Spoke Tension Meter, Hows This Work?

[Brucey]

Colin's ratio of 80:120 (equivalent to 66.6%) tallies well with the percentage figures I quoted earlier for a 135mm OLN 8/9/10s wheel.

Measuring a ratio of 1:3 is just as unbelievable as the earlier "10% less on the NDS". It would seem that something is seriously amiss with those measurements.

cheers

[MikeDee]

IMO, you should use some form of threadlocker on the NDS spokes to keep the spoke nipples from loosening. Spoke tension alone is inadequate for a wheel with 130mm OLN hubs to keep NDS spoke nipples from loosening.

I haven't found NDS spoke loosening to be a problem on wheels I've built, including quite a number of 11s 130 OLN wheels over the past few years, and I don't use threadlocker. Indeed, the first thing I do with nipples is to open the packet, drop them into a sieve, submerge them in oil, let them drain, and store them in a bottle until use.

When building a wheel, you need to set the spoke tension about 10-15kgf higher than the final value you want, as installing and inflating the tyre will cause the tension to drop by about this amount.

I'm a trendy consumer. Just look at my Pixel 2 using hovercraft full of eels.

The higher the spoke tension the lower the fatigue life of your rims at the spoke hole because fluctuating stresses are closer to the yield point of the material. I don't think using high spoke tensions (greater than the rim manufacturer recommends anyway) to keep the spoke nipples from loosening when threadlocker works splendidly for that task, is wise. Wheel Fanatyk says going beyond 100kgf is unnecessary for conventional 32 spoke wheels.

[Brucey]

some riders on some rims will not only loosen NDS nipples that are not threadlocked in 32 spoke wheels with 100kgf on the driveside, the NDS spokes can start breaking in odd ways, presumably because the spokes don't just go slack, they start to buckle in compression. This happens when a strong rider gives it 'full gas' eg when sprinting uphill out of the saddle. When it comes to building lightweight wheels, there are very few hard and fast rules; what works for one rider won't work for all.

Rims crack for various reasons and in various ways. One of the commonest failures I see on all-weather training wheels is that the rims fail because of stress-corrosion cracking. Rims that fail this way can crack up at very low mileage; they just need to be sprayed with salty water once or twice and then parked in the shed. High spoke tension and salty water are all that is needed; fatigue per se may not be important in this case. Such failures are greatly deterred by some additional corrosion protection, eg running in waxoyl around the rim eyelets etc. Rims can fail in this way even below the tension limit set by the manufacturer but failures are made even more likely if the tension limit is exceeded of course. These failures often involve multiple cracks and even chunks of the rim coming away from round the spoke drillings.

Another mode of rim failure is seen in wheels that don't have enough spokes/tension in them to withstand the weight loads. It is unlikely that a wheel with 32 spokes with 100kgf on the driveside will fail in this way but if the build is any slacker than that it is quite likely with a heavier rider. In this case the clue is that there are radial cracks (often only radial cracks ) in the rim, usually starting at the drillings. In extremis some cracks may start at the rim lips. Again this is made more likely if there is road salt on the rim.

Somewhat counterintuitively the cyclic fatigue loadings in the rim can increase in a slack build; the reason is that not only do the spoke tensions vary as you might expect, but the rim sees additional bending because it is no longer so well supported by the spokes. The two stresses can add to one another in certain places. However I don't think this happens too much unless there are at least some spokes that are going completely slack in a wheel. IIRC someone has done some FEA work which suggest that between lateral and torque loadings when climbing, it is the 'pushing' NDS spokes which are most likely to go slack first. Sprinting out of the saddle on a climb is potentially even more violent than the condition modelled, I think, so it may be that there is even more slack running going on that the model suggests.

cheers

[MikeDee]

some riders on some rims will not only loosen NDS nipples that are not threadlocked in 32 spoke wheels with 100kgf on the driveside, the NDS spokes can start breaking in odd ways, presumably because the spokes don't just go slack, they start to buckle in compression. This happens when a strong rider gives it 'full gas' eg when sprinting uphill out of the saddle. When it comes to building lightweight wheels, there are very few hard and fast rules; what works for one rider won't work for all.

Rims crack for various reasons and in various ways. One of the commonest failures I see on all-weather training wheels is that the rims fail because of stress-corrosion cracking. Rims that fail this way can crack up at very low mileage; they just need to be sprayed with salty water once or twice and then parked in the shed. High spoke tension and salty water are all that is needed; fatigue per se may not be important in this case. Such failures are greatly deterred by some additional corrosion protection, eg running in waxoyl around the rim eyelets etc. Rims can fail in this way even below the tension limit set by the manufacturer but failures are made even more likely if the tension limit is exceeded of course. These failures often involve multiple cracks and even chunks of the rim coming away from round the spoke drillings.

Another mode of rim failure is seen in wheels that don't have enough spokes/tension in them to withstand the weight loads. It is unlikely that a wheel with 32 spokes with 100kgf on the driveside will fail in this way but if the build is any slacker than that it is quite likely with a heavier rider. In this case the clue is that there are radial cracks (often only radial cracks ) in the rim, usually starting at the drillings. In extremis some cracks may start at the rim lips. Again this is made more likely if there is road salt on the rim.

Somewhat counterintuitively the cyclic fatigue loadings in the rim can increase in a slack build; the reason is that not only do the spoke tensions vary as you might expect, but the rim sees additional bending because it is no longer so well supported by the spokes. The two stresses can add to one another in certain places. However I don't think this happens too much unless there are at least some spokes that are going completely slack in a wheel. IIRC someone has done some FEA work which suggest that between lateral and torque loadings when climbing, it is the 'pushing' NDS spokes which are most likely to go slack first. Sprinting out of the saddle on a climb is potentially even more violent than the condition modelled, I think, so it may be that there is even more slack running going on that the model suggests.

cheers

Brucey, almost all rims are double walled. Only the cheap ones are single walled. Thus, while NDS spokes could go slack, they are not in compression/buckle. The slack spoke just gets pushed into the rim cavity.

[Brucey]

Brucey, almost all rims are double walled. Only the cheap ones are single walled. Thus, while NDS spokes could go slack, they are not in compression/buckle. The slack spoke just gets pushed into the rim cavity.

if they buckle in the right way and slide through the eyelet, yes. But in used wheels they very often do not do this (presumably something to do with the spoke crossings in some cases, dirt in the eyelets in others perhaps); it is quite normal for nipples to bind in the rim, and for there to be a fair amount of movement between the hub and a loose spoke before the nipple moves around lengthwise in its seating.

In any event I have seen plenty of wheels that broke spokes this way and (as I mentioned upthread) a good test for this is try the wheels without threadlock to start with; if the NDS nipples back out pdq then you know you are sailing a bit close to the wind.

cheers

[gxaustin]

Measuring a ratio of 1:3 is just as unbelievable as the earlier "10% less on the NDS". It would seem that something is seriously amiss with those measurements.

I thought I'd explained that. The 10% figure was based on the difference in reading, which when fed into the very non - linear relationship between reading and tension, gave the 3:1 ratio. It is that relationship which appears to be wrong for my tension meter. However, the wheel is true and reliable so why worry?

[Samuel D]

Which tensiometer are you using, gxaustin?

[Brucey]

"why worry?" -well if there is that large an error in the tension ratio, there may well be a large error in the absolute tension values too. If the wheel is correctly dished the actual tension balance will be around 65%. It rather begs the question 'which measurement is most wrong; the DS or the NDS?' Your wheel could be built with too much or too little tension all round.

I wonder what is happening? Looking at the calibration chart

https://www.parktool.com/assets/doc/product/TM-1_conv-table.pdf

it is difficult to see what could be going on; even if the chart is used wrongly or something I can't see how you would end up with 3:1. I feel sure there is a lesson of some kind to be learned here and if we find out what it is it may help others in the future.

cheers

[gxaustin]

Your wheel could be built with too much or too little tension all round.

Not really:
1) the wheel is reliable (as I have stated more than once);
2) It compares with other wheels by squeezing the spokes and taking readings.

How can it be worse than having no guide to the uniformity of the spoke tension? The readings are repeatable - my money is on the conversion table.

[gxaustin]

Which tensiometer are you using, gxaustin?

It's X tools - which is a copy of Park Tools et all. The principle is so simple that I doubt the device it at fault. There are 3 pegs, one pivot and a spring. There are quite a few which are closely related

[Brucey]

the calibration of spoke tension gauges is rather sensitive; they can be in error because of offset or 'gain' errors, as well as scatter arising from other sources.

Because of the way they work small differences in either spoke straightness or spoke diameter add to the 'offset' errors and these in turn add to any 'gain' errors which might arise from (say) the spring rate being slightly wrong in the spring. Most gauges have a screw adjustment for the spring so the load on the fulcrum can be set to be 'correct' at one point in the throw of the tool but this is an offset adjustment; if the spring rate is wrong then it will only read correctly at that one point; at other tensions or on other spoke thicknesses the gauge could still be miles out. The spring rate goes as the wire diameter of the spring cubed; tiny differences here will render the usual calibration chart nigh-on useless.

In addition to the above, any slack in the pivots or binding/inconsistency where the tool touches the spokes will lead to large errors.

Because of the way the tool works small changes in spoke diameter can lead to large errors. In the range of interest (~100kgf) a 0.1mm difference in spoke diameter gives a difference in the reading on the gauge of ~1.5 increments (on the park tool) and this is 'worth' about 20kgf of tension. Since nominally 14g spokes vary in actual diameter from just over 1.9mm to 2.0mm then you can easily get a 20kg error here even if the gauge is properly calibrated.

Comparing the gauge reading with a 'known good' wheel will give you some confidence I suppose but this is still subject to spoke wire diameter based errors and anyway you don't need a tension gauge to make that comparison any more than you do when comparing tensions on one side of the wheel.

With a typical tension meter based on a park tool design, if you want to be really accurate, you should set up a dead weight test using a perfectly straight) spoke from the batch you are building with (and the desired tension) and use that to verify the meter's calibration.

BTW with some rims if the tension is 'correct' (i.e. near the maximum on the drive side) with an inflated tyre, when the tyre goes soft/flat in storage, you may get a situation where the tension exceeds the manufacturer's recommendation and SCC of the rim is a danger.

cheers

[Mick F]

I haven't got a tension gauge, I go by "feel". It works very well indeed.

In all my years of cycling, I've only ever broken a spoke once .................. on the rear wheel, and not one I'd built, but bought as a complete wheelset from Ribble.

[gxaustin]

Comparing the gauge reading with a 'known good' wheel will give you some confidence I suppose but this is still subject to spoke wire diameter based errors and anyway you don't need a tension gauge to make that comparison any more than you do when comparing tensions on one side of the wheel.

Very true and a micrometer or decent caliper helps to confirm the spokes are alike.
I still like to use the meter to check uniformity because after a while spoke squeezing makes my hands tired.
We know now that a tension meter is subject to all manner of potential pitfalls. So is a camera's exposure meter but most people use one - usually accepting it's readings without question. Surely the point is that they both work best when used advisedly - and success is fairly easy to establish in either case. In my case the half dozen or so broken spokes I've had over the years have all been in wheels I haven't built myself. If I built loads of wheels on a regular basis I might not feel the need for a tension meter, but as it is I choose to use it.

[531colin]

If you own a tension gauge, and use it, why not use it properly?
So far, you have said the driveside and NDS spokes have a 10% difference in tension, or a three times difference in tension. Neither of these is even remotely possible.
If, as you say, a 10% difference in the gauge reading equates to a three times difference in tension, then small differences in tension on the same side of the wheel won't even register a different deflection on the gauge.
Why don't you scan in the table that converts gauge reading to tension, and tell us what readings you work to?

[gxaustin]

If you own a tension gauge, and use it, why not use it properly?

I use it to suit my purpose, not yours.