Upholstery Conservation

Burlap in place over springs

Original straw stuffing in place

New Burlap tacked over stuffing

Original Spanish moss in place

Muslin covering

Final Results

Upholstery Conservation

A few weeks ago, one of my marquetry students stopped by the shop and asked me if I wanted a chair. In his pickup was this chair, made around 1880 or so, with all the original upholstery intact, although in very poor condition. One foot was broken off and a carved finial was missing.

I cannot say "no" to free antiques.

After the repair and refinishing, I began to conserve the upholstery foundation. Note the previous post on this site where I am taking it apart.

The method used to upholster this chair is amazing. During the 1870's the art of upholstery reached a very evolved state, and the taste in furniture was for elaborate and complicated upholstery designs. Not only was the fabric and trim exotic, but the method of tying springs to achieve comfort was very advanced.

The opportunity to examine and conserve an example, which has survived for over a century was too much to pass up. Here was another chance for me to sit next to the master who created the upholstery on this chair and learn from his work. Layer by layer, as I took it apart, I understood what he was doing and why.

The selection of stuffing, for example, is revealing. The initial layer of stuffing is straw, which was not expensive, and provided a rather hard edge when stitched. On top of that was Spanish moss, a medium expense, designed to provide a medium layer of resiliency. On top of that was curled horsehair, one of the most expensive materials, chosen to provide a softer top. On top of that was a thin layer of wool cloth batting, which prevented the hair from sticking through the show fabric.

Of course, the burlap and muslin was rotten and torn; that is usual. The proper conservation method is to replace the burlap and muslin, while conserving all the original stuffing, and adding stitching to hold it all in place, like it was done originally.

This series of photos shows the work to restore the seat and sprung back, up to the muslin. The seat has a wire front edge, so the stuffing must be sewn to that, as well as on the sides of the seat to hold an edge. The back is interesting, in that it contains 4 small springs in the center of a stitched edge. As you can see, I needed to re stitch the edges and sew the springs to hold them in place.

The last photos show the chair in muslin, ready for show cloth. I haven't covered the armrests yet, but they are in place.

Sitting in this chair is like sitting in a catcher's mitt...you fit so comfortably you do not want to get up.

Italian Empire Dining Table Project

We are still working on projects in these difficult times, but two things have changed. We are accepting more diverse types of jobs these days, like gilding, sharkskin/parchment, upholstery, etc. And we are not able to produce the continuous flow of work we had previously enjoyed for several decades.

In the past we normally had a minimum of 6 months backlog of good work sitting in the shop. At the same time we were putting jobs out the front door, new work was being bid and brought in the back. Therefore, we could work full time and never see the end of the work. The importance of this backlog of quality work, properly bid, meant that I could continue to bid new jobs at the best price.

It also meant that I could bundle projects with similar activities for each week to gain maximum efficiency. One week might be spent hand sanding all jobs that were at that stage. The next week the shop could be cleaned and all those projects would begin the finishing operation. After that, I could spend a week upholstering, since the shop was clean. In a small, two man shop with not a lot of space, this method of processing the work was very profitable.

These days we still have work, but the backlog is down to 6 weeks instead of 6 months. Also, we do not have the quantity of similar jobs, since we are accepting more diverse projects, and it is difficult to do the same activity for a full week. The result of this is that the shop needs to be reset more frequently during the week for each type of project. Much less efficient.

At the same time, a large project, which was not a problem in the past, can completely block the flow of work.

We are building a large Italian Empire Table and matching chairs for a client. It has taken us longer than we estimated, and the client has been patient. (Note the recent post: It Always Takes Longer...) Although it has been in pieces all over the shop for several months, we can easily visualize it in our minds. The client, however, can only look at the parts laying here and there and wonder if we know what we are doing.

Lately, the parts have been coming together. The bases are nearly assembled, the gold bronze mounts will be shipped from Paris next week, and the top is completed and sanded, ready for the polishing to begin.

The top has been on my workbench for the past month. You can imagine how that has interrupted my work space. It takes two men just to move it, so I cannot do it alone. It has been completely fabricated using only hand tools. All the joints were made with hand planes and the top and back of the top were surfaced using a sequence of hand planing and sanding methods.

The single mahogany board was cut into four pieces and the edge was cut in for the Greek key using a basic old fashioned router. Not the router you plug into the wall, but a chisel mounted vertically in a piece of wood, set to a 2mm depth. Using this tool around the top, I was able to remove all the wood evenly for the Greek key to fit flush with the solid top. Then I added a solid mahogany edge to protect the veneer from damage.

The reason I started this post talking about the flow of work is that I need to make several dovetailed drawers and I hope to have this Empire table off my workbench and on its own base soon. I miss my workbench; I haven's seen it for a month!

Simple Example of Vector Clamping

I was just looking at the photos I posted of the chair with all the clamps. Unless you go through a sequence of applying the clamps it is difficult to fully understand why each clamp is placed where it is. After dozens of clamps are in place it just looks like a mess.

I may be accused of using more clamps then necessary in my repairs. That may be simply because I have collected more clamps then a single person should have over the years. However, it is also because I study each repair carefully to fully understand where all the forces need to be applied.

If the part is shattered or previously repaired badly, it is necessary to take apart all the fragments and clean the glue off of each using small chisels or a small toothing plane iron I keep for that job. I use an Optivisor to see more closely what I am doing and I am extremely careful to not remove any wood.

Then it is important to begin the repair by gluing several small parts together, wait overnight, and continue adding fragments day by day until the repair is complete. Sometimes with chairs I have more than a dozen fragments to assemble just for one joint.

In other cases, the repair is simple and easy. Just remove the modern glue, clean up the wood surface and add protein glue.

In the case of this American Empire center table, pictured above, the two rear legs had come loose. Since the top of this table is marble, there was a lot of force on the joint, which was originally held by three dowels. The owner had had some person try to glue it with the "strongest glue on planet Earth" with horrible results. Part of his problem was the glue and part of the problem was the failure to understand clamping pressure on this joint.

Note the central axis of the joint, perpendicular to the face of the plinth, extends above the toe of the foot, out in space. There is no way to put a clamp there, unless you add a block. That illustrates very clearly what I intend to show about vector clamping. Just by adding a pine block of wood on the foot I was able to create a purchase for the primary clamp. The red clamp is the primary clamp. Since gravity also creates a force downward on the clamp in this position, I added a second, orange, clamp to compensate. These two clamps are all that are required to achieve a correct vector force on the face of the joint.

The entire operation took 15 minutes and made me a profit and the owner of the table happy. I have always guaranteed my repairs for life. Sometimes the project will return for more repair but never in the place where I repaired it. I can say, with some satisfaction, that I have never had a repair failure in my career of restoring antiques. Do it right and forget it.

Eaten Alive!

I hate bugs that eat wood. Build a house, spend a lot of time and money and what happens? Termites move in from next door and consume your equity in short time.

I once visited an elderly client who lived alone in a very large, rich house. She was the kind of person who, in her late 80's, would wander around in this 10,000 square foot house all day, barefoot and in her nightgown. She called me about bugs in her clock.

When I arrived, I walked into the entry and noticed a beautiful early Georgian tall case marquetry clock. I didn't see any bugs, so I removed the bonnet to look at the works. When I pulled the bonnet forward a kilo of live subterranean termites fell to the floor and started crawling all over the place.

I was amazed at what I saw. Inside the bonnet was a large colony of termites, which usually don't eat hardwood furniture. Moving the case from its position at the wall I discovered a single hole in the floor, under one of the feet of the clock. The termites had eaten through the oak floor, up into the foot and continued along the body of the clock into the top, where they set up shop. The entire clock was crawling with bugs and I put it into several layers of plastic bags before I removed it for fumigation.

At the same time I had to remind the elderly lady to keep away, as she insisted on walking all over the bugs crawling over the floor.

Generally, hardwood furniture is attacked by a pest called a powder post beetle. They leave small exit holes in the surface, which serve to provide a convenient place to remove their waste, called "frass". Frass can be distinguished from fine sawdust in that it feels like small round pieces of sand, where sawdust feels more like baby powder.

There is another bug, which has a more morbid reputation. The Death Watch Beetle is so named because during the 18th and 19th centuries, when a person was on his death bed, his friends would sit through the night next to him and quietly wait his demise. Often the only sound in the room was the loud noise created by the Death Watch Beetle, as it ate the furniture.

The Death Watch Beetle is a much larger insect than the powder post beetle, and the size of the exit holes is the clue which animal you are looking at.

It is fairly easy to determine if something is actively infested. One way is to use a stethoscope to listen for the sound of the bugs as they have dinner. The easiest and quickest way is to use sun light or a strong flashlight and look at the inside edges of the exit holes with the raking light coming from the side. The holes that appear fresh indicate activity. The holes that are filled with dirt or wax or oxidized are old and the bugs have probably moved on.

These bugs are hard to kill. The eggs can survive two weeks in a vacuum. The Getty has developed a method to use nitrogen inside a bag which is able to keep the oxygen out. If the bag keeps all the oxygen out for two weeks the bugs will die. Even a single molecule of oxygen gives them a chance to survive.

I do not recommend heating or freezing antique furniture, as strange things can happen to the wood, glues and finish. Even Vikane gas is not enough to kill the eggs; only the adults.

Methyl Bromide gas applied at a certain concentration for 48 hours is guaranteed to work on all these bugs all the time. I have a company with a sealed container, where I put the furniture every two weeks to have this treatment done. It is the only way I know, and, since Methyl Bromide gas is a serious problem for the ozone layer, it is soon going to be eliminated as an option.

When I asked the customs agents why they don't require fumigation for furniture beetles as the antiques arrive from Europe, like they do for pests on fruits, they answered: "We don't care, since that bug is already here."

The bugs are going to win this war.

Vector Clamping

Question: Can you have too many clamps?

I think you know the answer...

A lot of woodworkers like to read books, magazines and posts about their craft. There are literally hundreds of authors who discuss dovetail drawers, finishing methods, the best tools, etc. I am no different. I have bought books that I was sure would be poorly written and contain misleading information just so I could read them and be proven right. I have many old and out of print books which also were essential for my career. I not only have all the magazines, but I have the promo leaflet for FineWoodworking which was sent out before the first issue.

In all this material, there is one aspect of furniture which is difficult to address: proper clamping methods. Sure, there are stories on vacuum bags, some on veneer presses, a few on specific clamping issues, and so on, but so far I have not seen an article which clearly addresses what I call "vector clamping."

I am sure that my mathematics, geometry and "proper" education in those fields lets me think differently about clamping, but at the same time it is said: "If you scratch the surface of a woodworker, you will find an engineer." This statement was proven at the first SAPFM dinner so many years ago. I found myself sitting at one of the tables in the front with Underhill, Breed, and a dozen other veterans, in a room with about 200 members. The speaker asked the group, "How many of you are making a living at traditional woodworking?" Most of those sitting at my table raised their hands, in total about 10% of the group. Then the speaker asked, "How many of you are engineers?" It was a clear majority.

So I assume that, when I use the term "vector," many woodworkers will know what I mean. A vector is a little arrow which shows the direction of the force under consideration. So a clamp applies pressure in a direction perpendicular to the face of the clamp. Of course, this pressure is not constrained in a single linear vector, but forms a cone of force as it leaves the surface of the clamp. As an example, my veneer press screws exert pressure over 81 square inches, or a 9x9" area, and I placed them on the press 9" apart to form a grid.

The primary purpose of discussing vector clamping is to illustrate how I restore antiques, which very often have curved surfaces. Tripod tables, cabriole legs, bent wood windsors, and many, many traditional forms of furniture rely on curved elements, which often break. These breaks can occur at a joint, like the leg on a tripod table, or in the grain of the wood, like the foot breaking off of a cabriole leg.

In many cases these breaks are treated in a way which makes my life miserable. The repairman will take some synthetic glue, epoxy or "the strongest glue on planet earth" and paint it on the break, then frantically search for a way to clamp it before the glue sets. The result is a mess, and I am constantly trying to undo the damage, scraping off the glue and starting over.

I usually will tell the client: "If you had not had it repaired before, I could do it for half the cost."

Assuming you have a fresh break with good wood surfaces, the first thing you need to do is determine the center of the break or joint, and imagine a vector line perpendicular to that point in space. (Editor's note: see comments below)  If you are able to put pressure with a single clamp on that vector line, it will clamp the repair properly. That is all there is to it.

Now, it is often the case that the vector line which pushes directly perpendicular to the center of the break goes off into air, away from the wood surface. Like the curved leg on a pedestal table, the line goes off the curved top edge of the leg, and a clamp will not gain a purchase there. That is why you need to create a proper place for the clamp to press.

Select a softer wood than the object, like pine, poplar or other wood. Shape the wood to exactly fit the surface of the curve, and leave a length of wood for another clamp to hold it in place. Clamp this new piece in position and locate the vector line on it. Draw a perpendicular line on the new piece which crosses the vector line at 90 degrees and cut to this line with a saw or chisel. This will be the place to put your clamp for the repair.

In every place where you need it to press directly on the vector line for the repair, you will need to cut a softer piece of wood and clamp it in place. Do this before you reach for the glue. Now your repair has little pieces of soft wood clamped all over it in places where you need to make the repairs. Grab your glue and put it on the repair, place the pieces in position and use a single clamp to press the vector line together.

It is rewarding to see a single clamp pull a complicated joint together, without slipping to one side or another. When done, all the soft wood pieces can be kept in a box for reuse in similar jobs which will certainly happen.

I do not need to stress that you use animal protein glues...

Typical Upholstery Project

Of all the traditional organic materials used in upholstering antique furniture, perhaps straw is the most difficult to conserve. Straw was one of the earliest stuffing materials used, going back easily to the 16th century. It is commonly available, essentially free, and has fairly good initial resilience. However, it compacts over time and breaks down, so that after a century of use it is in poor condition.

Compare that with the best material, horse and hog hair. Even after a century of hard use, curled hair retains its shape, and is much easier to clean and reuse.

I have focused on conserving straw during my career, since it is a challenge and since it is often discarded by other upholsterers, so that a majority of antiques are being converted to modern, synthetic stuffings. Did I mention I hate foam and staples?

I just started a parlor chair from around 1885 which has its original straw foundation. This chair is from the decade after Turkish upholstery was the fashion. Turkish upholstery represented the high point of the upholsterer's craft in the 19th century and exotic and complicated forms of upholstery were produced, using expensive fabrics, trims and accessories.

This chair carries that style of upholstery on a decorative frame, with spring stuffed back panel and seat, and large arm cushions which curl around both sides of the arm support. These photos show the original stuffing foundation as I took it apart. The first photo shows the chair without its fabric. You can see the basic burlap covering on the seat, stitched with a wire front edge. The arms are also stitched. The back panel has been removed and is resting in place.

The second picture shows the back panel structure. The springs are surrounded by a hard stitched edge with straw stuffing. The horsehair is placed in the center of the top.

The third picture shows the chair laying on its back on my work table. The seat stuffing has been cut away from the springs and is folded back to make it easier to remove it. This stuffing is very fragile, since it is just straw and the burlap is rotten. I use care to keep it intact, vacuum it and transfer it back into its original position during the restoration. You can see the difference in the burlap which covers the springs now (replaced by an upholsterer some 50 years ago) and the torn burlap which is original to the stuffing, holding the straw in place.

The last picture shows the original springs and the method of tying them in place. This pattern is called "8 knot" tie, since each spring has 8 knots. Note the cords tie both the tops of the springs as well as all the centers of the outside springs to hold them in place under load. This is the way I was taught to tie springs. There is also a great deal of broken straw debris which sits on the top of the webbing. Fortunately, this chair was not used much after the burlap became torn. With most upholstery, and straw in particular, if the seat is used after the supporting burlap tears, the stuffing quickly becomes chewed up and damaged beyond repair.

This chair has been properly restored once before me, by an upholsterer who knew what to do, probably 50 years ago. He removed the stuffing exactly as I have done and removed the jute webbing from the bottom of the chair. He added new jute webbing, and then sewed the base of the springs to the webbing to hold them in place. He tied the tops of the springs using Italian cord and added new burlap to the top, stitching it in place to the wire edge and springs. Then he carefully returned the original stuffing to its original position. I will do the same, except I do not need to treat the springs, as they are still in good shape.

One of the clues that I am not the first person to work on conserving this chair is the chalk numbers indicating which corner blocks go where.

When you can feel the springs come up into the seat, stop using the chair. Take it to a traditional upholster and save the stuffing. It's the right thing to do.

I will post more during the restoration so you can see how I do it.

Diversity Survival Mode

When I arrived in Paris in 1991 to enter ecole Boulle, I was introduced to many important craftsmen in the field of furniture making. They would ask me what I did for a living and I would mention that I did everything: restoration, conservation, creation, upholstery, weaving splint seats and rush seats, veneering and marquetry, turning, carving, polishing, making hardware, fixing locks and keys, replacing period glass, and so on, thinking that I was impressing them with my diverse talents.

I quickly learned that this approach was counter productive. The philosophy of craft in Europe is that you spend your life in one field. In fact, you should spend your life focusing on one aspect of that field, so that you can truly master the trade. You know the mantra: Jack of all trades, master of none.

So, I changed my introduction. If I was introduced to a carver, I said that I was a carver. If I was introduced to an upholsterer, that was my trade. And so it goes, to quote Vonnegut.

I did not start out to be a "furniture conservator in private practice" as my listing in the AIC directory states. I first started out as an upholsterer. That was because, in my neighborhood, there was an 85 year old upholsterer who moved here from New York and set up shop. I would visit and spend hours talking with him, watching him as he worked, methodically tying springs, stitching horsehair stuffing and spitting tacks. He was old, but his hands were strong and agile.

I thought, "I could do that." And so I did. It was very rewarding.

When customers asked me where they could get the wood repaired or refinished while the upholstery was changed, I said, "I can do it." I started to get more jobs, since I was able to do the complete project. My entire career has been successful because I am not afraid to take on jobs which require adapting and learning new methods. I should say, discovering old methods which are almost forgotten.

This typically American philosophy of craft has kept me in business during the past years, when others are closing up shop. I can adapt to whatever job walks in the door.

Currently, I am doing a lot of upholstery. I still find it interesting and rewarding. I am one of the last who spit tacks, tie springs by hand (8 knot Italian cord) and stitch horsehair.

The trade of upholstery took a wrong turn in the 1950's, when foam and staples became available. Hundreds of years of using organic materials were abandoned in favor of the new materials. Horsehair, Spanish moss, kapok, down feathers, cotton batting, excelsior, straw and other materials were removed from antique furniture and thrown out in the trash. Foam was added in place and, not only was the shape of the upholstery different but the comfort that those earlier materials provided was lost.

As it turned out, foam deteriorates rather quickly. Early foam from the 50's lasted about a decade. Later foam was improved and lasted several decades. Modern foam is supposed to last much longer, but will it? Horsehair lasts a century and more and still retains its shape.

Perhaps the most serious loss has been to the actual craft of upholsterer. Being an upholsterer in the 18th century was more prestigious than being a cabinet or chair maker. The trade of upholsterer included aspects of what we call today, interior designer. The upholsterer advised the client on fabric selection, bed "furniture", drapes and carpets and diverse textiles. All these materials were very expensive and the value of the actual upholstery on the chairs and sofas exceeded the value of the woodwork.

Upholstery is one of the few trades you can learn by undoing and studying antique examples. To restore upholstery means to take it apart, layer by layer, conserving the stuffing and springs, and replacing the jute webbing, cord, burlap and muslin with similar materials. Properly done this work can fully restore the original comfort and look of early seating furniture. As older, traditional upholsterers retire or die, this trade is in danger of becoming obsolete. The pieces which survive with their original foundation are rare and need to be properly conserved.

This week, if you ask me what I do for a living, I will respond: "I am an upholsterer."