Part 1 of this sucker discussed what a "true wheel" is and what we do to try and get them to that state. This next part is about what a stable wheel is, and why that can supersede some of the stuff from part 1, and why we think that hand building is still the only way to achieve this.
True, round, and dished out of the box is one thing. The next few steps of that wheel's life - having a tire installed and being ridden for the first time - bear out the importance of a stable build. In this post from a while back, we showed what happens to a wheel that is relatively true but hasn't been stressed after some stress goes into it.
A wheel that's "true" but hasn't been thoroughly stressed and balanced is likely to become less true as soon as it is stressed. Simply putting a tire on imparts a lot of stress into the wheel. Especially with modern tubeless rims, a clincher tire exerts a lot of force radially inward along the spokes. This makes itself known by the spokes losing some amount of tension. Some rims compress so much that the wheel loses a shocking amount of tension, where deeper and heavier aluminum rims, and well-designed carbon rims, exhibit less of this.
Because aluminum is the same strength along any way you might stress it (isotropic), the only way to combat compression is by making the rim section deeper (a 28mm deep rim versus a 21mm deep rim, for example), or to increase the wall thickness of the section. Going to a deeper section will generally be more effective than adding wall thickness, but you do see some strange stuff if you build enough different rims. Because carbon is stronger in some directions than others (anisotropic), the plies of the laminate can be engineered to counteract compressive load without adding much weight at all. This is one of the things that we do very much like about carbon rims.
Once the tire is on and the rim is compressed and you go for a ride, you put a torsional stress on the wheel. Any "hard spots" in the spokes, any amount of spokes bedding into the hubs, any amount of spoke lines from hub to rim getting straight, and any amount of tension imbalance between spokes will all adjust toward their stable state. When this happens, the wheel goes out of true/round/dish.
One of the profound moments I've had as a wheel builder was pretty early on, when I stopped thinking about manipulating the rim's circumference to create a round wheel, and started thinking about placing the hub's axle in the middle of the rim's round shape. It's a subtle semantic shift, which probably 100% of people reading this won't get, but its actually significant. Finding that spot and putting the hub there is the first big step.
From that point, you add tension to the spokes to pre-load the structure with hopefully enough pre-loading that the wheel won't ever be stressed beyond that. But that's a multi-step process. You have to add spoke tension gradually around the wheel. If you were to screw in the nipple as far as it needed to go on each spoke in succession, that wheel would be a pretzel and if it wasn't strong, might actually break or at least permanent deform in that process. So you take small numbers of turns (or even increments of turns) around the wheel in several passes to avoid ever putting too big a change into the wheel with any one adjustment. While you're doing this, you need to stress the spokes to work out any hard spots in the spokes (and I'm not a metallurgist and if there is one in the audience please feel free to light me up for incorrectness, but this phenomenon definitely exists), and to seat and align the spokes to where they want to end up.
That's too long already, so I'll have to put a pin in it there and save some for later. I'm screwing around in CA later this week and into next, escaping winter (though I've done a better job of embracing winter than in recent years), so should have lots of chances to finish this saga.