Post 11 in a series describing the design and construction of a dining table based, in much of its structural system, on an unusual Ming Dynasty side table. Previous postings can be found in the 'blog archive' at the right of the page.
Today was a big day as I finally had the custom shaper knives in my possession. I had been holding off doing more work on the principal rails, or aprons if you like, as the shaping cuts with this curly grained bubinga I figured are pretty much one-shot affairs. I had decided that taking the cut in one pass was the route most likely to avoid any tear out, as the deeper the cut the more material is present to give support on the exit side of the cut.
So, a good chunk of today's activity was consumed by setting up the shaper for these cuts. One of the rails had moved a slight amount necessitating a little tune up with the 60˚ plane:
Then set the insert knife cutter head to the required height:
Now, you'd think that an aluminum extrusion such as these shaper fences, would be flat and straight, however I checked carefully and discovered that the Powermatic did not have dead flat fences. The cast iron fence supports are carefully aligned and flat (after some tuning), but the aluminum is simply not quite flat. I find that sort of thing incomprehensible, considering that probably one of the most important features, period, on a shaper would be fences which are co-planer, flat, and 90˚ to the table. It must be some sort of manufacturing challenge I guess, though companies like Martin seem to be able to execute the basics well, and then some. That's the point that riles me a bit - the basics - forget the bells and whistles of digital readouts and lexan safety shields, I want an accurately made fence. What's so hard about making that? Why do I have to spend hours fiddling until I get something which should be so fundamental to the machine set up properly? It's like that classic line out of Star Trek, "I'm a doctor, not a ___":
Anyway, incredulity and grumbling aside, I am most grateful to have a nearly new shaper available for use, that's for sure, and with all this practice in setting up machinery, perhaps I could look into a career as a mechanical engineer.
I set about building a sub-fence, which I was able to shim with some paper here and there to achieve the flat, 90˚ reference surface I was after:
Here's a view from the front:
I then fitted a second sub-fence atop the now-flat and square base, to create a zero-clearance insert of sorts. I ran the machine and backed the fence into the cutter until I had the opening I wanted:
The next kerfuffle I had was with the stock feeder. You'd think the purpose of a stock feeder was to feed wood, but this one was really uncooperative. It's nearly brand new, and the wheels aren't glazed over or chewed up- they just seem to be of a particularly un-grippy urethane (or whatever the material is) compound:
I fiddled and fiddled with that feeder, waxed the tables and fence, etc, and in the end had to settle for the feeder's mere assistance in moving the wood, as I did at least 50% of the pushing to move the stick past the cutter.
To test the set up, I ran a piece of Yellow Cedar through and it came out perfectly:
Now the moment of truth, my crossing of the Rubicon as it were, my walk down the Valley of Death. Well, maybe that's a tad over-dramatic, but let me tell you it was a little stressful. Wood like this doesn't grow on trees ya know.
I was more than a little relieved when all went well with the shaping. Three and a half hours getting ready with the shaper, and 5 minutes of cutting and they were done. There was next to no tear out and the pieces came out looking even more fabulous than I could have imagined:
The photo really is not doing the wood justice here I'm afraid. There's still some more profiling to be done on the front faces, but that is the main piece of work taken care of. Whew! It's called living on the knife edge my friends!
-- okay, okay, that's enough word play for one day!
This profiling cut out will also help to equalize the stresses in the pieces from all the drilling out on the backsides I did a couple of days back.
And on to the next thing- the center rib. One change i have made in the design from the earlier posts in this thread involves the center rib, which now will have an exposed portion on the table surface rather than being hidden below. The reason for this change concerns wood movement. I ascertained that a 20" wide bubinga edge grain panel, given a moisture content swing of 4% could move 1/8". So with two panels, the total amount of movement to be accommodated is 1/4".
Actually, I believe it will be less than what is indicated by the wood movement calculations, I do believe, because these are done on the basis of un-finished wood samples. Surely it is the case that wood with a Tung oil (or other) finish will somewhat inhibit moisture transport compared to no finish at all? I am skeptical, given past experience with this material, that the top panels will move as much as the calculation indicates, however I don't want to take any chances either. Wood movement can lift foundations of houses and certainly wreak all sorts of havoc with a table if allowances are not built in. After thinking this issue over for a couple of weeks, I consulted with the client, then sent him drawings of various possibilities and we talked it over.
You see, the panel movement can be taken care of in 3 different ways:
the panels can be fixed at the outer edges of the table so that all movement occurs in the middle.
the panels can be fixed at the middle so all movement occurs at the edges
the panels can be fixed centrally and allowed to move both at the middle of the table and at the edges.
In the end, the solution that the client preferred was #1 above. So, I will be installing the table with a 1/8" gap on either side of the central rib to allow the panels to swell in the summer months to close the gap.
The result of this design change is that the central rib is to have a 'T'-shaped protrusion on the top to capture a tongue on either side from the top panels. To keep this 'T' adequately stout at the base and allow for movement, I designed it to be 1" wide on top.
Time to have a little fun with the router:
A while later, the piece was taking shape - here's a view with the 'T' shaped rib in process, over on it's side:
Some time after that, the roughed out 'T' is really taking shape, now oriented topside:
And later on I had the desired result:
The protruding part will be a male dovetail tenon eventually, one every 15" or so down the stick:
The target depth to accommodate the 1/2" thick table top panels was .5050", and I obtained that, +/- 0.0020":
The idea is that the rib will protrude 0.005" from the table surface, which is enough for a couple of passes with the plane to clean it to finish.
Next time I'll be digging into the legs and getting into all sort of trouble. I hope you'll stay tuned and, as always, thanks for coming by today. Post 12 is up next.