Sunday, 15 January 2017

The scapradekijn for Castle Muiderslot, part 2 : more panel carving

Part 1 of this series of posts showed the start of the making of a scapradekijn for Castle Muiderslot. This second part will continue with the carving of the middle and lower panels of the hanging cupboard.

We are busy working on the scapradekijn at Castle Hernen. Bram is working on with one of the front boards on the workbench, while my board is clamped on a sitting bench.

The middle panel 

The basis of the middle panel. During construction some adjustments were made in the top and bottom lancet windows.

The carving of the middle panel also started by testing each part on a pine panel in order to get a feel for the technique and to decide whether it good good enough for the real panel. The routing of the basic parts was followed by drilling holes and carving. We initially had planned a carved rose in the centre of this panel, but then the people who had ordered the scapradekijn announced that they wanted a heraldic shield somewhere on the cupboard. The only reasonable place for the heraldic shield was the rose, as this did not disturb the design. Just after we had made the routing jig for the shield, the project leader of the Muiderslot project changed her view and wanted a more neutral cupboard without  heraldic shields. The rose was back in our design. Luckily we had not started yet with routing and carving, which would have made this last minute change impossible. Even more lucky, the heraldic shield of the jig was slightly larger than the rose, so we still could use the jig. 

The same router jig as for the top panel was used for the upper lancet windows of the middle panel with an insert added to adjust the routing area. The jig rests against a wooden rule (clamped separately with two clamps), allowing the jig to slide to the next position. This way only one dimension needed adjustment. the lines on the jig allow for correct placement on the panel.

Left: the result after routing the upper lancet windows of the middle panel. Right: the heraldic shield jig against the wooden rule. The heraldic shield is attached to the sides; the jig had to be turned over to allow this attachment 'line' to be routed away.   

The lower lancet windows of the middle panel were routed similarly as those of the top panel, but using a different insert. This produced a different window type. Also, for the central board one of the windows was lowered to allow for the lock plate.

The sequence of of carving the top part of the lancet window. The panel was clamped upside down in a double screw vise as this worked easiest. (1) The window just after routing, bevelling of the edges has been done. (2) Rounding of the curve was done with a half-round file. (3) Then a hole was drilled with a Forstner bit. (4) The hole was enlarged with round files. (5) The opening was sawn very slowly (carefully) with a small back saw.

The four crosses - looking like shooting slits in a castle wall - were made by drilling four holes, enlarging them to an oval form with a fretsaw and small needle files. Finally, the central hole was drilled. 

Two stages of a 6 mm thick oak test piece.

The setup for drilling the holes. The oak panel was clamped to a pine board in order to produce clean drill holes. 

For filing I clamped the oak board overhanging horizontally on the workbench; I found this the most stable filing position. You can also see the heraldic shields, that have to be turned into a rose. 

Left: The central hole is ready to be drilled. Right: the finalised cross.

The upper lancet windows were fully open in the original design, but we decided to fill it in to make it more attractive. Several designs were tested on the pine test board. Construction of these windows was done by drilling and enlarging the holes with needle files.

Left: The test panel with different options for the upper lancet windows. The first option on the left was used for the actual panels. Also written on the test panel is the drill guide, showing the size of the Forstner bit and the number of holes necessary. Right: the pine test panel was not very satisfactory, so we also drew the carving with a pencil on the oak panel.

The finalised small lancet windows.

The most difficult part was the creation of the rose. Testing a rose on pine was dramatic: everything broke of. A test on oak gave a better result, yet carving had to be done very carefully taking the grain into account. First, the heraldic shield was made round with a gouge. Then the rose was drawn with pencil and carved using small gouges and a carving knife. The petals were also partially cut free from the board. After the rose was carved the openings on the sides were drilled and enlarged with the carving knife and half-round files.

Left: the oak test rose, with the holes drilled for the side openings. Right: the panel with the heraldic shield - to be the rose.

Those who look carefully will see that the petals of the two central roses are arranged differently. When you look at original medieval panels, you will notice that such differences are commonly found. When panels look too similar, they are likely neo-gothic (19th century).

The lower panel

The routine for the lower panel was the same as for the previous ones. Only for the middle 'crosses' a new jig was needed. The upper and lower lancet windows made use of the old jig with an inserts to make it smaller. The upper lancet window and the 'cross' window were the easiest to carve: just drilling five (or three) and enlarging them with a round and triangular file. For the half-cross windows a carving knife was used to reach the sharp corners.

The basis of the lower panel. During construction some adjustments were made in the lancet windows.

The pine test panel, showing the different options for the carving.

The mdf board containing the jig for the rose also has a smaller version (left top corner) 
that was used for the 'cross' windows of the lower panel.

The upper lancet window and the 'cross' window were drilled with Forstner bits, 
while the oak board was clamped onto a pine underground. 
The upper part of the lower panel, before drilling the lower lancet windows. 
The V-form of the lower lancet window was filed to a point using a half-round file.

For the lower lancet windows fist the holes were drilled, after which the oval and the 'fish' bladders  were sawn with a fretsaw.

 The lower lancet windows finished.

One of the lower panels finished, but unoiled.

Bram working on the middle and lower panels of the front board at Castle Hernen.

All carved boards in a row in the workshop: two side panels (still under construction), 
three front panels and two pine test panels.

Wednesday, 28 December 2016

An octagonal medieval folding table at the Musee de Cluny

The foldable table from the Musee the Cluny, Paris. Cl.22795. Photo from internet.

The Musee National du Moyen Age also known as Musee de Cluny in Paris, France does contain an elaborately carved medieval folding table. While photos from the side are easily found on internet, photos with details of the construction are not. We were at the museum in April and able to take some photos of this medieval table. What we noticed was that the table was very worn - it had suffered much during the more than 500 years of its life. The oak was also very darkly coloured, much more than visible on 'regular' photos. As the room in which the table stood was also sparsely lit, it was difficult to take good photos. Nevertheless, they will show some additional information on the folding table. 

The octagonal table top consists of three boards. These were fixed together with loose tenons that were pinned. You can see the pins for the tenons on the photo.

The table is dates from the last quarter of the 15th century, first quarter of the 16th century. It has a height of 75 cm. The width of the table top is 90.5 cm, while the width at the bottom is 79 cm. The table was made in France. When taken apart the table consist of five numbered parts (two feet, two panels and one table top) and four pins to attach the table top to the pedestal. The board of the tabletop are fixed together with a loose tenons that are fixed with two pins at each side of the tenon

Most of the openwork tracery is missing parts. Unlike the scapradekijn for Castle Muiderslot and the rood screen, these panels are carved at both sides as both sides are visible. The panels were also constructed of several parts as can be seen by the lines on the photos. 

This photo shows how the tabletop is connected to the pedestal panels. 
There is one such construction in each quarter of the table.

The ornamented end of the pedestal panels.

The feet was carved but the elements have mostly disappeared. Likely these have been animals, 
such as lions, dogs or dragons that are often found on late 15th century furniture.

It is hard to figure out what this worn piece would have looked like.

This remarkable photo is from the 'Mobilier domestique vol. 1. Vocabulaire - typologique' by Nicole de Reynies (ISBN 2-85822-461-7) and shows the table with its different parts in a semi-detached way.

Sunday, 25 December 2016

The scapradekijn for Castle Muiderslot, part 1 : the start of the project

In May this year we were asked if we would like to make a scapradekijn or hanging cupboard for castle Muiderslot (also known as Amsterdam castle for tourists). We thought this was a nice opportunity to display our joinery craft to the public, as well as to do more elaborate carving and open tracery work. The task was daunting, as there was a deadline set for the end of the summer holidays. It became even more challenging when the castle management took about a month to consider and accept our offeeggler. For us the proposed time was too short - we have other regular jobs and a family to go on holiday in the summer - but luckily we could extend the deadline to early October. Anyway, as soon as we got the green light, we were able to start the project.

The hanging cupboard in the Museum for Angewandte Kunst in Cologne, Germany 
that served as the basic model for our scapradekijn.

Before the green light for the project, we already made a design of the hanging cupboard. The scapradekijn had to meet a number of conditions. It had to be 'open' as the Muiderslot wanted to display some (recreated) medieval things in the cupboard and these must be able to be seen. Also the cupboard had to be sturdy and idiot-proof (as some tourists would like to try out things that were not allowed). Our design was largely based upon the hanging cupboard in Cologne, but for the backside we used a grooved panel construction as this is more sturdy than a nailed backside.

The design for the hanging cupboard with measurements. We used different 'layers' in Photoshop to swap, fit and adapt the carved panels. Some of the carved panels and the ironwork were derived from the Cologne cupboard. The designs of the carved panels did change a bit as we tested some variants first on a (pine) board, and then used the one we liked most on the final oak panels.

We bought our oak panels ready to use. They were pre-planed, however we had to use a planer-thicknesser to get them to the appropriate thickness. Using power tools, such as a router,  as much as possible for the project was necessary in order to complete the project in time. Also the use of repetitive forms in the design allowed us to speed up the construction. The carving of the panels and the router jigs were first tested on a pine panel, before we used it on the oak panels. This way we also learned what strategy was the best and how the hand carving could (not) be done.

The top panel

While I will mention each 'panel' design separately, they were more or less worked on simultaneously. The photos will thus show the different stages of the different panels next to each other. Like the Cologne cupboard, the panel boards have different thicknesses. The front boards have a thickness of 11 mm, the side boards are 16 mm because they will also contain a groove for the shelves. Approximate half of the thickness was routed away for the deeper layers of the panel, either with use of a jig or freehand. Note that we carved the panels only carved at the front; the backside was just flat board, as this is not visible for anyone. This was also done in medieval times.

Front and back of two panels of a church rood screen dating from the 15th century. Musee des Arts Decoratifs, Paris. You can see that only the front of the panel is carved and the backside is just flat board.

Left: The router jig for the top panel. The lines on the panel and the panel itself are square (at the top and bottom) to allow the jig to be easily aligned. The round form is only for the top panel, the bullet form was also used for the middle and bottom panel. The router had a ring set at 5 mm, which was also the distance of the slope that needed to be carved. The clamped square itself acts as a fence to slide the jig to the next position. Right: The router jig with two inserts: one to produced a half circle and one for the open 'windows' cut through the board. For the latter one, routing was done in two layers; first at half depth without the insert, then through the board with the insert. (except for our first board panel when we had not figured out that this was smarter to do.)

Left: the routing setup for the top panel of the board. Right: the test panel after routing.

The router was fitted with a copy ring that left 5 mm between the router bit and the outside of the copy ring. This was also the space we needed to create the bevel. Thus, the router jig could also serve to draw the line for the bevel.

After routing the remaining work on the panel was mainly handwork. We used a carving knife as our main tool as it the most flexible use and a selection of other carving tools. Very useful were a low sweep fishtail gouge and an abegglen knife for corners that were hard to reach with normal gouges. For setting a straight bevel on the open windows, the use of the shoulder knife proved very useful as this provided a long stable cut. A cordless power drill with different sizes of Forstner bits was used to create the starter holes for the carving. To create rounded surfaces also round and half-round files and needle files were used.

Bram working on the first panel board at the Historic Open Air Museum in Eindhoven. On the right photo he uses the shoulder knife (or is it rather an armpit knife...)

The sexfoil pattern on the top panel was drilled with a Forstner bit using a cordless drill. At first we thought that using a paper jig (left photo) with the drill starting points on it was handy. However, the paper jig did not fit well into the deepened round, leading to misplaced drill holes. After that, using a pair of compasses to draw a circle on which the starting points should be placed, and a protractor to define the exact spot on the circle proved to be more exact. Right: a top panel halfway ready.

Some different stages of the top panels of the three front boards.

We also used to test panel to test how visible something inside the cupboard would be. We placed some of our medieval glassware at shelve height behind the panel. This pine test panel also shows the different options for the top of the bullet windows, e.g. the size and lay-out of the holes. We used the second option from the left for our cupboard.

Bram is busy with rounding the bullet window with a file (left) and carving the bevel/creating space for the trefoil.

Left: The deepened parts for the trefoil and quatrefoil were routed freehand. The surrounding area was pencilled black so it was easier to distinguish between what could be routed and what not. Right: A finished and oiled top panel.