Friday, March 2, 2018

Dark Chocolate and Sponge Cake (5)

The last month has had some rough patches. Our son aspirated some bits of almond, and this culminated a week later with a hospital visit, a bronchoscopy where 3 pieces of almond were retrieved, and a hellish 2 day stay in the pediatric ward. That was followed a week later by our son bringing home from daycare the norovirus, which caused him to puke up a couple of times and be generally miserable and get no sleep, and then I few days later I got it, and vomited close to 30 times. It took about a week to feel normal again from that. Good news is hat once you've had it, you are immune for a further 4-9 years. Not something I ever want to have again, let me assure you.

So, shop work is a little behind schedule. I've been plugging away the past couple of weeks on the latticework for this cabinet, which is on a hexagonal pattern. I've found the milling machine to be an exceedingly useful tool in the pursuit of accurate results with cutting the joints. With a standard orthogonal lattice, one can make spacing errors in cuts out and it often doesn't really matter - if one rectangular space between lattice bars is 0.05" longer or shorter than the one above or below, say, it is not something noticeable or of any consequence. It goes together fine, and looks fine.

However, with hexagonal patterns, you really can't deviate in your spacing if you want things to go together well.

One of the initial milling tasks was the creation of a gauging block, using the same 0.625" 4-flute carbide end mill as would be used for making lap joints, to mill a slot in a block like so:

This slot then could be used to gauge finished width of the lattice bars, like so:

Any fine adjustments to the lattice bars were made on the planing beam:

Once the sticks were dimensioned and finish planed all around, which took days, I moved on to rough cutting the notches on my chop saw:

I cut them in gangs of four:

Here's one set after rough notching:

Then I took advantage of my mill's rotary table, first by setting my DRO's '0' on the x-axis and '0' on the y-axis to the exact middle of the table. Then I attached a piece of MDF to the table and milled grooves so as to form a fixture for notching the lattice bars, or kumiko.

The grooves in the MDF made for a tight fit with the bars in place, so getting them out without a problem proved a challenge. I found that my Knipex plier wrenches, some new toys of late in my box, were excellent for this task:

I'll be doing a review on the Knipex plier wrenches in the near future, once I've had more time with them. So far I'm quite impressed.

My initial idea was to set up a partial grid work and then place insert pieces into the grid to serve as indexing points:

Then kumiko were fitted (indexing bars removed) and some notching could be completed, with some MDF packing pieces fitted between bars to preclude spelching:

The kumiko, once notched on one side, could be removed, the index bars fitted to the grid, and then the kumiko replaced onto the bars, like so:

The packing pieces were then refitted and the intermediate notches cut.

Trouble was, this approach didn't obtain the correct result, as the notches on the back side had to angle in the opposite direction to those on the front side. Doh! I had an incorrect image in my head as to how the notches were oriented. Fortunately, I'd only cut a few short pieces of kumiko before I became aware of the issue. No great loss.

Also, the repeated cuts just ate through MDF, which is a material I detest generally, especially the dust, and cutting it over and over accelerates wear on my end mill. So, I ended up changing my approach after milling those few small bars, and adopted a gang-cutting approach for the trenches on the one side of the bars for the rest.

For the notches on the opposite side of the bars, I rotated my table to the correct position and offset the mill head 1/2-way from the spacing I had used to make the fixing grooves in the MDF bed.

Unfortunately, milling the grooves on the backside of the bars could not be ganged up for cutting, and I had to mill them one by one, for hundred of notches. This simply ate up vast amounts of time.

And, despite precautions, some kumiko broke going in and out of the fixture, or had unexpected milling tear-out on notch corners, so I lost about 25% of the bars that way. Other stock, when re-sawn, jointed and planed for kumiko, warped unacceptably, so it was also rejected. Waste was close to 40% in total for these pieces I would estimate.

In the end though, I had my pile of sticks through the preliminary notching stage:

These sticks will be shaped on my Zimmermann Profile sander so as to accentuate the 'woven' look.

Here, I've done a test shaping on a pair of rejected kumiko pieces:

Another view:

The lap joints came out about as well as possible I feel. The curved treatment will be applied to both front and back faces. The advantage to doing this is that the sticks can have their arrises slightly chamfered without any gaps occurring where the sticks lap.

At this point I'm more than halfway through the process of shaping the sticks, and will be able to assemble the lattices soon. I'll save an accounting of that step for the next post. Thanks for visiting this time, and hope you'll drop by the Carpentry way again.