First Set of Napkins Hemmed

Well, they didn't hem themselves, either, but the first set is all ready to be sent to the recipient.

On the theory that if you're going to go to the trouble of hand-weaving napkins you shouldn't do machine-stitched hems, these have been hand-stitched. Luckily, in a former life I sewed a lot of my own clothes, so a hand blind stitch goes pretty fast. Not as fast as the machine, but then my machine can't do a true blind hem stitch.

Here are some images of the first set. The first shot was scanned at high resolution and then cropped way down, so you can see the degree of deflection of threads that occurs when a densely sett 3-color echo weave is wet-finished and pressed. The three warps, although not completely independent layers, still push and shove during finishing and sometimes the layers that want to be in front really bend to get there. Look at the light threads close to the base of the feather, and higher near the tip, and you can see the light warp actually crossing over the peach warp in places. (Click the image to open a larger version first.)



The next shot shows the four napkins, from left: grey weft, gold weft, rust weft, and brown weft. If you scroll down a couple of posts, you'll find pictures of the coned yarns before weaving.



This warp was 19 inches in the reed, and shrank to 16 after finishing. I thought mercerized cotton would never shrink almost 16%. I had calculated 10%. Hah! Luckily, 16x16 is an accepted size for dinner napkins, so I'm still in the ballpark and don't have to call 'em luncheon napkins!

Problem Warps, and Other Sorrows

In a comment on the previous post, Leigh said "And I thought only novice weavers like myself had problems like this. I'm not sure if it's comforting to know that professional weavers also have these problems, or discouraging that I'll probably never stop experiencing it."

To which I can only reply that the problems never go away, you just get better and quicker at solving them. Murphy's Law guarantees there will always be knots and breaks. You'll get so you don't let them get to you - it's just part of the process, not the end of the universe.

I used to just dread to have broken ends, or knots that needed to be repaired, or a warp that was mis-sleyed and needed to be fixed before proceeding with weaving. Now, it's just another ho-hum.

I keep a stash of film canisters that I weight with washers. I wrap the replacement end around the canister, thread it appropriately through heddle and reed, pin it to the woven web, and then drop the canister off the back beam of the loom such that the snap-on canister lid holds the replacement end at an appropriate altitude. I can see when I've woven enough that the canister has risen to the level of the beam, and take a little break from throwing the shuttle to walk around to the back and let the canister back down to just clear of the floor.

Side note: I took a film canister to my local hardware store and looked for the largest washer that would fit into the canister, with the smallest possible hole in it (looking for maximum mass in the smallest package). The hardware guy asked what I wanted 100 washers for. I tried to explain, but his expression got blanker and blanker. Nowadays, when I go on a shopping expedition like that, when they ask why I want whatever-it-is, I just say "You don't want to know." At that particular hardware store, when the guys see me come in the door they just become invisible. Abracadabra, they're gone.

For most warps, it takes 4 or 5 washers per canister to get the right tension on the replacement end (maybe 6 if it's a selvedge end). If there are lots of knots or breaks, when I run out of canisters and washers, I raid DH's workshop. He's a wood worker and all-around handy repair person, so there are lots of things in the workshop that make good temporary "accessories."

If I'm weaving a run of scarves or shawls or placemats or napkins, at the dividing point between items, I'll bring the replaced end back up to the web and cut out the temporary replacement. I find that if I tie the two together, it's very quick to just pull it through and weight it off the front beam with the same film canister until I've woven an inch or two of the new item.

When I was learning to weave, my teacher told me that her own teacher had a unique method of getting students accustomed to repairs of this kind. She'd carry a pair of tiny scissors in her pocket, and would cruise around the room and arbitrarily snip a warp end while the student was weaving. Sounds a bit cruel, yes, but my teacher says it worked. She got used to making repairs early in her weaving life. However, I have to admit my teacher didn't repeat the process with her students!

Napkin Project Moving Along

On a happier note, the napkins, while not weaving themselves, are at least moving along fairly smoothly. Here's a shot of the current one on the loom:



I'm weaving 3 sets of 4 napkins each. Within each set, there will be 4 different weft colors and 4 different treadlings, a mix-n-not-quite-match sort of combination.

This is the first time I've woven cotton this fine. I've used finer silk for warp, but by comparison the silk is like iron - you can abuse it, treat it roughly, and it just smiles. The cotton is far weaker and I'm spending quite a lot of time repairing broken ends. Here's what the back of the loom looks like with replacement ends hanging off the rear:



Part of the problem at first was that because it's 3-color echo weave, it's a dense warp, which makes it hard to get a clean shed. The metal tips of the fly shuttle were a bit rough, and when the shuttle hit a thread caught in the middle of the shed, it would grab the thread and break it instead of just sliding under or over it. A little sand-paper fixed that.

Then, I had trouble with a tad too much draw-in, so the selvedge ends became worn and kept breaking. On went the temple, and the selvedges haven't given me any more trouble.

Now what's happening is that two of the colors of warp thread tend to just drift apart, not necessarily anywhere near the shuttle race. It's probably not quite as tightly twisted in the ply as the third (darker, mill-dyed) yarn. The weaker thread is the undyed yarn and the mid-tone yarn pictured in an earlier post about this project.

Oh well, I know how to swap out broken ends. And I'm getting lots more practice!

Adventures with the Ahrens Fly Shuttle Throwing Mechanism

Some time back, Bonnie Inouye posted a message on WeaveTech with information about a fly shuttle throwing mechanism that she had seen and photographed at the studio of Jim Ahrens (one of the original developers of the AVL looms). Bonnie indicated in her message that only a small movement of the hand was needed to send the shuttle across the race. You can find her original message in the WeaveTech archives, and a photo in the Files section of the WeaveTech site.

Since my husband is a woodworker and an engineer, I asked him to try to duplicate the Ahrens mechanism for my loom. It didn't quite make it into production by Christmas, but before New Years he installed the gizmo on the loom.

First, for comparison, here's a short video showing the standard AVL side-pull fly shuttle throwing mechanism in action:



You'll note that I use my whole forearm and a fair amount of body english to throw the shuttle, so the stress on the body is not limited to my hand and wrist. I try to use the motion I was taught for fly-casting - keep the wrist fairly straight and use the forearm instead. With some degree of arthritis, I'm really anxious to avoid repetitive small-motion stress. (You'll also notice that there's no warp on the loom; at the time, I'd threaded the heddles but not the reed. However, this doesn't really affect the outcome of the experience, as I'll discuss later.)

It may not be immediately apparent, but the cord attaching the handle to the castle is closer to me by about 4 inches than the standard AVL setup. This is because my monitor is mounted in a cradle under the castle, and we wanted to be sure that the cord cleared the monitor to avoid accidental damage. So Mike bolted a piece of wood immediately under the castle, above the monitor, with an eye-bolt on the end, and the fly shuttle handle cord is attached there instead of directly to the castle. These mods were done shortly after I got the loom in 1997

By the way, I really love having the monitor right in front of my face! I don't have to turn to the side of the loom to see where I am in the draft. The PC itself is mounted on a board on top of the castle. Initially, I worried that the constant jarring motion of the loom would wreak havoc on the disk drive, or destroy the monitor, but I've been using the same monitor since I bought the loom, so my fears were unfounded. The PC has been replaced once in that time, but only to have a slightly faster processor, not because of any disk drive damage. The AVL dobby box is also on top of the castle, having moved up from the original position beneath the castle. I never have needed to see the dobby box display, so the location change was no big deal.

Here are a couple of pictures showing the modifications. One shows the pivoting handle, which is attached to the castle so that it clears the monitor (although it does obscure the display somewhat). The other shows the pulleys that route the cord from the pivoting handle to the picker. Initially, the pulleys were mounted higher, about 6 inches below the rocker that supports the overhead beater arm, but in early testing it felt to me like the direction of pull was too vertical, when at the level of the picker, the travel has to be horizontal. So Mike moved the pulleys down about a foot.





And, here's a short video showing the Ahrens fly shuttle throwing mechanism in action:



Bottom line: It's been removed from the loom, and I've gone back to the standard fly shuttle throwing mechanism. After using it for several hours, I felt that the effort required to throw the shuttle was too great. It wasn't just a little tilt of the hand, it was a very strong, sharp motion. My hands and wrists were definitely sore after that amount of time. It was an effort to get the shuttle to travel all the way into the box. Often, it didn't quite make it all the way, and had to be tapped gently to get it all the way to the full extent of the picker so it would be in position for the next throw.

With the standard setup, I can start the motion gradually and pick up speed before ending the stroke. I can also vary the position of my hand, and it's usually farther away from me (toward the monitor) for the right-to-left throw and closer to me for the left-to-right throw, because this is more comfortable for me. With the Ahrens mechanism, the position is absolutely fixed. It's also higher, since it has to clear the top of the beater. The standard AVL shuttle handle can be lower than the top of the beater in rest position, because I typically move the throwing hand closer at the same time as I move the beating hand closer.

After Mike removed the Ahrens mechanism, and I started to weave with the standard mechanism, I found that I was throwing the shuttle way to hard, so the shuttle rebounded out of the box. That told me that I'd been using a lot more effort with the Ahrens mechanism and that body memory of the amount of effort required carried back over to the AVL mechanism. It took a while to re-learn to just use enough force to get the shuttle across the race and into the box and no more. Note that this was with the same loaded shuttle and the same warp.

I don't want to imply that the Ahrens design doesn't work, only that our implementation didn't, at least, not for my hand and wrist. We tried to make the parts as close in dimension as we could to the photos we had of the Ahrens design. When I first tried it, without a warp on the loom and with no filled pirn in the shuttle, it seemed like it would be a workable amount of effort. However, with a loaded shuttle and the added resistance of warp friction, it just didn't feel good to me. Part of that is because I've been using the standard AVL fly shuttle since forever (almost 10 years now) so that any change means retraining both my body and my brain, and part is that my hands aren't particularly strong.

I believe that AVL is working on their own version of the Ahrens throwing mechanism, and it may well be that their clever engineers can fine-tune the design so that it really is "a small tilt of the hand" motion. I look forward to hearing from Bob Kruger that they've come up with a smoother operating version!