Holywell Glass Studio, St Alkmunds

Me in hundreds of jumpers because it's freezing standing outside Holywell Glass Studio where I worked on a placement.

Holywell Glass is a studio in the rather scenic Cathedral city of Wells in the South of England. It does not do any design work of it’s own but concentrates on restoration, a field I knew nothing about. It is run by Steve Clare and had a workforce of eight not including myself, although all of them were never there at once. It was a very nice studio and the staff were all very friendly and welcoming. My experience was a positive one. I saw Steve only briefly while I was there and was introduced to the studio and told what to do for the first few weeks by a very nice woman Emma Russell who hadn’t worked there for very long but had done a degree at Swansea and had a lot of technical expertise certainly in comparison to myself.

I was in Wells for a relatively long placement lasting two months. I worked on five different projects while I was there. St Alkmunds, Grisaille panel, Compton Pauncefoot panel, tiffany style lamps and Arundel window. From September 29^th when I started until the 11^th of November I only got to work on one project, St Alkmunds, because that was the big commission and the whole studio was only really concentrating on that. From the 11^th St Alkmunds was reinstalled, so I got to observe some of the other work that had been put on hold in the studio. From the 11^th till the 20^th, my last day, I worked on four different projects. I enjoyed the diversity of their nature and feel this meant I learnt a lot from them.

St Alkmunds is a church in Shrewsbury near Birmingham. The East window of this church was designed by Francis Eginton (b1737 – d1805) in 1795. It bore the image of the Virgin Mary surrounded by clouds with a bible, a crown, a goblet and other religious illusions. It was many different shades of brown with reds, purples and yellows but was still quite a dark window. It is quite an unusual piece because Eginton had tried to make it look like an oil painting, something that was popular at the time, and so it had no leading at all. It was held together by a metal structure and over time this structure had warped. The joints holding it together were being pulled in and this broke the glass. Some of the glass was really very badly broken. Only one panel from the bottom two rows of glass survived the weight of the rest pushing down upon it and could be salvaged. Although some of the panels had been mended and some were replaced the whole window had never been taken out before. Ministers of the church had painted cold colour quite badly onto the glass at different times and some of the broken panels had been plated to keep them together. No-one was sure exactly when this had happened but it was a long time ago.

I arrived after the glass had come to the studio although barely any restoration had been done so I came early on in the process. Each panel was photographed, cleaned, inserts cut and fired, resined, cold coloured, photographed, copperfoiled and photographed before it could be re-installed. When I arrived the first set of photographs had been taken but everything else still had to be done. Most of the work was quite straight forward but labour intensive. The window had about ninety panels and several had two or three layers of plating. Some had to be completely repainted and the project ended up being a few weeks behind schedule.

This is a broken panel that is being cleaned

The first thing I was to do along with my colleagues was to clean the panels and prepare those that needed it for resining. The glass was cleaned with fifty percent acetone and fifty percent water (50/50) or using synperonic a soapy cleaner. These liquids were applied again and again as the surface was gently rubbed with cotton buds and fiber glass brushes to remove the thick sooty residues that had built up over time. Panels with a lot of cracks could take hours to clean so it was a long process. Some of the cracks had been puttied over in the past as a way of holding them together and it was very hard to get this off again. A surgical scalpel had to be inserted under the putty and hammered to break it off. Even after it had been soaked this temporary measure still often resulted in more of the glass breaking.

Some of the glass had been plated as a way of holding the cracked pieces together. Some kind of animal glue had been used for this and it was near impossible to get off and left some of the glass pitted. To remove the glue sections were soaked in pure acetone and pulled at from time to time to help loosen them. This did work a little but really stubborn bits still wouldn’t come apart. The most broken panel was the surviving bottom row panel and it had cracked into miniscule pieces. About five of these small pieces were still stuck onto a plated piece even after the rest had been wrenched off and to remove the remaining parts it was put in the kiln. The first time this was tried the glass seemed to come apart but was left to cool down and so stuck back together. On the second attempt they were separated hot and triumphantly slotted back were they belonged. Nobody was sure if this would work but it worked beautifully.

To resin a panel back together the glass must first be completely clean. After cleaning it as before the crack lines need extra cleaning. Gary Flex an abrasive block is rubbed along crack lines to remove dust. 100% acetone is then used with cotton buds, going over these cracks until it is completely clean. Sometimes while cleaning you’ll notice hairline cracks, these cannot be resined because they will just continue to travel so they must be broken then properly resined back together. I broke them using grozing pliers. The glass was badly annealed and under a lot of stress and so when it cracked it would curve in unusual directions sometimes at right angles something I’ve never seen glass do before.

This panel has been taped together ready for vinyl

Once all the cracks are dealt with Magic tape is cut into tiny slivers and stuck along the crack lines holding them back together. Some of the glass was quite warped so it was difficult getting it to match up again precisely. This is very important however, for aesthetics as well as strength. If you rub your finger along the surface of the glass you should not be able to feel the cracks if they have been lined up properly. This needs to be done really tight too so light can’t be seen coming through them badly.

Once it is taped the panel must be turned over. Clear vinyl is cut to the shape of the cracks, so it is larger than them but not too large. The reason we don’t cover the whole thing with vinyl is because it’s sticky and we don’t want it to remove loose paint. Caution must be taken before applying it because, stick it on the side that has the least paint. The vinyl should be placed down slowly, taking care to minimize trapped air bubbles. It should then be rubbed all over, especially on the crack lines, with the back of a scalpel or something similar to make sure it’s stuck down properly.

Dental modeling wax called ‘Anutex’ is used next for support. It holds to glass in place while the resin dries. It is cut a little smaller than the vinyl but also follows the crack lines. The reason it is cut smaller is because it leaves a residue on the glass that is hard to get off. The vinyl will protect the glass. When the vinyl and the wax are cut there will be some areas that the edges of vinyl will meet vinyl, and wax will meet wax. Either because the vinyl/wax is too small to cover it or because it’s a tricky shape. When this happens make sure these joins are not taking place over complex spidery cracks because the resin can seep up between the joins.

This panel is in the process of being resined.

The wax is hard then it comes out of the packet. Once it is cut to shape is can be heated with a hairdryer so it softens. Once soft it can be moulded to the shape of the glass. The wax will then harden again and give an even surface that the resin can be applied to. It will support the glass so the cracked pieces will stay at the same level to one and other and not sag.

Once all of this is done the resining can begin. Holywell used 2020 Araldite by Huntsman found at www.araldite.com . It is described as ‘a transparent epoxy ideal for bonding glass and ceramics’. It comes in two parts, part A and part B. Part A contains bisphenol A (epichlorhydrin) and part B contains isophorone diamine. The resin is only activated when both parts are combined. To use mix ten parts of A with three parts of B. For example ten grams to three grams making thirteen grams of liquid. We measured it out using electric scales and a dish to mix it on collecting the liquid with separate pipettes. Put each liquid in a separate part of the dish incase you make a mistake and then mix it at the end. Mix well. The glue should dry completely within 24hrs if it’s warm enough (the studio was really cold so sometimes we’d leave it overnight on the light box for warmth) but is only workable for about an hour and a bit so if your doing a lot of resining it’s better to mix up less then mix more, rather than mix lots but have it set halfway through.

The resin was applied in tiny dots along the line of the break. The resin will fall down into the crack and run along the crack line. One the dot of resin deflates into the crack another dot is added in the same place until eventually after a few times the crack is full and the resin has nowhere to run so the dot remains. All the dots must stay really round like this at the end of resining so that when the resins almost dry (around 20hrs approx) they are easy to ping off. It is done in dots rather than just painting lines because that would make it hard to remove and also may leave a visible residue. The residue can be removed a little with acetone but other chemicals are available that remove araldite although the studio had none of them.

Once the panel is dry and dots have been pinged off with a scalpel the vinyl and wax can be removed. When held to the light you can see whether or not gaps remain between dots. Sometimes the resin does not run properly and you need to resin the other side as well. It is hard to tell on heavily painted areas but what you are looking for is darker patches that catch the light, that glint at you. Those areas still need resined. The resined areas become invisible because the resins refractive index is the same as the glass, meaning the light passes through it in the same way. It is therefore very effective.

Resin can be used to fill small missing sections and shales where glass is missing, but for bigger areas a new piece of glass must be cut and painted to match before being inserted into the panel. Conservation diagrams are made for every panel showing where the cracks are, where resin infills are and where new painted inserts are. This is so in the future if it is restored again somebody in theory will still have these records and be able to pass them on.

On the 21^st of October after a lot of cleaning and resining had been done we started cutting pieces to be used as painted inserts and Emma Russel and Dan Humphrys both employees, began the job of painting new panels where the old ones were either missing or badly replicated. This was extremely difficult. I had a go and wasn’t very successful at it. The problem is all the original paints and enamels are no longer made so thousands of paint combinations have to be gone through to make new ones that look the same. I think this is a skill that must take years to master, I do not have a good enough knowledge of all the different paint types and companies to be able to do it and even the experts really struggled with some of them.

I cut shapes for inserts into the glass out of equally thin glass and got it to the right shape by using a grinder. They were later painted with a mix of enamel and tracing paint, fired, and resined into position. When I rather unsuccessfully tried to match the colours myself I used red for flesh, umber brown and bistro brown combinations of tracing paint along with red, purple and blue Kansa Craft enamels. Mixing these together in different quantities, different thicknesses with different brushes creates different effects, but I still never got it to look exactly like the original. Emma also suggested mixing powdered flux into the paint to make it thinner/more transparent while retaining the strength of colour.

Once the inserts were properly painted, fired and resined into place they, along with the rest of the crack lines and resin infills, need cold coloured. I always thought one needed a special substance for cold colouring but in reality we used ordinary acrylic paint. In particular we used Daler Rowney ‘cryla’ artists heavy body acrylic colour. This was preferable to oil because it dries quicker. It was combined with slow drying gel and glaze mediums also by Daler Rowney. It is important to use the most expensive acrylics because the qualities better so it will last longer. This is mixed to a colour similar to the glass and painted on in dots with a tiny brush to block out the light and hide the cracks. Anywhere glints of light sneak through the acrylic is applied. We did however leave areas where the paint had simply worn away because we wanted to conserve them and thought it ethically wrong to touch them up with cold colour.

This shows some of St Alkmunds panels laid out on a lightbox to be resined.

Once this was done it was time to have the panels photographed. From the 27^th of October till the eleventh of November I photographed all of them. They needed a photo of the front with reflective light and then with transmitted light, before capturing the back with reflective light. All surrounding light had to be removed from the image using black paper cut outs and each was photographed with a ruler to show size. Once they were photographed separately they were copperfoiled and photographed in plated layers.

The copperfoiling was very straightforward. Plated panels had silicone glue smeared along warped edges so pressure wasn’t put on them before being copperfoiled together. This helped to keep them in the correct alignment, helped for transportation and was a preventative measure to stop moisture getting in between the glass and creating mould in future. They were not solderered so they would remain flexible to stop the glass breaking as the building moved.

We made up spring clips to hold the glass into the metal frames in the church. This was done by putting a dot of silicone glue onto a clip before sticking a ‘D’ section of rubber onto the top of it. When dry the rubber will cushion the stainless steel clip, which would otherwise be pressing into the glass. Several of these were needed for each panel of glass so we put a lot of them together.

On the 11^th of November St Alkmunds window was re-installed. This was done up until I left although I was not permitted to be involved in this. I found that to be a real shame because I have no sight experience and even just to watch something like this being installed would have been of huge educational value to me. Instead myself and one other employee were left in the studio when everyone else went on sight. I will make a point of visiting this church under my own steam in the future if I can find it because I feel like I have unfinished business with it and having spent such a long time on the project, not to see what it looks like reinstalled is very sad.