Tuesday, February 28, 2006

gentlemen, start your chainsaws

The Single Block Classic is now underway! Lars was disappointed not to hear a starting gun and a "symphony of chainsaws" at 9:00 AM. I guess this sort of thing starts slowly. A little over an hour into the event, we are still waiting for the zoom boom to get around to our site (#17) and re-orient our ice block to our specifications. We placed some timbers on the ground last night and leveled them this morning so that the block will be mostly level but leaning just slightly away from the side we will cut. We're hoping this will minimize the chance of having 450 pound slabs fall on us!

I'm typing this on a PC in the Ice Art warming hut. It is a good way to exercise my fingers out of the -13°F outside air. It's a shame I don't know how to type with my toes.

Sunday, February 26, 2006

on my way

After packing an unbelievable array of footwear, I'm leaving for the airport and will arrive in Fairbanks at 12:50 AM. I had some time to work on a set of custom spade bits today, but I'll have to finish the job at Belfair.

Saturday, February 25, 2006


After spending about 18 of the last 22 hours working on design finalization and template layout, the templates are finally in the hands of the Kinko's in Fairbanks. We'll be able to pick them up on Monday. The reason it took so long was that I had to do some significant last minute design changes. One of the main purposes of the plywood escapement was to experimentally determine the amount of time the escape wheel has to turn from one tooth to the next. I had estimated half a second, but it turned out to be about a third of a second. This small difference caused the anticipated required sled weight to increase from 103 pounds to 220 pounds!

Fortunately, I was able to drastically reduce the weight to 83 pounds by scaling down the escapement and the last couple gears in the gear train. Unfortunately, it was the kind of change that affected many other things. Quite a few parts had to be re-drawn, some extra parts for the gear tooth cutter had to be worked out, and I had to start over my template layouts from scratch.

Meanwhile, Lars has finished his home-made lathe and is proceeding with his bandsaw for slab cutting. We had been counting on being able to rent a 48" bar chainsaw in case the bandsaw doesn't work, but that plan has fallen through. Now all our eggs are in one basket, and our second best option is a smaller bar chainsaw that would be rather suboptimal. Go, Lars!

Friday, February 24, 2006


I got to bed early tonight after being short on sleep for a couple nights in a row, but I ended up waking up around 4:00 AM anyway. There is so much to think about!

Lars finished his mobile drill press. I can't wait to see it. He ran into some obstacles with his lathe, but it sounds like he is getting past those. Lars is also working on a custom band saw for cutting our big slabs.

I'm trying to finalize the design. There aren't many major changes to make, but there are a whole bunch of little ones. I need to get them all done so that I can finish the template layouts and get them to the printer.

Thursday, February 23, 2006

extended forecast

It is starting to look like the weather will be favorable with highs around 15° to 20°F. This is slightly cooler than ideal ice carving conditions, but it certainly could be a lot worse. Too warm, and the ice melts; too cold, and it gets very brittle.

Steve Munk, an old friend and roommate, writes that "the roads have been nightmarishly icy all week. Like skating rinks all over town." I'll have to bring my skates.

In our correspondence this week, Steve helped me get over my sticker shock at the cost of printing a couple dozen 5' by 3' paper templates at a Fairbanks print shop. It turns out that the cost isn't really that much more than they pay at ARSC. Steve also writes, "I have a suggestion. Make the bushings out of ice. It's slippery. ;)"


escapement testing

I finished building a pendulum and escapement out of plywood this evening. I haven't added any way to power the escape wheel, but just turning it by hand is enough to tell that the design works. I was afraid that the design required too much precision when cutting out the parts, but the thing worked on the first try with parts cut by hand with a jigsaw.

Wednesday, February 22, 2006

how to build a gear tooth cutter, part one

Okay, Lars, this rather lengthy post is for you. I've sent you a number of parts for the gear tooth cutter in the mail, but these first steps you can do before the package arrives if you have time. My prototype was constructed primarily out of wood. You may decide it would be better to use steel or have other ideas for improvements, but I'll assume for the sake of expediency that you'll build it just like mine.

Precision during construction significantly affects the machine's ability to cut precise gear teeth. You should make sure your table is level before starting so that you can level various components throughout the assembly.

First, build the frame out of two-by-two lumber. The frame should be assembled exactly as in the photo (except for the diagonals which were not correct when this photo was taken) with 48" and 21" horizontals, and 27" verticals. I used 24" verticals, but you should build yours taller in order to accommodate our increased gear thickness.

Diagonals should be added to keep the structure rigid, but the position of the side diagonals must be correct. They can terminate at the front no more than eight inches from the top of the frame. The side diagonals should be attached to the verticals so that there is a gap between the diagonals and the side horizontals.

Next, build the spindle bar rails. The spindles will be mounted in top and bottom spindle bars (two-by-twos) that will run lengthwise through the frame. The spindle bar rails will guide the spindle bars, allowing them to move freely lengthwise. The spindle bar rails will also be adjustable front-to-back, allowing various gear blanks to be positioned nearer or further from the cutting blade.

The left spindle bar rail is made from two 27" two-by-twos separated by short two-by-two spacers at the top and bottom. Unlike mine, your spacers should be no more than two inches long.

Place the rail along the left side of the frame so that it is held against the frame by the left diagonal.

Then carefully attach a short two-by-four (mine was 9.5" long, but I recommend about 14") such that the rail is held perfectly upright as it is adjusted from front to back. There should be a small gap between the rail and the table so that the weight is held by the two-by-four. Drill a pilot hole near the bottom of the rail, but do not drill it through the frame horizontal. This is for a set screw (see two photos up) that will be used to hold the rail in various positions.

Make the right spindle bar rail the same way as the left one, except that the two-by-four must extend 8.5" from the front edge of the rail. You might want to make your two-by-four few inches longer than mine toward the rear for improved vertical stability of the rail.

The reason the right spindle bar rail has a longer two-by-four is to support the guide rack table. This is a 21.5" long, 7.5" wide board supported by two-by-fours at both sides. Use a short length of two-by-two as a foot at the rear of the left two-by-four. Note that you cannot add a foot at the front because the table must be able to extend over the front frame horizontal when the spindle bar rails are positioned forward for cutting smaller gears.

The left two-by-four needs to extend far enough to the rear so that the bottom spindle bar can ride on top of it. Hopefully you won't have to, but I had to chisel the top of mine down a bit due to a bowed bottom spindle bar. With the spindle bar (or any two-by-two since you haven't made the real spindle bar yet) in place, add little front and rear guide blocks to the top of the two-by-four. These will keep the guide rack rigid with respect to the spindle bar path. The rear guide block needs to be pretty short.

Attach a 12" two-by-two to the top of the guide rack table along the front so that the edge facing the spindle bar is parallel to the spindle bar and exactly 22.8 cm from the center of the spindle bar.

Make the bottom spindle bar from a single 61" two-by-two. Drill 1/4" holes into the top of the bar about 1" deep (not all the way through) for the blank spindle and guide spindle. The holes should be made along the center line of the bar and should be perfectly vertical. The blank spindle hole should be 26" from the left end of the bar, and the guide spindle hole should be 13" to the right of the blank spindle hole.

Add the tension bar perpendicular to the bottom spindle bar. It should be 18" long, attached about 1/2" to the right of the center point between the two spindles.

The tension bar needs a little cut-out at the spindle bar end in order to move freely above the little guide block on the rack table support. You should also add a foot to the tension bar, precisely sized to level the spindles. The foot can be no more than 3.5" away from the spindle bar or else it will run into the rear frame horizontal when the spindle bar rails are positioned toward the rear for cutting larger gears.

Make a little tension slider that slides freely along the tension bar. Drill a 1/4" hole about 1" deep into the slider for the tension spindle.

Mount the slider on the tension bar and add rubber bands that pull the slider toward the end of the bar.

The top spindle bar is identical to the bottom spindle bar except that the spindle holes go all the way through, and there isn't any of that tension bar nonsense. You'll need to sand down the ends of both spindle bars in order to allow smooth motion through the spindle bar rails.

Tuesday, February 21, 2006

one week to go

Next Tuesday morning the Single Block Classic begins. Lars and I are making preparations at a feverish pace. I took apart my gear tooth cutter and sent most of the parts other than lumber to Lars this morning. Last night I made a full-scale, plywood escape wheel, but I won't have a chance to make the pendulum and test the complete escapement until tomorrow night or later. Some last minute design modifications are required, including thickening a couple gears and several minor frame changes.

Monday, February 20, 2006

broken teeth

I built a load tester to try to find out how much force the gear teeth could take before shearing off. Unfortunately my results were somewhat discouraging and not entirely conclusive. The first tooth supported 65 pounds (on a 60° ramp, the same angle as the clock's ramp) but failed with 220 pounds. The second tooth failed with 150 pounds. The third tooth supported 150 pounds but failed with 220 pounds.

In ideal circumstances, such as a complete lack of friction throughout the gear train, I estimate that the sled will have to weigh 103 pounds. We may be able to do some things to reduce the required weight a bit, but I'm not certain how much. To be safe, I would prefer to be able to count on the teeth supporting a sled of 150 pounds.

The test was non-ideal for several reasons. The gear was only 9 cm thick whereas the gears experiencing the most force in the clock are designed to be 10 cm thick. The gear was not cut very well, mostly because my lack of a drill press prevented me from using a very precise method of aligning the gear blank on the gear tooth cutter when I flipped it over to the second side. The quality of the ice was marginal. It had quite a few air bubbles, unlike the arctic diamond we will be working with at the competition. The load tester was hastily designed and constructed, so the gear wasn't seated perfectly against the racks, and I believe the load was uneven across the length of each tooth.

Even taking into account the imperfections of the test, the results were a bit troubling, so Lars and I have decided to increase the thickness of the first two gears in the gear train from 10 cm to 15 cm. The test also points out that we absolutely must make our best effort to cut the gear teeth precisely and align all of the axles correctly.

Sunday, February 19, 2006

how to make an ice gear

This is the process for making a gear out of ice using the gear tooth cutter. Start with a slab of ice. This one was about 10 cm thick.

Use liquid water to paste a paper template to the slab. I thought that a paintbrush would be good for this, but it turned out that just dribbling water on the paper was easier. A spray bottle might work well too.

Following the template, cut out the gear blank. I used a hand pruning saw which is good at cutting ice but not particularly precise. Ideally, the cut would be perfectly perpendicular to the surface of the slab and exactly following the outer circle on the template. Use a 1/4 inch spade bit to drill the center hole and two holes for the sprocket assembly.

Mount the appropriate sprocket and guide pinion on the guide spindle. Clamp the corresponding guide rack to the rack table. For this gear, I used a 24 tooth guide sprocket and a 20 cm guide pinion.

Mount the appropriate sprocket on the blank spindle. For this gear, I used a 20 tooth blank sprocket.

Any sprocket will do on the tension spindle. Make sure that the chain is properly engaged with all three sprockets and that there is enough tension to keep it from sagging.

Mount the gear blank on the blank spindle. It should be seated on the sprocket assembly so that they turn together. Place the top spindle bar through the bar rails and onto the spindles. Now when you manually move the spindle bars from side to side, the guide rack and pinion cause the sprockets on all three spindles, as well as the gear blank, to rotate. At this point, you will probably need to adjust the guide rack position to the left or right in order to allow your first cut to follow the template. Your first cut should be with the gear spindle to the left of the saw blade.

With your left hand, provide a little clockwise tension on the gear blank. Make your first cut with the circular saw by pulling it down through the gear blank and then back up until the saw blade is clear of the blank. Move the rack to the right (you'll soon get the hang of how far), provide clockwise tension, and make your next cut. This process allows successive cuts to follow the involute tooth profile.

After finishing each tooth, manually advance the guide pinion with respect to the guide rack to set up the next tooth for cutting. For this gear, I advanced two guide pinion teeth for every one tooth on the gear blank.

After making your way completely around the gear blank, you are halfway done. Now you must flip the blank over and repeat the process on the other side to cut the other edge of each tooth.

After finishing all of the teeth with the gear tooth cutter, remove the blank from the cutter and use an iron to remove the template.

The area halfway between each tooth will be left a bit rough by the gear tooth cutter. Use a small drum sander or a round file to finish those spots.

Friday, February 17, 2006

on the phone

I spoke with Lars today for the first time in six weeks. He has already raided Fairbanks dumpsters, and he answered the phone while salvaging a motor from a discarded washing machine. It will probably end up driving his lathe. He's also building a steel work table, a slab cutting rig, a copy of my gear tooth cutter, and his mobile drill press. It would take me months to do all that stuff, but, knowing Lars, he'll probably be looking for new projects by Wednesday.

I finally found a place that sells an adjustable spade bit that goes up to a 5 inch diameter, but at $169 it probably isn't any more cost effective than a handful of large forstner bits. Lars is going to see if he can improve on my design (I'm sure he can) and make a few custom spade bits instead.

the nerve of some wives

Not only am I not allowed to operate my circular saw in the middle of the night, but now I can't leave the garage doors wide open in sub-zero weather to create realistic practice conditions, inadvertently allowing pipes to freeze in the pump room! Can you imagine?

Thursday, February 16, 2006

custom spade bit

I made a proof-of-concept adjustable spade bit tonight. It worked pretty well on ice, although the poor design resulted in the cutting blade rotating around the machine screw. This caused the bit to bore a conical hole rather than a cylindrical hole. Even so, it was an effective proof of concept.

I also tried my jigsaw on ice for the first time. It was terrible. In general, it seems that power tools capable of automatically clearing snow from their cutting blades (chainsaw, circular saw, spade bit) rock while tools that harbor packed snow (twist bit, jigsaw, hole saw) suck.

he's back

Lars is home from Russia and "looking forward to building MACHINES!!! Like Morpheus would say it." Now the fun begins.

Tuesday, February 14, 2006

two weeks to go

The gear tooth cutter is working well, although so far I've only had a chance to use the full process on plywood blanks which splinter quite a bit. It doesn't look like I'll have enough time to make a complete working model of the ice clock, but I will at least make a model of the escapement.

A matter of much debate at the moment is the method we will use for putting the axle assemblies together. The two main techniques under consideration are the sandwich method, two separate axle pieces bonded to either side of a solid gear, and the through-axle method, a single axle piece inserted through a hole in the middle of a gear. Until this week, I had been leaning toward the sandwich method because I thought it would be much easier to get a good waterglue (Lars's term) bond between flat surfaces than between cylindrical surfaces. The biggest drawback of the sandwich method is probably the difficulty of getting the parts aligned correctly.

Now I'm leaning toward the through-axle method, mostly because of Lars's plans to build a mobile drill press and an ice lathe. Those tools should make it much easier to get good cylindrical waterglue bonds, eliminating the only major drawback of the through-axle method.

about Lars

It's hard to imagine a better partner in this crazy endeavor than Lars Hansen. I met Lars at the very beginning of our first year at UAF, and we became fast friends. I like to think that I'm a bit like Lars, but Lars is, uh, more so. Allow me to explain.

Lars is always making something. Now a brewer of excellent beer, he made wines in his dorm room at Nerland Hall. A few of his wines were pretty good, including a strawberry wine and, surprisingly, a batch made from a bag of frozen, mixed berries. Most of his wines were terrible, however, because he was forever experimenting with strange and new concoctions. Two of his more dismal failures were banana wine and rhubarb wine. Always one to make the best of any situation, he would pour the undrinkable batches into a stainless steel bowl and place the bowl outside to freeze on cold, Fairbanks nights. After skimming off the ice a few times, the remaining everclear was bottled for future use as a mixer or as a combustible for one of his famous party tricks.

Over the last few years he has built his own house, called Belfair, near Fairbanks. The four story, plywood palace features a secret door; a room with a dome ceiling; a dual axis, steel framed, spiral staircase; a room ("the West Wing Theater") suspended over open air between the house proper and a nearby hillside; and a fifty foot long, two story wood shed. During early phases of construction, there was a half-pipe in the living room.

Lars is strong. He used to walk around the dorm on his hands, even up and down stairs. When his high school (junior high?) gym teacher noticed his muscles, the teacher asked, "Lars, do you lift weights?"

"No, I lift rocks!" he replied enthusiastically.

Lars is enthusiastic about, well, absolutely everything. While visiting Belfair last year, everyone in the house settled down for an afternoon nap one day. After we all woke up, Lars said, "Isn't it great to be awake again?" He meant it. Late at night at Nerland Hall he would knock on my door and say, "Let's go slogging through the snow!" You simply can't turn down an invitation like that, so I'd put on my slogging boots and my warmest clothes, and out we'd go out into the sub-zero midnight just to find the nearest, untouched field of three foot deep snow and trudge through it as hard as we could. You can count on getting plenty of exercise when you hang out with Lars.

His tolerance for cold weather is unparalleled. People used to stare in amazement as he ran from class to class in -20°F weather wearing shorts, birkenstocks, socks, a jean jacket, and a sheepskin hat. Partly he's just tough-skinned, but I think the main reason he was capable of running around like that was his amazing metabolism. He eats more than anyone I've ever known yet is skinny as can be. When friends didn't clean their plates at Lola's, Lars would scrape the remnants off every plate onto a big pile on his own and chow down.

Terrifyingly thrifty, Lars never lets anything go to waste. Generally this is quite admirable, but I have witnessed the dark side of the tendency from time to time. Once Lars brought a home-made rhubarb pie to a party at school. It was grey. I asked, "Lars, was that rhubarb, uh, used?"

"Yes!" he proudly replied, "it was left over after making the rhubarb wine!" I'm sure I was the only person to eat a whole piece of that pie (except perhaps Lars, himself, who is capable of superhuman feats of ingestion). Never have I suffered more just to be polite.

Every day spent with Lars is another story. Sometimes the story is bizarre, sometimes it is touching, and sometimes it is downright amazing; but it is always a story worth telling. I want the ice clock to work, and I can't imagine anyone more ingenious and fun to work with while "freezing and suffering" (his words) through the competition. Mostly, though, I want this experience to be a good story, whatever the outcome. With Lars along for the ride, it can't help but be a story worth telling.

Thursday, February 09, 2006


"When I was carvnig the sea turtles out of the river ice in тйгйль, I only had a leatherman to work with. . ."

Ah, it's good to hear from Lars again. He's back online, though infequently and with a slow connection. Still, it's good enough to exchange a few emails for the first time in more than a month. He has some great ideas for tools and techniques. One of my favorites is a plan to mount his drill press on a carriage assembly so that we can drill precise holes in the middle of large parts.

functionally complete

The gear tooth cutter is now functionally complete and ready to cut its first gear. The starter gear, although imperfect, moves pretty smoothly as part of the rack and pinion mechanism. I have a gear blank mounted and ready to go, but it is getting too late to be operating the circular saw. Impatience!

Wednesday, February 08, 2006

three weeks to go

I've been under the weather for the last three days and haven't made much progress. I managed to finish my starter gear tonight. It isn't perfect, but hopefully it will be good enough. Sandra Skibsted is traveling from London to film a documentary of this year's Ice Art. I've corresponded with her a bit, and it looks like we'll get to be part of the project. I haven't heard from Sharon this week, but Lars should be back in another week or so. The weather is warming up in Fairbanks (highs above 0°F!), and the days are getting longer by several minutes each day.

Sunday, February 05, 2006


The prototype gear tooth cutter is less precise than I would like (which is one of the reasons I'm starting to call it a prototype). I've tried to do the best job I can making the parts, using threads to center the sprocket assemblies, for example. Unfortunately I don't own a drill press, and that has been the biggest factor limiting my precision. It sure is hard to make holes through metal in exactly the right spot with a hand drill. I thought about buying a press, but so far everything is good enough for a prototype.

The other reason I'm calling this a prototype is that it will probably be quite expensive to ship the whole contraption to Alaska. It looks like it may cost $25 each way to ship the circular saw alone (a $38 item that is most likely available for sale in Fairbanks). Instead of shipping it, I think I'll ask Lars to build a second gear tooth cutter after his return. I'll refine the prototype as much as possible, and then he can make the parts more precisely in his shop. (He owns a drill press and is much more experienced with metal in general.) I'll just bring a few key components such as the sprockets and starter gear.

starter gear

Well, I didn't meet my goal of having my first ice gear made by the end of the weekend, but I came close. I am more than halfway through cutting my first wooden gear, but it has been more challenging than I thought. The trouble is that my gear tooth cutter requires a gear as part of the mechanism to move the gear blank you are cutting, so I have a chicken/egg problem. I had supposed that I could carefully use a plain wooden disc instead of a gear, but I gave up on that idea after a couple failed attempts.

I ended up printing a paper template, pasting it to the wooden blank, and manually positioning each of several cuts per tooth. This method is slow but fairly precise, and I am able to duplicate each cut across every tooth by manually advancing the chain along the sprockets. One edge of each tooth has been finished, and I have started on the other side. I think this will result in a usable starter gear, but, if all else fails, I can always have someone make one for me.

Thursday, February 02, 2006


Word from Sharon is that Lars, still waiting for cargo, has been practicing cutting ice blocks out of a nearby river. He's still saying he'll be back in mid-February. I haven't practiced with ice for several days as I have been spending all available time on the gear tooth cutter. My goal is to have made my first gear out of ice by the end of the weekend.