Thursday, February 24, 2011

Drawer-Front Experiment

I needed to make two hanging cabinets for my laundry room.  Badly.  Not exactly blog-worthy.  But I decided to try out an idea I had for drawer fronts, and that is blog-worthy (I set a low bar :)
First, the final product so you know where it was going...

Okay, so what...
Ah, I see.  My photography skills couldn't capture the subtle triangle shape of the drawer front, but if you see them in person, it does make for a nice light illusion.  If you are there in person, please do some folding.
The drawers are really trays 18" deep on 14" full-extension slides (which are cheaper than 18" 3/4 extension slides ;)  This is very handy for getting at the stuff in the back.
If the fronts were just cut in to very shallow triangle profiles, it wouldn't be that interesting to discuss.  The top edge of the fronts are shallow triangles you see in the last picture, but the bottom edge is 1/8" thick the whole way.  While it's easy to cut, there are a couple caveats to mention.
First up, you need very straight-grain nearly quarter-sawn stock or the monster twisting bevel will make for some ugly grain; the straight grain is why these drawer fronts don't project they are cut unusually, but do that with the light.  Here you see the drawer-front stock with the trays clamped up.
Since you are cutting a series of triangles at first, you can squeeze a lot out of one board; each offcut is another drawer front.  I made the triangles a little longer than I needed, but ultimately, I think you should shoot for 3/4"-1" longer; remember that the sides of the triangles need to be cut back to where they are 1/8" thick plus cut losses.
Cut the triangles out on the bandsaw.  "Your eye is your fence."  To help assure I was holding the stock square, I blue-taped two Woodpeckers clamping squares together then pressed them to the back of the board.  Worked well!
After the first two cuts, each cut gets you another drawer front.
The final results of this first cutting pass.  Uhh, the reasons for the burning are the subject of another entry... some things shouldn't be cut on a bandsaw.
Cutting, part deux.  We now need to cut in a twist.  Idea is to cut the fronts again.  The bandsaw blade tracks the top of the drawer front (so... it remains a triangle; no change) while the bottom is forced to cut 1/8" from the backside.  My solution was this tiny 'resaw fence'.  It has a point and is beveled down to about 3/16".  I clamped it 1/8" away from the blade.  To do it again, I'd set it 3/16" due to material processing.
'Resaw' the front like I explained: track the top of the blade against the edge of the triangle not cutting it, but letting it cut a twist into the front so the bottom is 1/8" thick.  The difficulty here is (looking at this picture), you need to tilt the stock back towards the resaw fence as you advance to the apex.  That twist makes the bottom want to move away from the fence.
This is the result!  Solution? Start the feed from the apex to the outer edge.  Near the end of this cut, you will need to pull the stock through from behind the blade and use a sacrificial block to push the rest through.  GO SLOWLY on this cut as you are, in a sense, forcing blade drift.
I initially thought to use a block plane to clean up most of the problems.  I then remembered where I put my electrons.
I used my Festool Rotex RO-125 sander in disc mode initially for fast cleanup.  Note: use the semi-soft pad!  While I love the hard pad, using it here would 'facet' the twist.  Try not to round off the ridge from the apex to the base: that's what helps the visual.  Follow up with passes in random-orbit mode.
I hit them all with a sanding block with the grain, going to P180.
So... clamping these to the trays will be a chore: these are beveled faces so clamping pressure will move them.  My solution, as it is for all drawer fronts, is cyanoacrylate glue, or instant glue.  I spread regular Titebond II on the front of the trays leaving 4-5 thumbprint-sized areas glueless.
When you have the spacer for the tray ready so you can tilt up the drawer front, put a good dab of cyanoacrylate in each of the thumbprints.  Tilt up the drawer front, slide it to one side then slide back into position.  This gets the moisture in the Titebond II into the cyanoacrylate, which cures through moisture.  Once you slide it back, seconds later it is 'clamped' in place by the instant glue.  Now you can apply clamps and they won't move the drawer front; these clamps are for the Titebond II.  Someone will think, "but it is instant anyway!".  Maybe where you're from.  Here, in Arizona, it is so dry that "instant glue" has taken 5-10 minutes to set for me.  Literally, I have to brush water on the adjoining piece before assembly.
The whole cabinet and trays were sprayed with 4 coats of General Finishes Polyacrylic water-based finish.  This stuff is a dream to work with: flows out very well, can be brushed or sprayed easily, is tough as nails, smells nice, and you can easily get 3 coats on in a day (the recommended minimum).
After the first two coats, I hand-sand with P320 or P400 since a water-based product raises the grain.  After the third coat, I hit it with a white ScotchBrite to make it glass-smooth.  You could stop there, but I added a fourth coat and only lightly hit it with the ScotchBrite in places.
What remains is to add a 'vanity panel' underneath.  These are above a washer/dryer pair and things on those machines tend to vibrate around and possibly fall off the back.  The panels will be far in the back attached to the cabinets, but made to stop items from falling back there.
The divider between the cabinets was an offcut of maple I had that I decided to shape with a spokeshave and rasp (for the end).  It matches the long pieces mounted to the wall with a rounded end.  Thought it made them look a lil better.

1 comments:

  • Anonymous said...
     

    Cool. I've spent this week carving twists by hand using a drawknife, spokehsaves, scrapers, and sandpaper. I wish I could cut them with the bandsaw, but my confidence on the saw is not at that level. And my twists may be more extreme than yours.