Spoke length calculators; a mixed bag...

[tatanab]

Hub dimensions - 37 centre to flange, 35 PCD

[amediasatex]

In theory the spokes are 2mm too long. In practice they protruded about 1mm beyond the nipple, so a 294 would have been 1mm inside the nipple, about the end of the screwdriver slot. Spokes used were DT, nipples were ACI Alpina.

I'm happy enough that this shows why we need not be bothered within a mm

All well and good until you come to a situation where you have to build a deep rim with hidden nipples where the only access is with a driver from the outside, or where you have to use inverted/internal nipples.
Likewise if you are mixing an matching spokes/nipples and that extra mm or so can make the nipple or spoke thread bottoms out and you need to extend the thread.

Also, you could be storing up problems for later, seized nipple + chimp swinging on a spoke key leading to rounding the flats off for example, so in you go with the screw driver from the other side to find a spoke poking out of the slot and no chance! OK so you have to cut the spoke and replace, no biggie in a workshop but it's still a wasted spoke and can also mean on-the-road repairs might be trickier.

It can be even worse the other way, 1mm or 2mm too short and you might think you've built an OK wheel, it might even last a while, but without the nipple head being fully supported by a correct depth spoke it puts a lot of extra stress on the nipple head, brass nips will often last a while in this setup, but Alu ones will start popping heads in very short order!

That's not to say I haven't rounded up and used longer spokes from time to time on my own wheels, so it's not that I disagree, you just have to make sure you understand the implications of your choices.

----------------------------------------------------

Thanks for taking the time to review and comment on all the options Brucey, nice to have such a review in one place! I've been using a combination of wallchart/Spocalc/maths for the last 20 years* but recently donated my chart to the local bike charity workshop that I volunteer with as an extra resource for their teaching classes as I've been using EDD for the last few years with satisfactory results but rounding up generally when I have an inkling it's not 100%.

* I was taught by a grumpy man in a workshop who very patiently taught me how to do all the measurements by handing me a box of about 20 different used hubs, and making me document them all, then the ~50 odd used rims in the cupboard, then making me do all the calcs and then having me build wheels in various different lacing patterns according to his mood, unbuild them, and then do it again differently until he was happy I knew what I was doing and could build something he'd be happy to ride before he let me loose on actual customer wheels. I don't think my thumbs have ever fully recovered but it set me up pretty well and got me hooked on wheel building I did get to build some pretty funky and posh stuff for the local racers (Road and MTB**) which I miss a bit, but modern factory wheels are getting more and more proprietary these days...

**Some of the most horrific stuff was for late 90's early 00's 'freeriders' who were demanding some pretty crazy tough wheels for what they were up to. Getting wheel stiff and strong enough to survive 20-30ft stair gap-jumps (often poorly executed) at 30mph was a challenge!

[tatanab]

This is how I started in 1978. A slide rule type spoke length calculator from Ron Kit. In those days most wheels were built with rustless or chromed spokes, stainless being only newly available and expensive. Rims, such as Weinmann Alesa were not box section so an over long spoke did not matter because it was filed down in situ. A horrible job.

I build only for myself, and no wheels with fancy patterns or low spoke counts. On the other hand, I have built everything from tandem wheels to track racing wheels - but all for me and so built "good enough", which is certainly not as perfect as a shop would do them.

[mig]

i really admire the skills and knowledge of a good wheelbuilder.

will these ever be lost amidst the ever growing factory built stuff?

[amediasatex]

will these ever be lost amidst the ever growing factory built stuff?

Diminished and not as widespread maybe, but not lost. There are still decent builders out there and all the material/books/resources are available for anyone who wants to learn how to do more than swap a spoke on Mavic's latest model.

[Brucey]

it is worth mentioning that CJ helpfully posted this calculator chart on the CTC website for 700C and 26" (559) rims, but the current regime has moved (or more likely removed) it so I have found it on the wayback machine;

http://web.archive.org/web/20120724225309/http://www.ctc.org.uk/resources/About_the_Bike/SpokeDiagram-622-559.pdf

I have added this link to the lead post in this thread too.

It is a little confusing at first to look at but all becomes clear, with a little application of grey matter. It doesn't say so explicitly but the assumed ERDs are ~614mm (for 700C) and ~551mm for 26" (559) rims. Use corrections to the chart to get different spoke lengths, i.e. if the actual ERD is 10mm smaller, your spoke lengths are 5mm shorter than the chart predicts. The chart also appears to assume a flange spacing of ~50mm which is OK for rear hubs but not (wider) fronts. If the flange spacing is nearer 70mm then adding 1mm to the predicted spoke length would be a better prediction.

[edit see here for fuller explanation of charts http://web.archive.org/web/20120527030929/http://www.ctc.org.uk/DesktopDefault.aspx?TabID=3524 ]

Somewhere (I can' find it!) there is a graph that CJ produced of rim section depth vs ERD. This typically gets you to within ~2mm of the correct ERD so is a useful sanity check if nothing else.

cheers

[NATURAL ANKLING]

Hi,
This is rare, Brucey has flaws

I am no daily wheel builder but have learnt much to cure my frustration in these post in last few years.
Like losing your first bike, you don't make same mistake with dud tables / online calc etc on hubs
I measure every thing myself now, all you need is a cheap digi calliper from Lidl.

Would be an advantage to be able to easily buy 1MM increments.
With all the double checking you need to do and calc consulting, it can still be out a tad out............unless you are building same wheel week in out and you know that size fits perfect, how easy is that.

So a fraction under slot is always fail Because you can get marginal spokes as in all wheels under the slot, I find in some wheels that there can be even a 1mm diff in location of spoke in nipple (that wheel, let alone that build standard), this is more apparent in diff between inner / outer spokes same wheel side with different bend angle at elbow, more so depending on which under over lacing you prefer

If faced longer or shorter whether your stock / determining on paper, longer can depending on parts used means that spokes can bottom out in thread a fraction after protrusion, you may tell us that using quality parts give good protrusion with out early fouling on odd spoke / Brucey never does that

I admit on repairing / replacing my skip rubbish spokes for green project, I will pick spokes from my used stock and then look for a nipple that looks ok, trying not to forget to check that the thread will foul early.

Some nipples I find will wind down the plain shaft easy some don't like it and complain, this is of course extreme case of matching a odd spoke up repairing a junk wheel.

IIRC you say never under slot with aluminium nipples, and maybe extreme angle to spoke hole? But marginal must be OK with brass nipples.

[amediasatex]

So a fraction under slot is always fail

I wouldn't necessarily go as far as to say 'always', but I'd be unhappy letting a wheel go out to a customer like that, and it is less than ideal...personally I'd rather the spokes were a little long than a little short, I'd sooner deal with extra threading or protrusion past nipple issues than the risk of popping nipple heads. Doubly so with Aluminium nipples though. Alu nips built properly should have no issues at all, but it is easy to cause issues by not doing it properly with them.

I once thought I'd risk it on a wheel with Alu nips just to see, a bit of experimenting is always useful, even if it does just serve to prove a point...the spokes were literally about 0.5mm below the slot, result? 3 rides before a nipple failed. I replaced it just in case it was a one off, managed one more ride and then lost 3 nipples within a few miles. Point proven, never again!

using quality parts give good protrusion with out early fouling on odd spoke

haha, in theory this should be true, real life is not always so black and white though!

[Brucey]

I agree, nipples can pop, alu nipples especially, if the spokes are too short. I've seen enough brass nipples fail that I know that it is a bad idea with those too.

By comparison a little 'thread crunching' on a spoke that is too long is usually fairly harmless. I don't think the nipple is liable to fail through crunching unless it (or your spoke key) is in some way of questionable quality. I don't think I have ever seen a spoke fail through 'crunching'.

The main worry with spokes that are a bit too long is what an undressed protruding spoke is going to do to the tube if the rim tape is a bit iffy, or sags or something.

cheers

[amediasatex]

BTW, this:

and maybe extreme angle to spoke hole?

+ Alu nipples often results in an apparently OK wheel, often for some time, but eventually can cause the nipples to crack lengthways along one of the flat faces, which if left then propagates around at the top and leads to you losing one side of the bottom half of a nipple, often swiftly followed by the other half. It can take some time for this to happen though.

I vividly remember one local racer coming in to get his wheels trued after a bit of a ding and I noticed immediately that most of the nipples were starting to crack down one face, he was absolutely adamant he didn't want to pay for a rebuild and asked me to 'do my best' while watching as I systematically broke nipple after nipple by wedging my fingernail in the crack and flicking the bottom half off...to be fair the wheel was still holding together on just the nipple heads, but he soon agreed to the rebuild

[RogerMusson]

[Aside; spokes are conventionally measured from the inside of the elbow bend to the other end. Spokes also settle in the hub, and stretch elastically. In practice the first of these things would make the spokes be too short

If you say the spoke settles in the hub then this would make the calculated length longer rather than shorter.

(since the calculation is from the drilling centre, i.e. with a 1mm error built in) and the latter points would make the spokes too long.

The calculators calculate to the edge of the hub spoke hole where the spoke elbow rests (and where spoke lengths are measured to), that's why most of them ask for the hub spoke hole diameter. So I'm not sure what this 1mm built in error is that you talk about, it doesn't exist.

https://www.wheelpro.co.uk/spokecalc/

Does not support spoke gauges that are bigger than 2.0mm.

It is compatible with DT Alpine and Sapim Strong. For elongation purposes only the central 2mm diameter is required (as explained in the help guide). What spokes are you using?

Not sure if the prediction of spoke stretch is that useful, since it assumes a specific spoke tension,

It assumes you build the tight side to 120Kg tension. It then determines the tension in the lower tension side, and calculates elongations based on the spoke length, spoke diameter and Young's Modulus. Building your wheel to slightly higher, or lower tensions, will have a negligible effect. How is all of this not useful? If you want to see what the calculator is doing, calculate a spoke length, then click on the Log.

and it doesn't allow for all the variables associated with the way the spoke fits into the hub, which is at least as important.

What are these other variables? They don't exist. Nor do they exist in all the other calculators.

These spoke lengths are slightly shorter than many, (and he suggests you round up) so I'd suggest that you measure ERD with the spokes screwed flush with the top of the nipples before you enter values into this spoke calculator.

Okay, this is the reason I chimed in on your post. You have given out totally BAD ADVICE here. I designed and wrote the calculator, I've also thoroughly tested it by building many test wheels, using different spoke diameters, cross patterns, wheel sizes. Please measure the ERD as I say (and explained in the calculator help guide).

Does not support fractional crossings.
Does not allow arbitrary drillings

Because I don't encourage building wheels that require these inputs. The same reason I didn't include 2:1 Triplet lacing. I did however put in the calculation for straight pull hubs, and made it simple for calculating offset rims.

All the other calculators use the basic theoretical spoke length formula. The exception is the DT calculator that takes into account spoke stretch, but using a very basic technique that subtracts fixed amounts based on the type of spoke selected.

Cheers,
Roger

[Brucey]

thanks for your comments Roger.

[BTW the 'bad advice' appears to make about one and bit turn's worth of difference with the spokes I use.]

cheers

[Brucey]

regarding 'spoke crunching' i.e. when the spoke pokes out the top of the nipple, past the point where the nipple spins easily.

I've just done some experiments that seem to be quite interesting. The photo below show some of the results.

Thread Karno's....?

The top two spokes are DT 14g stainless ones each with a nominal 10mm length (factory 56tpi) thread. The bottom one is a 14g stainless spoke with a thread rolled on using a Cyclo rolling head.

DT 12mm brass nipples on the DT spokes typically ran on 17 or 18 turns to flush, and then +4 turns (i.e. almost 2mm) further without binding. [This may vary slightly with batch of spokes/nipples and/or manufacturer.] A new DT nipple on the Cyclo thread span on OK until flush, and then started into a 'soft crunch' presumably because the thread is more tapered than the factory threads are, owing to the different method of making the thread.

The top spoke had a DT nipple 'crunched' another + 2-1/2 turns onto it. This was done bone-dry and did not require a great force to achieve it; estimated at less that 1/5 of the torque at which a spoke is liable to shear (small enough that it might not be noticed if the spoke is under tension when the crunching occurs). This left a small amount of brass swarf in the spoke thread, and some tiny scuffs on the shank of the spoke. Presumably there would be less of both has the nipple been lubricated. Compare the crunched thread with the uncrunched thread on the middle spoke.

The 'crunched nipple' span on to either the crunched or uncrunched DT spoke easily, to +6-1/2 turns. An uncrunched nipple span onto either DT spoke as it did before, i.e. to +4 turns. The crunched nipple was a bit draggy onto the Cyclo thread, so that it would not spin on easily using finger pressure, but when using a spoke key there wasn't an appreciable amount of drag.
When measured, the Cyclo thread had an OD which was ~0.002" smaller than the factory threads, and presumably a shallower root to match. The latter would be likely to cause interference with a slightly damaged nipple.

[BTW Threads rolled using other equipment, (eg a Phil Wood machine) may also vary in size and quality. I had some cut and threaded using such a machine recently, and the threads were at least 0.003" undersize, yet the LBS concerned has not had a brass nipple pull off (not in many thousands of spokes) and they build their wheels pretty tight. I think that the usual distribution of loads and strength of the thread on spokes does not behave as per engineering texts; in all likelihood typically, more threads share the load with a shallower engagement than normal; note also that a shallower/tapered thread in the spoke may leave the spoke itself slightly stronger than it would be otherwise, perhaps lessening the chance of spoke failure near the nipple.]

The conclusion is that in this case 'crunching' 2-1/2 additional turns (making a total of +6-1/2 turns, or over 3mm protrusion) has primarily damaged/altered the nipple and not the spoke. Probably the number of good threads in engagement (including those in the slot region which may not contribute fully) has gone from 17 or 18 to about 15. This is extremely unlikely to cause any problems IMHO (past the usual ones where the inner tube is at risk due to the protruding spoke). Further 'crunching' would be possible, with further loss of thread engagement being the primary effect.

Non-factory threads may vary in size and fit onto nipples. With either 'crunched' factory threads or nipples screwed onto (slightly tapered) non-factory threads, there is a prevailing torque that may help some nipples (eg NDS ones) not to unscrew in service where they might otherwise tend to do so, unless threadlocked. I have seen wheels built deliberately to exploit this effect.

Whilst I would nearly always prefer to see spokes finish somewhere within the depth of the nipple slot, it seems that a protrusion beyond the nipple is usually pretty harmless, even if it involves some 'crunching'.

cheers

[RogerMusson]

With either 'crunched' factory threads or nipples screwed onto (slightly tapered) non-factory threads, there is a prevailing torque that may help some nipples (eg NDS ones) not to unscrew in service where they might otherwise tend to do so, unless threadlocked. I have seen wheels built deliberately to exploit this effect.

Sapim do this with their SecureLock nipple. They deform the nipple thread to give extra friction.

https://patentscope.wipo.int/search/en/detail.jsf?docId=WO2013131564

[Brucey]

I've also seen (years ago) a heavily dished rear wheel built where the NDS nipples were deliberately slightly 'crimped' in order to help retain them. I think that this was done after the wheel was built, judging from the alignment of the marks on the nipples. It seemed to work OK, but I have not heard of anyone recommending this practice.

cheers