Popular Woodworking’s Megan Fitzpatrick published an article some time back about a kitchen cart she built with a wink toward the workbench designs of Andre Roubo. I thought it was clever, so I adapted the idea for a serving bar. Love the bar… hated finishing it twice!

At 48″ long, 19 1/2″ wide and standing 40″ tall, this bar is
just right for serving coffee or cocktails.

The features that delineate this piece from typical furniture styles are the bulk, the joinery, and the flush relationship between the legs and the top. The result is a chunky, functional, industrial-looking bar that anchors my coffee kitchen/wet bar with authority.

This piece was my first foray into working with quarter-sawn oak. I have seen lots of uninspiring QS oak furniture and molding. For years I dismissed QS oak as antique-y and cliché. I realize now that very little of the QS oak we see in the world has been finished to emphasize chatoyance.

My house is predominately Arts & Crafts style. It is hard to avoid oak if you want to make an Arts & Crafts statement. When I decided to furnish my coffee kitchen/wet bar with a signature serving bar, oak was an obvious choice. I also wanted a piece worthy of several estate sales, so I set about building something special, starting with stock selection.

I invested the treasure required to procure a stack of 8/4 QS red oak boards. I spent hours arranging figure. The table-top is solid oak, 3” thick, glued up from individual planks roughly 6/4 thick as one would a workbench. Each board was placed in the line-up so that the figure elements would be balanced and symmetrical across the width of the table top. Boards with robust ray fleck are spread out evenly between boards with fine, sparkling figure and boards with medium figure. The arrangement is intended to produce the effect of shimmering water, with light dancing evenly across the top.

The legs are veneered on the plain-sawn edges so that QS figure can be presented on all 4 sides (a biological impossibility for solid wood).

The lower tray is comprised of QS oak slats also arranged by figure, like the table top.

The joinery shows Roubo’s influence. The legs are flush with the top. The top is held flat by a large mortise and tenon typical of breadboard ends, with a dovetail on the outside board at all 4 corners. The dovetails keep the breadboard ends tight to the top while the top expands and contracts on the tenon across its width. There is no hardware. Through mortises anchor the legs to the top and the stretchers to the legs down below. Truly, all this piece lacks is a couple vises and it is ready for service as a workbench.

The dovetailed breadboard ends are real and functional.
This top shrinks and swells about 1/16″ with the
seasons and the ends tell the story.

I loved building this piece. When I built it, I had ample time and energy and patience to maintain my focus on creating something special. I loved working out the design details and executing the joinery. Although I had to hide more mistakes than I’ll ever admit, I am proud of the build. But not the finish, and that’s what this article is about.

Finishing is my weakness. Joy comes to me from bending wood to my will with steel, not brushes. Coming to grips with finishing has not been fun for me. It continues to be a long journey fraught with denial and do-overs.

I read this rule of thumb somewhere: When you complete a build and turn toward finish work, you’ve reached the half-way point. That hit me hard, and I didn’t want it to be true. Accepting this rule of thumb may have been the first time I began thinking of finish work as woodwork. Admittedly annoyed, I began reading Flexner and others hoping to diminish the fear and loathing of starting the finishing phase on my projects.

I had always been a rogue finisher. I expected products to yield the results promised on the can, and I had no patience for test sticks and drying times. It never seemed reasonable to spend hours and hours, sometimes days and weeks, working out how to finish a project before actually finishing the project. It seemed so inefficient… and there is the irony. I have learned the very, very hard way that knowing what your finish is going to do before you apply it to your project is the most efficient approach, by far.

Which brings me back to my serving bar. I spent weeks working out the design. I spent hours selecting and arranging stock for maximum impact. I spent weeks milling and joining and assembling. When everything was ready for finish, I grabbed my trusty can of General Finishes Poly Gel Java Stain and ruined the piece.

My heart fell as figure-flattening stain smeared over my project. My mind raced with “Maybe if I…” ideas. In the end I could do nothing. My oak did not shimmer.

I felt defeated and needed to attend to other matters. I accepted my shitty finish as the best I could do at the time and pressed the bar into service. I knew I would have to circle back one day and fix it.

Three years passed. During that time I learned more about dyes and layering different types of varnish. I learned about sealing and splotching and glazing. I learned to make test pieces. The time came to refinish my serving table. To build confidence, I re-read several articles on emphasizing figure in a variety of woods. I found some of the QS red oak cut-offs from the original build and selected pieces with a variety of figure representing different boards I used in the project. I veneered several pieces to a scrap of poplar and began experimenting.

After a couple weeks of daily experiments, I found how to get maximum chatoyance together with a pleasing color. I confirmed that my recipe of dyes, sanding grits and topcoat works equally well on different types of figure. Keeping in mind I’m working with QS red oak and not it’s flamboyant white cousin, I found a finish with depth, richness, subtle chatoyance, and a silky feel that invites caressing.

It is brutally difficult to capture chatoyance in still photos but here are three before-and-after photos.

First finish flattened all chatoyance.
The new finish draws out the quartersawn figure.

My recipe is this:

Sand to 320 and thoroughly clean the pores

Dye with TransTint “Golden Brown”

Sand most but not all the dye off with 220 along with any raised grain and thoroughly clean the pores

Dye with TransTint “Honey Amber”

Sand most but not all the dye off with 220 and thoroughly clean the pores

Dye again with TransTint “Golden Brown”

Lightly sand with 320 just until silky smooth and thoroughly clean the pores

French polish (shellac)

Wax, wax, wax

This serving bar is used for cocktails. Much hand-wringing went into finishing with shellac. Shellac is promptly eaten by bourbon on contact. My several coats of hard, Carnauba wax give me some protection, but this is still a compromise top-coat. Here is my rationale: Adding a coat of poly would increase durability but render repairs far more difficult. I know there will be dents and dings and I’d like them to be easy to fix. I am going to give Shellac and wax a try and see how it goes. I can always add poly later, but once you go through that gate, you are out of the garden forever.

Hindsight is an unforgiving yardstick. With its benefit, I see the idiocy of applying a finish to this bar without testing first. Knowing what I know now, I am gob-smacked that I would ever have done such a thing. I’ve heard we learn more from our mistakes than our successes. Having finished this serving bar twice is a poignant testament to that.

Leave a Reply

Your email address will not be published.

*