Saturday, January 7, 2012

Why Use Reversible Glue?

I just finished cleaning up the glue room. Yes, I have a special room for glue. Not for using the stuff, but for cooking it. Over the past decade, the demand for Old Brown Glue has increased exponentially, from a few bottles a week to hundreds a month.

I never thought I would be in the glue business. I only wanted an organic, reversible protein glue to use in my restoration of antiques. I always had the hot glue pot, but there are a lot of jobs where having a slower set time and more liquid glue is an advantage. So, I began to experiment with modifying the Milligan and Higgins 192 gram hide glue that I always use to see if adding urea would lower the gel point.

The exact formula was eventually perfected, and these days the Old Brown Glue takes care of the majority of repairs in the business.

It is not obvious to some woodworkers why it is important to use a glue which is reversible. After all, if you put it together when you build it shouldn't it stay together...forever? Never happens. As the popular expression goes, "If it breaks, it will be thrown away". Never heard of that? Well, just consider the ratio of surviving antique furniture (made with reversible glue) to the amount of surviving modern furniture (made in the last century with synthetic glue).

The reason I started my business, as a furniture conservator in private practice, using only organic finishes and reversible glues is because I wanted authenticity in my restoration. I knew that the old pieces have all been repaired at some time or another, either due to accident or misuse. I found it much easier to take apart and repair those areas where traditional glue was used and much more frustrating to repair damage which had been repaired with synthetic glue.

The old protein glue usually was crystalized, due to extreme loss of moisture, and could be easily scraped away with a toothing iron, leaving a clean original wood surface. Where plastic glue was used, I needed to carefully carve away the glue with a chisel, which usually removed some of the wood surface, leaving the joint damaged. As the plastic glue tended to seal the wood pores, it was necessary to remove all of it, since synthetic glues do not bond to themselves. Any residue of protein glue was not a problem, as the fresh application of hot protein glue would completely bond to the earlier surface. Protein bonds with protein.

Synthetic glues, in general, are made of some chemical which changes into a different chemical with the addition of a catalyst. Epoxy is well known, and you always have part A and part B to mix together to activate it. Other glues activate when in contact with moisture or another chemical. All of them have a similar feature: they do not easily reverse once activated.

Animal protein glues are similar in their working characteristics to water. Water is sometimes solid, sometimes liquid and sometimes vapor, but it changes back and forth from one state to another simply as a function of temperature. Animal protein glues go from solid to gel to liquid and back again as a simple function of both temperature and humidity. It takes both. In fact, Old Brown Glue can be frozen without affecting its quality, unlike synthetic glues.

In the dry state, either as pearls, granulated or solid sheets, animal protein glues will last for thousands of years, assuming they are stored in a cool dry place. If they are in a cool but humid place, mold with begin to attack them. When cold water is added to the protein, it immediately begins to absorb the water and expand, gradually turning into a gel. The speed and amount of water absorbed indicates the quality of the glue. In the gel state mold will begin to form on the glue in a few days, damaging the glue. That is why it is important to heat it as soon as it gels.

Raising the temperature in a double boiler changes the gel to liquid. Cooking it as much as you want between 140 and 160 degrees keeps it ready to use and prevents the mold. I cook my glue in the glue pot about 12 hours a day, every day, all year long. From time to time I add water to keep the viscosity rather thin. On cold glue in the morning I add cold water. During the day, when it is hot, I add hot water. Keeping a plastic lid on the glue pot helps reduce the evaporation. I always keep a stainless steel meat thermometer in the glue to monitor the temperature.

Protein glues set initially by loss of heat, which only takes a short time, and then fully cure by loss of moisture, which can take overnight. Old Brown Glue sets by loss of moisture, so it has a much longer open time and cannot be used for hammer veneering or rubbed joints. On the other hand, it has a very liquid viscosity which allows it to penetrate deep into fine cracks. It also cleans up easily with cold water, using a sponge, toothbrush or paper towels, and doesn't damage original finishes on fine antiques.

The secret to reversibility of protein glues is that you need to both hydrate and heat the surface of the glue. Heat alone will not do it. Once you figure out how to get water and heat to the glue it will liquify. It doesn't matter how old the glue is. I can take apart 18th century joints and apply a hot wet rag compress to the joint and easily remove all the old glue. Doesn't take long and conserves the wood joint as it was originally made. Add fresh glue and clamp. Easy.

There are other advantages. Last week I finished cooking a large batch of Old Brown Glue and my wife came in and said she wanted to go to lunch. I said that I just needed to stir the glue once more and I would join her. Soon after we were sitting in our local eatery across from each other and I noticed a strange look on her face as she looked at me. Looking down at my favorite Pendleton shirt I was shocked to see gobs of glue running down the front onto my pants. I guess I had been in a hurry to stir the glue and splashed it all over me.

It was a very unpleasant lunch. I got all Monk on myself, thinking everyone in the place was staring at me. I was sticking to the table, the napkin, the chair. We quickly ate and I returned to the shop, taking off my shirt. By putting it in hot water I was able to completely remove all traces of the glue. When it dried you never knew anything was wrong.

Imagine if that was a toxic, plastic, non reversible, synthetic, modern glue..."If you get that stuff on your Pendleton, it will be thrown away!" Surely you've heard that old saying!


Anonymous said...

Excellent write-up. In my previous comment (which I hope inspired this post) I mentioned I'd tried alcohol to crystalize the glue (based on google search results) Any more comments on that technique or should that be avoided and only use heat and water? I am usually a little averse to water based on grain raising, but maybe because the showing parts will already have finish on them it won't matter?
What sort of time and volume of water are we talking about? (rag/sponge/cup/bucket) (seconds/minutes/hours)

W. Patrick Edwards said...

I have several 19th century period references which mention techniques for removing veneer. One says to coat the surface with linseed oil and place near a fire. This method uses the oil to transfer the heat through the veneer to soften the glue, which I suppose still has a moisture content sufficient to liquify the glue. Another method mentioned is to place the board with the veneer side face down on a bed of wet sawdust and leave overnight. This method works fine, but it will warp the board, and the veneer needs to be pressed between newspaper to dry out flat.

Conclusion: prolonged exposure to cold water alone will reverse the glue, but I only use this method to salvage materials, not for delicate conservation of fine antiques.

I prefer steam, where conservation of the finish is not an issue. Usually the finish is shellac so it can easily be repaired if damaged, however.

I also have successfully used a special hammer which I have just for this purpose. It is made with a paper head. The paper is saturated with glue and makes a firm surface which doesn't damage the wood. By tapping repeatedly on the joint I have been able to break old joints open without damage. You need to be patient and understand how much force and from what direction will work.

Aside from that, you need to be creative to get the water and heat to the glue, where joints are not easily accessible.

In some cases, I have also used alcohol, but I find water and heat methods have proven easier and quicker.

Anonymous said...

Excellent info. Thanks for the time spent to help me understand what it is that is going on. Once a principle is understood (in this case how hide glue works) I can apply it creatively as the piece dictates (as you have noted) Thanks for shining the light on "forgotten by most" techniques.

W. Patrick Edwards said...

Je vous en prie. C'etait un plaisir de vous aider.

Anonymous said...

In addition to the bent lamination, I'm thinking my next piece will have sheets of copper applied to two wood surfaces...somewhat large surfaces, say 18" X 48". Is Old Brown Glue right for this and if not what you suggest?
Thanks, Paul

Joe said...

I think you should be able to, if you etch the copper surface with garlic first.

W. Patrick Edwards said...

We did a lot of study on this issue in France, with serious lab analysis and accelerated aging techniques. We were working to understand the best ways to glue metal to wood, since we were concerned with restoration of Boulle surfaces.

The primary physical problem is that wood and metal move in opposite directions during environmental changes. Heat causes metal to expand and lack of humidity causes wood to shrink. Cold causes metal to shrink and excess humidity causes wood to expand.

Therefore, the decorative surface is at risk of lifting during these changes.

For that reason the early European cabinetmakers generally used fish glues to attach exotic materials to wood surfaces. Fish glue has much lower sheer strength than animal protein glue. It "relaxes" its grip during changes in heat and humidity and allows the surface to slip slightly, while still holding it down. We also did research which confirmed that using a fresh clove of garlic on the metal surface prior to gluing provided a better bond, as the garlic worked to mitigate the surface oxidation of the metal. Finally, the glue side of the metal must be toothed, using a toothing blade or hack saw blade, scraping grooves in the surface to increase surface area for the glue.

The biggest problem with Boulle restoration occurs when repairs are made using nails or epoxy. By fixing areas of the surface with these repairs, there is no allowance for surface movement, and the rest of the surface will quickly detach during heat/cold and wet/dry cycles.

Not to mention the problems with removing nails and epoxy during restoration. By contrast, fish glue can be easily removed with cold water.

To answer your question, use fish glue not Old Brown Glue for metal to wood bond.

Anonymous said...

Thank you very much. You're a wealth of real information.