minutes, and then inspect for bubbles. For the rest of us, it is just a matter of mixing the 2 parts together as thoroughly but as gently as possible and trying to minimise the formation of bubbles.

The other type of bubble is caused by air being trapped in small spaces when the Silastic is poured in. This is a significant problem with scrolls as they have these nice little undercuts in the turns of the scroll that will readily trap air. I will pass around one of my early attempts - you can see small bubbles which are in the mix and these are not really a problem, but the 2 big ones that are in the turns of the scroll are obviously not acceptable.

The way to get around this is to do a two part procedure. The first step is to mix a small amount of Silastic and brush it onto the scroll. Wait until this starts to set, then put it into the box and add a second batch to do the rest of the mould - and that is what I will do now...

The health and safety requirements are not especially stringent - Dow Corning recommend good ventilation and eye protection as the main requirements. It is probably also a good idea to wear gloves.

Dow Corning also suggest stirring the base before use - especially if it has been stored for a while. Note that ALL stirring should be done slowly and carefully to avoid adding more air bubbles. So we weigh out a small amount of the base - 100 g ..... and then add 5 ml of the curing agent. Stir gently and make sure that the catalyst is mixed through. I have read different ideas for the best shape of stirrer to use - some people suggest round, others flat. My feeling is that the flat stirrer mixes more thoroughly, but gets more air into the mix. By the way, the Silastic S is good in that the curing agent contains a pigment so you can easily see when you have complete mixing.

Dump all of this mix into the mould box. Now, leave it until it starts to get quite thick but can still be brushed - it will depend on which Silastic you are using how long this takes, but it can be an hour or more. Then paint some onto the side of the scroll using a disposable brush. Don't go too close to the edge anywhere as it will still probably flow down over the sides. You just want to make sure the scroll turns are as full as possible without it running everywhere. Now you need to wait until it has really started to set, so that it is no longer flowing. I prefer to hold it and then I can keep tilting it around to make sure the Silastic stays exactly where I want it - occasionally I even turn it upside down. Keep some cotton buds handy in case it starts to go over the edge. You may also need to add some more to the central turns of the scroll to keep them nice and full, and you could also paint some onto the side of the pegbox, and go closer to the edges as it thickens. You will need to keep an eye on it until it can be left clamped with the Silastic side up.

Again, we can't wait so I will go onto the next step. After a few more hours, you can safely set up the violin properly in the mould box and pack some play dough around the neck cutout. Now, mix the second batch - this time 300g of the base and 15 ml of the curing agent and slowly pour it in a thin stream - this will help to remove some of the larger bubbles. The level should go past the centre line of the scroll. You can chock up one end of the box if you want to make sure that the level is nice and even on the head, or adjust the violin if that is easier.

Now, it is just a matter of waiting for the specified setting time and then de-moulding. I think it best to remove the mould from the box first - untape the bottom of the box and remove it with a parting knife, then break the contact at the sides and slide out the mould. The Silastic is then flexible enough to carefully work it away from the head and ease it free. I think this order of removal puts the least strain onto delicate areas such as neck grafts.

Whole Scrolls

I will just make mention of the technique for making a mould of a whole scroll. I made a slightly taller box (75 mm high) for this but the rest of the technique is similar - except you will of course need to fill the pegholes on both sides. After doing the mould on the first side drop in a some locating pins like these - they are just pieces of cello soundpost tacked onto a strip of rib (I made up two pairs). Cover the ends in Vaseline, and let them hang into the Silastic while it sets - they only need to go in about 5 - 10 mm. These will leave holes to make locating pins for the other side of the mould.

When the first side is set, ease out the pins. Then carefully paint ALL the Silastic surfaces with petroleum jelly, making sure you also work some down into the holes. The Silastic will stick to itself, so make sure you have covered everything.

Build a wall of play dough to fill in the neck cutout, then mix another batch, and pour in the other half of the mould. The result should be two nicely separated halves, that can be accurately re-assembled with the locating pins. It is best to make the join line off centre so that the central spine is well-defined.

Plaster Cast

Making a plaster cast from the mould is quite easy - you can just keep experimenting till you are happy with the result.

The mix for very absorbent plaster moulds with low strength is 110 parts of dry plaster by weight to 100 parts of water (eg. 110 g of plaster per 100 g or ml of water). For high mould strength but low absorbency, use up to 150 parts of plaster per 100 parts water. In between proportions will give intermediate results. We need low absorbency with higher strength, so the 3 parts plaster to 2 parts water ratio is the best.

I have also been using a product called Patternstone 'H' which is available from dental suppliers. It is very fine-grained and gives a beautiful smooth, silky finish. The proportions for this product are 32g water to 100g Patternstone. It costs around AUS$60 for a 22.5 kg bucket. It sets in 30 minutes, and has an average linear expansion of 0.24%.

As I said earlier, I like to use the mould box to support the mould when I take a cast. It helps to keep the flexible mould perfectly square and avoid any chance of distortion.

Mix a suitable amount of plaster and wait for it to start to thicken, before carefully pouring it in. I found it best to pour it slowly straight into the eye of the scroll and let it fill from there to avoid trapping any air.

Making a cast of the whole scroll is a bit more difficult. I use rubber bands to hold the two sides of the mould together. You can't just pour the plaster in and leave it because there will be air trapped at the uppermost turns of the scroll - you can see it on this example. You can experiment with tilting as you pour, rolling it after you pour (put your thumb or some play dough over the hole) to try to get the air to the top. You can also try a two stage procedure - partly fill the scroll turns with plaster, then after they have begun to set carefully assemble the mould halves and pour in a second batch.

Acknowledgements

I have been enormously impressed by the support and help that I received from Dow Corning. They were very willing to answer questions and provide information and advice, and they seemed to be particularly interested in the "niche" users of their products. They have very kindly made up these information folders containing a lot of technical data, supplier lists and supplementary material. I would particularly like to thank Edmund Cueno from the Sydney office, who kept up with all my leaps from one product to the next.

I would also like to thank David Butler who is a Picture Frame Conservator at the Art Gallery of NSW. David regularly uses Silastic 3481 in his restoration work and he helped me get started and gave me a lot of good advice.

Conclusion

There is not a lot of information around about these techniques, even less about their applications to stringed instruments. I expect those of you who do start to experiment will come up with lots of useful ideas and tricks. Especially if you start doing plates or whole scrolls. I will be putting a transcript of this talk onto my web site and if people want to keep me informed of any ideas or new techniques or problems, I will be happy to also add them in a postscript to the article. Our web address is www.abcviolins.com and you can email me through that site.

Finally, I have made up about 50 plaster casts of a 1713 Strad scroll so that you will each be able to take home a sample to inspire you. I should point out that I did not take this off the original instrument, but from another plaster cast that was given to me by Alex Grant. I have made them up in the Patternstone plaster so you can get an idea of the qualities of this product. So at a convenient break please come up and help yourself to a folder and a scroll.

Supplementary Information

ARLEC ELECTRONIC KITCHEN SCALES - these cost around $90 from a local hardware shop. It measures in 1g increments to an accuracy of +/- 0.5%.

PLAY-DOUGH - available from shops like K-mart, etc for around $10.

To make your own there are numerous recipes on the internet. We used one from www.recipegoldmine.com

1 cup flour
1/2 cup salt
2 teaspoons cream of tartar
1 tablespoon vegetable oil
1 cup water
Food colouring (optional)

Mix together flour, salt and cream of tartar. Add oil, water and colouring. Cook, stirring for 1 to 3 minutes or until thick. Knead almost immediately. Keep in an airtight container in refrigerator. This makes a nice, soft play dough and will last up to a year.

CASTING PLASTER can be obtained from most hardware shops.

PATTERNSTONE 'H' came from Ainsworth Dental Supplies in Sydney (02) 9519 7223.

Postcript:

1. At the conference Hieronymus Koestler mentioned that he used a mould support that was "scroll shaped" - rather than being a rectangular box, it was shaped to the head of the instrument and this significantly reduced the amount of silicone rubber required for each mould.

2. Hieronymus has since supplied the following information about the materials that they use for mould making at his workshop in Stuttgart.

The mould material is Elastosil M 4601 from Wacker silicones. (Order No. 9100001432)

To fill the pegbox and pegholes they use a 2 component putty (1.6 kg box), but only the white compound is used without the addition of the hardening fluid. (Order No. 18000)

You can find out more about Wacker Silicones at www.wackersilicones.com.
In Australia the Elastosil M 4601 is supplied in a 1 kg pack (0.9 kg of base and 0.1 kg of catalyst) and costs around $58 + GST. It is available from:

Barnes Products in Sydney at (02) 9793 7555 or www.barnesproducts.com.au
or FGI in Melbourne

or contact Wacker Chemicals Australia Pty. Ltd.
Suite 3, 11 Leicester Avenue
Glen Waverley
Victoria 3150
Phone (03) 9802 6100
Fax (03) 9802 6311

© Alan Coggins 2003

---