Saturday, February 21, 2009
Friday, February 20, 2009
News Flash: McKennawerks goes high-tech!
Thanks to the technical expertise of Generica, My Lady Wife, I'm beginning blogging this from the laptop down in the bowels of the McKennawerks. Gen, with the able assistance of our computer guru friend Gregor, were finally able to get the wireless router to recognize the laptops, so we can compute while sitting in front of the warm woodstove, instead of at the main computer which lives in the coldest corner of the house. Apparently wireless routers don't like dial-up, and it took some doing to convince it that, yes, while we are still stuck in the middle ages courtesy of Verizon, we really do want to use more than one computer.
I've also added a hit counter just for the heck of it. I looked at adding a guestbook as well, but after checking out some on other blogs it appears that the comments are mostly banal, off-topic, spam, or obscene. I'll just stick with google's comment widget. And of course you can email me at McKenna(one word)werksatgmaildotcom. I don't normally bring anything of a sensitive electronic nature down here due to the ever present dust and potential for (fragile and expensive) things to be accidentally knocked to the concrete floor, but we're in a holding pattern because we've started to steam.
So it's time to talk about some of the specifics of steam bending wood. It requires one hour per inch thickness of wood, plus (in this case) bringing 3 gallons of water up to a rolling boil. Fine Woodworking has a good book on many different bending techniques. If you're interested in some in-depth knowledge about how the pros do it you should check it out. I've not steam bent anything in ages and ages, I much prefer cold lamination with glue for bending to steaming. The steaming part of steam bending is easy and straight forward: where the problems arise is in the bending part.
To bend wood with steam we need:
--patience of the saints
--a source of steam
--wood to be bent
--a form to bend the wood over
Patience. While wood, especially green wood, is flexible to an extent, it doesn't like to bend and stay bent. Depending on many factors, especially thickness and grain of the workpiece and amount of curve, you can expect a failure rate of 50% or more. This should elicit no more than a muted, "Darn." If breaking wood is going to send you into a cat-throwing, obscenity yelling, jumping-up-and-down rage, stick with lamination.
Time. You need an hour per inch thickness of wood, and it doesn't start until a good head of steam is built up. Then you have to get the wood instantaneously into the form while maintaining a serene calm.
Safety. For safety for this project I wore high leather boots in case of a spill: I didn't want boiling water splashing my ankles and soaking into sneakers. There was a fire extinguisher close to hand, and I wore good leather gloves throughout the process in case I needed to grab something hot.
As in metal casting, though, if something starts to "go" it's usually a lot smarter to back off and let it and clean up the mess, than to risk a burn by grabbing at something. Eye protection, of course, is essential.
Steam. For the source of the steam I'm using an old turkey fryer that I actually bought with steam bending in mind. Several years ago we held an early spring outdoor event where we made a big pot of lentil soup (and roasted a pig, yummm). Since then the cooker has been sitting at the back of the shop awaiting the day I finally decided to steam bend.
For this bend I started by heating about 3 gallons of water with a hot plate, but it took a long time to get it up to a low boil, and couldn't really get it up to the rolling boil we need. So I switched to a propane camp stove. This was an improvement, but if I weren't doing a one-off project, I think I'd bite the bullet and fill up the propane tank for the gas grill and use the turkey fryer stand.
For the steam box I used an old piece of 4" thin wall pvc drain pipe. I cut a plywood plug for one end and drilled out a hole for the inlet pipe near the bottom edge of the plug. As water condenses from the steam in the steambox it will run back into the boiler. Again, avoiding a closed system, I left the irregularities in the edge of the plug from cutting it out with the sabre saw in place. The pipe clamp you can see around the end was due to concern that the heat might soften the plastic to where the friction fitted plug would slip out. It's just helping the pipe retain it's shape--it is not clamping the plug in place.
The piece of pipe I had lying around was too short for the 9' long poles so I extended it with flexible aluminum clothes dryer hose.
Wood. The wood I'm using is (probably) ash. Oak, ash, and hickory are widely considered the best woods to use steam bending with. Don't bother with pine-it'll split. The thicker the piece, the longer it needs in the steambox (one hour per inch) and the more likely it is to split. The tighter or more extreme the curve the more likely it is to split. Also try to use pieces with long straight grain, it's more likely to crack where the grain 'runs out' to the edge of the wood. It's easier to stream green wood. Kiln dried wood does not steam well. If you're stuck with MegaLoweMart lumber, I'd strongly suggest considering glued up laminations.
Form. The form for bending was made from scrap lying around the shop. There is a good distance between points where the blocks push against the piece to be bent. If we were making furniture or boats where precision is a prerequisite we'd make the bending form almost continuous along the bend, but such precision isn't needed here, and would not have been bothered with by our hypothetical medieval craftsman.
To use it we need lots of clamps. Lots and lots of clamps. If you don't have them there are work arounds to clamps that I covered in an earlier post. The other thing is to make sure you do a dry run. Have all the clamps laid out to hand. Have a hammer ready to hand to help tap along the piece if necessary to overcome friction against the mold. It's really better to do one piece at a time, but since one piece was so close to the desired bends, to save clamping and drilling holes in the bending form to take clamp ends I went with doing both pieces at once.
To do this I placed small pieces of 1/4" dowel between the pieces as spacers,and used cable ties to hold them together in rough alignment. Once the boilers had built up a good head of steam they were ensconced in the steam box. Then we wait.
After a tad over an hour (the pieces are a tad over an inch thick) I pulled them from the steam box and stuffed the front ends under a cross piece and forced the back under another cross piece at that end. Then frantically began clamping. The starboard piece went in fine, but the port was a little stiff. Unfortunately, after about 20 minutes in the form there was a sudden crack! and the port piece split at the compound curve where the harness hitch would be.
This is not a total disaster, as I have plenty of wood to make another pole from, and I can cut pieces from the split pole to use in other parts of the cart.
Post mortem analysis shows that I left the split piece a little wider than the other intending to trim it down to size after bending--I should have made it the same width as the other. Secondly, when I was clamping it there was more resistance than the other so I should have aborted the clamping, let the wood cool for 24 hrs and started over. Impatience is often my besetting sin. Also, I probably shouldn't have done both at the same time. There's a lot of truth in the old saw, "if you don't have time to do it right, will you have time to do it over?"
As I said above-expect splits. I'm disappointed it didn't go perfectly, but it's not going to impact the deadline.
Thursday, February 19, 2009
It occured to me while sitting there waiting for the poles to finish steaming that there is another argument in favor of steam bending the poles.
In period illustrations of both wheelbarrows and plows you can see arching handles or shafts. Some of these are apparently cut to curves from wider boards, but some appear either steam bent or cut from tree stock of the correct curve. Thus, we can reasonably infer that curved and compound curved pieces of farm equipment would be recognizable to our period craftsman, and considered as a solution to minimizing the weight of the cart without enlarging the wheels.
Here's a few illuminations to illustrate my point.
Tuesday, February 17, 2009
Man, I'm tickled pink! I got so pleased with the way the poles are turning out I had to drag Gen down to the McKennawerks in her jammies just so I could share with somebody. I thought the shop was still too cold for steaming, but I went ahead and built, scratch that, cobbled together, the bending jig. As you can see from the below pictures, the first pole I cut is nearly perfect just cold clamped in the jig, and the second pole is very nearly there, too.
Monday, February 16, 2009
As previously noted we're faced with drawing conclusions from period iconography of horse carts to design our dog-cart. The Lutrell Psalter has several representations of wagons, but all of them have straight shafts. Should we not use straight shafts as well?
There are what I believe are several logical reasons not to.
First, let's be honest, the curved shafts are more aesthetically pleasing, and if I'm going to all this work I want something that looks cool. What can I say.
Second, in line with the conceit that this cart would have been built at an on-site workshop of a monastery or latifundia it's not unreasonable to assume bent poles. If we are to keep the bed level, and not spill our goods on the ground we cannot use straight shafts. They would require larger wheels, with a consequent increase in weight. Remember, we're not looking at a cart to be pulled by a heavy draft horse or ox, but by a dog. The load is limited. Every ounce more in cart weight is an ounce less of potential useful load. The cart would be made as light as possible to make the load as great as possible.
Third, the craftsmen of such an estate would have had ready access to coppice. It took me only a few minutes walking around to find two suitable poles. For my full-size Djebjerg Wagon project I have several pieces of very complex curved natural growth seasoning. It took me about an hour wandering through the woods to find the two trees I used. If it's this easy in 21st century mid-America to find suitably curved standing wood, how much easier must it have been in a more heavily forested and coppiced medieval Europe. Additionally, our hypothetical craftsman would not have been under a deadline--he could have bent trees and allowed them to grow to shape as was done for boat building.
So I think it's not inconceivable our craftsman would have elected to use smaller, sturdier, lighter wheels, and use poles that curved to a useful height and distance.
Unfortunately, finding perfect poles is not as easy as finding fairly good ones. To get what we need is going to require more than careful shaping with the drawknife. Thus we'll need to steam bend the poles to get the desired compound curves. We know that steam bending was done in period by furniture makers and boat builders so the technique is appropriate. And I don't think it's stretching our conjecture too far to posit that our craftsman would have used it rather than spend large amounts of time searching for just the two perfect pieces of wood.
Today I cut the second pole and shaped it as with the first one. Since it didn't match exactly with the old one we'll correct the differences with steam.
Right now the two poles are clamped together hand bent to roughly the intended shape. This is as much as could be achieved with hand pressure, clamps, and carving. The back end is clamped flat to the table and some curvature induced with blocks and weights. The tips of the poles are about eight inches higher than the back ends. We can fiddle with the axletree some, but this puts our axle about 11" - 12" high and the poles about 19" high, a good hieght for a drafting dog.
Sunday, February 15, 2009
I started off by removing the bark which peeled off easily in long strips (a clue to its being ash). Then I set it in the B&D workmate. Most of the time the workmate is a catchall for odds and ends, but for working long stock, especially irregular stock, it's the bee's knees (Do bees even have knees? If so, why are they considered so exemplary?).
I then used the drawknife to flatten one side of the pole.
Here you can see the line at the wide end of the pole. Metal calipers are massively useful for shaving wood to the right dimension. I set the jaws to about 1/2 the desired width at the base, and marked down the pole to either side of the centerline. Some places I needed to do a lot of shaving, others not so much.
Once one side was as close to what I wanted as drawing and planing would get it (allowing for the anticipated shrinkage, of course) I ran the pole through the jointer/planer to make the sides parallel. I then ran one of the perpendicular sides across the jointer to get a reference perpendicular edge. The pole is intended to be about 1 1/8" thick and 2.5" wide for about the first 3 feet where the bed will be, then taper to about 7/8" or 3/4" x 1" at the other end.
At this point I have one pole rough cut and ready for steam bending. Here you can see the natural bend at the front end of the pole compared to a straight edge. I'm going to exaggerate this bend when I steam it.
Tomorrow I'll get the other pole set up and make the jig for clamping the poles to shape after bending, and put together a steaming box.
This is a very rough idea of the final shape I'm shooting for.
In the cold, dismal morning the new owners of my Frisian/Percheron gelding, Oberon, came out to the stables to pick him up. I've had to face the fact that I've buggered up my back so badly over the years that horseback riding has had to move into the category of 'used-to-do.' Though he's had draft training, I haven't: and it's a big and time consuming skill set to develop. Even if I had someone to teach me around here, I probably couldn't manage it, and trying to teach yourself something involving 1800 easily startled and fast pounds of "I'd really rather be back in the barn eating hay, thanks just the same" is definitely not smart.
So though I haven't been able to do much with him in the past few months; and logically, paying feed, vet, farrier, and stable fees for a pasture ornament--no matter how pretty--just isn't in the cards, it's still a wrench to say goodbye. He's going to a good home where he'll be spoiled rotten and played with regularly, which is important to me, but still...
When I got back home I figured I'd get to work on the spokes for the dog-cart wheels... nothing like makin' sawdust to cheer a fellow up. I decided the wheel diameter should be about two feet. Why? Jest seemed right--it's one of those aesthetic things. Thus the spokes ought to be 9" between hub and felloes, which gives a wheel diameter of 9"+ 9" + 4" + 1.5" + 1.5" = 26" ( spoke length + spoke length + hub diameter + felloe thickness + felloe thickness ). Allowing for shrinkage, and tenons into the hub and wheel, meant I needed to cut and split enough oak billets to get 16 spokes (12 spokes plus 4 spares for when I screw up the tenons, count on it folks, screwing up at least one tenon is a given) each blank about 13" long. The length of oak I had that came down in the storm should have given me enough billets, to quarter by splitting, 20 blanks, though I figured that I'd probably only get two quarter sawn blanks from some of them due to grain run out, knots, twist, etc..
No such luck. After cutting the trunk off the stump in the cold and drizzle (the whole day it only rained while I was outside cutting, sheesh) I hauled the log down to the shop and cut 5 billets. Then I grabbed a wedge and sledge hammer and went to work splitting. I was unhappily surprised that the healthy appearing oak (it only came down because a much larger white pine further up the hillside fell over directly on it) was infested through the center with wood boring beetles.
I now have some split oak for the firewood pile and a growing concern about some big oaks near the house. Said wood pile btw being freshly delivered after the last storm when we were beginning to run low, is made up of some nice oak that I spent an afternoon cutting into smaller lengths to fit in our small woodstove, all of which are therefore now too short to use to make spokes.
However, I'm led to rethink splitting the spokes from billets. In discussion with a wheelwright many years back (eight and eleven twelfths, to be precise--it was on my honeymoon) who makes reproduction wagon wheels the old fashioned way as a docent in a living history museum, he mentioned that sometimes spokes were shaved down from saplings, rather than being carved from billets, in wheels intended for light use vehicles, like fer instance, dog carts.
Don't get excited--the dog carts he meant were very light, 19th century, horse drawn carts with a high seat over boxes/cages used to haul hunting dogs to the fox hunts. The thing is, I have a surplus of small oaks (less than 4" diameter). It might prove more economical of time and resources to use some of these saplings, which would also be good forestry as many of them are growing too close together and to more mature trees.
It would also make logical the construction of a springpole lathe. A drawback is, since I never bought a lathe also I never bought gouges and chisels appropriate for wood turning. IIRC Roy Underhill (a personal hero) did a show where he made a springpole lathe and the first thing he did was turn handles for gouges and make more gouges.
So it's entirely possible I may decide tonight to cut a few saplings and shave/lathe the spokes. The more I think about it the better I like the idea. For one thing, it would let me use lengths of the trees I'm going to have to cut anyway to make the poles.
I've also started working on the molds to cast pewter buckles and fittings for the harness. A dog-cart pulling harness looks just like a bridle for a horse, so I can use examples from a book on period horse tack fittings found in London--and hey, bridles I know!
So now it's off to help my Lady Wife tape sheetrock for a while...
Friday, February 13, 2009
I used a roto zip with the routing attachment with a 3/8 straight cutting router bit to shave the rough billet into a cylinder 12 1/2" long and XX inch in diameter.
I then marked up the cylinder with reference marks to make two hubs. The hubs will be separated after the coves are cut. I'll do this on the miter saw so that I get nice square cuts.
In order to turn the coves I used a trick usually done on the table saw, but here carefully done with a circular saw. The saw is set so that the blade is slowly lowered at a slight angle to the centerline of the cylinder into the workpiece which is turned by hand and the angled blade carves out an ellipsoidal cove.
Here you can see two guide blocks screwed to the top of the jig to accurately position the saw.
After cutting the hub blanks the ends are sealed with paraffin to prevent them from drying out too fast and checking (cracking). I also recommend storing them in the fridge until you're done working them to slow the drying process.
After that it's off to the drum sander to clean them up a little. Tomorrow we'll dress up mortises and cut the spokes from more of the red oak. I'll also take some more detailed pics of the jigs.