Reversibility...Priceless!

Windsor Bench on the Bench

I have used different kinds of animal protein glues for over 40 years in all types of applications.  I began using these glues since I was concerned about proper conservation and restoration of period antiques, and that was the only kind of glue available to those pre industrial wood workers.

I thought, "If it worked for them, why change?"

One of the most important rules in restoration and conservation of early objects is that the repairs must be "reversible."  That is because it may happen that subsequent research into certain objects may suggest that the current repair was not done properly and needs to be removed so that more appropriate work can be done.

This applies to glues and finishes.  I remember very clearly the summer of 1978, while I was attending the Summer Institute at Winterthur.  They have in their collection a magnificent high chest made in Boston around 1740 by John Pimm.  It is without a doubt the finest example extant of its type, and is distinguished by the superb preservation of its decorative surface.

I had visited Winterthur many times before and always took some time to stand in front of this piece and examine its surface carefully.  You could see the marks of all the owners in the small scratches and marks of use which were clearly visible, and yet the surface was as beautiful as the day it was made.

That summer, when I went into the collection to visit this piece, I was shocked.  It was covered with a high gloss surface, which reflected the light and made it impossible to see the character of the surface.  It looked, frankly, like a piece of formica.  I asked around and was told that it was a new DuPont finish that was being used for "conservation" of the surface.  I was assured that this new finish, should it be determined to be inappropriate, could be removed by a "special solvent" that was designed for this purpose.

There is no reason I can think of to "experiment" with this new finish on one of the most important objects in the collection, and this example of improper conservation represents a dark period in what is otherwise a very conservative museum policy.  I note that this finish has now been removed and the Pimm highboy is supposed to be back in its "original" condition.

When it comes to glues, reversibility is just as important.  All furniture experiences damage at some point in its life, and it is essential that the glues can be reversed to allow for the structure to be taken apart and repaired.  From time to time a client will bring in a modern copy of an oak chair, which was made in some other country with epoxy holding the spindles in place.  I have to tell them that there is no way to repair it, except to start over.  You just loose money with cheap furniture.

All protein glues, on the other hand, are completely reversible, by using heat and moisture.  You can take an Egyptian chair and reverse the glue.  It doesn't matter how old the protein glue is, it will always reverse with the proper amount of heat and water.

Like water, which can change from solid (ice) to liquid (water) to gas (steam) simply by changing the temperature, animal protein glues can easily change from solid to gel to liquid and back again as a function of heat and moisture.  All you need to do is understand whether to add heat alone or water alone or heat and water together, depending on the condition of the glue, and you can make it do whatever you want.

Most people are concerned about "reversibility" of the glue meaning that it will "fall apart" given the right conditions.  In fact, normal furniture construction has a finish, which protects the wood, which, in turn, protects the glue from exposure to moisture.  Heat alone will not do it.  Properly cured glue can stand extreme temperatures, when there is no moisture present.  On the other hand, long exposure to cold water will soften even the oldest glue and turn it to gel.  By long exposure, I mean soaking in water for several days.  Normal furniture doesn't have to worry about that, except for floods.

A case in point happened this week when a chairmaker in Australia contacted me about a glue up problem.  He has been using Old Brown Glue, which he finds perfect for Windsor chairs since it is easy to clean up and has a fairly long open time.

Go Soak Your Head

This time of year it is winter in Australia, and the shop was not heated.  He warmed up the glue and started to put his bench together, but when he went to put on the crest, the glue was cold.  As he tried to seat the crest on the 17 spindles it stopped short.  On one end it was 6mm from seating and the other end 11mm from seating.  Nothing he could do would move it.  He tried hitting with a mallet but it wouldn't budge, and he couldn't get it off.  That glue grips when it sets.

He reached for the heat gun, but that only dries out the glue. Nothing worked, so he contacted me.
I told him that he needed to first hydrate the glue.  The only way was to soak the crest, allowing the water to penetrate the wood fibres and thus reach the glue surface inside the joint.  Once that was done, he could simply heat the joint and it would open.  

He wrote back a few days later with a report of success.  He had soaked the crest and then heated it up, and, in his words, "The crest rail slipped off without a blow."  Problem solved.

200 years from now, if someone wants to remove that crest rail, it will still be possible, since it was put on with animal protein glue.  "Reversibility" is what keeps antiques alive for future generations.

Ready To Glue Up Second Time

Lecount Clock Door

Edwards Clock #2 Private Collection

I have written before about my business plan.  To clarify, it is not so much a "plan" as a "philosophy."  You see, when I decided to work every day of the week, and not take weekends off like the "normal" people in the world, I quickly realized that I would eventually need a break.  That is why, for the majority of my career, I have worked daily for 3 months and then taken a month away from the bench for a "business" trip.

Sometimes my business trip is to Europe or the East Cast to buy stuff, meet new and old mentors and just enjoy the world of museums, historic houses and antique shops.  Sometimes my "business" trip it to the cabins I own on the Madison river in Montana to enjoy the world of the brook trout, eagles and elk.

That said, I also realized early on that there would be no "retirement" for me.  No "pension" or other means of support, except Social Security, which I now am receiving in very small amounts.  Therefore, I needed a plan to resolve the fact that I would work at the bench until they pried the chisel from my cold dead hands.

What I decided to do was calculate the exact amount of income I would need to survive a day of life, and then make that amount every day.  If I missed a day of work, then the next day would require a double amount to keep even.  In general, I need to work for paying clients at least 4 hours a day, every day for life.  Assuming that I arrive at work at 7 am and close up the shop at 7 pm normally, that gives me 8 hours of "retirement" in my workshop, which is fully equipped to provide me with tools and materials for anything I want to do.

This method of budgeting my time has allowed me over the years to create some large projects which involve hundreds or even thousands of hours of unpaid time.  Essentially, I work for a living and, in my retirement, create amazing objects of marquetry covered furniture.

Last week I visited two of my favorite clients, who have generously supported my efforts to keep alive the traditions of this craft.  Together, they own several of my best pieces, including the jewel cabinet I made some years ago, covered with ebony and 32 different marquetry panels.  They also have the second clock I made, with works by David Lindow, and I took some photos since it is very similar to the case I am currently making for Lecount.

The bonnet has pierced work, backed by silk, which allows the tone of the bell to escape and keeps dust out.  I cut the fretwork on my chevalet, and it went very fast.  Here is a photo of the top:

Bonnet with Fretwork and Rope Carved Columns

Note the marquetry in the upper door.  I guess I like birds, as I seem to always include them.

Upper Door of Clock #2

Here is the marquetry on the base.  More birds.  I really like the end grain molding, even though it is a pain to carve.

Base of Clock #2

Now I am making the case for my 5th clock, which will contain the Lecount works from 1690.  I always start my clocks with the backboard.  You can review my earlier posts about this process. Making the clock from the bottom up allows me to enjoy it standing next to the bench (as encouragement) and provides me with actual measurements for the bonnet when it becomes time to make that part.

The bonnet is the most complicated part of a tall case clock.  It has a mask, which fits closely to the face.  It has side windows which allow the viewer to see the works.  It has a glass door, which pivots with the carved columns as the offset pivot point.  It needs to slide onto the case molding and be held by guides so it doesn't fall forward.  It needs to be closed from dust, and it needs to allow the bell to be heard.

But I am not there yet.  Today I was able to rough sand the door itself.  Over the past week I took some time to cut in more yew wood oyster veneers and glue them to the oak door.  After that, I laid the marquetry panels over the oysters and marked the outline, so I could cut away the yew and place the panels.  When that was dry, I carefully cut in 2mm boxwood string inlay banding around all the parts.  Then I glued in the olive cross band edges.  I still need to make the ovolo molding which will be glued to the edges and cover much of the cross band trim.

Here is the door on the bench during sanding:

Lecount Door On The Bench

So, I placed the door on the case and took this picture this morning.  It allows me to see what this clock will look like when it is assembled.  My next job is to start cutting and attaching all the molding on the case, both lower and upper, as well as the molding around the door.  That will allow me to install the hinges and lock to be sure everything fits properly.

Lecount Standing Proud!

After that, I can start making the bonnet.

Painting In Wood Packet

Guide To Species and Green and Mahogany Layers

I promise to make a video soon about this method.  Those of you who have issue 109 (February 2008) of the magazine Woodwork, will be able to look up the article I wrote then about it, with good photographs.

To illustrate how I make a veneer packet for a picture using the Painting In Wood process,  I took apart the packet waste, layer by layer, after I finished removing all the pieces I needed for the final picture.

Most important is the initial layout and proper orientation of the individual species of veneer.  I use the four corner edges of the design as a locator.  Each layer, no matter how roughly put together, must have all four corners match exactly with the drawing.  From that reference, I can confidently locate the elements inside the design so there are no errors.

Note that I usually have at least two different species for each flower, and sometimes three.  That gives me a colorful flower, and, since I keep each of those colors for each flower as I cut them out, I can change my mind when I put together the final picture on the assembly board.  For example, if I have a flower with a light and dark selection, I can decide to have a dark flower with light edges or a light flower with dark edges, depending on my mood.  What is not used is discarded, but not until everything is done.

At the top of this post is the layout design with my notes on the woods I selected to use.  The yellow highlighted areas are isolated pieces of ebony background, so I don't forget to keep them. The layer missing from this photo is the ebony background, which I used and can be seen in the last post about this.  Note the green and mahogany layers are usually full sheets, since leaves and branches are all over the design and too much trouble to try to place individually.  The pieces of veneer tape are left over from taping all the layers together securely.  Always tape each layer one by one to be sure nothing moves.

I try to use the smallest piece of veneer possible for each element of the design.  These pieces are usually small scrap pieces from previous projects, so, in the long run, this is an efficient use of surplus material.  In between each element I place a cheap scrap veneer of the same thickness as the others to fill up the gaps and prevent the blade from tearing up the veneer when I saw.  Since all the veneers I use for these jobs are sawn material, I can use the 1.5mm Sipo wood that I have on hand for the front of the packets.  It is cheap and easy to saw.

Layers 3 and 4 Face Down

Looking at these two layers, note that they are face down, so you can see the woods more clearly.  The white wood areas are the Sipo and not important.  Going down the left side first, you can see the two woods I selected for the Tulip, the two different yellow woods for the next flower, the two woods for the next flower, and so on around the packet.  There are no woods for the second tulip on the top right as they are included in the first and second layers sown here:

Layers 1 and 2 Face Down

The first two layers in the packet are shown here.  Now you can see the two woods I used for the top flower, again on the right hand side the two woods for the tulip and so on.  In the center are the two woods for the blue bird, and at the bottom are two woods for the central leafy motif.

Note that I can select the grain direction for each species to correspond with the design.

It is true that all this wood is thrown away, after I select the single piece from the 7 or 8 layers which I intend to use.  However, I gain a lot of time, as the cutting is very fast.  For example, I might take 4-6 hours to construct a packet of this size, and another 8 hours to completely cut it out.  So in two days I can go from design to assembly, and produce a picture which some may think is complicated.  I just think it's a fun and rewarding way to pass the time.

To see the final picture check out the last post.

Back To Work On Lecount

Marquetry Panel For Lecount Case Door

I have been rather busy lately with the business of business.  I knew that when the jobs and housing markets turned around the phone would ring.  Well, it has, and I am looking at a shop full of antiques that need my attention.  Also, I had two weeks of teaching which is always a nice change in routine.

The Lecount project has been calling me...literally.  The clock works sit on a stand in the corner of the shop and chime precisely on the hour, reminding me that they need a case.  The amazing thing about this set of works is that they run for 30 days!  And they keep time to the second.  Not bad for a set of brass works made by hand around 1690.  How many mechanisms made today will still be running perfectly in 300 years?

During the last week we had a holiday, and holidays mean that I can work without interruption.  So I returned to the chevalet and cut out the remaining panels for the door of the clock.  At the same time I thought I would make a video of the process of building marquetry face down on an assembly board, which is the French method.

I have written before about this method, but I find that using words to explain it really is confusing.  Watching a video is much better.  So I selected a simple flower example to demonstrate.  I should note that the only thing that didn't happen during this video was the phone or door bell ringing.  What did happen was the Lecount clock chimed, the mail man arrived and Bridget kept barking, Patrice stumbled over the stool while filming, and the garbage truck drove by the shop, making it impossible to hear what I said at the end.

All in all, a very professional shoot.  Watch it HERE

I am using sawn veneers, which are 1.5mm thick.  The building process is very easy and you just place the pieces in the picture cavity face down, holding them with hot protein glue.  The glue allows for some adjustment if you do it quickly, so you can move pieces around slightly for a good fit.  Later, if you find a mistake, you can easily reheat the pieces and remove them or replace them.

When the panel is assembled, I will use diluted protein glue and fine sawdust to make a mastic to fill the saw kerf, again working from the back side (glue side) of the marquetry, so the front surface remains clean.

Here are some photos of the process of building a marquetry panel using an assembly board.  The first photo shows the gluepot, assembly board, simple tools and a tray with the parts.  Note I have just completed the first panel and am ready to build the second.  The paper design goes with the parts in the tray for the second design.  All the parts in the tray are laid face down, and the paper design is inverted left to right so I can follow the design easily.

First Panel Done, Ready For Second

Here are the parts in the tray, carefully laid out according to the design.  All the pieces have been burned in hot sand.

All The Pieces Laid Out

Here is the ebony background which has just been laid into the hot glue on the stretched paper of the assembly board.  It is important to place the branches and leaves in place first, so that the background is properly located before the glue sets.  Putting the individual flowers in place takes longer so is done after the branches and leaves.

Sawn Ebony Background Glued Down

Here is the drawing for the flower at the top of the design.  It is made up of lots of pieces, and each one is unique, having a walnut piece with cut engraving lines and a box wood piece which is the lip of the petal.  This type of flower is typical of work done late in the 17th century.

Flower Design Inverted Left To Right

Here are the pieces for the flower, carefully burned and laid out in a precise location in the tray.

Flower Pieces Ready To Install

Here is the flower assembled.  You are looking at the back (glue side) of the flower, which will be filled with mastic to complete the process.

Flower Assembled Face Down

Here is the final picture, glued to the assembly board, ready for mastic.  The entire assembly of this picture took less than an hour.  Much less.

Ready For Mastic