A lovely rainy day to show some wheel testing. First up, a comparison between a 24 hole Pacenti rear, and a 28 hole Pacenti rear. We've also included a 28h Stan's Alpha 400 as a foil.
In order to test deflections, we hold each wheel in the frame fixture as shown, and apply a calibrated downforce by hanging a weight (or mass, if you prefer) from the blue rope with the red stop ball in the foreground. Deflections are measured at the places where each red arrow points.
Different wheels are known to respond differently. For example, a very stiff rim without enough spokes would deflect significantly downward at the front arrow (the 0* point on our table below), and significantly upward at the rearmost arrow (the 180* point). It would typically not deflect much at the 90* point (arrow to the left). A soft rim would deflect downward quite a lot at 0*, upward quite a bit at 90*, and downward a bit at 180*. This is why rims hitting brakepads is often not a sign that your rims are not stiff enough, but rather that they are underspoked.
The magic of system design is to build the wheel to minimize deflection all around. In a race wheel, you want to be just at the point of diminishing returns in order to minimize weight, where in a training wheel you might want a bit more margin for error in the system. The softer (edit) the rim, the more spokes you need, simple as that. A softer rim will "sag" more between spokes. I think of the distance between spokes as the "unsupported span." On the 24h Pacenti rim, the unsupported span is 75mm, on the 28h Pacenti it is 64mm, and on the 28h Stan's rim, it is 65mm. The unsupported span per given spoke count decreases with increased rim depth - a 24h Rail 52 has an unsupported span of 68mm, which is closer to the span of the 28h Pacenti than it is to the 24h Pacenti. The stiffer the rim and the shorter the unsupported span, the fewer spokes you typically need - up to a point. That point is where the rim "overpowers" the spokes, and you get the "tilted" deflection where the rim goes down hard at 0* and up hard at 180*.
It's not super easy to see in this photo comparison, but the distance between spokes is just about 11mm more on the bottom photo. In case anyone wonders, we are big fans of the Pacenti SL23. It's a nice, stiff rim that's well construction and relatively light. I often ride them, and am currently using them to test a new hub we're evaluating.
As you can see from the data in the table below, the 24h build comes closest to overpowering the spokes. It deflects downward the most in the front, and upward the most in the back. The actual amounts are worthless in comparison to tests not done on this rig, but from what we've learned these characteristics get you close to where we don't want to be. You start to put a lot of stress on the rim at the nipples as the rim "works" against the spokes, and this is often borne out by the failure of "underspoked" wheels coming through cracking at the spoke holes. It's not necessarily the static load that gets them, it's the dynamic load transmitted through the deflection.
A last point to make is how rim depth affects this whole calculus. A deeper rim makes for a better spoke bracing angle. As you move the hub and the rim closer together, the angle that the drive side spokes makes from hub to rim becomes more acute. This is always a good thing.
The stated deflection increments are thousandths of an inch. .177" is very roughly equivalent to 3/16." A 25 pound kettlebell is used to create the deflection load.
There are a few other things going on here, like how critical each spoke becomes as you have fewer and fewer (a not great builder plus few spokes often equals disaster) but we are running out of time here as it is.
|Pacenti SL23||Pacenti SL23||Stan's 400|
|Hub||WI T11||WI T11||WI T11|
|Spoke Type||CX Ray||CX Ray||CX Ray|
|Distance Between Spokes||64mm||75mm||65mm