Brendan Devitt Spooner makes a rosewood dining table

THE INSPIRATION for this dining table came when I was lucky enough to buy a large quantity of Borneo rosewood (.Melanorrhoea curtisii). With 35 cube of it tucked up in an already overcrowded workshop, and an exhibition on the horizon, I decided to use some of the part-machined planks for the top of a dining table. At this stage I was still in the dark as to the timber's workability and finished appearance!


Until someone comes up with the ultimate table design, that needs no legs or underframe to interfere with sitters' legs, a table must have supports. This can range from a single pedestal with radiating legs, through the standard four legs at the corners, to the highly extravagant lifelike forms of John Makepeace. But, whatever structure is chosen, the main function must be the support of the top. It would seem rather pretentious to design an underframe that looked fabulous, but was incapable of supporting six pairs of elbows.

My design was based around the triple tenon joint that links the two ends by a single rail.


The first stage in making this table is to start with the top. Five of my planks measured 2743 by 203 by 50mm (108 by 8 by 2in) and had already been part machined.

Bearing in mind that to sell a dining table off the exhibition stand, a sensible size is deemed necessary, I decided to make it 2133mm (7ft). This also helped with the physical problem of moving the table around - at about 651b a cube it is no lightweight!

After re-machining a face side and face edge, and then thicknessing to 45mm (l%in), the boards are then shot by hand for gluing together. Sometimes the surfacer plays ball and you end up with edges that need no further attention - but not always!

I do not rely on glued edges, and always use birch ply loose tongues. Using an old 6mm (!4in) slotting cutter on a spindle, I find the birch ply is a much tighter fit.

Twin stopped grooves are worked on each edge stopping at 25mm (lin) from each end. Before going ahead with this process, it is wise to mark out the end curves so that you can plot the groove ends.

• BRENDAN DEVITT-SPOONER is a former teacher who turned to furniture-making and design in 1987. Largely self-taught, he takes a delight in using solid wood and has a preference for English timber

RIGHT; A modern variation on a traditional theme - the trestle table taken a stage further

RIGHT; A modern variation on a traditional theme - the trestle table taken a stage further

« Plywood tongues

H Borneo Rosewood top

right: The Borneo rosewood, sandwiched between the oak, visually connects the two woods

Ply tongues

With all the grooving done, the next stage is to prepare the ply tongues. Because it is such a tight fit, I reduce the thickness of the ply slightly, until it is a push fit. When the glue is applied, the ply and timber will swell ever so slightly -enough to prevent a calm assembly.

Because of the length of the grooves and gluing surfaces, it is best to only glue-up two planks at a time, which prevents the glue from beginning to set and makes it easier to manhandle.

The gluing boards are placed in a row of T-bar cramps - don't forget to protect the timber and cramp faces by placing paper between them.

When you come to glue-up the next stage, check the edges are still true. The cramping forces may have altered it and it will need to be attended to - this is where it starts getting heavy!

alignment when gluing. Using a few short pieces of ply, the ends can be dry assembled to see how they look.

Main joint

Before gluing, the main joint has to be tackled, which is, essentially, quite straightforward. The real concerns are accuracy in marking and cutting.

Remove most of the waste from the mortices by drilling it out on a pillar drill, and then cutting to the line using a chisel. Don't forget that the outside face mortice will be slightly wider than the inside, to make space for the wedges.

The tenon cheeks are cut on a bandsaw and the remaining parts are done by hand. It is a good idea to make the tenons longer than needed. A small chamfer is worked around each end of each tenon which makes entry easier and also prevents the tenon from creating any breakout in the mortice. Getting the fit right is important - it is one place the client will look at and judge your work by.

With all the joint work done, the long curve on the main rail can be sawn and cleaned up. A 13mm (/iin) radius is formed along the curves, using a router and a bearing guided cutter.


With all the major work done on the underframe, it can now be sanded and finished. Using a pad sander, work from 100 grit to 240 grit and then finish with two coats of Danish oil.

The two ends and main rail can now be glued together. Have the appropriate wedges already prepared and, with the joints held tight with sash cramps, carefully drive in the wedges. Note that the force exerted by the wedge is across the grain. Hit them in too much and you will end up with left: Subtle curve adds some lightness to a substantial table. Note wedged tenons of the stretcher going with the grain -but there's plenty of strength here!

below: Tongues and sandwich construction clearly visible. Note fixing blocks are cut all the way through

"Getting the fit right is important - it is one place the client will look at and judge your work by"

Final surface

In a couple of days time you will have a large top that needs planing, scraping and sanding. Unless you know a friendly firm that has a thicknessing sander, the next two to three days will seem very long.

With quite a twisted grain in places, I found my trusty Record No 80 scraper invaluable. The sanding process needs no elaboration. When you think it is finished, lightly damp the surface, which highlights any undetected scratches that can then be removed before the finish is applied.

At this stage I like to leave the top for a while. If you are worried that it may cup slightly whilst unrestrained, screw two substantial bearers to the underneath, where the frame will eventually sit. If these bearers are made longer than the width of the top, it helps protect the top if it is leant up against a wall.


The design of the underframe was based on previous tables that I have made - the major difference being that I did not have any 50mm (2in) oak, but I did have a lot of 25mm (lin), so rather than glue two 25mm (lin) boards together, to achieve the required thickness, I decided to sandwich a piece of rosewood in between them, see drawing.

The oak is machined up at 19mm (Xin) thick, and sawn to 177mm (7in) widths for the wings, and 159mm (6Min) widths for the ends. These ends are loose tongued. When dry, the whole lot is thicknessed again, down to 27mm (IVitin).

The infill pieces, the rosewood for the outsides and oak for the insides, are thicknessed to 19mm (3/!in) and, using a vast number of G cramps, the six laminated sections are glued-up. Because the final dimensions are based on the outside parts of the sandwich, leave the infill pieces protruding slightly. Then, when dry, they can be planed to size.


The curve on the outside of the wings is a matter of taste. It is a good idea to mark all four wings with a French curve, and then position them in the final form to ensure that the grain directions look flowing. Using a bandsaw to remove the waste, keep the waste parts for pressure pads when gluing.

The vertical posts between the laminated sections are cut from 75 by 50mm (3 by 2in) pieces of rosewood. After initial surfacing and thicknessing, the angled faces are made by canting over the table sawblade to 12.5° and passing each piece in both directions, keeping the wider base part up against the fence. The angled faces are then surfaced and put aside, ready for grooving.


All the posts and sandwiches are then grooved for loose tongues. Apart from making a strong joint, it also helps for or

161 Y

300 -i



21mm radius

a split that wasn't intended! Knock them in with a hammer until the gaps have all closed up, and then just a little more.

Final glue-up

Whilst this assembly is drying, the curves on the wings can be cleaned up using a spokeshave, and the radii formed using the router. The radii on the posts also need to be formed. Because the corners are not 90°, these can be done by hand using a plane, and then sanded. Be careful not to sand into the joint surface.

The final glue-up is probably the hardest in that you do not have many 90° corners to pull up. I find that it is easier to glue the posts to the wings first, using the waste from the curve, suitably padded to aid a uniform pressure.

When dry, the wing assemblies can be glued to the main frame. With careful positioning of the cramps, and lots of softwood blocks, it should not cause too many anxious moments.

Last stages

The last stages in completing the underframe lie with filling the voids on the bottom with appropriate pieces of oak, and then cleaning up the bottom surface. This is necessary as it allows the table to slide on carpet.

A radius is worked around the bottom to prevent a ragged edge being formed. After spending a lot of time sanding the bottom, protect this surface from scratching by placing on an old piece of carpet.

The sections have to be cut out to accommodate the blocks which hold down the top. Because the void in the wings was 75mm (3in) wide, I made all the blocks this width and cut out 25.4mm (lin) deep sections on the four wings, and at each end of the main frame - these are then levelled using a straight cutter in a router.

Because the frame and the top will move the same way when subjected to varying humidities, these blocks can be screwed directly to the underframe, and have holes rather than slots for screwing into the top.

Back to the top

With the underframe completed, attention can now return to the top. Firstly the ends need to be shaped. Mark out the curve required and prepare to cut. Because of the weight of the top, it is a good idea to leave the table top where it is and use a jigsaw, if you have one - alternatively you can use a bandsaw. Unfortunately I didn't have a jigsaw, but with the judicious use of lots of people and odd roller stands I managed to rough a fair curve.

With a template of the curve in plywood cramped to the end, the edge is cleaned up, using a bearing guided straight cutter. The ends are sanded, starting with 80 grit and finishing with 240 grit. The radii on the edges are then formed, using a 13mm (Min) bearing cutter set at less than full depth, in order to give a definite line around the edge.

Before attaching the top to the underframe, the underside of the top is given two coats of Danish oil. The top is held down with 38mm (IMin) no.12 brass countersunk screws. The top is oil finished.


I was impressed with this psuedo rosewood - the colour is rich and it worked well, even though it was hard and quite abrasive. The only downside is that most people at exhibitions seem to prefer lighter colour timbers for dining tables, so I have still got one extremely heavy Borneo rosewood dining table for sale! ■

above: Plenty of French brown oak

Weight of history


John Lloyd on the heavy side of traditional furniture-making

THE MOST COMMON type of early dining table, often listed in 15th century inventories, was of tresde construction. In these, large boards of oak (Qiwrcus sp.) or elm (Ulmus procera) form a detachable top which rested on a number of supports.

These tables were often removed after meals, which is of course the perfect end to a perfect meal -manhandling a huge slab of oak must be a wonderful aid to digestion!

The other form of dining table around in the 15th century was the 'joyned' table, which consisted of - usually turned -legs which were 'joyned' by means of rails and stretchers with mortice and tenon joints.

These were also of very heavy construction and had the benefit, if you were a diner, in that they were not of 'knock-down' construction; however, the huge bulbous legs that were popular from the late 16th-century meant that you certainly knew about it if you got the leg!

My design brief for this project was for the trestle form of table which was made in the early 16th century; it was to have curved supports at each end linked by rails at the top and a single stretcher passing through the uprights and secured by wedges.

The size of the top was also prescribed, dictated by the size of the room and the number of people required to be seated around the table.

French connection

The wood decided upon was oak, at my suggestion of the quarter-sawn brown variety.

The construction of the table in terms of joints and so on was, therefore, going to be fairly straightforward; the challenge turned out to be in sourcing the oak and its subsequent handling while converting and jointing.

My first lead for the timber was a local estate which had, apparently, two butts of brown oak tucked away.

When I arranged to inspect them it was found that they had mysteriously vanished - it was suggested that an estate worker may have cut them up

"It was suggested that an estate worker may have cut the brown oak up for fence posts because the wood was a funny colour!"

for fence posts because the wood was a funny colour!

Luckily the French have a much less haphazard approach to the management of their forests; it seems that their standing trees are listed, and if an oak tree is brown oak it is recorded as such.

Consequently the wood was imported from France, and an afternoon was spent sifting through several butts of brown oak - no mean feat when the boards you are sifting are over 50mm (2in) thick, 480mm (19in) wide and 4 metres (12ft) long! Having picked the butt with the best figure and most even colour it was transported to a local sawmill, to be converted into slightly more manageable pieces.

above: Plenty of French brown oak

BELOW: End cleats are loose-tenoned into stopped grooves

BELOW: End cleats are loose-tenoned into stopped grooves

ABOVE! Three tenons at each end join trestle-ends to feet and bearers

RIGHT: Top rails are dovetailed into the bearers - cut the tails first, then mark them onto the bearers

"To keep stress levels to a minimum I like to have as much as possible under my control when gluing up"

Table top

The boards selected for the table's top are from the centre of the butt -these have the best figuring - and to give a symmetrical appearance each was split into two 25mm (lin)

boards and arranged in three pairs around the centre line of the top.

Having planed and thicknessed the boards there is just the small question of jointing them. I do not possess a planer with tables the length of Brighton Pier so this meant resorting to my trusty No. 8 jointer, and a great deal of heaving big lumps of wood around.

Balancing these huge boards on their mating edges and, with a light source behind, peering along the join to find the high spots was followed by more heaving, a bit of judicious planing, yet more heaving and peering until the perfect joint was attained!

To keep stress levels to a minimum 1 like to have as much as possible under my control when gluing up. To this end, five dowels or biscuits were used in each joint to keep the top surfaces in register and the boards glued up in pairs, taking care to protect the freshly-planed edges from indentations from the cramps.




"Because of the length of the stretcher and the rather confined space in my machine shop, I cut the shoulders of the tenons by hand"

Biscuit jointed top


Loose tongue

Rails with dovetail joints rail with loose wedged tenon

Foot rail with loose wedged tenon



0 0

Post a comment