furniture making

Building Gragg’s “Elastic” Chairs — Harvesting the Wood 1

Much like the recently concluded account of replicating the early 19th Century legislator’s writing desk, over the next few months I will run a series of posts recounting the c.1810 Gragg “Elastic” chairs I am currently replicating in the shop.  Though I have posted on Gragg chairs several times over the life of this blog I am hoping this series will renew your interest in them.

The story for these chairs (I am building two commissioned by clients and another for myself) goes back to a couple years after the completion of the original replica I finished in 2011.  I published the proceedings of the project in that year’s American Period Furniture, the flagship publication from the Society of American Period Furniture Makers, you can read that article here.  I am still pleased with that article but en toto these posts should be dramatically more complete in both my descriptions of work, with the continued refinement of technique and process,  and my reflections on the challenges of making these chairs.

While reverse-engineering Gragg’s methods I learned a lot about not only how he worked, a combination of sinuous elegance and brute force, but also the kind of wood that would be best suited for making these iconic chairs.  Like Gragg I confirmed the material that suited the project best was something like air dried red oak from very large trees (he used red oak, white oak, ash, and IIRC hickory).  I remain so fascinated by the form that I continue exploring other routes to the same end point and will comment on those paths as appropriate.

A bit over a year after finishing my first full-blown Gragg my neighbor Bob told me he was having several  over-mature oak trees removed from his yard.  Though still mostly-healthy their decline was underway and it was only a matter of time before they became an imminent threat to his house.   When the lumberjacks arrived I asked if the boles could be sectioned into five or six foot pieces and left intact on the ground for me to harvest.   I had to be elsewhere for much of the day so I did not get to watch them bring these behemoths down.

When I returned home later that day I inspected the bolts with anticipation, and spoke with Bob (and his teen-aged grandchildren who were lounging around) that I would be arriving after dinner to begin the harvesting.  I’m not suggesting the yutes did not believe me when I showed up with a sledge and a bag of steel wedges and wooden gloats and told them I would be turning these 5,000 pound hunks into manageable pieces from which I would eventually make chairs, but truth be told the looks on their faces indicated they did not believe me.  Over the next couple of days their eyes bugged wide more than once.

Writing Desk – Finis

The light at the end of the tunnel was glowing brighter with every day.

I created an archival label for the desk, describing its antecedent heritage, the making process, the patrons, and me as the maker.  I glued it into the drawer and with that the creation of the desk was complete.

After that the only tasks were the glamour (?) photography, including the obligatory selfie, and delivering it to its new home.

All it took was almost two years from the initial client inquiry, my suffering and recovering from a broken hip the first year and a broken arm the second, a pile of prototyping, and a ton of hand-work as the project mandated.

I was not displeased with the ultimate outcome.

Throughout the project and its aftermath I assiduously acquired enough vintage select genuine mahogany, stored mostly underneath my planing beam, to make another half-dozen of the desks.  All I need now are the clients to make it happen.  If you know anyone who wants one, let me (or them) know.

 

A Cabinet For Mrs. Barn — The Wood

One of the challenges of “downsizing” is finding furniture to hold your stuff and fit in the new, smaller space.  This was particularly nettlesome for us as we moved from a 16′ x 24′ bedroom to one that was half that size.  For several years we had been searching antique shops for a small cabinet to hold some of Mrs. Barn’s clothing (fortunately we are blessed with two large closets here) but never could find something to fit nicely.  There ws only one place in our little bedroom for such a cabinet, and the size was very specifically determined by that space.

Instead she did something that had not occurred in more than three decades of connubial bliss — she asked me to make a piece of furniture for her.  Not just a piece of furniture but one made from vintage chestnut, a wood she likes very much.

So I did.

My source of vintage chestnut was unsurprisingly the old shack that my brother, nephew and I tore down three summers ago, a grand week of camaraderie whose ending punctuation was me laying in the hospital with a broken hip.  The shack had been a living quarters, probably in the 1930s according to the newspapers pasted to the wall.  The entire roof structure was chestnut, and the walls and floors were white oak.  The haul was pretty impressive.

In the intervening time the wood was stacked nicely in the lower log barn awaiting my attentions.

Up next – prepping the stock and joinery

Writing Desk – Polishing and Assembly

Once the finishing process was essentially complete it was time to rub out the surfaces and glue-up the desk.  Since my goal for the surface was to present an unfilled, lightly-worn, ancient-but-well-cared-for appearance and character I rubbed down all the surfaces to accomplish those ends.

My typical procedure for this undertaking is to abrade the surface with ultra-fine steel wool saturated with paste wax or conversely a 4F pumice/paste wax blend, incorporating or succeeded by the inclusion of tripoli/rottenstone to leave subtle traces of “ancient schmutz” in the crevices.

In this instance I mixed up a paste wax/rottenstone blend to use with the steel wool, and proceeded to rub every exposed surface until the outcome arrived at the destination I was aiming for.

The “glow” and subtle grain of the resultant surface was so glorious I had to just look at it for a while.

This comparison of a rubbed-out (L) vs. raw surface (R) is demonstrative of the transformation.

This is the address I was looking for.

With the finishing mostly done I glued everything together with hide glue, and suddenly there was a complete piece of furniture staring me straight in the eye.  There were only a couple of final steps to take, applying another round of pad finishing to the writing surface before rubbing that out, and buffing off the rottenstone-laced dried paste wax.

Battles With Bears

Some days you eat the bear…

 

And some days the bear eats you.

As I dive ever deeper into producing Gragg chairs, currently two on commission and another for myself with hopes of including this item in The Barn Store in the coming months/years (and even entertaining thoughts about offering a Make A Gragg Chair workshop as soon as the summer/autumn of 2020; first, I have to make one from start to finish in a week myself), I realize all the more how close Gragg was coming to the limits of what can be accomplished with wood as the raw material.  Given the extremity of the bends involved, including the main serpentine element with two 90-degree four-inch-radius bends within a foot of each other and a180-degree four-inch-radius bend for the arms, working out the routine is a critical process.

I recall the first time I tried this almost a decade ago on the original Gragg repro prototype, in front of my Smithsonian colleagues no less, I broke every single piece of the ash I took from our prized lumber inventory in the conservation studio.  Every.  Single.  Piece.  Not to mention breaking several of the bending form elements which had been woefully under-built.  It was not my proudest moment of professional accomplishment but drove me on to get it right.

Even now I test the boundaries, trying riven and sawn kiln-dried oak (at the bottom of the pile in the truck), trying ancient/recycled but possibly air dried oak salvaged from a derelict weaving loom,

trying oak I harvested several years ago but have kept from seasoning fully, first cross-stacked outside and now residing in the basement/first floor of the barn,

and finally using oak I harvested this year.  Through it all my batting average kept improving.

I’ve heard my friend Bruce Hoadley tell the story of a small manufacturer who was plagued with broken elements resulting from very tight bends after steaming.  The punch line was that after going to watch the definitive practitioner for making the identical items, the manufacturer said with a smile something to the effect of, “He breaks most of them too!”

I encountered the same thing with the Gragg chair elements.  Some stock bends like taffy, some breaks like crystal.  I am moving more toward the former than the latter, but it is exasperating all the same.  At least the failures make good kindling.

Yesterday afternoon my success rate was 92% with only a single stick making it into the failure pile.  By observing the character of the pieces being bent, the stresses of the bending itself combined with the addition of bending straps, this is an outcome I can live with.

Still, the pile of kindling grows, just not as fast as in the past.

I am definitely gaining ground on the bears.

Gragg Chair Video Session 4 – Fitting and Assembling I

Our latest day of video recording dealt with the beginnings of assembling the pieces into a whole chair, including fitting the individual pieces together to fit the master template based on the many Gragg chairs I have examined over the years.

Once we moved on to fabricating the cross pieces things got fussy.

Chris wrapped up the day by taking some detailed footage of my shaving beam, my primary stock preparation tool for the project.

Writing Desk Finishing – Color and Craqueleur

With the desk surfaces well-sealed with lemon shellac and the finish foundation built up with garnet shellac it was time to wrap up the color work while simultaneously imparting the craqueleur that would reside under the final, polished surface imparting the “look” of an ancient but well-cared-for surface.  The process was simple, one that I have employed before and takes advantage of the properties of the materials found on the shelves of finishing shops.  Well, at least on the shelves of my finishing shop.

As I have written previously I am loathe to contaminate the raw wood with colorant that cannot be easily removed.  Hence my distaste for pigmented “stains,” dyes, or chemical treatments.  I find these techniques to be insufficiently control-able for precision finishing, preferring instead to introduce any coloration into the finish system itself.  Not only is this much more easily controlled but comparatively effortless to undo if the target is missed.

My common terms for this kind of in-finish coloration is “toning” if the coloration is included in the film forming material itself (i.e. the varnish), or “glazing” if the coloration is imparted via a discrete material in-between coats of finish.  In this project I was able to blur the lines in these concepts and add an additional feature to arrive precisely where I wanted to go.

In short, my goal was to provide visual unity of both the color/tone and impart richness to the texture to replicate a finish that gave the appearance of being well-cared-for but 200 years old.  It was not to be a grain filled, brilliantly glistening “French” polished surface, that was simply not appropriate for this project.  I wanted the surface texture to be presented subtly for both the wood grain and the cobwebbed craqueleur within the finish film itself.

Here is a brief recitation of the technique I used to accomplish this.  It depended on understanding the nature of materials and their means of forming films (or not).  There was no magic elixir, but rather an exploitation of those materials.

I began with my glaze formulation, which first consisted of solid acrylic resin beads dissolved in hot mineral spirits to a 25% solids content.  This served as the backbone for the glazing solution and the governor for my desired solvation limits that in turn controlled the craqueleur.

To this I added some oil/resin varnish in the proportion of 1 part acrylic resin solution to 3 parts of varnish, followed by stirring in asphalt as the primary chocolate-y brown colorant and a dab of sienna artists’ oil paint to get the right amount of reddishness, thinned as needed with naphtha.

I slathered this on to the surfaces that had been well-built-up with the shellac base.  The glaze was evenly distributed with a well-worn vintage (and thus soft-ish) bristle brush, pulling off glaze in places where it was imparting too much unnecessary coloration, smoothing it out where the color was correct.

The final step was to soften any striations by whisking the surface with a badger blender or goat hair hak-e brush just before the glaze dried too much to be manipulated further.  The entire process for a particular work area from slather to done was probably about 90 seconds.  This was then allowed to dry overnight.

The next morning to lock it all in place I spritzed it with a light coating of sprayed shellac.  This resulting surface might have sent an inexperienced finisher screaming for the hills but I had a huge smile on my face because the result was exactly what I wanted to achieve.

Let me explain.

The oil-resin varnish and artists’ paint in the glaze had stiffened but were not yet fully hardened, so the introduction of of a more polar solvent (in this case alcohol in the sprayed shellac) to the glaze would cause the glaze film to become imbibed with the solvent, swelling the polymer matrix into a crinkled texture.  On the other hand, since the acrylic resin and asphalt are not susceptible to the same solvation effect they remained calm, otherwise the surface would not have been a variegated craqueleur, it would have been velour.  And, with the glaze film being ultra thin yet bonded to the underlying shellac via the adhesive properties of the acrylic resin component I did not have to worry about the whole laminar construct coming apart.

(NB – I discuss all these properties and effects much more thoroughly in my upcoming book, A Period Finisher’s Manual).

I waited another day to make sure the solvent from the shellac mist coat had diminished sufficiently and abraded the surface very lightly before building more finish.  At this point my main concern was adding too much additional alcohol to the system and causing even more oil/resin film swelling, so rather than brushing on shellac to build the finish I padded it on.  Remember, my goal was not a glistening pad-polished presentation surface.  I padded the finish simply because it was the best way for me to deposit film-forming material while controlling the solvent encroachment.

After enough new finish was added on top of the craqueleur, three applications as I recall, I abraded the entire surface smooth, in essence leaving the craqueleur embedded down in the finish but still presenting a smooth surface that I could polish out.  I built up another half dozen padded applications to the writing surface after the piece was assembled then set it aside for a fortnight before final polishing, assembly and detailing.  As you might surmise the craqueleur was so subtle it was impossible for me to photograph.  It has to be observed in real time and in real space.

It is worth remembering the beginning point for all this color work.

Now it was time for the polishing, assembly and detailing.

Stay tuned.

 

 

 

 

Gragg Video Shoot Session 3 – Dance Me A Jig

Our recent day of video recording for the Making A Gragg Chair was pretty much all about bending forms/jigs for the components of the chair.  I showed how I make the forms and why I make them the way I do, including a discussion of “spring back” and how that affects my design and layout of the forms.  Somehow that one topic managed to consume the entire day of work in the studio.

These three pictures were ones I took when I was prepping for the video shoot, but I replicated the process upstairs in the studio with the camera running.  In them I am making some form stock, the 1-inch thick double laminated sheet I make from two pieces of 1/2″ baltic birch plywood.  The finished product is what I use to cut the actual bending forms themselves.  I begin with two pieces of 24′ x 48″ x 1/2″ plywood and slather PVA on one of them, then screw them together with decking screws and fender washers.

Here is that panel after finishing along with another double laminated panel made from two pieces of 3/4″ baltic birch type plywood that serves as the ground for the bending forms.  Once they are dry overnight I remove the screws and washers and they are ready to go.  I got all of this material from the closest big box home center.

Here are some images from the session in the studio.

Desk Finishing – I

With the construction finally completed it was time to dive into the part of the project that had no down side, the part where it was all fun all of the time — finishing.  In my back-and-forth with the client I decided to shoot for a non-filled-grain look, a general tonal consistency, and with craqueleure underneath the the final surface.  In short my overall strategy for the process was to have the desk looking like it was “old but well-cared-for,” looking new was not part of the plan, nor was “antiquing” nor “distressing.”

Since the desk was small and convoluted I intended to finished almost all of the pieces prior to final glue-up.  So the full glory of it was not revealed until nearly the last step when all the pieces were put together.  Finishing each piece separately required a careful masking of all the surfaces that were glued together in that assembly.

The first step was to lay down a few seal coats of garnet shellac brushed on in three or four applications of a one pound cut.  This allowed me to get a good sense of how I needed to shift the visual coloration and tone of the pieces to bring them into harmony.  I would make no attempt to make each and every piece identical to the other, that was nonsensical given the variety of grain directions, etc. involved, but simply to bring everything closer en suite.

For the most part the starting point for this shifting was accomplished via the addition of dyes to the garnet shellac being brushed on.  I am loathe to employ chemical stains to bare wood as part of the finishing exercise; they are simply too difficult to control for an equal outcome when the starting points are different, and when they do not behave their reversal (removing the surface of all the wood) induces foul language in the shop.

By building the finish with dyed coatings the depth of the wood’s beauty soon emerged forcefully.

I built the finish steadily, five coats becoming ten, then fifteen.  I smoothed all of the flat surfaces by scraping with razor blades and the curved surfaces with pumice pads and before long I was ready for the home stretch.

But the real magic of final blending of the coloration, along with imparting craqueleure, was achieved in the next step.

Stay tuned.

Writing Desk – Final Pieces of the Puzzle

With the light shining at the end of the tunnel, in hopes that it was not an oncoming train there were but a few more elements to fashion and fit and the desk construction would be complete.  Sort of.  In fact the final assembly did not take place until after the individual pieces were finished, but more about that later.  This final construction phase would perhaps be better named the final fitting phase.  This included the two top sections, one the writing surface and the other the horizontal back piece all cut from the same mahogany slab, and a little extra enhancement.  The back (horizontal) piece of the top was thinner than the writing surface by 1/4″ in keeping with the original.

The “construction” of the top pieces was ordinary to the point of mundane.  They only had to be cut to the right size for the places, planed and scraped with a rounded edge planed onto them, then fitted together in a butt joint.

Before the top was glued onto the sides with glue blocks it was the proper time to make the drawer stop.  I used a preposterously simple technique I have seen in many small-ish pieces of early 19th Century styled furniture like this replica.  A simple flat head wood screw served the purpose perfectly, with the additional benefit that as the dimensions of the case or drawer changed over time to to moisture cycles and any resultant compression set, the depth stop could be easily adjusted infinitely.

On final little whimsy was to add a hidden “whisky shelf” immediately above the drawer.  Since the top was slanted and fixed there was a fairly spacious void above the drawer, and it was not uncommon for a small shelf to be included inside to that a flask of libations could be kept there.  I made a “U” shaped shelf that could be accessed when the drawer was pulled part-way out, then tacked the shelf on to the fixed drawer guide rails with a spot of hot animal hide glue.

And now it was on to the finishing.