Apothecary chest

[iclark]

Derek,

Just a thought: you might consider using your prototype drawer as a backing board when you are tracing the curves for the drawer fronts. That should cut down on the minor left-right rocking that is so hard to avoid when there is little surface area to hold the piece being marked.

If you put the prototype drawer in first, it is harder for the piece being marked to rock. It might be even easier if you use double-back tape to attach the drawer front to the prototype. Then you can just reach around the back and hold it in place (or clamp it) while you trace the curve.

[Derek Cohen]

Thanks for the suggestion. My plan is to build all the drawers up to the drawer bottom stage (that is, before it it inserted), and only then mark and shape the drawer front. With the attached drawer sides and back, it should be pretty stable.

Regards from Perth

Derek

[hbmcc]

Derek should be done by now.....

[Derek Cohen]

Derek should be done by now.....

Yeah, right!


Regards from Perth

Derek

[Derek Cohen]

I was in two minds whether to post this, but since the method is a practice, it would be great to get feedback, since the strategy I have come up with is complex. Can you do this another way? 

Each row has 4 drawers, and these will be shaped to match the bow across the chest. 



At the start, the drawer fronts are to be left straight. This maintains the reference sides. The ends of each drawer front have been bevelled to match fit the bow of each drawer blade.

This is a fitted (practice) drawer front (posted last time) ..



The drawer side has been dovetailed to the obtuse angled side (again, details in my previous post:  http://www.inthewoodshop.com/Furniture/A...kend8.html)  ...



The need now is to dovetail the acute angle ...



This is where it gets interesting. It you look at the lines drawn on the drawer side, if made coplanar with the drawer front, the dovetails will need to me cut at an angle. That is much too complicated, and likely to be a poor fit.

Then, if the baseline is cut square (as usual), the dovetail will end up in the centre of the side (and not extending up from edge of the board).

The only way I could come up with for a fit that simplified the tail board was to rebate the pin board, so ....



The rebate needs to be as deep as the drawer side (for a flush fit), and square to the side (so the baseline of the tail board fits flush).

The first step is to mark the baseline ...





On the piece above, you can also see the rebate markings.

The rebate is now cut parallel to the side ...



Remove some of the waste with a chisel ...



Now that rebate needed to be both straight and flat. It needs to be an equal depth along its length.

It could have been chiselled, but that is less efficient. A shoulder plane as this would not ensure a square shoulder without extra work to create an absolutely square edge for a tight fit. In the end I came up with this idea to plane it using a LN Edge Plane. 

A spacer was attached to plane to the 1/4" depth ...





The finish was spot on ...

[Derek Cohen]

Transferring

The rear of the tail board, with blue tape used to create a fence ...



Tails on pins ...





The socket shoulders are deepened to create a socket that undercuts the baseline ..



Because the angle was so difficult to chisel, a trimmer router was used to remove most of the waste ...



... before the remainder was removed ...



Coming together

The fit ...



The angle ...



This is a rough idea of what it will look like once the drawer front is shaped ...



The two sides that must be made for all drawers ...



Regards from Perth

Derek

[BpB123]

Thanks for the amazing details/instructions!

[stav]

This is probably why you don't see these chests with curved fronts.  It's just so complex an average woodworker would not attempt it.  I'm looking forward to every installment of this build.

[giradman]

Thanks for the amazing details/instructions!

1+  Dave

[hbmcc]

An excellent walk-through of the process! Thanks Derek. 

You have the needed side depth at mating to accomplish the geometric of positioning the tail board so don't really need the wedge (false face) you notched. Just allow tolerance for adjustment at the tailboard. I can't fault the direction you propose. It's a mechanical confirmation, and redundant safety check.  

I can nit pick details: I can't see those acute corners surviving as shaped in the test pieces. Are you looking a few moves forward?