Persistence in the face of turning disasters

I resolved to finally finish my nemesis bowl, whether it liked it or not, and so I firmly grabbed it by the base. I planned to figure out a way to deal with the chuck marks later.

After truing the rim AGAIN, I took my time and peeled away the central column with facing cuts from the bowl gouge. It had been my attempt to take a shortcut by cutting through the column near its base that had caused my massive catch last time, so I just grit my teeth and did it the slow way.

A chunk of loose end grain came flying off while I trued the rim, and it left an interesting shape. As I hollowed, I lost another piece.

In this photo you can see how I tried to fair the trued rim back into the out-of-true bowl body. It came quite smooth with a little sanding. And you can also see that another chunk came tearing away. I do love my helmet.

That's about half the rim gone. I won't be reversing this to finish the foot! Turning half a bowl is interesting, as you can see right through the side of the bowl as it rotates. It's a new perspective on the tool.

I finished the hollowing and smoothing of the bottom of the bowl with a round scraper, and then resorted to hand sanding to remove the substantial waviness left in the remaining wood. I also used the 120 grit cloth abrasive as a shaping tool to refine the jagged edge left by the breaks.

Very interesting, and well balanced by accident. There's more hand sanding to do, but I want to solve the problem of the broken foot next.

Throwing bowls across the shop

I had daydreamed about turning bowls from chunks of tree for years before I finally bought a lathe. I made myself learn the basics, of course. But the first time I caught a city maintenance crew felling a big tree in my neighbourhood, I scored a six-foot log. It was a dead or dying Manitoba maple with characteristic streaks of pink and red through the heartwood.

I chainsawed and split it into half rounds and set them aside to dry, but couldn't wait to turn some. I chose an odd crotch piece, wedge-shaped and massive. Any sane turner would have called it firewood, but I didn't know better and thought it would make a fine first green blank.

Look how nice and new that lathe is!

The chalk marks are where the wood was binding on the lathe bed. I just kept whacking chunks off until it would turn. When the blank was finally free, I switched on the lathe and immediately switched it off, as it was doing its best to walk across the shop from the eccentric weight distribution. (Ever wonder what's in a vibrator? A motor and an eccentric mass, not quite as shown.)

Apparently I got it roughed down somehow, and the blank sat on my shelf for seven years. I had already sawed out a wedge of splintered splits, and you can see the wax on the branch end grain.

I had tried finish turning it once already, and had more or less trued it up and begun flattening the rim when I had a catch that sent the bowl flying off the screw chuck and across the workshop. Enough time had passed that I was ready to try again.

I grabbed it by the foot, used the tailstock to adjust it to run as true as I could, and then turned a tenon for the chuck into the central column I had left while rough hollowing.

Reversed and grabbed by the tenon, I could true the figure on the new centre. After some sanding it started to look pretty good.

Time to cut a new foot. My chuck manual says that the minimum recess diameter for an inside grip is 2 3/8 inches, so I dutifully used that dimension. I was justifiably worried about the skimpiness of the foot, but foolishly ignored my misgivings.

In truth, it looks like the chuck will grab a much smaller inside diameter, maybe as little as 2 inches, and I will ignore the manual and use the smallest recess that works in future. But in this case, I went for it, had a catch in the hollowing, and sent the bowl flying across the shop again, with a chunk of the foot missing. My fault, of course. I set it aside until I am not angry with it again.

Finishing Bikes and Trees 2

With the turning complete, I had to figure out how to get a clean, straight slice through the work on the bandsaw. First, I made a sacrificial V-sled the hard way by grabbing a nasty old cedar table leg out of the burn pile, jointing and squaring it, and then taking angled cuts out on the tablesaw. Completed, it cradled the turning securely and let me use the mitre gauge to set the cut angle.

I also took the time to carefully reset the blade tension and guides so that the blade stayed quite straight and did not wander. I was rewarded with a very clean cut. I can't see any blade tracks in the photo, and it sanded dead flat in just a few strokes. (To keep flats while sanding, I lay a full sheet of sandpaper on a flat table, lay the flat of the work on it, figure out how to hold the work to avoid any rocking, and then take straight light strokes back and forth. It usually works out very well.)

A quick test fit kept me excited about the project. I like this:

But the sphere was now clearly oval, as the kerf of the cut was much thinner than the sprocket. I measured the difference with calipers and marked the excess with a round marking gauge.

Doesn't that look precise and sophisticated? I think the Fine Woodworking magazine next step would be to construct a shooting jig for a low-angle bench plane and take sweet shavings of end grain off until the marked line was just kissed. But I'm not that good.

Instead, I clamped a belt sander upside down on the bench and did my best to sand down to the line more or less flat. More less, in fact, which led to a lot more hand-sanding to restore the flat. (So maybe the shooting jig is the right thing, really.) But it got there.

I carefully put the sprocket right where I wanted it, and then traced the inside onto the wood. I had started out with the idea of a loose tenon, and then dowels, but drilling into the top was looking like trouble. I decided to use brads as pins, which let me progressively refine the orientation of the gear by adding brads.

Once the sprocket was locked in place, I snipped the heads off the brads and carefully positioned the top. Once I was certain it was where it needed to be, I pressed down to mark the pin locations, and then drilled at the marks. With a tiny drill bit, I could just hold the top to prevent it tipping on its round.

The press fit was very firm, and so I completed the assembly without glue after finally hand-sanding the turning with the grain. With a rubbed coat of walnut oil, it turned out very nearly as I had hoped:

I don't think that this will be the last of this sort of thing, as my head is exploding with variations on this theme. It's fun to be inspired!

Turning Bikes and Trees 2

Two of my lifetime passions are cycling and woodworking, bicycles and wood. I am inspired to look for interesting ways to integrate them. Here, for example, is a slightly rusty freewheel sprocket and a nice quarter section of black walnut.

My intention was to turn a sphere and bisect it with the sprocket.

The first thing to do is rough the wood down. I marked and punched the centreline I wanted at each end of the log, then bandsawed away a big chunk of waste and mounted the blank on the lathe. That's the roughing gouge sitting on the wood. It chips and peels the wood away very quickly, and doesn't rely on the work already running true for a good cut. You can just pass back and forth, sending chips flying. Eventually you will have a cylinder.

I wanted to start with a diameter similar to the gulleys of the sprocket's teeth, so I set a pair of calipers to that dimension. The parting tool is lying across the calipers, as I planned to use it to establish the diameter.

I laid out the sphere with three pencil rings, intending to leave the centre one and turn down to the outside marks. Partway there, I noticed a large split developing at the tailstock end. I stopped, flooded the crack with cyanoacrylate glue, and tightened a hose clamp around the end. (I removed the hose clamp before resuming turning! It's just posed on the lathe like this.)

I had time to think, and decided that I had better make myself a half-circle template if I was to have any hope of achieving a sphere. As it turns out, the diameter I'm eventually looking for is about 2.5 inches, and I have a hole saw in that size. I made a quick plywood template and used an orange pencil crayon to mark the highest spots.

Mark, turn away marks, mark, turn away marks... it didn't take long to get quite close to spherical. By this time my vision had changed and I wanted to leave the sphere on a simple footed post, and then set the sprocket at an angle.

I had used a square scraper to peel down the post, and now switched to using it instead of the gouge on the sphere as I crept up on the shape.

I sanded to 240 grit on the lathe and then parted the base off. I made a hash of the parting off, and the bottom of the base was indeed undercut but also badly scored. I mounted a sanding disc in a Jacob's chuck and cleaned up the base, doing penance for my turning error.

Now I had to figure out how to slice through the sphere in a straight line at the right angle. I didn't want to use the tablesaw, because I don't have a thin-kerf blade and don't want the sphere to become squashed. So I decided that meant a jig on the bandsaw and sanding to fit.

Welding a scrap steel cowbell

Before I finished folding the cowbell closed, I took a hand file to the edge of the back that will be welded to the bottom. I wanted it to lie nice and flat. I also put a slight bevel on the corners so that they wouldn't bind during final bending.

I haven't found a way to effectively clamp the bend while closing the cowbell, so I laid the top of the cowbell flat on my jointer table and tried to push the bend over further with a length of angle. I persuaded the joint into position with clamps, then locked it down with a big welder's C clamp. (G cramp, UK. Wotcher!)

My welding is improving. The major mistakes I made this time were in positioning the work and retouching the weld. I laid the cowbell on its side with the weld up, and the heat caused the sheet steel to sag noticeably. It would probably be better to position the cowbell on its mouth so that the side is supported while welding. The excessive retouching didn't help because of the heat buildup.

However, I did get a complete, solid, gap-free weld all the way around on the first try this time, which pleased me. Strong first, pretty later, I guess.

The next step is to drill out the mounting bracket. I put a 3/8 twist drill in the drill press and laid the cowbell on its top face (which is coplanar with the top of the mounting bracket, unlike the bottom). I eyeballed centre, put a bead of cutting oil in place, and slowly went straight through the bottom and then the top of the bracket, ensuring perfect hole alignment. Then I swung the table out of the way, stood the cowbell on its mouth on some blocks, and drilled a similar hole in the back for the thumbscrew.

I used 5/16 18 TPI rod and nuts for this cowbell, but as long as your hardware is sturdy and all fits together it doesn't much matter. To receive the screw, I welded a flanged nut onto the inside of the back of the bracket. A short bolt and second nut keeps the flanged nut in place.

I managed to grab the camera in time to catch the glow. It fades fast in the Canadian winter!

Finally, I wire-brushed the cowbell on all sides to remove the enamel and welding grunge. I like the brushed finish.

My local hardware store doesn't have thumb screws, so I used threaded rod and a wing nut. I have noticed that the world looks different now that I think of metal as something I can modify like wood.

I welded the wing nut on to the end of the rod, trimmed the rod to length with a cutoff wheel, and beveled the end on the grinder.

As a bonus, it looks way cooler than bright plated hardware store stuff.

This cowbell lacks the two distinct notes of the first one, which was 6 inches long and made of 23 gauge steel. At 5 inches and 20 gauge, it has a single ringing note that is higher than the last one, but still pitched lower than my storebought cowbell. Here's audio:

Folding a scrap steel cowbell

Someday I may feel justified in buying myself a sheet metal box bending brake so that I can do this right. In the meantime, I worked around the missing tool by making a brake jig for each bend with a couple of lengths of slotted angle. I clamped a length of angle along the bend line, front and back, and used the benchtop or another length of angle to push the steel over. The angle spreads the force and prevents the metal from folding, bending, or creasing anywhere other than intended.

I find that to get the fold right on the line, I need to place the outside angle edge right in the middle of the line, and the inside edge just under it. That is, the inside angle is set perhaps 1/32 inch below the outer one, to account for the thickness of the material and the radius of the bend.

The sequence of bends is also important. I folded the back over first, and then made the long folds, starting at the over flap and finishing with the under flap. It would be very difficult to fold the back last.

I had to persuade a couple of the long bends to square up a bit with an anvil and ball peen hammer.

The last bend can only go to about 45 degrees before it binds on the guide angles. That's OK, because I still need to brush off all of the enamel in the general area of the welds.

The mounting bracket will be folded over after the cowbell is welded closed.