Table top marquetry drawing

Posted on 6 August 2013 by J. Leko

One of the key elements of any marquetry embellishment is the drawing from which it is cut. A few weeks back I managed to make a full-scale photocopy of the table’s top.

Pierre Ramond’s Masters of Marquetry, vol. III contains a detailed line drawing in reduced form. Luckily, it is printed on a fold-out page that doesn’t cross the spine of the book. Subject matter copied from near the spine always results in distortion! This drawing together with measurements of features in the original top allows computation of a scale (enlargement) factor. These full-scale copies are important because adhered to veneer packets they become the templates from which the marquetry is cut.

The table top marquetry is the only portion which will require a chevalet. Parquetry adorning the aprons and interior compartments can simply be cut with a chisel and straight-edge.

In school, I was always advised to come up with independent methods to validate any result. Having already constructed the prototype top, it should fit precisely on the bounds of the photocopy if my calculations (and drawings) were correct. Happily, they did! 🙂

On the importance of drawing…

Posted on 14 July 2013 by J. Leko

Oeben Original Back View

Rear view of the SketchUp model

One of the things that makes this project complex are the mechanisms on which the table operates. Although I have several pictures of them from my trip to the Getty, I still needed to draw everything out. This is necessary not only to have them manufactured, but also to ensure that everything fits together and operates smoothly. Despite these reasons, it’s a major pain in the posterior!

SketchUp to the rescue!

SketchUp is a fantastic design tool/laboratory. By determining proportions from the aforementioned pictures, I created scale drawings of each of the mechanism parts. While the urge to send these individual part drawings off for manufacture was great, better judgement prevailed. Luckily, this averted a “situation”. Instead, I methodically fit the pieces in place in the carcass model, and made several “discoveries” along the way!

The top rack and guide are offset toward the outside of the carcass.

The upper racks are offset toward the outsides of the carcass least they interfere with the main box travel. Here’s surprise number 1: if I mount the upper racks and guides flush with the inside of the carcass, they will interfere with the “main box”. Checking my photos, sure enough, the original makers moved the racks and guides outward to make the room necessary.

Initial model of the upper rack (it’s upside-down). It’s not tall enough!

Also, I needed to extend the height of these racks so that the top cleared the main box. My photos provided the multiple aspects necessary to deduce their dimensions. However, they can be unreliable sometimes. Small errors propagate. “Installing” the racks into the SketchUp model allowed their height to be tweaked, and the top to clear the main box to the desired level.

Fixed the upper rack. Now the top will clear the main box!

This process also helped me discover a quirk with the lower rack installation. The lower guides mounted to the carcass sides extend beyond the inside face of the rear panel, that is, the back. Although I’d seen this before, I had never really considered the rationale for it. I figured that the makers were being lazy. Turns out it has a critical purpose. It accommodates the movement of the lower rack (the one fixed to the main box) when the table is closed. I didn’t realize it, but the back panel in the original is mortised just for this reason. The evidence was there in the photos the whole time. I just didn’t see it until my attention was drawn by this.

Note how the lower rack projects into the rear panel? The original is mortised out for just this purpose.

The final parts to be modeled are the main spring barrel mechanisms. The chief consideration with these involves the force required to drive the main box/top assembly when the latch is released. Too great a force, and it could jolt the table possibly knocking it over. Too little force, and the assembly could get stuck. My solution is to build the prototype main box, I already have the prototype top, and get their combined weight/mass. From this, we should be able to determine the appropriate tension, or at least get close, for the spring drives. I’m betting that this is the same process that Oeben’s craftsmen used in their time too. Except for SketchUp! 🙂

Stocking up

Posted on 26 June 2013 by J. Leko

Over the last two weeks, the prototype materials have been arriving. I wanted to post some pictures of them for you. Recall that the prototype will be made of poplar. It will serve as the construction test case, especially for fitting and fine tuning the mechanisms. If my experience with the drawings is any indication, I anticipate that I’ll encounter a learning cliff on that last part! More on that in a later post…

Based on the SketchUp models I’ve discussed previously, this is more than enough poplar for the carcase. It should cover everything necessary from the ten inch wide panels that form the bottom and intermediate dust frames between the main box and drawer, to the massive 12/4 thick side apron pieces, and literally everything in between. Forming and shaping the Louis XV style legs will be an educational experience. I can’t wait!

Finally off to the left rear in the image above, are two more pieces of poplar for the prototype. Along with this, I’ve managed to pick up some materials that will be used in the final piece, specifically the marquetry. The massive reddish block near the center back is bloodwood, in front of that to the right are three pieces of satinwood, and two padauk boards. To their left is a chunk of Spanish cedar (which smells wonderful!). All of these will be resawn into 1/16 to 3/32 inch thick sheets of veneer which can be combined into packets to create the final designs. You will also note three sheets of commercially sliced veneer towards the front. These were used to assess the proper species for the parquetry on the side panels: kingwood, tulipwood, and bloodwood.

A problem arose while researching the parquetry species in the original table. One source, Ramond’s Masters of Marquetry, states that the parquetry is composed of tulipwood and bloodwood, while the J. Paul Getty museum claims kingwood and bloodwood. The figure of tulipwood and kingwood can look similar, especially in small pieces such as those used on this table. Judging by the present appearance of the table, and the knowledge that woods change color with time, I’m going with the Getty on this one…

Taking it from the top…

Posted on 21 June 2013 by J. Leko

Every project has to begin somewhere, so prototype construction commenced with the top because:

I had the ideal board for this element already in my rack.
As a single “unattached” unit, it would be quick to complete.
It can be used to gauge the layout of the other carcass pieces.

The top is substantially flat with breadboard ends to help keep it that way! In the original (and final version of this project!), this surface is the substrate to which the marquetry is adhered. Wood changes shape with the variation of moisture content in the air. This construction technique assures that the panel field will experience minimal “cupping”. To perform this task, the breadboard ends are pegged in place through enlarged holes in panel tenons. This leaves the edges of the panel field free to protrude beyond and within the extents of the end boards. Cyclically throughout the year, the panel field will be wider than the end boards, while at other times the reverse will be true. Rarely will they ever be flush – a necessary trade-off. The profiled perimeter is based on the measurements, and drawings made during my visit to the Getty museum.

Here is how it is done:

The process begins with the selection of a suitable board, which is cut into appropriately long pieces.
These are hand planed down to the requisite thickness.

Several boards are glued together to form a panel.

On this panel, “tongues” are created on either end using a moving fillister; matching grooves are ploughed in the remaining boards, the breadboard ends.
From the “tongues”, the majority of the material is removed except for three tenons, one at the center and the others on either end, leaving a much shorter “tongue”, approximately 3/8 inches wide. This tongue is what effectively holds the panel flat through the changes in seasonal humidity, held in place within the groove of the end boards.

The breadboard ends have been pegged in place through elongated holes in tenons to accommodate seasonal movement.

With the panel complete, the table top profile is laid out. Most of the waste is sawn off.

Chisels are used to pare the remaining material to the layout lines.

The profile is then scraped, filed, and sanded to fair smooth curves.

SketchUp modeling

Posted on 27 May 2013 by J. Leko

Finally, I’ve completed the 3-D SketchUp model for the carcass. Since I tend to use models like this in the studio as my “plans”, I wanted to create as accurate a representation of the piece as possible. Everything you see, even the parts you cannot, has dimensions associated with it.

I started from “scratch” by re-creating the top from the dimensions and drawings I made at the museum. The rest of the model proceeded from there. In the end, it contains nearly all of the details of the original table including (some) joinery.

I find SketchUp to be an extremely useful program, and an important design tool. Representations like these help me plan out the piece, determine the sequence of steps I’ll follow as I’m building it, and aid in the purchase of materials.

Just for fun, I thought I’d include a picture of the original model I put together as I first contemplated this project, after seeing the Getty video for the first time…

It’s a bit crude, isn’t it?! 😉

Want to get a jump start on SketchUp? Popular Woodworking Executive Editor Bob Lang is teaching a five-day intensive 5-9 August. Check out the details and sign-up here.

J. Leko Furniture Maker

On re-creating an Oeben mechanical table…