Hondo6

Wanted to pick the C&V mental collective's "hive mind" here, as I'm very new to building wheels from components.

I'll be building a new wheelset in the (hopefully) not too distant future, and have everything I need to lace it less the spokes. This includes purported dimensions on the hubs and rim ERDs, so I've done preliminary calculations for spoke lengths. (I will of course physically verify these dimensions against actual measurements and re-run the calculations prior to "pulling the trigger" on any spoke order).

Preliminary calculations from Karl Stoerzinger's online version of Freespoke and UBI's online spoke calculator agree. Given this, I'm fairly sure I'm entering the data correctly in both (the data is auto-populated on Freespoke, as the hubs I'm planning to use are in that version's integral database).

Three of the spoke lengths are slightly over an even mm (2XX.1mm or 2XX.2mm). Being inexperienced, it seems to me that I should round these down, especially given the fact that spokes stretch slightly when under tension.

However, the fourth calculated spoke length is just under an even mm - specifically, 2XX.9mm. Would I be best served to round these up or down?

Thanks in advance for any advice.

Classtime 

Round up to make building easier. Rounding down can make it more difficult in my experience (using the Roger Musson method).

NVFlinch 

You're way overthinking this. In my experience spokes can be within 2mm and still work. Rounding up or down is OK, although I would rather round up, and in cases where the spokes are within 1mm of each other I order all the same length.

Any good calulator takes into account spoke stretch, which isn't much.

repechage

.1mm is theoretical, unless you have your own spoke threading machine. .2 is probably also.
you did not mention much to give any real world opinion.

in general, 5 speed 1mm-2mm less on the drive side.
6-7 speed, 2mm less on the drive side. The reason on all that is to let the spokes do much of the dishing work for you, one will end up with a wheel that provides for maximum adjustment later.

dddd

With alloy nipples, I don't round down at all, since they might break.

I usually test-fit the chosen parts (spokes and nipples) before starting, to see how easily the end of the spoke passes through the head of the nipple without the threads binding too much.
I then might drill out a millimeter or so of threads (using a drill-stop tube on the drill bit) to free up any threading that might threaten to interfere (if the spokes are perhaps chosen more than a millimeter too long).

Spokes cut on a shop machine that is well along it's service interval may tend to bind up in the nipple at full draw during the last turn of the spoke nipple. This can be frustrating and has on occasion led me to use looser-fitting alloy spoke nipples just to allow full tightening.

Some double-walled rims leave the head of the spoke nipple touching the tensed rim tape, making the spoke length more critical, in which case it seems ok to err on the short side unless using alloy nips.

Some rims are relatively intolerant of higher spoke tensions or of much in the way of pre-stressing, including Velocity Aerohead, single eyelet Module E (as well as some modern vintage-look rims). I like to check the web or check with the rim maker about what spoke tension can be used (with what spoke count) to prevent cracking (or "warp" instability) during or after building.

Hondo6

Sorry, I was asking a general question concerning how best to round spoke lengths when the calculated spoke length is not a whole mm - e.g., whether rounding up or down to the next whole mm would be better, and when one should do one vice the other. That's almost always going to be needed, since spokes typically are only available in whole mm lengths. I didn't realize more specific info was needed; my bad.

Here are the specifics: hubs in question are Shimano HB-1055 and FH-1055, both 36H. Rear is 7S. Lacing will be cross-3. Rims are of relatively recent-manufacture (Alex Z1000), single-walled, with specified ERD of 605. Calculated spoke lengths from both online spoke calculators for the rear hub are 292.9mm for NDS and 291.1mm for DS - or approx a 2mm difference. Calculated spoke length for the front hub is 294.2mm. Nipples will be brass, manufacturer yet undetermined, most likely 12mm but possibly longer depending on what's readily available at time of order. Since this one will be more-or-less a learning exercise, plan is to use 14G straight gauge stainless spokes - if I screw it up, I'd rather it be a $45 "oops" than a far more expensive one.

Hub measurements and rim ERD will of course be confirmed by physical measurement and calculations re-checked prior to placing any order.

Hondo6

Thanks - don't plan on using alloy nipples on this build - or any time soon, for that matter. But I'll keep that in mind if I ever do.

Good advice; thanks again. I'll probably do that after ordering the spokes and before starting the build.

Andy_K 

I agree with this advice and reasoning.

More controversially, I would recommend straight gauge spokes. Yes, I know that butted spokes can build a wheel that is both stronger and lighter, which is a rare combination, but straight gauge spokes are easier to work with as an inexperienced wheel builder. I have a theory that the likelihood of lost tension over time due to not having fully taken care of wind up will more than cancel out the strength benefits of butted spokes. Also, wheels built with straight gauge spokes are plenty strong. The weight difference is also fairly minimal.

That's just my amateur opinion, of course, and I have never seen an expert wheel builder endorse such a view.

Hondo6

Pretty sure you're wrong here, arguably on both points.

Best I can tell, both Karl Stoerzinger's online version of Freespoke and UBI's online spoke length calculator are considered excellent spoke length calculators.

Neither appears to account for spoke stretch.

This is taken directly from Freespoke's online documentation (the "How to Measure" Tab on the home page):

As I noted above, in the test case I ran both came up with identical results for the same hub/rim data. Since UBI's online calculator comes up with the same numbers for the same hub/rim combo, it's fairly obvious that it doesn't account for spoke stretch either.

When you think about it, why neither calculator accounts for spoke stretch becomes obvious. Spoke stretch depends on spoke length, spoke tension, spoke geometry (e.g., diameter, whether butted or bladed, etc . . . ), and spoke material. The calculators produce spoke length, but neither asks for the other three items - so neither can accurately estimate spoke stretch when tensioned.

I got curious, so I decided to do a bit of research and see where that led me. Turns out that spoke stretch is pretty easy to calculate, once you know what type of spoke you're using and how long it is. Once you know that, the stretch becomes a linear function of the spoke's tension.

I assumed the simple case I plan to use and "did the math" - multiple times, to ensure I didn't screw it up. I used a 294mm 14g plain-gauge spoke (2mm dia, 1mm radius) and assumed 304 stainless for the material. The formula for stretch (deformation) is given in the Section Summary part of this link (see the 2nd equation listed under Hooke's Law).

The initial length (Lo in the formula) is 294mm.

Young's Modulus (Y in the forumla) for 304 stainless is given as 190 to 203 GPa. To calculate max possible stretch, I used the lower figure (the equation for deformation divides by Young's Modulus).

The diameter of a plain-gauge spoke is constant, so the spoke's cross sectional area (A in the formula) is a constant (PI / 10^6) throughout.

For tension (F in the formula) , I used a value of 110kgf (not unheard of, especially for rear DS spokes), converted to SI-standard units of force, Newtons - 1,078.73 N.

Plugging in the numbers gave a stretch at that tension (110kgf) for a 14g plain-gauge 304 stainless spoke of length 294mm of a bit over half a mm - 0.53125+mm, to be precise. Half that tension would yield half that much stretch.

Half a millimeter isn't huge, but it's also not insignificant. I had no idea that spoke stretch would be that much.

unworthy1

I avoid alloy nipples, using best quality brass nips I can find.
I use as many spoke calculators as I can find (often plugging in numbers to as many as 6 for any given build) cause tho most are very close there's always some "interesting" outlying variation...then I average the results.

Despite this I have had too many wheels build up with slightly loose spokes, especially on the rear wheel where calculators seem to have trouble with the different length requirements for dishing, and so:

I always round DOWN.

YMMV

Reynolds 

If rim and hub dimensions input is correct, you shouldn't have any problems using 293/291 and 294 spokes IMO.