Warnings: This project involves cut glass, very hot water and tiny glass pieces. I don't tend to use gloves, and I'm not sure if gloves might cause more problems than they solve because of loss of traction/grip. I don't tend to cut myself when I do this, but that doesn't mean it's not possible to do so. You will also be grinding glass. If you are cutting bottles infrequently, using the manual grinding methods described here, you will be inhaling little to no glass dust. If you do this a lot, if you use a dremel or other high speed power tools for grinding, you will kick up much more dust, and you will need to use a suitable dust mask. Finally, I have never had a bottle fail in such a way that glass flew any distance. However, the little flecks of glass that fly off during the scoring process can go a couple of feet, so you will want to use safety glasses or at least sunglasses to protect your eyes. I don't even want to think about getting one of those little things out of my eye.
Much later: Hey! I was a finalist in the Green Design contest! I had actually written this as an addendum to the lantern instructable mentioned above, and didn't really see it as prize material. Thank you to everyone who voted for me and to all the contest judges! I'm super psyched!
Step 1: The Right Tool for the Right Job?
I just did an Autodesk 123d mockup (this is my first time using the tool, and I literally just did this moments ago, so be gentle - it is a pretty damn cool tool though) of my jig. To hold the bottle cutter in place, I cut notches into the vertical board as I need them. This set-up has allowed me a fair amount of versatility over the size of the bottle I cut and where I make the cut. I would not call it a precision instrument. The scoring tool is a basic glass cutter from a hardware store. They run about $3.
(However, the lumber for the jig was entirely free: Big box stores (like Lowes and Home Depot) cut wood to their customers' specifications. The customers then take what they want and the big box store worker throws the "waste" into a scrap pile or bucket. I've found that if you ask nicely, and don't take more than a couple of boards, they will give these to you for free. Don't be greedy! Don't make them feel like they're giving you stuff they could otherwise sell to you! It probably helps if you approach the worker with a cart containing other stuff you are (or appear to be) buying.)
Edit: A quick note on the jig design. A reader below suggested a wheeled design, which I like, but in the end didn't go with. My reasons for this are flexibility, simplicity and security. The wedge shaped slot here can accommodate a wide variety of bottle/jar diameters without any moving parts. With the base peg in place, the bottle is supported/secured in virtually all directions except up. It is far from a perfect design, but it is easy to build and fairly efficient. However, for those who will be cutting only beer and wine bottles, the wheeled version of jig may be more efficient than the one I show here. Here are a couple of alternative designs:
Mine could be called a variation on the last of these. I wanted it to be able to cut at many different heights, hence the wedge with an unfixed cutter/scorer. I cut slots in the vertical board to hold the cutter at the place I want to make the cut. Step 3 will show this in greater detail. I'm also including a picture of a small San Pellegrino bottle and a large mayonnaise jar I recently cut. The former has a 2 inch diameter, the latter, a 4 inch diameter. A wine bottle falls in between these two, at about 3 inches diameter. This flexibility is one of the advantages of this style of jig.
Step 2: The Bottle!
Step 3: The Score!
A note on cutting: I've gone from heavy scoring of the bottle to a much lighter touch. I do try to make sure that the score line is uninterrupted around the entire perimeter of the bottle. It's also important to make sure the score line meets itself as closely as possible. It can be tough, because the cutter's wheel does tend to waver a bit as you score the bottle. This is the reason I started drawing the Sharpy line: if I follow that, I don't generally go too far off track.
Step 4: Like Water for the Elephant ... in the Room
Nearly every instructable/online resource describing this method says to alternate between "almost boiling water" and "cold tap water." I realized just recently that I have no idea what that statement really means. Does almost boiling mean 210˚ F? 180˚? What temperature is "cold tap water?" I wanted to codify the obvious thing that wasn't being addressed: what's cold; what's hot?
Here's what I found out: if I run the tap for a few minutes, the water temperature steadily drops to about 70˚F. There is some variation, but it doesn't get much colder than this. Early on, I tried using ice water, for the cold water part of the cut, and it was a disaster.
The same has been true (though to a lesser extent) when I use boiling water, or water just off the boil. I've found that water from 175˚F - 185˚F works best ... however ... only in conjunction with straight cold tap water. It's easy to forget that you aren't cutting with hot water or cold water, but rather causing thermal stress by using both. Not enough thermal stress, and no cut will propagate along the score line. Too much thermal stress, and the crack will ignore the score line and wander freely.
This means that the difference between those temperatures - 105˚F-115˚F - is optimal. If my tap water ran at 60˚F, then I would want to bring the temperature of the hot water down to 165˚F-175˚F.
There are other factors at play here as well. The type of glass, how it is annealed, its thickness, etc, all affect how it responds to different temperatures. You can probably rule out this method in trying to cut pyrex, whether it's lab-grade borosilcate or consumer-level annealed. My suggestion is to start out with jars or bottles that you don't care too much about, find the temperature of your cold tap water, and add 110˚F for the hot water temperature. If your cold tap water is significantly below or above 70˚F, you will probably need to do a little experimentation to find the best temperature interval.
Step 5: The Kindest Cut
However, you will probably have to make repeat applications of the hot and cold water to propagate the cut. I generally start with the hot first, then the cold, hot, cold etc. until the bottle splits. Lately I've been placing a silicone hot pad underneath to stop the cut off piece of glass from getting damaged when it falls away from the part I'm holding. Alternatively, you can do this in a PVC utility sink. You could use cloth or a piece of polyurethane foam, but I wouldn't recommend it. It's hard to see here, but little tiny shards of glass pop off the score line, and I imagine that eventually your cloth or foam would become quite unpleasant to touch.
The last picture here is of the cut bottle. As you can see, it's a pretty clean cut. However, it's still gonna need some work before you would want to touch the cut edge with an unprotected finger - let alone drink out of the bottom part.
Step 6: The Inside Story
Step 7: Grindin'!
I use plain 220 grit sandpaper here. It's super-cheap and you can get it at the dollar store, Ace Hardware, True Value, etc. Unless you're taking off a long jagged bit from a bad cut, I wouldn't use anything below 120 grit, as little pieces of glass are going to be forever flaking off from the edge and making your life miserable. Anything over 320 grit is going to take forever to do the bulk of the grinding, but you might want to finish up with something in the 400-600 range to make the lip as smooth as possible. I don't find sanding pads/sponges to be particularly useful. They tend to fold around the rim of the glass and scratch the glass in places you would like for it to be scratch-free.
The other thing to avoid at this point is setting up unwanted vibrations in the glass, as these also tend to cause chips and breaks in the glass. If you hold the paper in one hand, and the bottle in the other, there will be no unwanted vibrations, but it's not a very ergonomic arrangement, and its gonna get old real fast. I find placing the sandpaper on top of a piece of non-corrugated cardboard (think cereal box, notebook cover or thick poster board), turning the bottle upside down and twisting the bottle back and forth against the paper is the most effective and least painful approach. If you hear a screeching sound from the glass while you're doing this, you need more dampening. Try putting some sheets of newspaper under the cardboard.
The pictures here are of the process, as well as the progress at the 1, 2 and 5-minute marks. At this point, it's clear that there are some dips that are going to take a while to grind out. Basically, you'll have to grind down everything else to the level of the dips, and then keep sanding until it's all even. However, as we're not going for perfection here, we can work with the dips and just try to make the glass as smooth as possible, without worrying too much about the edge being perfectly even. It's worth noting that these dips and the unsanded glass are not particularly sharp or dangerous at this point.
You can work on the dips by folding the sandpaper around your finger and rubbing it back and forth inside and over the dip, but this has the same drawback mentioned before: it's tiring. I prefer to take a piece of wooden dowel (or even a reasonably smooth stick), and wrap the sandpaper around that. Doing this allows you to apply more pressure, and apply it more evenly.
At this point, the glass is drinkable - not pleasant, but drinkable. The human lip is pretty sensitive, and while the edge is not a hazard, it's going to feel uncomfortable. We can solve this by beveling the edge. To do this, keep working with the sandpaper covered dowel, but this time, keep it at roughly a 45 degree angle to the glass. Do a couple of passes around the outer and inner edges of the rim.
Now, wash the glass. You're done! The whole process, from score, to cut, to grind, to bevel should take between 15 - 30 minutes for a cut without any significant jags.
Green Power Science: I don't know if he originated the water method, but he's certainly done a lot to popularize it. There is definitely an infomercial-like feel to the video, and he is trying to sell you something, but there's good information too.
The Encyclopedia of Working With Glass: this was referenced in the Wikipedia article on cutting glass (which I strongly suspect the above gentleman, Dan Rojas, wrote). I got the book from the library, and it really is the bomb.