Reversing Disaster Part 1

This video is brought to you by Squarespace. From websites and online stores to marketing tools and analytics, Squarespace is the all-in-one platform to build a beautiful online presence and run your business. All right Bob, well let's see your painting. Wow, huge difference from what it looked like when it came in. Yeah. And I see that the chair cleaned up. So the most important part of the painting, the thing that was really, really critical. Should I explain it? Yeah. Bob's painting, believed to be by the artist Adrian Brower, depicts an interior tavern scene filled with people gathered around playing cards, drinking, and celebrating. But there's no joy on the back of the painting. An ill-fated conservation attempt to flatten the piece by adding a large cradle has resulted in problems. The cradle itself isn't even that good. It's not attached in the corners and it's just an assembly of wood. It's not well bonded to the panel. The joints are starved and there are many gaps. And the presence of this cradle is really the problem that we have to address. And from the front, once we switch to raking light, we can start to see those problems. The paint has puckered and is lifting throughout the painting. Wood panels need to move with humidity changes. This panel had a slight bow to it. And in an attempt to flatten that panel out by adding that cradle, they effectively immobilized this piece of wood. It was no longer able to expand and contract and move with humidity changes. Yet humidity changed. And that tension built up in this panel and had to go somewhere. And then the panel split. Because when the panel is unable to deform as it would naturally do otherwise, it splits. And now we have all of these checks and cracks and splits throughout the painting. In addition, the paint layer itself has puckered and lifted up where that wood has then contracted. All of this is compounded by the fact that on the surface of this painting, there is a very heavy coating of a synthetic substance. But none of these problems appear to be insurmountable. Yes, they will take labor. They will take creativity and effort. But we are up for that challenge. Because what lies beneath this damage is a beautiful, stunning painting that's just waiting to be set free from all of this lousy work. Looking at the painting under UV, we can see some obvious retouching here circled in red. This is fluorescing very strong and sitting on top of the surface coating. On the side, circled in yellow, we can see some of the tests I made and the original paint layer underneath. But down here on the bottom and on the top, highlighted in green, are areas that I know have retouching because I can see the fill-in on the edge, in addition to the paint coming over the edge of the board. But I can't see it fluorescing. It's not reacting to the UV. And that is a big question. In addition, we know that something is going on with the toilet. But what? That remains a big mystery. While our job is to undo this work and reverse this disaster, it could have been avoided by doing things right the first time. And while you may not be conserving paintings, if you ever have to build a website, do it right. With Squarespace, a successful website is almost a foregone conclusion. From creating an archive of beautifully designed, mobile-friendly templates, to eliminating the need to understand or write code, and developing a platform that's as simple to edit as drag and drop, and clicking to add features whenever you like. Be it a gallery, an e-commerce solution, search engine optimization, online scheduling, and many more. And Squarespace knows that just wanting it doesn't make it so, so they have robust help features, if in case you need them. So head over to squarespace.com for a free trial, and when you're ready to launch, go to squarespace.com slash Baumgartner to save 10% off your first purchase of a website or domain. Before I can do any work on this painting to address the multitude of problems, I need to take some precautionary steps so that no further damage happens. And here, I need to inject an adhesive into and underneath the voids where the paint is lifting. Now, as I explained, this painting had suffered at the hands of somebody who thought they knew what they were doing, but inadvertently caused more problems through the actions they took in an attempt to stabilize this panel. The addition of that cradle caused the wood to not be able to move, which caused the wood to split, and the heavy coating attached to the paint layer created this flexible area where we have voids now. Once the adhesive has been heated up and thinned out to a consistency that will pass through the needle, I can inject it underneath. I'll have to puncture the paint film, but that's a small price to pay if we get some stability out of it. I'm going to flood this joint until the adhesive comes out the other holes that I've made. That's how I know that I have enough adhesive in there. I'm not looking to be terribly neat or clean here, I really just want to make sure that there's a ton of adhesive underneath this lifting paint. Some of it will be absorbed by the wood, and some of it will sit there and slowly dry for the next step where I can try to stabilize this. It's slow going, and I'm hopeful that this will work, but there's no guarantee. Sometimes the wood absorbs the adhesive completely, sometimes the adhesive just isn't thick enough, and sometimes the bond between the paint layer and the coating that's on top of it is stronger than the bond between the adhesive and the paint layer that I'm injecting. Once the adhesive has had a couple of days to dry, I can come back with some silicone release film and my tacking iron and begin reactivating that adhesive. Once the solvent evaporates, we're left with just the adhesive. The solvent thins it out so that it's easier to maneuver and easier to handle. And then by applying some heat, I can reactivate that adhesive. So a little bit of heat applied to the surface, and then I'll press it. It's not terribly, terribly hot, so it's not shocking on my hands, but pressing it is the best chance I have of securing that paint down. And you can see here, it looks like it did a really good job, so I'll continue on with this stabilization technique. Now, in the past, I have had both tremendous results with this approach and not-so-great results, and unfortunately, it's one of those things where you have to do it to find out if it works. Unfortunately, I don't have any other method. If this were on canvas and it were flaking, I could apply the adhesive to the back of the canvas and allow it to penetrate through the canvas, and then use the heat table to bond all that paint down. And I've done that many times, and that works phenomenally well. But because this is a solid wood panel, I can't take that approach. So this is the best option that we have. I'll continue tacking down this paint layer and hope that it holds. I'm optimistic, but as I move on, I'm also very, very cautious, and my optimism starts to get a little bit hedged. While it is bonding back down, I'm nervous about the surface coating and about how strong it is. It's not cracking, it's not flaking, it's not moving at all, and that leads me to believe that it's really, really well bonded to the paint layer. And if that's the case, all I can hope is that this adhesive is stronger than the bond between the top coating and the paint layer. If that proves to be the case, then the paint layer will bond back down to the ground or the wood. If not, well, we'll see. And you can see a lot of the adhesive oozes out, and that's totally okay. That's extra adhesive. Once the adhesive is dry, I can come back with some solvent and remove the excess from the surface. This is mostly just to have a clean surface and to get rid of adhesive that I don't need, but ultimately, all of the surface coating will come off. The next step is to face the painting with Japanese paper, and I'm doing this for a couple of reasons. One, because the paint layer is really unstable, and two, I'm going to be handling this panel pretty aggressively, and I want to protect that paint layer. Yes, there is a very thick coating of, well, I don't know what on it that will protect it, which is why I've left it on, but this paper will be extra protection. And if any paint does become detached, this paper will hold it in place. So I'm applying the adhesive directly to the paint layer itself, or at least to the painting itself. The old surface coating is protecting the painting. I'll lay the paper on, and then I'll apply more adhesive, and this ensures that we have a really, really good bond between the painting and this paper. And then I just let it dry, and we say goodbye to this wonderful little tavern scene. At least for now. It'll be back later on. After allowing the adhesive to dry for about an hour or so, I can begin thinking about the next step, which is removing this cradle, or this assembly of wood that is pretending to be a cradle, because realistically, it's not actually doing anything beneficial to the painting. It needs to come off because it's one of the sources of harm. I'm clamping this panel into my bench using a piece of felt to protect the face of the painting. I don't need a ton of pressure here. I just want to make sure it stays vertical. And then I can take a palette knife, and I can start working the edge. This cradle wasn't ever really well bonded to the panel. In fact, there are multiple voids in the glue joint, and you can see just by inserting this palette knife and giving it a little bit of wiggle, I can start breaking that glue joint and separating this cradle. There are many ways I could remove this cradle. I could use my router sled and turn all of this oak into sawdust, and I considered that. But before doing that, I want to see if I can just simply break this joint and separate it. If this works, there's no need to use power tools, and I'm optimistic that this will work because, well, it is working. Whoever glued this did a lousy job. And I have said this so many times, I feel like it might need to be a t-shirt or something, a sticker that I put on my door, that sometimes lousy work is our best friend. Had whoever did this work prepared the panel well and used a proper adhesive and used enough of the adhesive, this just simply wouldn't be possible. There would be an incredibly strong joint between the cradle and the wood panel, but because they didn't do that, including not even cleaning the back of the original panel, that glue is mostly bonded to old grime and dirt, and there's not enough glue. So I can simply break this joint and hope that the cradle separates. Thus far, I'm having pretty good luck. This glue is really brittle, it's not a flexible adhesive, and so just by tapping in this palette knife, it just breaks it pretty easily. And I'm monitoring the front of the painting to make sure that there isn't too much flex on the panel, and ultimately, it seems to be responding really, really well. This isn't how I thought I was going to have to do this. I thought I was going to do this and it would fail, and then I'd have to break out the router sled and make a whole bunch of sawdust, but I'm pleasantly surprised at how easily this is going. So sometimes, sometimes we get lucky. And in this case, at least up till now, I feel like the stars are aligning. After working it for a bit, I can take a hammer and see if I can just tap it loose. Again, I'm optimistic this will work, but not confident that, well, it was that easy. You see, the cradle comes right off. And you can see the cradle wasn't attached to itself, even though those joints were mitered. There were no biscuits, there were no dowels, there weren't even any nails, so it wasn't a unified cradle. It's just a bunch of pieces of wood stuck to the back. Again, somebody thought they knew what they were doing. Maybe they had seen a video on YouTube and they were mimicking that. But seeing it and mimicking it doesn't make it good, and this is a prime example of bad work. Now, even though the ends came off really easily, these crossbars seem to be attached, either with glue or with dowels or biscuits. So I'm taking my Japanese pull saw and I'm just cutting through this joint. You can see right where my blade is now, it looks like this was a biscuit. There's a telltale sign of a biscuit slot being cut into these crossmembers. This isn't terribly difficult work, it takes a little bit of time, but the Japanese pull saw is really great because it's a very thin kerf, so it doesn't make a very wide cut, and it's very sharp, so it works pretty efficiently. Once those have been cut through, at least I hope they've been cut through, I can return to the bench and start removing them just like I did the other ones. Again, I'm very pleased with how this is going. The amount of time it's taking is much less than I thought, and everything is coming off pretty easily. You can see here, another piece, just tapping it loose, comes right off. Yep, there were biscuits, and I'm glad that I cut them, because otherwise this wouldn't have come off. On to the next side, and well, same thing, just tapping it a little bit breaks the remaining glue joint that I didn't break with the palette knife, and this piece of wood comes right off. Now we have just the remaining crossbars on the panel, and just like I removed the outer pieces, I will continue to work these and remove them. And just like the outer pieces, these come off very easily, tapping in the palette knife and breaking the glue joint, and then they separate. And here's what we're left with, a panel free of lousy cradle. You can see the areas where the pieces of wood were glued down, because they're clean. What I'm also noticing is that the panel's been scored with a knife. You can see all of these vertical cuts they made into the wood. This was an attempt at creating a tooth for the glue to bond to, but that's not helpful if the wood isn't clean. Had this wood been scraped, or had it been sanded, this glue joint would have been really strong. Simply running a knife over the wood and creating some score marks doesn't help. Whoever did this didn't know a lot about gluing or wood or, well, anything related to painting conservation, and so this is the mess that they left. You can see some areas of old repair. It looks like there was a split. And when I take the panel and flex it, yep, you can see lots of checks. These were caused by the panel being under tension and flattened by that cradle. This panel has a slight curve to it, and by flattening it, they either caused those cracks, or they immobilized the wood and made it so that it couldn't move with humidity changes. And anytime wood is immobilized across the dimension it wants to move in, it's going to find another way to move. And in this case, because it couldn't bow or warp, it split. So the attempt to keep the panel flat inadvertently caused the checking and the splits. What I'm doing now is just removing all of the grime and old glue from the back of this painting. And I'm not using a card scraper, I'm not sanding it, I'm using my scalpel, and I'm just trying to slice off and chip off all of this stuff. There's glue, there's wax, there's grime, there's dirt, there's all of this stuff that should have been removed in the first place, but, well, we've seen the quality of work that was executed on this painting, so this is no surprise. Now, this took surprisingly long. I thought it was going to be maybe 15-20 minutes, but it took, oh, about an hour to remove all of this stuff. Now, I certainly could have pulled out my sander or card scraper and gone over the whole back, but that's too aggressive and it would have damaged the wood. So the scalpel was the right choice, even if it took a long time. Because what we're left with is a panel that is largely free of old glue, old grime, old dust, old dirt, and we can see the original wood. Yes, it has damage, but that is what it is. Whoever worked on this painting in the past made those decisions, and we just now have to deal with them. But I'm very happy with how the back of this panel looks now. So now I can turn my attention back to the front of the painting, and I'm going to remove this facing, because it's time to start thinking about cleaning off this painting, removing that old surface coating. This is a solvent-based adhesive, and similar to the water-based adhesive I would use, it can be reversed, except not with water, with a solvent. So I will apply the solvent to the face, let it sit for a little while. Once the adhesive has activated, then I can begin peeling this paper off. And one of the nice things about using this solvent-based adhesive as opposed to the water-based adhesive is that the paper, the kozo, doesn't disintegrate like it does with the water-based adhesive. And one thing that I'm noticing is that as I peel off this paper, some of the surface coating is coming with. You can see it's very smooth and shiny on the back of that paper. This leads me to believe that maybe there are multiple coatings of whatever is on the surface, and just the top coating is brittle and peeling off. There is some excess adhesive that I will remove with solvent. This solvent has no effect on this surface coating. I tested it just to make sure, so I don't have to be terribly, terribly careful. I just need to get all of that excess adhesive from the facing off. Takes a couple of minutes, but ultimately it's not really all that problematic. And once I have it cleaned, then I can start thinking about the next step, which, surprise, is actually not cleaning the piece off. I thought I was going to do that, but then I realized that I had to do all of this structural work, and that keeping the old surface coating on the painting right now would act as, well, sort of a facing. It would protect the paint layer as I work with the clampinator. I designed and built this clamping table many years ago for a very large panel that was split into multiple pieces. I let viewers name it, and they decided to name it the Clampinator. It's much better than Clampy McClampface. What this allows me to do is clamp this panel in place while preserving the natural curve of the wood. The first step is injecting an adhesive, here using hide glue because it is reversible, and it bonds really well to these old panels, into the joints. And I'm going to flood them, make sure that there is a ton of adhesive in there, because I don't want any of these joints to be starved. Because, as we just saw with this lousy cradle, a starved joint isn't a good joint, and it can break really easily. Plus, I can remove all of this excess adhesive later on, so I'm not terribly concerned about any of it getting on the back or the front. And remember, the front of the painting still has the old surface coating on it, so the paint layer is protected. Flexing the joint allows the adhesive to travel, and then, once I'm confident that there's enough adhesive in there, I can place it into the clampinator. And here is the real magic of the clampinator. Because this panel is bowed, if I put lateral pressure from both sides to close that joint, the panel will continue to bow and eventually snap. If I tried to put cauls on the top and the bottom to hold it from snapping, those cauls would keep it flat, and I wouldn't get a good purchase on the joint. So the clampinator allows me to preserve the natural arc of the panel, while still exerting pressure on the sides. So once everything is aligned and in place, I can start adding pressure with the clamps on the side of the painting. These are what's going to hold the glue joint together. Once there's enough pressure on them to start closing up that joint, I can start moving all of these vertical posts, and these will be aligned on either side of the crack or the split. And I can press each one of them up and down to make sure that the panel is bonded flush, even if it has a curve to it. So you can see I'm feeling with my finger and my fingernail to make sure that this ridge is really smooth, and we don't have a ridge. I'll raise the table up. It has hydraulic lifts on it, which makes getting under it easier as I get older and tired of sitting on the floor. And I'll do the same with the bottom. So the combination of pressure on the top and the bottom makes sure that that joint where the split was is held in perfect registration. The clamping pressure on the side makes sure that that joint has adequate pressure to be bonded well together. And because there are so many of these verticals, I can adjust them accordingly. Sometimes this wood is warped in multiple directions. Once I've checked it with my fingernail to make sure it's really smooth, I can let it dry overnight, and then come back the next day and repeat the process all over again. This painting had three or four, I believe, checks or cracks, so I have to do this three or four times. Flood the joint on the bottom, flood it on the top, get it into the clampinator, align everything, and then begin exerting pressure on the sides. Once there's enough pressure on the sides, I can then use the And you can see the curve of the panel. That's what I want to preserve. Without that curve, we wouldn't get a good joint, and this bonding process would fail. And you can see it here again. The panel has a natural curve, and the clampinator allows me to clamp this glue joint tight, preserving that curve. That's the whole point of a clampinator. It's a big, complex machine that allows me to deal with the imperfections of old panels. With all of the structural work complete and the panel stable, I can begin thinking about removing this surface coating. And I tested every single solvent, solvent gel, enzyme, detergent, cleaner that I have in the studio, and nothing made a difference. This surface coating was resistant to all of them, which means the only way to remove it is manually, with a scalpel, by scraping. And this is not my preferred method of removing a surface coating. It is incredibly tedious. It's nerve-wracking, it's stressful, it's incredibly detail-oriented work, and it's very, very slow. But unfortunately, the decisions made by the person who applied this have resulted in this being the only way to remove this coating. And you can see under magnification, it's not one coating. It's multiple layers of a coating, which leads me to believe that this is something that was sprayed on, something purchased at a hardware store that was built up layer by layer. The good news, however, is that this painting was never fully cleaned. There's a very thin layer of old varnish on this painting. And that thin layer of old varnish is the only thing that's standing in the way of this being a forever permanent coating and something that I can remove. Remember when I said bad work is sometimes your best friend? Well, this is another case in point. So I will continue to chip away at this surface coating, revealing the paint layer underneath. You can see that there's a slight yellow tint to this surface coating. That's the old varnish. And I've got some work cut out ahead of me, because even though this is a small painting, this is very, very slow. So join next time to see if I do get all of this surface coating off and what lies beneath.