So while I'm having a reasonably boring day and Mike and the CFO are frantically setting up the storage space and shipping wheels out, I thought I'd share the first of several treatises on how various things are made - in this case, spokes. These, by the way, come from a series of lectures that I've given to help my wife fall asleep. The all time record was when I started telling her about how thread pitch in a screw equates to mechanical advantage and "SNKK!SNKK!ZZZZZZZzzzzzzzz." Mission accomplished in about 27 seconds.
So anyway, the vast majority of quality spokes on the market are stainless steel. These start as huge spools of wire, typically 2mm wire. As the wire is unspooled, it is fed into a series of rollers in order to straighten it and make it "forget" its rolled up state. The rollers start out pretty wide but gradually get tighter and tighter until at the end of the roller process the wire is pretty well straight as an arrow. As it's coming out of the end of the roller process, the wire is cut into "raw lengths" to begin the journey into becoming a spoke.
The longest common spokes are a bit over 300mm long, 300mm being a metric foot. A Mavic Open Pro rim, which is a very low profile rim, has an Effective Rim Diameter (ERD) of 608mm. This is the diameter of the inside of the rim plus the thickness of the inner wall (and eyelets, if applicable) of the rim. Deep section rims have lower ERDs - the rim of our 85mm clincher has an ERD of 523.5mm. When you lace a low profile rim with 3x crossing (which is usually how you lace a low profile rim), the spoke travels tangentially from the hub to the rim, plus one end of the spoke is at the outboard edge of the hub while the other is in the center of the rim, so depending on the flange diameter of the hub, a spoke for 3x lacing will be a bit longer than 1/2 the ERD of the rim. Radially laced spokes are roughly 1/2 of ERD, minus 1/2 of the flange diameter. The spokes we buy vary between 281 for the 2x lace front wheel in an FSW (flange diameter of 30mm and 588 ERD) to 247 for the RFSC85 front (radial laced, 523.5mm ERD rim, 30mm flange circle). All of which is a REALLY long way of saying that they chop the spokes into raw lengths of a little over a foot.
Once the spokes are into raw lengths, they get heads formed on them. Basically, one end of the spoke is held in place, a nice man says "now this won't hurt a bit" and then they smash a stamp onto the end of the spoke. Voila! The spoke has a head.
Next, we need to decide whether this spoke is going to live its life as a straight pull or a j-bend spoke. If it's a straight pull, skip this step. If it's a j-bend, like our spokes, it gets a 90 degree bend right up by the head. The positioning of this bend is CRITICAL! Too close and the head will not fit into the hub flange hole. Too far and the bend will be cantilevered out and unsupported by the hub flange. You want the head to fit nice and snug on the hub flange, which is why the bend position and the hub flange thickness are both so important.
Next, we need to decide what kind of butting the spoke is going to get. Our spokes are double butted, which means that the center section is thinner than the end sections, and the end sections are the same thickness as each other. Triple butted spokes are thickest at the hub end, thinnest in the middle, and in between thicknesses at the rim. Triple butted spokes are mostly used in high load applications. Straight gauge spokes are one thickness straight the way through.
Why do spokes get butted? For one it makes them a little bit lighter, for another it makes them a little bit more aerodynamic (the center sections get thinner), and most importantly it makes them stronger. The actual butting process (it's very similar to swaging, if anyone's familiar with swaging) consists of running the spoke between two rollers that aren't completely round. As the spoke enters, there's enough space between the rollers for the spoke to pass. Then the rollers tighten up, so the center of the spoke gets smushed between them. Then the rollers open up a bit for the other end to pass through. There's a metallurgical thing going on there that if I claimed any knowledge of it would be time for you to go get the noseclips and hip waders - the BS would get thick. But the center of the spoke being thinner makes the spoke springier and more resilient. It doesn't point load at the end as much. It fatigues less under cyclical loading. It's all good.
Finally, the spoke is chopped to final length and threaded. The threading process that spokes undergo is pretty unique, as the threads are rolled onto the spoke and not cut into the spoke. If you go to a hardware store and get a bolt that is partially threaded, and spin a nut onto it, the nut will stop where the threading stops. The unthreaded part of the bolt is thicker than the threaded parts - the threads are "valleys" cut into the spoke by removing metal. On a spoke, the threads are "rolled" onto the spoke, so no material is removed and the ridges of the threads are the widest part of the spoke. There are two reasons why this is important. First, all of the material is kept on the spoke. It isn't weakened by having material removed. Second, it's important that the spoke be able to thread past the inboard end of the nipple. You seldom see threads sticking past the end of a nipple on a well built wheel. It happens (it happened to me last night*), but not often. This is because the threaded part of the spoke is usually just shorter than the length of the nipple. A well sized spoke will thread all the way into the nipple, with the first spoke thread hidden by about a millimeter or three of the inboard end of the nipple.
This isn't the way all spokes are made, but if you read this and understand it, you know a heck of a lot about how most spokes are made.
* Why I Had A Thread Stick Past a Nipple Last Night (and why that isn't as dirty of a statment as it sounds): Sometimes even the US distributor of the premier spoke brand runs out of a certain size of spoke. No problem. They sometimes have blank, unthreaded spokes that they can chop to length and thread, or they can cut down and rethread a longer spoke. We ordered some spokes in a unique size (for us) to build up a set of RFSC38s with a PowerTap for a customer. Sapim USA didn't have 259mm spokes, so when I ordered them they told me it would take a day to make some up for us. This is the second or third time that it's happened, and each time I've noticed that the spokes which were threaded especially for us had a little bit longer threaded area than stock spokes. Since the threads are rolled on and no material is removed, it's no issue. All of the spokes were really close to showing threads past the nipple, but one spoke hole must have been at a little bit of a high spot in the spoke bed (it can vary by a tiny little bit), and the nipple was sitting ever so slightly more "in" the rim. That spoke showed a thread past the nipple. The critical thing is that the spoke threads far enough into the nipple. If the spoke doesn't thread far enough into the nipple, the head of the nipple is not as strong as it should be and can eventually collapse. Your spokes have to be significantly too short to do this, and most spoke lenght calculation programs seem to err on the long side to prevent this.
Now do you see why my wife is so well rested?