Exhibition Box Pt 1
Monday 08 June 2009
The 3D and 2D lattice construction of Robert Ingham's 275-piece box must make it one of the most complex ever made. In the first of a 3-part series he explains how he achieved it
This box came about as a result of being approached early last year by Peter Korn who runs the Center for Furniture Craftsmanship in Rockport, Maine, USA, to take part in an exhibition promoting the work of 25 international box makers. I felt privileged to be included with top British makers Peter Lloyd, Andrew Crawford, Chris Pye and Ray Key, to show our work at a 3-year travelling exhibition which starts in Rockport.
The event will be further enhanced by the production of a book entitled Boxes and Their Makers, to be published by John Kelsey and the furniture historian Dr Oscar Fitzgerald of Fox Chapel Publishing.
Design metamorphosisI mused for quite some time about the design and was inspired by a chest of drawers I made a few years ago.
Lattice Tower drew inspiration from Japanese expression, with a strong emphasis on squares and rectangles. The box that evolved was a scaled-down version, but instead of containing drawers thi one has a lid and interior partitions for jewellery.
The 3-dimensional contribution of the lattice construction is apparent on the lower part of the corners of the box and is continued 2-dimensionally up the front, back, sides and top with the contrasting colour of another timber.
This led to a drawing which in a simple form was sketched on squared paper to examine the proportions. I drew the front corner full size and was quickly influenced by the fact that the squares of the paper measured 5 x 5mm; times three, the width of the lattice strips, worked out at 15mm and this became the unit dimension for everything that followed.
Once satisfied with the sketch I committed myself to a full-size working drawing on a piece of white melamine board which lived on my bench for the duration of the project.
Project overlapI was getting to the finishing stages of another piece when I produced the drawing and this overlap gave me time to tweak and further develop the design of the box. In fact the design continued to evolve during the progress of the box, but before I started to prepare some of the components I thought that it was too hard and too square.
It needed softening, so I decided to round the internal and external surfaces of the corner posts. This in turn led to the inclusion of a circle that connected the lattice strips on the top.
The construction of the components was going to be labour intensive. As with all my boxes, I wanted to make everything myself including hinges, stays and lock.
The setup and making of jigs was going to be time consuming so these considerations prompted me to make three boxes with different timber combinations. I have done this in the past and it has enabled me to be much more ambitious with the final permutations of colours.
Applying veneersThe lattice construction appears to form a frame around the more striking grain of the contrasting show veneer. Whatever is visible on the outside must be mirrored on the inside. In one box, the lattice is made of pearwood (Pyrus communis) - framing the stunning grain and colour of Macassar ebony (Diospyros celebica).
Stability of the finished product was high on my list of priorities so I decided to use MDF for the substrate on which the veneer is laid. I could have created the lattice pattern with veneers using a marquetry technique, but the colour and grain pattern of the veneer would not match the solid wood of the outer frame and corner posts so I inlaid strips into 1.5mm-deep grooves instead. I also wanted to enhance the pattern of squares, rectangles and strips with a contrast in grain direction.
In the past, I have cut the veneer and joined the edges by hand planing and using my edging jig, but because of the large number of junctions I took the risk and cut packs of veneers on my dimension saw.
Fixing a spelch board made of 3mm MDF onto the sliding table with double-sided tape prevented any breakout on the underside of the pack, and with a trapezoidal-tooth saw blade the cuts were every bit as good as they would have been with a plane despite the fact that 50% were across the grain - and all this was done in a fraction of the time.
The packs were held together with masking tape and held down onto the spelch board with the sliding table cam clamp, photo 2. The veneers were then glued edge to edge and held together with masking tape which was removed after the glue had set.
Making tapeless jointsThis is my method of making tapeless joints, photo 3. The fronts, backs and sides were narrow enough for the veneer to be glued on with cauls held in my bench vice - I have three vices on two benches.
Using Titebond III with its short pressure time, these narrow components could be veneered very quickly. I applied the glue with a comb spreader and used thick cauls made from offcuts of melamine kitchen worktop material.
To prevent any risk of glue that might migrate through the veneer bonding to the melamine, the contact surfaces were covered with brown parcel tape, photo 4. Because they were wider, the tops and bottoms were veneered with similar cauls in my platen press. This was a more convenient method, though it could be done in a bench vice with additional pressure from G-cramps.
sanding jigThe principal of sanding circles and arcs is quite simple. The component is rotated against the disc on a pivot inserted into the sanding table. A commercially made device enables the distance of the pivot from the disc to be adjusted mechanically, but because its surface area is too small to support the underside of the box lid I have made a larger jig from 6mm-thick MDF that is held onto the sanding table with G-cramps, photo 6.
The hole for the pivot is drilled at a distance from the disc, slightly shorter than the radius of the curve to be sanded, and a 6mm-diameter piece of metal rod is inserted into it.
A complimentary hole is drilled into the underside of the lid and this is then dropped onto the pivot so that the lid can be rotated against the sander to remove the unwanted material.
Drilling the holes in the corners of the lid is critical. This was made possible with my milling machine with the two horizontal axes being adjusted to tolerances of 0.02mm. Adjust your sanding table jig to increase or decrease the radius by slackening one G-cramp and gently tapping the MDF table until the right dimension is achieved before the corner is sanded to the right profile. To do this I made a dummy from an MDF offcut and sanded this until satisfied with the fit, photo 7.
Abalone insertsAt this point I decided to introduce some additional colour. I had intended to line the bottoms of the boxes and trays with a blue-green-toned crushed velvet, and this colour combination is present in abalone shell which is now available as a reconstituted sheet material that is popular in musical instrument making.
A square of abalone in each corner of the top seemed appropriate so I used my milling machine to create a recess for them, photo 5. This could have been done with an overhead, hand-held router but the degree of control that the milling machine affords is irresistible. Set up with the base plate, all three tops were done in a very short time. The abalone inserts were glued in place with slowzap super glue.
Once the glue had bonded I set about rounding the corners to match the inner curved surfaces of the corner posts, using a Hegner disc sander.
Recesses & inlaysThe next stage was to rout the recesses for the lattice strips that form the pattern on the surface in contrast with the Macassar ebony. For the top, the ideal setup involved router table and 15mm TCT router cutter, photo 8.
For the front, back and sides I used a combination of the router table for the long strips and the milling machine for the shorter cross strips, photo 9. The dimensions were set up using dial callipers in conjunction with the sliding rules of large and small combination squares, photos 10 & 11.
To prevent breakout from one recess to another I routed and inlaid all the grooves in one direction. The inlays were inserted and planed flush in succession until all were completed before the opposing directions were routed and inserted, photo 12. This is a typical example of planning a sequence of processes that are efficient and that take into consideration the problems of breakout when grooves cross each other.
The inlay strips were made from a piece of wood planed so that the thickness was exactly that of the width of the recesses.
These were then sawn to a thickness of 1.7mm on a Hitachi band-resaw; the stellite-tipped blade produces such a clean cut that I was able to use the strips straight from the saw, photo 13.
Machining top circleTo rout and machine the circle in the top I decided to treat myself to an engineer's rotary table which can be mounted on the milling machine table horizontally and vertically; I planned to use both functions in this project.
The first stage in creating the circle was to rout a recess into the box lid with a hand-held router. My Microfence circle jig was too large for the small radius of the circle so I drilled a hole in the supplementary base of my router to accept a brass pivot that was inserted into a hole drilled in the centre of the lid. This hole went through the lid so that the process could be repeated on the underside.
The depth of plunge was set to 1.5mm and the recess was routed. An abalone disc would be inserted to cover the hole on the upper surface and my logo would be fitted to do the same job on the underside.
With this setting I also routed a recess in an offcut so that I could test the fit of the inserts and dimension the ends to length after the profile had been produced on the rotary table, photo 14.
The block of wood from which the curved insert would be milled was glued onto an MDF carrier plate; I made two plates for each lid to make sure that I could slice off sufficient inserts. Mounting holes were accurately drilled in the carriers which were then screwed onto a centre plate fitted to the rotary table. The position of the cutter was accurately adjusted to produce the outer curve of the inserts and the rotary table was rotated to remove the waste.
The diameter of the curve was set with the precise adjustment of the horizontal axes wheels. Because 50% of the wood being milled away was against the grain I made several passes, with the final cut removing only 0.1mm.
Once the outer curve had been cut I moved the table by precisely 15mm and milled the inner curve, using a succession of light cuts; because I could zero the drive wheels of the milling machine sliding table I could be certain to finish with the final dimensions, photo 15.
Using the bandsaw, I cut off 1.7mm-thick slices; because the profiled section was still glued to the carrier plate I was able to do this safely. The length of each slice was slightly longer than a quarter of the circumference of the recess of the ring into which they would be inserted, enabling me to size them accurately before gluing them in place. I used the test block made from an offcut and planed the quadrants to length with a block plane pushed on its side, photo 16.In the next part, Robert assembles the box and cuts off the lid