Two-dimensional luminescent beans

I have rediscovered the excitement of two-dimensional space again! I usually get the urge a couple of times a year and usually as a ‘break’ when the 3D work starts getting too laborious .. which it tends to a bit too often! For these ‘beans’ I manipulated beetle-shell photos in PaintShop Pro and then exported them into Procreate on the iPad to further paint, add layers and refine.

beans_final02websize

beans_final02,websize.jpg

beans_final01websize

beans_final01,websize.jpg

beans_final03websize

beans_final03,websize.jpg

I hope I’ll be writing about the combination of PaintShop Pro and Procreate sometime later, because over the past few years these two have given me everything I could possibly need in terms of image manipulation and ‘physical’ touchscreen painting.

What I’m not yet sure about is whether these finals can actually be the final artwork, either existing only in digital form or suitably printed, or whether they are just detailed prototypes waiting to be resized and copied in real paint. I’m not happy with even the best quality printouts. On the iPad or promoted to a bigger screen (assuming I can correct the colour and contrast changes) they’re luminescent .. literally composed of light .. and surface doesn’t come into it! But even the best print is just a pale imitation of the third dimension it no longer has, and suddenly real surface is there .. but it’s without any character! The magical third ‘dimension’ in 2D work needs to be re-invented and re-introduced by hand. For the moment, for me at least, digitally created has to remain digitally viewed.

Making a panelled door in stencil card

Recently I was asked by a friend to cover for her on the ‘Foundation in Art & Design Diploma’ course at Central Saint Martins. The day was intended to deal with aspects of model-making relevant to a project the students are currently working on. Each is designing an enclosed space with particular emphasis on the doorway leading into it, so we took the opportunity to focus on doors and the different methods of simulating surfaces.There was no budget available for materials so I had to devise a short practical using whatever small leftovers I could spare. The most promising idea seemed to be working with stencil card since I had a lot of small pieces, and stencil card was available at the CSM college shop if the students wished to take it further.

making a panelled door in stencil card

So I spent a bit of time working out the easiest way to make the traditional panelled door above. I’ve already looked at layering stencil card to create the wall panelling effect below and I also discovered some time ago that stencil card could be scraped with sandpaper leaving a fairly convincing ‘woodgrain’ effect, but I hadn’t combined them much. Also, the panelling below was made by carefully marking out and cutting the layers separately, then just as carefully aligning them while gluing. This is quite demanding! .. I wanted to make it more achievable.

using stencil card for wall panelling and windows

The improved method involves four layers (but as yet only dealing with one side) and the only ‘graining’ done is on the top layer and on the bottom layer where the ‘panels’ are seen. Everything is led by the ‘second layer down’ .. the one shown first in the line-up below, on the left. This is the one which needs to be carefully measured, marked out and cut. These doors are 1:25 scale and I’ve rounded off the UK average for a traditional interior door as .. 198cm high by 76cm wide. If you want to be either very specific or if you’re working in feet and inches, it’s properly 6′ 6″ by 2′ 6″! What happens within that outline is more a matter of taste .. there are no similar ‘standards’ for the size or arrangement of the panels. I’ve cut the first piece of card according to what looks right, but also I’ve observed with 4-panel doors that the top pair are usually longer than the bottom and there’s most often a broader strip across the base of the door for strength. The long thin panel in the middle is not meant as a letter box but it could house one, and the handle or doorknob would be positioned roughly halfway up the door which makes it on average a little less than 1 metre up.

stages in making a panelled door in stencil card

The drawing below should print out on A4 at exactly 1:25 scale and if you’re using this design as a template only the first one needs to be traced or pasted, as I’ve said .. the others are just there to illustrate each stage of layering. It goes like this .. after the first is cut out it should be stuck down onto another scrap of stencil card leaving a small margin around it. Spraymount works well, as long as you don’t intend to treat afterwards with a spirit-based medium because this will dissolve the glue .. otherwise superglue applied with care (very thin lines or dots) works perfectly. Pva wood-glue will grip but not bond very well with the stencil card surface. Trim around the outline of the door using the top stencil layer as a guide then judging by eye cut out all the panel areas a little inside the top-piece outline all around making a little ‘step’ .. as illustrated by stages 2-3 above and below. It may take some practise to get an even strip but it’s too slight to measure/mark. I’ve used the smallest division on the 1:25 scale ruler as a visual guide.

stages of making door using stencil card

This piece is then stuck onto another piece of stencil card and the outer edge trimmed again as before .. before doing this the stencil card which comes underneath needs to be ‘grained’ first because this will show. For these examples I’ve used a small piece of 120 grit sandpaper to grain, pressing firmly down and straight along, using the edge of a metal ruler as a guide. Once all three stencil card layers are stuck together and the door outline trimmed around once more (stage 4 in the line-up above), the fourth and final layer comes on top. This one is applied differently though, in separate pieces. It has to be because the grain of each strip must follow its longest edge .. essential for a convincing look! The task becomes a bit like marquetry in wood, but much easier because the stencil card is easier to cut. I grained a much larger piece of stencil card first and cut the strips from it, and I made these a little narrower to form a final ‘step’ around the panel areas.

colouring stencil card door with ProMarkers

There’s almost no end to what one can use to stain or paint stencil card because, in spite of the linseed oil waterproofing, it will accept both water-based, oil or spirit-based media. I’ve detailed a number of these already in my post February 2015 The art of alternative staining where I’m working with wood, but all will work well on stencil card. In fact many will work better because although a fine-grained wood is often the best option for a good ‘wood’ look when it stains well, it can also be difficult to eliminate the scattering of light specks where the polish or stain has failed to penetrate. Generally stencil card accepts stain a lot better and more evenly.

For the two samples above I used Letraset ProMarkers. The alcohol ink in these covers well and dries quickly, though it stains so well that the lighter scratches tend to disappear. These are ideal if you want something subtle. The ProMarker ink itself dries matte but there is a very slight sheen from the stencil card.

staining stencil card with Marabu GlasArt

If you’d like more shine or even brighter colours another option is using Marabu Glasart glass paints above, or ‘vitrail’ as they’re often labelled. These are spirit-based and, in the case of the Marabu, can be diluted or cleaned up with white spirit. One has the choice of either a silky or a glossy finish dependent on how much is applied. Here for example I brushed the vitrail on thinly and also went over with tissue and cotton bud to remove the excess collected in the raised edges .. if I’d just left it the effect would have been more glossy. Vitrail doesn’t work well as successive coats, because like shellac a further coat just starts to dissolve the one underneath and the results could be patchy.

colouring stencil card with shoe polishes and wood-stains

As shown above, if you’re intending a worn or ‘distressed’ effect I would recommend either a liquid shoe polish (which are almost always water-based) or a water-based wood varnish. These will tend to sit more on the surface rather than staining, and with each of these samples I started to rub or gently scrape after only a few minutes, before fully dry .. achieving a properly ‘chipped’ look fairly easily. These are, from left to right, Wickes ‘Quick-dry Woodstain’ mahogany; Cherry Blossom brown shoe polish, Kiwi ‘Wax Rich’ black shoe polish. Stencil card will warp a little with water-based media but not as much as other cardboards and, once dry, it is easier to bend carefully back into shape.

Conventional wood-stains also worked well .. both spirit and water-based. The middle one has a light coat of Colron ‘Georgian Oak’ and to the right I have used a water-based ‘Dark Oak’ wood-stain from Flints in London. The spirit-based stain has remained fairly matte whereas the water-based dried to a slight sheen. Spirit-based stains will also infiltrate quickly to the other side, even when more than one layer .. worth bearing in mind if this will be seen.

colouring stencil card with shoe polishes and wood-stains

Lastly, for the pale sample to the left I tried Osmo Dekowachs ‘Transparent White’. This is a specialist wax-based paint I was using in Germany which I still have some of, though these paints are also available in the UK. Like Humbrol enamels I’ve found that these paints will fix on almost anything. The first coat of Dekowachs is always matte and one has to build up a shine with further coats.

Coating styrofoam with polyurethane resin

These are the latest forms I’ve been making for my .. not-quite-working-title .. Ridiculously Organic Construction Toy. For this I’ve been creating simulations of eroded rock and driftwood cast in resin, twisted Pvc branches covered in fake moss and lichen, corals, leaf clusters and strands of seaweed made from latex etc. But I also wanted to include some play elements which are more obviously scaled down, such as these brickwork ruin pieces. The best way of picturing the whole idea is to think of aquarium or reptile tank accessories and then imagine getting a large collection of these instead of a box of Lego. I’m still working on the question of how exactly the ‘construction’ is achieved .. i.e. how such components will be fixed together when playing .. but as part of the system I’m working on an artificial ‘mud’ which I’m hoping will solve part of it.

ruin fragments in resin-coated styrofoam

The forms above were cut/carved in regular blue styrofoam, textured using a heavy-duty wire brush and then coated in polyurethane resin. There’s a bit more to the ‘painting’ process .. something new I haven’t tried before .. but I’ll come to that. If properly done the method of resin coating makes the forms unbelievably strong! .. perhaps not enough to survive little children, but certainly any adult wear-and-tear.

Making a brickwork arch in styrofoam

These two photos illustrate other forms intended for the collection and the process of making them. I’ve described this method of form-making in more detail in Shaping styrofoam. The arch piece above started with a Pvc template, which I used to help sand a block shape. I found I had to make a separate drawing template (the one at the bottom) just in order to inscribe the brick pattern onto the styrofoam shape. Then I used the special diamond needle files pictured to scratch out the brickwork divisions at the right thickness. I wanted these pieces to be 1:12, i.e. usual dollshouse scale, but I’ll eventually use a mixture of scales.

Making a brickwork niche in styrofoam

To make the ruined ‘niche’ shapes above I also used the method I described in Shaping styrofoam of using a curved sander to create the concaves. I roughed out very deep channels for the mortar lines, because these will become partially filled with coloured resin .. and this is what gives the pieces unusual strength. I found it was better to make all the channels before attacking with the wire brush, because I made the pitted texture mainly by hitting or pressing with the brush. This peppers the foam with deep holes and it may fragment a bit too much if the channels are made afterwards.

diamond needle files

Here is a close-up of the type of file I’ve found to work best for detailing foam. These have a ‘diamond coated’ surface which has more of an effect on relatively soft materials than the other, cheaper, form of needle file which is just ‘toothed’, grooved metal.

wire brushes useful for texturing rigid foam

I usually use the smaller brushes pictured above when working with the more delicate polyurethane foam in Kapa-line foamboard, but styrofoam has a tougher surface .. the heavier wire brush has more effect. Importantly, the action in this case is not a brushing or sweeping one, it’s more hitting downwards and rocking around .. I call it ‘scumbling’.

styrofoam 'ruin' fragments

Now to get to the main point of this article! Of the polyurethane resins I most often use (Sika’s Biresin G26 and Tomps’ Fast Cast) I know that both can be used in the following way, but Tomps Fast Cast is best because it’s a little thinner, powder pigment mixes better into it, and according to Tomps it is designed to cure properly in very small amounts or in very thin layers. This is not the case with all polyurethane resins. I’m basically making a very quick-setting paint with it, and because it’s quick-setting it has to be done a little at a time. To dose both resin parts I use disposable plastic pipettes (which are available from a few places online) and usually work with not more than 2ml of each part at a time. I can normally manage to use up to 4ml before it thickens too much. Because there’s usually no time spare to clean the palette surface before it sets I use a ceramic tile which can be scraped clean afterwards. There’s always just enough time to clean the brush though, and this can be quickly done with acetone.

Coating styrofoam with polyurethane resin and pigment

Here I’ve dosed 1ml of each resin part together on the tile, added a small amount of powder pigment, mixed the whole together with a synthetic-hair paintbrush and used the same brush to paint the foam. Synthetic is best because the hairs will be rigid enough to push the pigmented resin into deep pattern, but full and fine enough to hold a lot of the paint. Powder pigment is the best form of colour to use .. strong colour, inexpensive, available .. and I usually find that it mixes better into resin than it does with water!

The polyurethane resin has no effect on styrofoam (unlike polyester resin), it will cure hard and ‘fused’ to the surface, and it’s done .. that is, it’s touch-hard and ready for further work .. in about 15 minutes! Whereas regular paint such as acrylic will infiltrate more and contract as it dries, polyurethane resin does less of both so there will be a little ‘smoothing over’ of fine surface detail. It will also be a gloss finish! .. which I don’t like, would never choose, and at the moment I’m experimenting with the different  ways of dealing with this. There is no matting additive for polyurethane resin, and regardless of which pigment or filler is mixed with it, the top surface exposed to the air will always be glossy. Obviously painting over with another matte paint, such as a good acrylic, is an option .. but polyurethane needs a lot of preparation if the paint coat is to resist a lot of handling and this is made difficult by such a patterned/textured surface.

One possible solution is to use my own version of cold powder coating. If you google ‘powder coating’ you will find that this refers to an industrial painting process in which fine thermoplastic powder is melted onto metal to create a durable surface. It’s very like the enamelling that you might have done at school, with coloured glass powder on a copper plate, melted in a small oven. My version does not require heat, and it’s perhaps more related to the model-making practice of scattering granules into glue to create a surface .. but it does share some of the surprising durability of these other methods!

crushed brick

Below is a close-up of the styrofoam ‘ruin’ forms after coating. I first gave the bare styrofoam an undercoating of resin mixed with black pigment, and then a second coat without any pigment, covering a small area at a time. While each portion was still wet I sprinkled a mix of finely crushed brick and sand onto the resin. I’m fortunate in that, living close to the Thames beach, I can pick up fragments of any colour of brick, illustrated above. Since these have already been broken down by the elements they are much easier to crush to a powder using mortar and pestle.

detail of brickwork surface done with 'powder coating' method

While working I could see that the particles were readily sinking into the thin coating of resin, and when the excess is shaken off after a few minutes the powdery top layer still adheres strongly. Polyurethane resin is a strong adhesive, especially if the dust or particles are porous and jagged. Having tested the strength of the surface once the resin cured I have little doubt that it is permanent. I still have to do some paint finishing on these pieces, emphasizing contrasts and colours and giving more ‘speckle’, but I have no worries about regular acrylic paint attaching itself on top. The greatest bonus in this particular case is that these pieces have a lot of the look and feel of real brick .. because that’s what it is!

 

The coalescence of putti in a summer sky

 

coalescence_basis1-5_1200

 

coalescence_basis1-4_1200

 

coalescence_basis1-3_1200

 

putti11-15_2_1200

 

putti11-15_1_1200

 

putti11-15_3_1200

 

Background

In preparation for an exhibition of my sculptural work next year I am planning to show a lot of my working sketches. In fact, I’m hoping that the exhibition will feature process just as much as final outcome, not only sketches but maquettes, colour/texture samples and even the raw materials, but at the moment I’m not sure how far I can take this. I’ve recently been trying out a new method of preparation and idea development, which first involves creating maquettes, photographing them and then using these photos to explore/develop form and colour digitally. Because digital material is infinitely adaptable and reusable it opens up all sorts of rehearsal/improvisation opportunities. It can also lay the basis for promoting sketchworks to finalised outcomes in their own right.

My new work on one of my favourite subjects of ‘putti’ is a case in point. The interest developed many years ago during a visit to the Bayerisches Nationalmuseum in Munich. Amongst many other truly emotive and tactile examples of Baroque sculpture, the museum had one of the best collections of nativity dioramas in the world. In a darkened, maze-like room thick with atmosphere I saw huge installations filled to bursting with carved figures. In many the richly blue skies were just as crowded, with colourful airborne beings .. many of them adult-looking angels, but just as many infants, and as I remember it, some were just fragments or, as if, in the process of forming .. like heads with wings, or clumps of flesh-coloured buds with golden petals, sprouts with layers peeling. This made a deep and lasting impression .. though a mainly formal and associative one. I don’t subscribe to religion, but I can be moved by the beauty such devotion generates.

So for the ‘putti’ sculpture I’m working on .. and have been nudging forward on-and-off for years .. I’m trying to recapture that thrill, trying to find a sculptural form which will suggest the physical simulation of something glorious .. but also ancient, and also strongly organic. For example the words ‘protean flesh’ spring to mind, and that’s the reason why I’ve preferred to keep to the title ‘putti’ rather than something more maturely angelic, because this makes me think of ‘putty’ and particularly the gorgeous, dark pink, rubbery ‘silly putty’ one could get when I was a child which seemed to have an innate life and will of its own and could become many things though only what it wanted to.

So I think it’s very fitting that I happen to be using digital material to find a way through this .. it’s very fluid, it can be breathtakingly spontaneous, all manner of variations can be fairly instantly and effortlessly previewed. Although in the beginning I fought against the intangibility, the fact that what I was doing did not really exist in any physical sense .. until it’s printed, and then it’s something else .. I think I’ve come to value that ethereal, ‘protean’ aspect. In a sense it has more allegiance to, or is in the same space as, what’s inside my head.

Technical

As I said though, I prefer to make something physical as a starting point, providing an anchor .. but something simple, no real pressure, it’s just raw material for transformation. These are the ‘putti’ forms I produced a while ago in response to the flying angels, and which I still want to use as a basis ..

putti originals

.. though over the years they’ve acquired a lot of experimental patination, because I haven’t been able to get the surfaces right yet.

older putti photes 2015

new putti photos Dec2015

To create the sketches I took these, or similar photos into PaintShop Pro where I could experiment with either softening or enhancing contrast. Eventually I found that the best basis for the effect I wanted was to enhance the contrast and deepen the shadows but change to an almost complete monochrome, to give more freedom when later ‘colouring in’. For this, the main ‘painting’ process, I exported the modified photos to Procreate on my iPad.

Procreate is a ‘painting’ application developed solely for the iPad. It has given me pretty much everything I’ve wanted so far from this kind of tool and I would strongly recommended it .. though I don’t know how it compares to others since I haven’t had to consider them. In either working colour gradually into the photo-basis or making alterations to the forms I found the brushes, blender and eraser nicely delicate. I did experience some frustration though, which I have not yet overcome .. feeling that I couldn’t fully judge what I was doing, compared for example with controlling the effect of real paint, pastel or pencil shades on paper. It also took me a while to realise that, for all the choices of brush or setting that digital painting offers, one has to choose a manageable handful of favourites and stick with them.

Working with epoxy resin

I’ve started working with epoxy resin and have put quite a few pages of useful information under casting in the Materials section, together with the write-up of a recent test. I’ve copied them here in full..

Definition and general properties

Epoxies tend to be stronger than other resins, certainly much less brittle on their own than polyester .. in other words, they have very good flexural strength! They come as two parts which are commonly mixed in ratios ranging 2:1-4:1 resin to hardener by weight. Also compared to others, epoxy resin generally has a very long pot-life (working time) i.e. even a ‘fast’ epoxy resin will still give c.15mins working time before it starts gelling whereas a regular/slow can take 100mins or more (the average would seem to be 40mins). A ‘fast’ epoxy may be demouldable in 8hrs and sandable after 12-18 hrs whereas a ‘slow’ may need 30hrs before it can be removed from the mould. Full cure generally takes 5-7 days.

The density is on average like polyester resin SG 1.1 (the weight in grams of 1 cubic centimetre of mixed resin). Viscosity is generally higher than other resins (i.e. its usually a thicker liquid) with an mPas of 1000-1400 being considered ‘medium’ for epoxy. The thinnest I’ve come across so far has a viscosity of 600 mPas .. compare this with polyurethane resins which can be as little as 50 mPas.

They are also much more adhesive (hence their modification as epoxy glues). They are usually more transparent and cleaner looking than many general-purpose polyesters. However, water-clear epoxies specifically for solid casting are not common, presumably due to the high risk of excessive heat build-up .. the water-clear epoxies commonly available are almost always just for laminating or coating. There are a great many varieties, but most share a relatively long pot-life and cure time compared to other resins. Epoxy resin is just as commonly used as polyester for the binding resin in fiberglass work but because epoxy is more expensive this applies more to industrial applications.

‘The chemistry of epoxies and the range of commercially available variations allows cure polymers to be produced with a very broad range of properties. In general, epoxies are known for their excellent adhesion, chemical and heat resistance, good-to-excellent mechanical properties and very good electrical insulating properties.’ Wiki ‘Epoxy’

In terms of working conditions, epoxy is almost odourless compared to polyester resins.. though this shouldn’t fool anyone into thinking that good ventilation is not as important!

Often resins marketed for sculpture purposes are tagged as ‘not Lloyds approved’ meaning they may not have the structural integrity or moisture resistance necessary for boat-building but are fine for sculpture. This works to advantage because they are often cheaper.

Uncured epoxy can be cleaned up with acetone, cellulose thinners or methylated spirit.

You may see the term ‘infusion resin’ applied to standard laminating resin. This simply means that the resin is of a suitably low viscosity to be used for vacuum infusion .. which is a process whereby instead of being brushed on resin is sucked into the reinforcement material under pressure. This eliminates the air pockets which may occur using brush-on methods and ensures stronger fibreglass.

EL2 laminating epoxy resin from Easycomposites

 

Advantages of using it

Because of its flexural strength it is ideal for the laminating or casting of load-bearing forms or those which will be subjected to stress. Many epoxies also have enhanced moisture or chemical resistance making them preferable for exterior sculpture (but see below re. UV exposure).

Because of epoxy’s adhesive qualities, good ‘wetting’ properties, the relatively long pot-life before it starts gelling and its toughness.. it is considered ideal for coating or laminating. It is a common constituent in special paint finishes; for the home-practitioner the longer working time makes it ideal if you want to mix your own resin paints.

Re. the above qualities, the resin can just be used as an adhesive and it has very good gap-filling properties when combined with fillers. Most epoxies will bond wood, metal and even quite a few plastics. For example it will usually bond with any plastic affected by acetone. Epoxy will bond well to cured polyester fiberglass but polyester won’t return the favour on epoxy. As with anything, if you’re serious, and you want to learn properly for yourself ..do a test first. Epoxy is more adhesive than polyester or polyurethane because it is able to form bonds with the substrate (the material being glued) at atomic level whereas other resins can only bond mechanically i.e. by gripping tightly.

Epoxies do not attack polystyrene, so may be an ideal choice for coating polystyrene or styrofoam forms.

Epoxy will take a variety of fillers, basically anything inert and free of moisture.. talc, Fillite, marble dust, metal powders etc. I’ve read that powdered/dried ‘ball clay’ mixed into epoxy will even make a good, clay-like putty. Because of the generous working-time it’s practical to mix resin with hardener first before adding filler .. this is usually advised with epoxy. This has an advantage because it means that the consistency can be judged as one’s adding.

Like both polyester or polyurethane resin, most of the epoxy resins available for home use need no special pre-warming in order to cure when mixed and will do so at normal room temperature i.e. c. 20C. Most however allow ‘post curing’, that is, accelerating the cure by heating at a moderate temperature for a number of hours.

The resin component has a much longer shelf life compared to other resins .. sometimes up to 3 years, although the manufacturers are usually bound to define it as 1 year. This is partly because the hardener part is more active, with a shorter shelf life. Apparently though, there are stories of unopened epoxies being discovered after decades and working ok!

Fully cured epoxy can be softened by heating to a temperature over 200F i.e. with a heat gun, but this should only be done in a well ventilated room.

Drawbacks

As one should expect, one pays for the advantage of a long pot-life by having to wait much longer for the cure and, as mentioned, with some it could be a couple of days before the cast can be safely demoulded. It also means that epoxy is not so suitable for ‘slush’ or ‘rotocasting’ methods especially by hand, unless you want to sit there doing it for more than an hour!

The resin itself doesn’t present quite the same health & safety issues as polyester and is considered little more than a possible irritant to eyes and skin. However, the hardener part is a different matter! It is classed as ‘corrosive’ and could be very unpleasant if it gets on the skin. It is also harmful by inhalation. Precautions need to be taken against skin contact and, as with all resins, good ventilation is essential!

In price epoxy resins average a little more expensive than polyurethanes, but a lot more than the cheapest polyesters (see example prices below).

Epoxy resins are pretty unforgiving when the mix is even a little bit out, for example the Technical Data Sheet for DX020 from Tomps states ‘The components should be measured to an accuracy of 2% or better’ .. in other words, more than 98% accurate! This can only be done by weight .. not recommended by volume! It also means that it’s not a good idea to measure out portions in separate cups and then decant one into the other when ready, because even the slight amount adhering to the cup could make a difference. Mixing needs to be obsessively thorough! .. not forgetting the sides or the bottom of the mixing vessel. When mixing silicone rubber thoroughly I usually recommend spending at least 3mins to be sure, and I would say the same for epoxy. Often it’s wise to transfer a thoroughly mixed batch into another vessel and mix again to avoid unmixed residues on the sides or bottom of the cup.

There are often warnings accompanying epoxy resins that thick layers .. especially massed volumes.. will become very hot during cure, causing increased shrinkage. Many epoxies are labelled as ‘laminating resin’ which often (though not always) means that they are not suitable for cast volumes. I’ve read the advice that, if smoke starts rising from curing epoxy ‘it’s likely that the epoxy is damaged and should be replaced’ .. and I would add that it should be swiftly but calmly taken outside! To avoid this happening, large and solid castings therefore need to be done in stages which, because of the long pot-life, can become a lengthy process! Although one presumably doesn’t have to wait each time until the layer is cured, I’m assuming it makes sense to wait until it’s at least cooled down but even this can take a while! West System recommends layers of not more than 12mm when working with their epoxies. There is usually no danger of excessive heat build-up or shrinkage when laminating thin layers with a reinforcement in the standard way.

Even the low viscosity epoxy types are considerably thicker than some polyurethane casting resins, so they are not a good choice for intricate castings. On the other hand, if the moulds are 1-piece and open the long pot-life gives a lot of time to pour and coax the resin to fill an involved shape. But partly as a result of higher viscosity, many small air bubbles are generated while mixing and these are persistent! Usually the extended pot-life allows taking the time to deal with these i.e. by skimming away, or passing a hair-dryer or heat gun over the top surface which will eliminate many. Bear in mind though that any heat applied will reduce the pot-life, though this will not be a dramatic reduction. Pouring the resin into an enclosed mould needs to be done very slowly and carefully!

Another problem involving bubbles occurs when using epoxy for coating a porous surface such as foam or wood. As the epoxy heats itself while curing this will expand the air underneath it, forcing it out to form bubbles in the resin. The only solution is to make sure that the original surface is completely sealed first. One way is to prime the surface first with a very thin coat of resin and let this set firm before applying a thicker coat.

Epoxies are particularly susceptible to prolonged UV exposure. For this reason when used in boat-building they are more often employed for the inner structure rather than the outer surface. Sunlight doesn’t just discolour epoxy, it degrades it. Deterioration due to UV is known as ‘chalk out’ in the case of epoxy paints or coatings. The usual fix is to coat with a ‘2k’ UV resistant varnish although I’ve read that this doesn’t solve the problem completely.

Although epoxy is ideal for fibreglassing there are some notable differences in method compared to polyester work. The major one is .. with the polyester resin I use for fibreglassing work (Tiranti’s GP) it doesn’t matter if one layer has been left to become fully hard before the next is applied. The second layer will bond firmly to the first, and I’ve certainly never experienced any instances of layers ‘delaminating’, that is, coming apart because of time intervals. Most of the guidance when using epoxy suggests the contrary. Layers should be applied while the first layer is still in the so-called ‘green stage’, meaning that although it may feel touch-dry it should still be possible to make an impression with the fingernail. Any later and the fresh epoxy will no longer be able to chemically link with it. In this event the hard epoxy surface needs to be sanded then dusted/cleaned, to at least ensure a good mechanical bond. Another difference is that whereas with polyester resin one can compensate for certain conditions simply by varying the catalyst dosage i.e. according to the volume of resin being used, the addition of fillers or the ambient workplace temperature .. with epoxy this is not possible because resin and hardener must always be mixed according to the set ratio. The only ways to compensate are either to have the choice of either a slower or faster hardener on hand to use, or to apply external heat while mixing or curing.

Working life

Here are my test notes from 25/5/2015, testing Polyfibre’s EL68 resin with EHA57 hardener (not the one imaged above). The mix ratio was an easy one.. 2:1 resin/hardener.

The product is +14 months old (bought March 2014 and unopened). A little, approx. 1g, poured in plastic cup then tiny amount of green powder pigment mixed in.. appearing to mix well. Topped up to 20g resin, then colour thoroughly mixed in again. Profuse creation of air bubbles! 10g hardener added and mixed together. Material is colourless compared to my usual GP polyester resin.. clear, though not quite water-clear. No noticeable smell. I left most of the resin left in the cup but I trickled some onto a polypropylene sheet to see how the resin would cure in small/thin amounts and to test whether polypropylene would be a good ‘releasing’ base.

epoxy resin pigment/mixing test May 2015

Surprisingly, no evidence of heat reaction for a long time, very mild heat felt from the bottom of the cup at 35mins. After 45mins mix is much thicker but may still be spreadable .. stronger heat from cup but by no means excessive. After 60mins firm gel in cup but the thin ‘spills’ on polypropylene sheet are still sticky like clear ‘honey’. After 90mins almost fingernail hard at cup centre but still soft at edges and flat spills are still unchanged.

epoxy resin cured enough to 'demould' from cup

Returned to at 7.30 a.m. the next morning (16hrs). Both pot and pools touch-dry and hard, no surface tackiness even on thin residue lining the mixing cup. On demoulding the cup contents the resin parted easily and cleanly, with complete surface reproduction, though it gripped noticeably more than polyurethane or polyester. There was a small area of tackiness around the rim of the cast piece (bottom of the cup) which could be due to not mixing 100% thoroughly .. even though I was consciously much more thorough than I normally am when mixing PU or polyester! The thin residue lining the cup above the mass came away intact with the casting and was strong but very flexible, no brittleness. The pigment colour was completely even with no grain.

The ‘spills’ on the polypropylene sheet were touch-hard and firm, with a beautifully smooth, polished-look surface showing no sign of pitting or clouding. But they would not detach even on extreme flexing of the sheet, remaining flexible. An attempt to prise up a small area with the tip of a scalpel blade merely curled and damaged the resin though it did detach. For the moment the resin is effectively stuck fast and I will have to wait longer before I try again.

After another day the resin ‘spills’ were no easier to detach. By getting a blade underneath the pieces could be popped off, but not easily. It shows the adhesive strength of epoxy because even superglue fails to cling as well to polypropylene! Epoxy is supposed not to be able to establish a bond though with plastics which are impervious to acetone, of which polypropylene is one. If I ever use polypropylene as a ‘releasing’ base for epoxy work I will have to remember to spray with a release agent i.e. pva (polyvinyl alcohol) or possibly hairspray (as I’ve heard).

excessive build-up of air bubbles in epoxy resin

Apart from this the two other most significant results were the lack of expected heat from the resin mass .. although small the piece is 2cm thick and av. 4.7cm across .. and the bubbles! I purposely left them unattended to in the cup to see whether they would pop of their own accord, which they didn’t. On the plus side though, they all rose to the surface. There were bubbles initially in the ‘spills’ but most of these could be forced out by ‘tamping’ i.e. jolting or shaking the sheet. I had to pop the remaining few with a cocktail stick.

As for the lack of heat, it may be due to the hardener part being a little past its shelf-life although the resin part is relatively inert and should be fine. Unfortunately I can’t check the results against technical guidance or MSDS since Polyfibre doesn’t provide either!

A little bit of history

‘Credit for the first synthesis of bisphenol-A-based epoxy resins is shared by Dr. Pierre Castan of Switzerland and Dr. S.O. Greenlee of the United States in 1936’ Wiki ‘Epoxy’

What was it about 1936? This was the same year that the first proper patents for both glass fibre and polyester resin were independently recorded!

Additional info

Colouring epoxy resin

Like other resins the general rule is that one can add up to 10% by weight if using a powder pigment, up to 5% if using any other liquid colourant. Similarly, water-based colourants are out .. but usually oil or spirit-based are ok, plus of course specially formulated resin colourants which are usually pre-mixed with a small amount of resin. In the test detailed above the resin ‘wetted’ standard powder pigment very well .. all of it dissolved very easily, there was no frothing of the resin, no graininess and no sinking of the pigment. For more info see this article;

http://www.westsystem.com/ss/adding-pigments-to-epoxy/

Dealing with the bubbles

Careful heating with hair-dryer or heat gun over the open surface of the mixed or curing resin. The heat source should not be too close, in the case of a heat gun about 30cm away. Another method is to put some methylated spirit in a small ‘mistifier’ bottle and spray a fine mist on the surface. The alcohol doesn’t adversely affect the resin and evaporates quickly, but acts long enough to reduce the surface tension and pop the air bubbles.

Methods of thinning

Thinning the resin itself could help a lot in the elimination of air bubbles when mixing. It can also help the resin to better impregnate a surface if the resin is being used as a coating, or to make it flow better into a complicated form. Apparently there are a number of ways of doing it, though I can’t vouch for them because I haven’t tried them myself. One method is to heat the resin! Epoxy changes viscosity, becoming thinner when it’s warmed. The recommended method is to heat up the two parts separately (whichever way you prefer .. but standing the cups in hot water would probably be best) and then mix them. As always, bear in mind that heating will reduce the working time and accelerate the cure. Note also that if using two cups for dosing the resin initially, once warmed both should be decanted into a third cup for mixing together to maintain the ratio. I noted from one info source that the temp should not exceed 115F (46C). I wouldn’t imagine that it’s a very good idea to heat it if you’re pouring a massed volume anyway though, because it increases the risk of the resin overheating with its own exotherm .. if one actually does have to be as careful as they say!

Apparently another method is either to add acetone (not more than 10% by volume) or methylated spirits (US ‘denatured alcohol’) at 15-20%. Adding solvent will affect the strength of the cured resin, but this may not matter too much with small castings. See this article for more informed advice:

http://www.westsystem.com/ss/thinning-west-system-epoxy/

Cold casting with metal powder

Works well with epoxy and the surface is not too hard to be successfully ‘cut back’ or buffed with steel wool. Apparently olive oil can be used to give an even patina.

 

What it costs and where to get it

Prices are dated, and adjusted to include VAT

EL68 resin/EHA57 hardener £26.04 per 1.5kg (1kg resin, 500g hardener. Specialplasters 5/2015) Specialplasters describes this as ‘a low viscosity epoxy resin for laminating and casting’. Mix ratio 2:1 resin/hardener by weight. Manufactured by Polyfibre. The hardener is described as ‘fast curing’; EL68 is a Bisphenol F type epoxy (Bisphenol F epoxies generally have a lower viscosity and greater chemical resistance once cured). Polyfibre do not currently offer an MSDS or further tech data on their website!

http://www.polyfibre.co.uk/index.php?sec=prod&prod=83

EL2 Laminating Epoxy £20.10 per 1kg (770g resin, 230g hardener), £65.93 per 5kg (Easycomposites 5/2015). A choice of hardener is offered; fast (12-17mins pot-life) or slow (95-115mins). Mix ratio 100:30 resin/hardener; medium viscosity when combined 1000-1400 mpas; clear; SG when combined 1.05-1.15. Easycomposites advises that applying it over 1mm thick in one pour could result in too much heat build-up unless the slow hardener is used in which case 5mm should be possible.

http://www.easycomposites.co.uk/products/epoxy-resin/EL2-epoxy-laminating-resin.aspx

DX020 £29.10 per 1.5kg (1kg resin, 500g hardener.Tomps 5/2015) Also described as a low viscosity laminating and casting resin. Pot-life 75-90mins; demould time 2hrs; Shore D 80-90 hardness after 5days cure. Manufactured by Atlas Polymers.

Epovoss Glosscoat £35.86 per 1kg (Tiranti 5/2015) from the website page: ‘A general purpose clear epoxy resin for casting, embedding, cold enameling and coating. The resin is very slightly straw coloured, but this is virtually unnoticeable in coating applications. Epoxy resin cures with a nontacky surface, is self levelling, nonshrinking and will adhere to most surfaces. Polyester Pigments may be used with this resin (5% maximum), and also a whole range of fillers.’ Mixing ratio 100:40 resin/hardener; pot-life 30mins; SG c.1.1; demould 12hrs.

Further info sources

http://www.westsystem.com

A leading supplier of high-grade marine epoxies. There is a lot of technical information and guidance on the site, almost all of it dealing with boat-building work, but even some of that is useful, i.e. if you want some expert advice on methods of coating with epoxy and getting a smooth, bubble-free surface have a look at this

http://www.westsystem.com/ss/bubble-free-coating/

The art of alternative staining

This was meant to be just another short ‘hidden treasures of the pound-shop’ piece recommending the value of cheap liquid polishes such as these below as alternatives to more expensive wood stains. I’ve used similar bottled shoe polishes many times in the past, particularly for staining model floorboards .. with very good results!

Poundland liquid shoe polish

But these ones currently in 99p Store are, as you can see below, unusually weak in terms of colour. I’ve made these tests on pieces of 0.8mm sheet obeche wood, which normally accepts stain very well. Each of the sample pieces featured in this post was made with a fixed procedure! Firstly I ruled three faint pencil lines and applied the first coat up to the top pencil line, leaving a little of the wood bare above. I then waited a minimum of 2 hours before applying the second coat up to the line below and then another 2 hours before applying the third. I chose 2 hours in between because this was the time it took for the pieces of wood to return to almost complete flatness after warping .. so I figured these were sufficiently dry for the next coat. As you may know, warping doesn’t occur with oil or spirit-based media, and with spirits the ‘drying’ is a lot quicker .. but I gave these the same intervals anyway.

Poundland liquid shoe polish samples

The polite word for these is ‘subtle’ .. the ‘Neutral’ had so little visible effect on the wood, even after three coats, that I didn’t even notice that I’d photographed the piece upside down! The brown and black did impart some even colouring as you can see, but even after three coats and rubbing afterwards there was only slightly more ‘sheen’ than the obeche wood has on its own.

But there are much better ones! The lighter brown sample on the left below is brown liquid polish from Tiger, the one in the middle is brown Cherry Blossom Readiwax    brand found in Poundland and the last is Kiwi Wax Rich Colour Shine also from Poundland but a while ago. I’ve overdone the photo exposure a little .. the colours are richer and deeper in reality. All of these gave a visible, regular sheen to the wood surface on the second coat, and more pronounced on the third.

liquid shoe polish samples on wood

Here are just two more, but also showing how best to apply the liquid polish. The sample to the left is the brown Kiwi Colour Shine shown in the photo, also found in Poundland, and next to it is one made with Clarks brown polishing cream, which is thin enough to brush on easily and infiltrate the fibres. The Kiwi polish gave a good sheen after three coats but the polishing cream remained matt. If you intend to cover large areas you could use the sponge applicator attached, but for these small ones it was easier to press out a small amount of polish on a tile and use a rigid synthetic brush.

Applying liquid shoe polish as a stain

All of the polish types I’ve featured so far are ‘water-carried’ including the polishing cream .. suspensions of pigment and wax in water. So one has to deal with the fact that the water will make thin wood warp! It can look rather alarming .. especially if as here the wood grain is perpendicular to the length of the strip. But as the piece slowly dries it returns to almost complete flatness again. As I’ve said, if using 0.8mm obeche wood this will take at least 2 hours .. but it will differ according to wood thickness and type.

thin sheet obeche warping with water-based polish

Although it might be an annoying way to work, I would recommend if you want to treat wood with anything water-based (including water-based commercial wood stains) that the pieces are treated like this i.e. stained and left to dry before sticking down. It may be logical to assume, when making a floorboard floor for example, that sticking all the boards securely on a base first will prevent the whole thing from warping when painted. It doesn’t .. in fact it can make it worse! I think what happens is that if something absorbent like cardboard is used as the base, when the wood on top is painted some of the moisture reaches the cardboard so it will also warp .. this is clear. But as both dry out the moisture which has reached the cardboard is trapped so that whereas the wood layer will dry the cardboard will take longer. The drying wood will therefore ‘set’ in this position. When this happens it is impossible to get the ‘composite’ flat again.

It can be a bit dangerous if one gets hooked on the satisfaction of making illustrative samples .. one looks around for more things to test! Samples don’t have to have any other meaning, and one isn’t going to be judged on their aesthetic value. Breaking free of both the conceptual and the aesthetic actually does have positive advantages sometimes.. it opens the door wide for serendipity! Otherwise I wouldn’t have thought of trying two other media I just happened to have around at the time .. GlasArt transparent glass paints from Marabu and Letraset ProMarkers. As it happened both worked far better than I expected, with the big advantage that neither of them cause warping because they’re spirit-based. The other significant advantage is the cost of them. If one only wants to stain a relatively small area it’s quite an expense to buy commercial spirit-based wood stain, or a combination stain/varnish. The smallest tins or bottles available are generally somewhere over £6.00. The smaller 15ml jars of Marabu GlasArt can be found for between £2.00 – £2.60 and Letraset ProMarkers are on average £2.00 each.

Marabu GlasArt paint as wood stain

The GlasArt paints work like a dream on wood! They can either be applied straight from the jar, as I’ve done with the sample to the far left above, or thinned with white spirit as I have with the other two. I used Bartoline Low Odour White Spirit and mixed roughly equal amounts of spirit and paint together on a tile. As before, the best brush to use is a flat one, with soft but rigid synthetic hairs. The first coat penetrates readily but dries relatively matt; the second has more noticeable sheen, and the third gives a regular satin finish. Whether thinned or used straight, the GlasArt paint penetrated the wood fibres better than I expected. For an even stronger shine I would recommend applying some neutral Kiwi solid shoe polish such as the one shown later. This has to be thinly brushed or rubbed in, left for about half an hour, then buffed with either a clean shoe brush or soft rag. I would recommend using a polish like this every time for scale model work, rather than using any kind of gloss varnish. Varnish is too ‘full on’ for miniature work, and it will emphasis every little imperfection. Often the sheen that the wood already receives is enough to suggest a polished look, but if one has to have more, polish can be far better controlled. It means for example that a floorboard floor stays more matt where it is difficult to buff i.e. along wall edges or in corners, which makes more realistic sense!

Marabu GlasArt used as varnish on wood

Letraset’s ProMarkers come in a wide range of colours .. 148 in total. Below I’ve used a few I had which were suitable as wood colours, but there are many more. In addition they can be easily ‘mixed’ .. that is, overlaid .. so the colour range is infinite! The ink is alcohol-based and it penetrates the wood fibres very well. But it is also ‘self-sealing’ .. there must be a fixative in addition to the alcohol, though there is no info on what this is .. so transparent layers can be built up and colours deepened. Here I’ve used, from left to right; ‘Ice Grey 1’, ‘Sandstone’, ‘Spice’, ‘Walnut’ and ‘Warm Grey 5’. Although the ink is thin, a build-up of layers will produce a sheen, though not as much as the GlasArt paint or the better polishes. Another advantage for small-scale work is that the ProMarkers come with both a broad and a fine felt nib. Because it penetrates so well I expected it to be difficult to draw a hard edge on the wood, at least for the first coat, but as you can see from the top edges it didn’t ‘bleed’ as much as regular wood stains.

Letraset ProMarkers used as wood stains

This could make it easier to ‘fake’ a more complicated parquet flooring pattern, just by using different tones or hues of marker. It can be fairly tightly controlled even on balsa wood, which is one of the most absorbent woods. Below I’ve compared the same colours on balsa, to the right, with obeche to the left. Balsa generally stains more evenly than obeche, especially when layers are built up, but obeche will give a richer and deeper colour.

Staining wood with Letraset ProMarkers. Difference between obeche and balsa

One last option .. let’s say, if you don’t want to get anything special at all for staining wood .. is just to use whatever transparent paints you’ve got such as watercolours. If you do use watercolours I would recommend painting before gluing down, as before. Don’t use either acrylics or gouache because these will not be sufficiently transparent .. they will obscure the natural grain of the wood, and the effect will be rather lifeless and artificial.

In place of watercolour though, I prefer to use regular colour pencils or pastels ‘dissolved’ using white spirit. This avoids the problem of warping, but I’ve also found that the colours are richer using spirit, and it means that so many more things can be turned into paints or stains. These small floor samples were made by firstly shading some pencil colour on the bare wood, then dissolving it with low odour white spirit and a synthetic brush. The dry pencil pigment dissolves readily and can be worked over and into the wood, but the ‘paint’ produced remains workable for much longer than watercolour or acrylic. It is basically a straight ‘watercolour’ technique .. just not using water .. and with the added control of adding the colour, not solely with a brush but also with the pencil-point .. a combination of painting and drawing. I’ve preferred to keep these samples subtle to show the nuances achievable, but one can go much richer and darker with the colours.

Staining wood by rubbing with coloured pencils and dissolving pigment with white spirit

I’ve used Karismacolor pencils here because these were my favourite and I still have them, but sadly Berol discontinued them long ago. Others who use colour pencils in their work say that Prismacolor or Polychromos are comparable .. rich colours, soft but not chalky-fragile! These are all oil-based pencils, which definitely work better for this technique, though I have found that any kind of colour pencil will work to some extent.

Old oak floorboard model sample

For the sample floorboard piece above, and below, I wanted to simulate the distinctive markings of old oak, but using obeche. I cut the planks and sanded the edges a little first, then ‘distressed’ them variously using a heavy-duty wire brush .. either pressing and pulling for deep scratches or just pressing and rocking for pockmarks. I then glued them in place, inscribing the floorboard ends with a sharp tool afterwards. I used the pencils to shade or cross-hatch on the wood in various browns, but defining some differences between the floorboards. When dissolving the shading with white spirit I made sure that most of the colour collected in the scratches .. conventional wood stain would most likely colour too uniformly, not differentiating enough. Much more variation could be achieved than I’ve shown here, by shading parts with pencil again and fixing with very minimal spirit.

Wood staining sample including old oak floorboard effect and solid shoe polish rub

But the pencil/white spirit technique doesn’t create much shine, so after I’d given the sample a few hours to dry I applied a thin coat of the solid shoe polish shown .. a ‘Dark Tan’ to give it a bit more warmth. One needs to wait for it to ‘set’ before polishing .. between 30-50mins. As one final test I tried the polish on an untreated piece of obeche and got a deep, satisfying shine .. though adding further layers of this polish did not deepen the colour.

Staining and polishing wood with standard shoe polish

 

Handy containers for small amounts of paint

One of my favourite paints is Rosco ‘Super Saturated’ scenic paint .. I like it because it has a more liquid consistency than most other acrylics, thins and mixes easily, adheres better than most to non-absorbent surfaces, covers incredibly well and dries almost completely matt! However, a major disadvantage is that it is only available in 1litre tubs .. really rather prohibitive in cost if you’re not a scenic painter and if you want a full range of colours! Added to that, the paint doesn’t keep very well once opened .. after about a year on average some of the colours start to smell, curdle or later still develop a thick skin of mould (but see postscript with solutions below). A couple of years ago I seriously splashed out on a number of 1litre tubs, partly because I intended to provide this paint to be used on my courses. Rosco has put together a sample box with small amounts of each colour .. costing around £50 if you can get hold of one .. but these sample pots don’t hold very much and once the seal is broken the lids aren’t quite tight enough to keep the paint inside from drying out.

Rosco SuperSaturated range

I needed to find better and more secure containers to transport portioned amounts. I’ve tried the smallest available food containers and for a while these worked quite well, but even these are a little bit bulky to transport in large numbers, if one wants to provide an interesting range. In any case none that I tried were truly ‘watertight’ so I had to bind them for transit with electrical tape. It also meant that the paint could only be properly dispensed from them with a spoon. What it did seem to solve fairly well was the problem of the paint ‘going off’ because I could fill a number of these to the top and thereby reduce the air contact. Nevertheless, I just didn’t like working with them .. too messy!

using food containers for paint

Another of my favourite forms of paint are the bottled acrylics from DecoArt, shown below. These are certainly not the richest colours but in all other respects the paint behaves surprisingly well for an inexpensive hobby-paint! Like the Rosco these acrylics are dense but liquid. But in this context it’s perhaps the bottles that have impressed me most of all .. they keep the paint completely where it’s supposed to be and allow it to be dosed in the smallest droplets needed. I’ve never had any leaks from these when carrying them around and when I’ve refilled empty ones with Rosco paint its appeared to keep for much longer.

DecoArt bottle acrylics

I tried to find identical ones online a while ago but couldn’t. But then, once ‘pound shops’ really started to get corporate and become a feature of almost every High Street I came across these in 99p Store, in packs of four as ‘travelling beauty bags’ .. or something like that .. containing shampoo, shower gel etc. They’re the same size as the DecoArt bottles, which hold a little more than 60ml of paint, but always square and made from a slightly harder plastic. I’ve never had the courage to use the contents!

travelling wash bag

But the bottles are almost perfect for paint! .. at least any paint that is liquid enough to be poured into them. Because they’re not quite as squeezable as the DecoArt bottles it’s best if the paint is cut with a small amount of water and .. if you’re familiar with the ‘ketchup bottle jab’, it will help at times! But their squareness is convenient when packing and, I suppose most of all, at 99p for four they’re cheaper than any options I’ve seen online unless you want to buy them in the hundreds.

using 'travelling bag' bottles for paint

As a postscript to this little piece celebrating small plastic bottles! .. I started trying to find out the best ways of preserving the Rosco paint and others that are predisposed to deteriorate and develop mould over time. It may not be simply a case of having to scrape off the mould layer .. the whole consistency of the paint or medium underneath may have altered and become useless. In addition the mould itself may be harmful to health, so it seems to me that prevention from the outset would be far wiser than making do with scraping the mould off as/when it occurs. What for example is it in the paint or medium that ‘goes bad’ in the first place? Is it simply solved by eliminating air contact? .. packing a layer of cling-film down on the paint surface when storing, or pouring a thin layer of distilled water, or even oil, over it? I was given these and other helpful suggestions when I posted the question on the Society of British Theatre Designers facebook page. A little later I received a very helpful confirmation from Jenny Knott, Paint Product Manager for Rosco Laboratories, which is worth quoting in full:

“Most mold/bacteria growth in paint is caused by introducing it into the paint.  Always use a clean, dry utensil to take paint out of the container.  If you use a wet stir stick to scoop the paint out, the water has bacteria in it so it will introduce bacteria into your paint.
If you have mold/bacteria on the paint, scrap all of it off then you can either put a couple of drops of pine oil floated on the top or plastic wrap over the top of the paint surface to seal it from the air.
Another thing you can do is to float distilled water on the top surface of the paint to protect the surface.”
…………………………………………………………………………………………..
In addition I would imagine that the colder the paint can be stored the better .. apart from freezing! I guess that these safeguards work best from the outset and that probably once the medium has become contaminated there’s really nothing that can be done to purify it again and that whatever can be salvaged probably has a short life, if it still works at all. I’d also imagine that all of the above applies to any water-based medium .. I know that it also happens with Idenden texture medium for example. I’ve just added one suggestion of my own which has worked for me so far .. decanting into smaller, separate, sealed containers.