John Hall (Gwin's Mechanic) on Spoke Tension

I did a little googling and no one should be surprised to learn that the late, great Sheldon Brown built just such a test fixture to perform just such a test, applying 25 pounds of lateral force to dozens of different wheels in different lacing patterns and spoke counts, and varying spoke tension to test the affect on lateral stiffness. The whole article is fascinating and worth a read: https://www.sheldonbrown.com/rinard/wheel_index.html

He found that, "Wheel stiffness does not vary significantly with spoke tension unless a spoke becomes totally slack." He also tests a bunch of other questions, like "Are front wheels stiffer than similar rear wheels," "Do dished wheels flex differently from one side to the other?" "How does spoke gauge affect stiffness?" etc etc.

Sheldon's test jig supports what Dave and Milton said- the effect of spoke tension on lateral wheel stiffness is not statistically significant in a laboratory environment.

So where does that leave us? For John Hall, Loic Bruni, and everyone else (me) who's convinced that they can feel changes in spoke tension, is that all in our heads? Or are we talking about a small effect in the laboratory that translates to tangible results on the trail? I don't know. It's possible that generations of racers and mechanics since the days of Nico Vouilloz have been imagining this effect, like performance-enhancing bracelets or magic vibration stickers, but I find it hard to believe that a myth like that would be so durable. But hey, it took road racers 100 years to figure out tire pressure, so who knows. Someone should test this on the trail.

So I will chime in with absolutely no hard data, but experience with building LOTS of wheels.  

The quick answer: Yes Spoke Tension Matters.

now, the deeper answer, is: It Depends and does it matter

So, one way that the spoke tension matters is durability/maintenance intervals.  A tighter spoke allows more consistent load on the spoke/nipple/rim interface.  This helps stop vibration and microscopic load/unload happening and therefore spokes loosening.  a looser wheel most definitely loosens faster and things go haywire faster. an evenly tensioned wheel at a higher tension will stay straighter longer than a lower tension wheel.  So from a durability perspective, this makes a massive difference.

Secondly, for anyone taht says that spokes don't lose tension under impact or load... I call BS.  Go build your wheel.  Check your tensions.  Now put a tire on there and inflate the tire. 25 psi, 40 psi, 60+psi on a road wheel, and tell me what your spoke tension does.  It drops.  The rim flexes inwards, and lessens the load on the spoke.  You already have a lower spoke tension and it goes even lower. 

Bracing angle/side load.  Again, a lower spoke tension will allow for more wheel flex.  A spoke does not pull in a direct line at a wheel.  Sponsel brings this up wtih bracing angle and wider hub standards. there is a side force put on a rim by a spoke.  A wider flange spacing makes for a stronger/stiffer wheel.  Lower spoke tensions provide less side force on a rim as the spoke is unable to pull the rim sideways more.    This is noticeable when building a wheel and stressing it to reduce wind up in a spoke.  When pushing on a rim from drive side towards non-drive it takes less force to move the rim, then it does from the other side.  It is easier to overcome the force of the DS spoke tension than the ND side. 

Also this gets into the spring/elasticity properties of the material.  Different spokes act as different "springs".  A straight gauge spoke is stiffer than a double butted (but weaker). as you are building a wheel it is clear that as tension comes up the straight gauge ramps up its tension WAY faster than a double butted spoke does.  The double butted spoke elongates or is more elestic....again allowing more flex in the wheel.  depending on tension, depends on where you are in this curve and what the spoke can allow for lateral flex.  For numerical data on this, look at the tension curves for tensiometers and what reading correlates to what spoke tension.

Now, what I will say that has a BIGGER impact on ride quality for a wheel (i am going to negate tires for this part of the discussion because as John mentions tire choice and pressure is massive), is rim material and construction. At the end of the day a high profile carbon rim will ride way differently than a low profile aluminum rim made out of a soft alloy. One material in and of itself is able to deflect easier than another due to construction geometry and material properties.  This will make the largest difference in ride quality of a wheel.

Finally part of this is taking a wide range of tensions.  20kgf is an un rideable wheel compared to 125kgf.  But John through out numbers in the 60s... that is MASSIVELY lower than any of us would be building wheels at.  Now, that might a random number he threw out there, but I can guarantee you any experience rider on a bermed track will notice the difference between 90kgf and 125kgf when all things are equal to the rest of the bike/conditions.  I can also guarantee that wheel would not hold up for what any consumer would consider appropriate (if alloy...carbon....bigger conversation on specifics).

End of the day, from experience. Yes spoke tension matters to the guys on the sharp end.  But everything is a give/take. and depends on differences.  Hell, elbows in/out depending on braking forces and drivetrain forces will change the stiffness of the wheel (direct from DT Swiss) and doing a mirrored build on the rear vs. a symmetrical build will creat differences that are measurable on a machine. But what durability are you looking for out of a wheel. that is a big part of all of this.

ok enough rambling about this, and sorry for no direct numbers to correlate, just experience in building hundreds of wheels.

What's missing from that test (according to my 5 second skim) is that basically all side loading on the trail will come along with a significant radial load

But agreed on the Sheldon Brown love

This is such a nerdy topic. Spoke tension has been discussed ad nauseam over the last 15-20 years and the discussion has been mostly inconclusive. I appreciate the input, but all I can gather from this conversation is that Gwin likes his wheels a little softer than I do - but there's seems no real fact-based reasoning behind any of this. On a scale from snake oil to physics nobel price the whole spoke tension conversation seems to fall largely under the snake oil category. There's a couple of research papers that suggest that spoke tension matters for grip and there's lots of papers that suggest that the actual range of adjustment in spoke tension for a functional mtb wheel is so miniscule that it creates only neglectable effects on grip and handling in the real world. DT Swiss' research for example falls under the latter category. 

Any links you could share for those research papers or DT Swiss' research? Apparently this topic has been debated ad nauseam for the last 15-20 years but some of us are just showing up to the party.

I'm not a wheel guru and I don't play one on TV...

As a visual learner I really dig this Park Tool video about spoke tension - https://www.youtube.com/watch?v=7vFlKI_Brts

I also like this one that I think I found on the Nox Composite's wheel building info/resource page - https://people.duke.edu/~hpgavin/papers/HPGavin-Wheel-Paper.pdf - it deals primarily with strength differences/non-differences between different lacing patterns but shorter in length than the next two links.

More academic/intensive items I have run across:

I haven't gotten through this one yet since it is about 120 pages with lots of maths - http://arch.library.northwestern.edu/downloads/fb4948624?locale=en

Also this thesis from someone at UC Santa Cruz - https://escholarship.org/content/qt2ng3f65c/qt2ng3f65c.pdf?t=qxq2gw

So is it safe to assume that lacing pattern has an effect on how compliant a wheel is? A 3x is stiffer than a 2x or do I have that backwards?

25lbs of lateral force is very low. The dynamic force from actual riding would be a lot higher. Looser spokes sure feel more flexy when riding to me. But all my experience of this is from the normal loss of tension that happens on the first or second ride on a new machine built wheel, and I don't necessarily notice it till all spokes are pretty loose. Or from trying to eek out a little more use of a rim that's seen heavy use and is in rough shape. In which case spoke tension is often uneven to keep the wobble in check, spokes might be damaged, etc. Not a controlled experiment in short. But it's not just that I feel like higher tension means stiffer wheel, "feels" is pretty susceptible to placebo. I had occasion to ride a 2.5 tire in a frame that really should have only had a 2.4 for about a month. I had to tighten spokes every second or third ride in order to minimize tire rub on chainstays. Hearing tire rub more often backed up the sensation of feeling more flex.

I only feel flex in a certain situations while riding: hitting a diagonal things (ruts especially), landing a jump with the rear angled out a little (as happens when you do a little whip and don't bring it back to completely straight), and occasionally in tight turns. 

One thing about this is that while riding we can feel how stiff or flexy (laterally) the back of the bike is. But when too flexy, we can't tell what is flexing (without changing parts), might be wheel, might be frame, might be a pivot coming loose, might just be tire squirm. Tightening spokes, changing to a burlier rim or frame the only way to know. And if you're a rider who hasn't experienced tighter spokes makes the wheel feel stiffer, then maybe feeling the wheel stiffness is being masked by a flexy frame or low tire pressure. 

And Like Zuest said, it's the interplay of lateral rim stiffness, spoke stiffness (material and geo determined), and spoke tension. A stiffer rim means a load is spread over more spokes, thus spoke tension is less critical. If a rim has less laeral stiffness then a load is applied to fewer spokes and spoke tension has a larger effect on wheel stiffness. 

I bet a donut that when they're running lower tension (at 60kgf or whatever crazy number) a spoke or three is actually going slack quite a bit. So I think there's a real effect. I suspect that the forces a world cup DH racer is putting through the wheels are a lot higher than what Sheldon or zee Germans tested for. 

This would be murder on wheels, every connection point would fatigue in short order, but the top guys choosing to go slack are probably using fresh wheels for every race run anyway.

Is the discussion of Spoke tension only applicable when talking about the same rim?  I have a set of Zipp 3zero Moto wheels and they are known for their compliance (I love them), I also have a set of Enve wheels on my XC bike.  Could a "soft tension" on my Enve wheels be a high tension on my Zipp wheels?  

Ideally yes because rim stiffness also play an important role in the complete wheelset 3 stiffness and strength parameter (vertical,lateral and radial). Without having actual stiffness number between the Enve XC rim and the Zipp 3zero Moto, it's not really possible to say if spoke tension could bridge the stiffness gap between the two rims.

It was either Troy or his mechanic aaron that basically said if the wheel wasn’t near failure at the end of the race run you either didn’t go fast enough or did not have a winning setup.  
 

now I think someone like Aaron or loic probably needs a much tougher set up than Troy or loris.  But I’d imagine they probably need to find the same sweet spot in their personal setup of balancing absolute performance and reaching the end of the track to find the top step.

Not really. Lowering tension will make the enve's less stiff laterally, but radially (up/down) it won't have a noticeable effect on stiffness. I'd avoid going low on tension if I were you, it'll make spokes be slack more often and thus lose more tension, the chances for some getting very uneven tension is and thus putting way more strain on fewer spokes is considerable. Big risk of breaking spokes or rim.

I don't plan on messing with the spoke tension, I actually like the wheels stiff on my XC bike.  It was more of a hypothetical question.  

Ah, gotcha. Theoretically you could make an enve have the same stiffness laterally (within a limited range of force) but radial compliance would be very different. Or you could alter spoke gauge and maybe material to match radial compliance (within a limited range of force) but then lateral stiffness would be very different. And one main thing about the zipps, in the marketing anyhow was roll or twist that's not something that could be changed by spoke parameters. So theoretically: not really. Practically: not at all. 

Yes lacing patern has a noticeable effect on wheel stiffness, the only problem is that due to large discs anything other than 2x or 3x can result in broken spokes since the angle of the force applied to the spoke during drive or braking is too high (imagine a 223m rotor with a radial spoke pattern, the force will be at 90 degrees to the spoke during hard breaking).

So yes, there is a noticeable difference, but in MTB you won't see anything other than 2x or 3x lacing patterns duing to braking forces.

A radial spoked wheel is laterally the stiffest (due to shortest spoke length), same as a wheel with the same rim and spokes and tension will be stiffer if the hub has higher flanges. 

It seems to me that the added compliance in the Zipp Moto wheels is the wide and thin rim twisting at the point of impact. The system of spokes at tension and the center of the rim where they attach is no different. The rim as a whole isn't a wet noodle. It's just able to twist and absorb rim strikes on a side to side basis as needed. 

Anyone have advice on which side they measure spoke tension on first?  Also do you want both sides tension to match?

Thanks!

Doesn't matter which side. One side typically has shorter spokes which will have higher tension (due to a lower bracing angle) than the longer spokes. This is due to offset from a disc brake mount or freehub. 

Offset rims and different flange lengths could actually get you symmetrical spokes, in that case they would likely be very close or the same.

Unfortunately not true. 
 

always take tension from the highest side. 95% of the time this will be drive side rear and disc side front.

Some offset rims will make it even but in reality I have RARELY found them to be exactly even except for front non disc wheels.

other side tension will be what it is. If you are at the correct tension on the tight side (rear drive, front disc) and your wheel is dished correctly (using a dishing gauge) then your less tighter side will be the correct tension. 

If you go the other way you will over tension the wheel (if running max tension) and crack rims (both aluminum and carbon) at the spoke hole.

The most important thing with build a wheel is EVEN spoke tension throughout. I would rather have a wheel 1mm out if it means I can get the tension dead nuts even on ever spoke per side. That will be the strongest wheel for sure. Releasing tension by stressing it multiple times during a build will also give you the most accurate readings and make sure you don’t have any twist in the spokes.

For earlier questions:

@TEAMROBOT  didn’t have the research, was talking with Shane Hensley, their head US race tech. (In regards to stiffness based on elbows in/out). But being a card-carrying Swiss…I have no doubt they tested it every way possible, and as long as it was measurable and repeatable (no matter how small) they would have gone with the strongest, not quickest, way to build a wheel.

Thank you for the advice!

A way to think about it is that you want the perpendicular force on the rim to be even left and right (i.e the force on the rim pulling it left and right across the ground). To acheive this due you HAVE to have different spoke tension on drive side and non-drive side. If you take any hub diagram, because of the disc brake and cassette the bracing angle of spoke to rim is different on each side, therefore to acheive the same force at the rim the spoke tension will have to be different (trigonometry, look it up). 

Again, MTB wheels are relatively boring, but you see a lot of interesting designs in road bikes, my old commuter bike rear wheel by Campagnolo has only 7 spokes on the non drive side (rear wheel) and 14 spokes drive side, because of the difference of bracing angle, the spoke tension is almost identical each side, but the 14 spokes drive side at a much lower angle equal the 7 spokes force at a much higher angle.

Zuestman is 100% correct that you ALWAYS measure the high tension side first (drive side rear, brake side front) to prevent overtensionin and potential cracking.

Some of the riders at world cups have spoke tension so loose you can almost shake them, but for a race run you throw the rim in the bin immediately after anyways.

The one thing i will say, is WC riders aren't the best data points about new tech and testing, they often just want what they are used to because it gives them the confidence to just hammer it.

My response was in the wrong context. I was thinking how measuring one side shouldn't affect the tension on the other side - therefore it doesn't matter. That's not what he was asking so thanks for the correction.

To add one more thought to the topic pre-tension and compression of spokes: on youtube there are a bunch of videos about so-called tensegrity structures: https://www.youtube.com/watch?v=76amOgNzdDQ. here the strings are pre-tensioned just like the spokes. When you put weight on the table the string in the center gets more tension and the 3 stings on the corners get less tension, just like the spokes in a wheel. It is also more intuitive to understand that not the amount of pre-tension determines the stiffness of the structure, but rather the elasticity of the strings. All 4 strings support the weight equally and act like one solid body. When the outer strings lose tension completely, the whole structure collapses. This is why the idea that top lever riders use wheels with low tension seems strange to me.

all 4 strings don't support the weight equally, The force in the middle string is equal and opposite to the force in the 3 outer strings, its slightly hard to vizualise on this setup, with what i'll call a 3 string setup its a bit more obvious. 

 Part of the reason world cup riders use low spoke tension is they want compliance in the wheel to improve tracking, its debatle wether lower spoke tension actually increases compliance, as Team Robot mentioned if the tension on the spokes doesn't  exceed the linear and elastic region of the material, the starting tension doesn't actually matter.

Second is, they check spoke tension a lot, so they can run lower spoke tension and just tighten it after each quali/race run, and check after practice laps, most pople don't check spoke tension each time they ride, and generally, higher tension means the spokes don't "undo" themselves as much, making it more reliable for non pros.

I'd argue spoke type isn't talked about enough, its much more important than tension in my eyes, as it determines the elasticity of the spokes fundamentally. 

"all 4 strings don't support the weight equally". I would argue that they do. The outer strings have each 1/3 of the pre-tension force of the middle sting, I agree here. However, when a weight is put on the table the outer strings lose their pre-tension obviously which requires a force acting in the direction opposite to the direction of the pre-tension. This force supports the weight. Another way to say it would be to state the if the outer strings would not support the weight, the structure would work without them. But it doesn't. All the strings are needed and the work like solid bodies. Pre-tension allows stings to carry compressive load like they would without pre-tension if they were unable to fold by some magical supports. That's actually the mind bending thing about these structures. You can use unstable structures like strings or spokes to carry compressive load using pre-tension. It's awesome 

Finally got around and through the topic, so yeah, late to the party.

My 2 cents (haven't built a lot of wheels, haven't tested things out much, only thought some details through) and I think most of it has been covered as well.

The wheel for sure hangs from the top spokes, no way around it. The spokes we use, the way they are laced, do not give a compressive support, so the bike is carried by the top spokes. What the bottom spokes do give is stability though. Spoke tension shouldn't give much of an impact on wheel stiffness, that is true, as the left and right side spoke are in equilibrium. Moving the rim sideways will require X newton to move sideways. That force will go towards additional tension in one spoke and will also detension the opposite spoke. As long as both are in some tension, the equilibrium will hold and the wheel will be taught. Once the detensioned, the rim will be less supported and only the stiffness of the rim will hold it in place. That's how I see it.

What preload gives you is a wider overlap of the two sides of spokes, meaning you will need more of a deflection to unload one of the spokes and losing tension. It has been mentioned that a properly tensioned, equally and highly enough, gives a strong wheel.

What I wonder (and what Zuestman mentioned) is how the cross section (and material at the end of the day) plays into it all. I always run 2.0-1.8-2.0 spokes as I'm not going to carry around additional, unneeded weight - on a 2.0 mm spoke the inner diameter in the thread has a cross section of even less than 1.8 mm if I'm not mistaken, which means the thread is the weakest point on a straight gauge spoke by itself, without taking into account the stress riser factor of the thread. So a straight-gauge 2.0 mm spoke is not any stronger, is heavier and, as alluded before, stiffer. A less stiff spoke will, at the same preload, give MORE of an overlap side-to-side than a stiffer one. Meaning it will support the wheel in sideloads for longer. It's not that it's stiffer, quite the opposite actually, but it will support the rim better in the edge cases. I kind of think an unsupported wheel is a compliant one

Here are two comparisons of preload and stiffness and the factor is where the lines cross the zero 'force' axis - that's where the spoke will detention:

Preload:

Stiffness:

This is just a theory though and I am thinking about it now as I type and I might find some flaws in it (starting off the typing I was more certain I have a viewpoint than now...). I also haven't tested any of it yet, but I have thought about trying out ~1.5 mm spokes. I doubt I'll notice any difference for what it's worth, but it should be less stiff. And It'd be interesting to see what Industry9's aluminium spokes that thread into the hub do, those might add some more stiffness even in bending and compression due to their much higher cross section. While aluminium is 1/3 as stiff as steel, the cross section at 4 mm diameter is 4-times as big as a diameter of 2 mm, so a 4 mm Al spoke is roooooughly similarly stiff in tension as a 2 mm steel spoke. In bending the added cross section changes things quite a lot though as the second moment of area is a power of 4, not two.
https://en.wikipedia.org/wiki/List_of_second_moments_of_area

Anybody got any spare EX511s and 350s lying around?  

If anyone needs it, I've made a Google Sheet covering the Unior spoke tension tool. I think it could easily be adapted to any other tensioning tool (by inputting the correct values in the Conversion tables sheet): https://docs.google.com/spreadsheets/d/1nfyM_6txzTd7xw_3Fj5-uF9OgFE1iM7…
It's shared publicly, if you intend to use it and modify it, copy it to your Drive. I've made it before Unior had any apps available and I wasn't going to write down hundreds of values by hand and looking at them when I could look at two lines and be done with it.

EDIT: watching the simulation video now. I suppose the spokes were simulated straight, not going one over the other where they overlap? I thought about setting up a simulation but had no idea how to cover the two spokes overlapping and I thought that is a crucial detail that would have to be covered...

I highly recommend using DT Swiss revolution (2.0 - 1.5) if you're running aluminium rims, I've found theyre the best spoke for EX471's. My theory (stolen from other mechanics) is that you should be looking to harmonize the wheel as a system, there is no point having a very stim carbon rim, with thin low tension spokes and exo tyres, you have a stiff rim, and a compliant tyre and compliant spokes, meaning when stressed laterally, the spokes won't work in unison, and the tyre is just going to roll over the rim.

Similarly, don't use "flexy" rims like a 471 (they are quite a bit flexier than 511's and the 541's) with 2.0m straight gauge spokes, you are simply going to lose the purpose of the compliant rim and dent because the spokes are forcing it to stay true.

I've been using 471's and revolution spokes for my wheels for a while and they are super solid, a few dents after a season of riding, but little to no punctures, heaps of greap and compliance when pushed hard.

The alternative is carbon rims that are way way stronger with a stiff wheel build where you don't need compliance to avoid denting and burping tyres because you run DH with cushcore etc. 

Which tyres do you run on EX471s?