We're on the home straight now! The base is going to be a a simple 3 legged affair inspired by
George Johnson whose work with expanding dining tables I have always admired. As a matter of courtesy I emailed George to let him know of my intentions and he couldn’t have been more helpful, so if you get to read this George, many thanks for your very useful suggestions!
I still needed to make sure that what I had in mind was going to look right, would be sufficiently rigid, and I also suspected that there would have to be a deceptively large number of subtleties that I hadn’t appreciated, so I ran up a mockup in softwood first before butchering some expensive cherry. I started with some simple templates run up in scrap ply so that each component would be repeatable.
Here’s my first attempt half assembled. .
- mock up
- (397.91 KiB)
Having mounted the table on top, I concluded that I was on the right track but both the legs and feet were too chunky, so I slimmed both of them down just on one leg and foot to satisfy myself. That seemed to work. Those with an eye for engineering may say that the arched feet will have an unacceptably high bending load where they meet in the middle, and also where the mortise and tenon joins the feet, so will that central joint be strong enough?
Tests showed that it would be!
- strength test 1
- (188.95 KiB)
- Strength test 2
- (399.5 KiB)
So is my joinery that strong? Maybe, but could be that the strength can be attributed to this little beauty made in 50mm x 10mm steel by our local fabricators
- steel brace
- (398.21 KiB)
This is let in to the underside of the feet, and the bolts go up into the legs, through the mortise and tenon and into the dowel nuts which will be covered with a plug. All the bending load is being taken by the steel plate and I don’t think it’s going to go anywhere.
Time to repeat in cherry. First of all I make some templates in scrap MDF and plywood, taking the shape from the mock up. All the joints are cut whilst the stock is still square to make it easier to measure and cut. The central join in the feet is cut on the SCMS. The steel plate rises at 6 deg from the horizontal, so I simply set the blade 6 deg from the vertical, and 30 deg horizontally and bingo - job done.
Oh don’t you just wish it was that easy
. Whilst that was it in theory, in practice I had to nibble at it on the SCMS 8 or 10 times until I got it right.
- cutting centre mitres
- (245.53 KiB)
- centre mitres
- (323.81 KiB)
Fortunately I had anticipated the problem and had cut the feet 30mm over length to allow me to have multiple “bites of the cherry” (pun). Doh!
I wanted the ends of the feet to be curved as though they were part of a circle, so I made a very simple jig. First I screwed a piece of scrap that was the same width as the mitre channel on my bandsaw table to the underside of a piece MFC. Then I screwed another piece to the topside with a shallow notch in the edge which determined the length of the foot. I tilted the table by 10 deg, clamped the jig to the table, put the point of the foot in the notch and simply rotated the workpiece through the blade as though it was the hand of a clock. Simples! Here it is in action on the mock-up version. The recess on the underside is to house the steel plate and cut on the router table.
- cutting foot ends
- (231.14 KiB)
Next I attached the template to the workpieces using double sided tape and shaped it with a bearing guided cutter on the router table.
- profiling foot with template
- (276.93 KiB)
The mortise and tenons on the legs were simply cut by hand, and the length and angle of the M&T shoulders determined from the mock-up legs.
Another template, drawn from the mock-up determined the extent of the chamfers that were taken off each foot and leg. Having drawn around the template, the bulk of the waste was cut off freehand with the band saw, and then taken down to the line by hand using a microplane.
- shaping feet with microplane
- (322.09 KiB)
Time for a glue up. I put a domino in each face of the central joint more to index the surfaces than for any other reason, loosely screwed and bolted the steel plate in place, and then clamped the whole lot together. I had experimented with several ways of clamping up, and the simplest and most reliable method is the one shown.
- clamping feet
- (257.43 KiB)
Before finally glueing the legs into the feet, I rough fitted the legs in place and bolted the tops in to their final position. The wooden blocks screwed to the underside of the table are temporary. I’ll beef them up in cherry when I’m happy with the fit. This went according to plan, so the legs were glued into the feet and the bolts through the steel plate up into the legs were tightened up and the hole for the dowel nuts plugged. The blue tape is to mask the wood around the joint so that any squeeze out doesn’t soak into the wood and affect the finish.
- glueing legs
- (248.1 KiB)
And all of a sudden that’s the woodwork finished! Yay! Time to think about a finish.
My original plan was to use a resin finish like a bar polish, or Rustins Plastic Coating which is absolutely bomb proof as on my original side table test pieces. However, whilst bomb proof, it is a little like putting wood in a plastic bag, and in the event of damage needs to be completely stripped and refinished. It also dries quite fast and I was a little concerned that it might be difficult to get an even finish on a large area. Time to consider something else.
In the end I decided to test a Polyvine water based lacquer, Osmo Raw, Osmo finishing oil and Fiddes clear hardwax. I still had the test pieces from when I was trying out various glues for the sunburst, so I divided them into 4 and used a different finish on each quarter.
To my eye at least, the Clear Lacquer turned the cherry a greyish brown, which I didn’t like. The Osmo Raw was the palest finish on the maple which was great, but it completely killed the colour of the cherry, making the finish look cloudy. The Osmo Top Oil gave the cherry a lovely honey colour, and the Fiddes was similar. However, the Osmo Top was the most yellowing on the maple.
In fairness to Osmo, they do warn against using Osmo Raw on darker woods, and they also advocate just one coat is enough on pale woods, which should then have subsequent coats of Osmo PolyX or Osmo Top Oil. I tried this over the Osmo Raw on the pale test piece and it barely darkened it at all. But how resilient to marking are they? Only one way to find out.
- damage test 1
- (286.42 KiB)
- damage test 2
- (298.35 KiB)
I put a small puddle of red wine, milk, coffee, tea and water on each finish for both the cherry and the maple, and then added a cup of hot (but not boiling) water which sat in a some spilt coffee. I left this for an hour before wiping it clean. All 4 finishes shrugged this off, so nothing to choose on resistance to marking.
So the final decision is Osmo Raw on the maple, followed by the Osmo recommendation of 4 thin coats of Clear Top Oil (Satin). Not only should this be durable, but it will keep the maple pale, and in the event of any marks in future it can be patched easily, unlike a resin or lacquered finish.
I masked the edge of the maple as shown.
- masking maple
- (281.49 KiB)
I then applied the Osmo Raw very thinly with a cloth, and as soon as I had finished applying it I removed the masking tape to minimise any detectable ridge caused by a meniscus at the edge. Then LEAVE IT ALONE UNTIL DRY! On a test piece I had tried feathering the edge with a clean cloth, but when the Top Oil is applied, even the slightest smear of Osmo Raw on the cherry shows through. So LEAVE IT ALONE! Just to be safe I threw away all cloths that had been used to apply the Osmo Raw so that I didn’t use one by mistake on the cherry, and had I used a brush for the Raw I would have used a different one for the Top Oil.
Once dry I ran a scraper very lightly along each edge to remove any high points on the maple, denibbed and then applied the Top Oil all over. After 4 coats, denibbing between each with a 160 grade pad, it just remained to re-attach the apron and legs and put it into its’ new home.
- finished 1
- (321.18 KiB)
- finished 2
- (283.71 KiB)
- plan view
- (303.28 KiB)
- side view
- (296.37 KiB)
- finished
- (273.43 KiB)
- ready for dinner
- (309.6 KiB)
I’m not unhappy with the result. It does the job it was intended to do and fits in well with the rest of the kitchen. This project took me further outside my woodwork comfort zone than I have ever been and the learning curve was not only near vertical, but I had great fun climbing it as well. Sometimes you’ve just got to go for it!
Anyway - if you’ve got this far and not lost the will to live, thanks for reading.