(SHOWN: Readout graph for Park Tool’s online wheel-tensioning app.)

Three of the Biggest and Persistent Myths in Wheelbuilding — including in Handbuilding

By Jake Brennand

In this article I want to expose some facades, misperceptions, and downright falsehoods (deliberate and otherwise) in the wheelbuilding space, including in the elite domain of high-end handbuilding. Cycling mechanics is a funny business. Its practitioners — as I’ve commented more than once before — can be truly great and dedicated individuals, but some are also testy, difficult, egotistical, and even rude sometimes, toward both colleagues and even more disturbingly toward customers. The “angry shop mechanic” is a real and highly regrettable thing. And like many skilled tradesmen (and to a lesser extent, tradeswomen) of a certain old-school mindset, some traditional bike mechanics are loath to evolve their understanding or share their information, which comes at the expense of an improved learning dialectic and better knowledge. What this attitude results in with wheels is a lot of salty conjecture online; hand-me-down supposed “best practices” in person (when these practices are, rather begrudgingly, handed down); and too often negative resistance to newer or empirically sounder information. If you’re blockheaded and just want to fit in with the stereotypical bike-shop crowd or at the bike park, then you probably won’t enjoy this article — and almost certainly you wouldn’t be a good fit with our brand. All is well, that’s fine. But if you’re fact-driven, open-minded, and you enjoy lots of technical detail, then welcome aboard — it would be our pleasure to help guide and realize your next custom wheelset. And as always, anyone reading out of pure curiosity is also more than welcome in this Hogtown Spokes wheelbuilding blog space.

Persistent Myth 1: Spoke length should be focused on with the precision of a board-certified anesthesiologist

Don’t get me wrong — it should be, but only until it shouldn’t be. Every single professional hand or even machine-builder aims for perfect spoke length as the goal, each and every time that they build a wheel. We certainly do at Hogtown Spokes. Because why would one do any less for a paying client? But, at the same time, in the real world there’s best-laid plans and then there’s dynamic reality. Parts can cooperate (or un-cooperate) unexpectedly. Hubs on the stand may require exceptional dishing efforts. Spoke flanges can react differently than expected with spokes, affecting the required length, or flange washers may be needed, also affecting length. Or perhaps manufacturer-supplied hub dimensions and rim ERD (Effective Rim Diameter) might be incorrectly listed (in the case of parts where you need to order or cut spokes for the client before measuring these data points by hand, for yourself). In the real world of wheels, the reality is that while perfect spoke length is always the goal and can certainly add to certain wheelsets’ durability, sometimes spoke lengths need to be less than perfect and length is not always the uncompromising longevity factor that some traditional builders and shop guys absolutely insist it to be. These guys hone in on spoke length, for whatever reason, with an almost fetish-like fervour; it becomes a cultish talking point.

As wheelbuilder and researcher Bill Mould and others on the handbuilding side of the industry have pointed out, if needing to be less than ideal spoke length should always be too long rather than too short — too short truly is unacceptable, even if the discrepancy is merely 0.5–1 mm from perfection. This is because spokes that are too short, whether attached using either Double-square/Squorx-head (collectively, extended-head, internally tensioning) nipples or conventional 12 and 14 mm nipples, will put undue stress on the heads of both brass and alloy nipple structures, leading to the highly stressed head section eventually being leveraged off against the rim under repetitive stress. In too-short builds, the unsupported gapped area between where the spoke stops inside a nipple and where the rim line/nipple head begins becomes a flashpoint for stress risers (see FIGURE 1).

FIGURE 1 — In the case, certainly, of conventional-headed nipples, what going too short does for the prognosis of nipple, spoke, and therefore of wheel. Too-short simply doesn’t last over the intermediate and long-terms. It’s a guarantied route to sacrificing wheel durability.

By contrast, spokes that are too long will merely pop out of the end of the nipple, and the nipple threads will take up this excess by starting to grip onto the unthreaded section of the spoke, below the nipple’s four-sided bore area (the externally visible and tool-ready portion; FIGURE 2). This, in turn, won’t weaken the spoke, because the threading attempts are simply equivalent to a really meek effort at putting new threads on the spoke — amounting to no more than microns of mere finish removed. (Spoke threads, as an aside, are always rolled instead of being cut; rolling is the soundest method that doesn’t weaken the steel substrate unduly by taking off a quite noticeable amount of material on such a small cross-section.) The friction thus created between the unthreaded section of the spoke and the nipple threads also becomes a new source of adhesion for the spoke against vibration-caused loosening while riding — kind of becoming a mechanical version of Loctite — and further adding to the durability of the spoke and therefore of the wheel. The only real drawback to going too long is that it looks untidy (if the spoke is more than, say, 1–2 mm maximum too long), and also that if the spokes haven’t yet reached full tension then excessively long spokes can become unusable in the finishing of a build. But the latter is extremely rare for professional or skilled builders, and the former can be addressed with unscrewing and hand-cutting a lengthy spoke to the needed length or by grinding or cutting a protruding spoke in a finished build in the case of single-wall rims (cruisers, fatties, moto-style MTBs), where the excess will be easily reachable with a cutting tool or file. There are solutions, friends.

FIGURE 2 — A spoke that’s too long will simply pop out of the top of the nipple, as an excess with no real consequences for the integrity of the build (beyond cosmetically). You can safely go up to, say, 2.5 mm too long, if needed in contingency situations (though you obviously don’t want to have to require this leeway). Note the circled attempt at new thread formation, in silver against the black oxide finish of the black spoke. These markings won’t affect the spoke, since they don’t abrade in use against anything (if they even abrade at all) except for the unthreaded and smooth bore area. Furthermore, these marking are simply finish, in most cases imperceptible in depth to the human finger in a blind touch-test. The markings are also protected against the elements (rain, debris, muck, etc) by the bore and tightness of the nipple fit. Too long isn’t ideal — but mechanically it’s more than acceptable. There’s no substantive drawbacks, unlike being too short, which is a big red flag in handbuilding.

The reality is that, as every materials-science student learns, the basic engineering shorthand for thread safety dictates that you need a minimum length of 150% (factor of 1.5) more than the base diameter of the fastened object, when using steel fasteners. So, in the case of a spoke nipple, with a roughly 2 mm-wide threading diameter, this means that in theory a builder only requires 3+ mm of threads to achieve mechanical locking. But, being safe and adding an extra 50% to the base calculation to account for the use of brass or alloy nipples, which are softer than steel, plus another 50% in order to address the thicker fasteners that this thumb-rule calculation is typically based upon, plus a final two-thirds safety margin for vibration in usage, this works out to a minimum of roughly 7.50 mm of threads needed for reliability. The complete math is as follows:

Thread Length Required (ThrL) = 2 mm x (150% + 50% +50% [or 2.50]) / 0.66 = 7.58 mm

And, honestly, this is a conservative estimate of what’s needed for security. When one factors in that most spokes contain 9–10 mm of total threading and 20 or more full threads, this gives a builder plenty of extra steel to lop off if needing to settle for “too long” with a particular problematic build. Note that since the thread pitch (distance between threads) of most spokes is approximately 0.45 mm, 7.5 mm of threads computes to just under 17 threads. Again, this is a really conservative calculation for what’s needed.

Contrary to what some of the angrier voices online may claim, occasionally and pragmatically going too long doesn’t make a builder a “hack.” It makes someone solution-focused and practical. Such “hypercriticism,” as an old professor of mine at the University of Toronto aptly labelled his more pedantic critics, is both petty and misunderstands wheel construction. There are far more important issues — such as perfect dish and balanced tension on each individual wheel side — for the handbuilder to fetishize.

Persistent Myth 2: Rim max tension should never be neared

This claim is simply wrong, and it takes away potential strength and durability from wheelbuilds. Unskilled or inexperienced builders — and even some long-time builders building for customers — act like the published rim maximum tension figure should never be neared, as if it’s some sort of yellow traffic light on the warning path to wheel failure and bodily injury. This is ridiculous. What skilled handbuilders understand, from experience and testing, is that the published rim maximum is much more a target to achieve than a prohibition to be avoided. It’s also, arguably, a quite arbitrary figure from the rim companies.

I don’t have direct evidence, but I strongly suspect that even in the case of no-nonsense, engineering-driven companies like DT Swiss this is the one area where science takes a back seat to the legal, insurance, and financial (combined corporate-risk) powers-that-be.

Let’s look at the circumstantial evidence for my theory. If it truly were the case that rim max tension is based upon scientific evidence alone, namely on yield and fatigue stresses in the hoops, then why is it that DT Swiss uses the same number for all of their rims? DT rims make use of the industry-standard strength-for-use designation known as an ASTM rating (formerly American Society for Testing and Materials, now ASTM International). DT gravel and road rims, for example, often come with an ASTM 2 designation, meaning moderate, predominantly road riding without jumps or obstacles. Surely, if science was leading, then these road and gravel rims would not be deemed capable of the same spoke-tensioning forces as their freeride and downhill cousin products. And yet, oddly, the opposite is asserted. DT is an exceptionally good rim company — we’re huge fans at Hogtown — but they bizarrely list the ASTM 2 road and gravel rims as capable of the same max rim tension, 120 KgF, as the burly and highly reinforced ASTM 5 freeride product or only slightly less hardy EX enduro series rims, rated at ASTM 4.

Hmm…

This makes no logical sense. It seems exceptionally likely that some initial lab testing probably went into arriving at this max number of 120 KgF (expressing a measurement called “kilogram force”/1200 Newtons) for DT rims, with the testing likely having been conducted on older road or perhaps cross-country mountain rims. However, perhaps then out of an abundance of enterprise caution, it would appear that this figure has never been iterated and updated with additional findings in the time since. It seems, on the whole, much more like a legal and financial call than an empirical scientific one; DT, quite understandably, doesn’t want to encourage foolish abandon and invite lawsuits from Joe and Jill Reckless in the cycling public.

But there’s just no way — putting aside corporate concerns — that a rim that’s substantively thicker and stiffer, in the case of the aggressive mountain rims, cannot withstand more tension than a spindly road rim. What this means, at a minimum, using DT Swiss as the case study, is that maximum rim tension is a fairly arbitrary figure with OEM rim makers — more specifically, it’s likely intentionally conservative and safe. This conclusion, in consequence, lends huge additional support to the idea that the rim max tension isn’t some weak flashpoint of failure in a wheel, but rather is a hardy target to aim at when wheelbuilding competently. This is because of the physics considerations always necessary when building up and correctly spacing out wheels, as I will explain in more detail below.

Does this make much sense to you? The comparatively spindlier XR331 road-gravel rim beside the much burlier ASTM 4-rated EX471 enduro and DH rim (used here, exceptionally, on the front of a gravel build, for more tire width). Both have identical rim max tension numbers, at 120 KgF/1200 Newtons, despite very different strength constructions and use intentions. If nothing else, it shows that tension max is basically an arbitrary and conservative figure and therefore offers additional, circumstantial evidence that the max is not something to fear. If it was, then you would see highly tailored prohibitive numbers on the sides of rims and in product packaging, truly data-based and specific to each rim design. If the EX471 really needed 120 KgF as its max rating in order to survive constant riding, then the softer, lighter, weaker XR331 would logically need, say, 100 KgF as the maximum rating. The math simply doesn’t add up.

What skilled builders understand is that the actual max tension figure for a rim tends to be much higher than what’s listed. Even under the fatigue forces of continuous riding, it’s probably north of 150–170 KgF (**but note that this is not a permission article for readers looking to go that high, not at all**). What is more, spokes are routinely subjected — and subjecting of rims — to tensions much higher than even these lofty estimates, for example during the critical de-stressing step in a hand wheelbuild, when spoke pairs can temporarily reach an estimated 200 KgF (2,000 Newtons) as part of stretching and conditioning the steel. Moreover, many carbon rims now list as a factory and warranty touchpoint a max suggested tension as high as 135–140 KgF — and this is with carbon.

But why go near the maximum from a mechanical perspective? Well, it’s frequently the case that “dishing,” or centering, of a wheel requires a builder to increase tension to the max or even slightly above the max at times so as to ensure a centred, flush wheel whose “low” flanges (Driveside [DS] Front and Non-driveside [NDS] Rear) are still high enough in tension to remain durable under long-term use. If a builder were to dish, a seminally important build step for any wheel, guided by a total aversion to reaching high or max tension it would often result in a centered wheel but one with unacceptably low average tensions on the low flanges — in other words, privileging staying below the rim max would end up ruining the strength integrity of the finished product overall. As pro handbuilders know all too well, it’s a wheelset’s low flanges where wheels are often vulnerable to losing spokes in use, more specifically the NDS rear in mountain wheels. This is most commonly the problem side with wonky offroad wheels.

Under-tensioned NDS Rear (or DS Front) spokes vibration-loosen over time. They can also cause rubbing abrasion at spoke junctions and stretch wear to a hub flange’s spoke holes. Most concerningly — and scientifically — of all, though, as Bill Mould has shown on YouTube, low-tension spokes have a greater range of cyclic (high to low) tension change during a revolution of a weighted wheel. This likely means that they are mathematically subject to greater fatigue forces over time. Singly or collectively, these vulnerabilities, over time, can result in premature wheel failure as they progressively get worse, eventually de-stabilizing a wheel in motion (this is when they break, and often dominoes-style). Thus, skilled handbuilders understand that it’s actually low-tension spokes that are more risky and even dangerous in dynamic application. Low-tension spokes are the true “tightrope wires” of the wheel world. Therefore, reaching max tension is often quite desirable.

At Hogtown Spokes, we won’t release finished wheels with low flanges under 70–75 KgF on average (700–750 Newtons), unless it’s absolutely necessary and defensible (and it rarely is). Skilled builders therefore aim for the max rim number, if dictated by the physics of dishing— we don’t avoid it. In DT Swiss’s own words, offered recently in a marketing release on Pinkbike for their latest freeride hoop, “the key to a well built wheel that lasts is on building it as close as possible to the maximum given spoke tension, something that is even printed on the rim stickers, while keeping the deviation of tensions to a minimum.”

True that, DT.

Even so, unfortunately, the myth continues to exist that max tension is to be avoided like the plague. It just isn’t — this is rubbish, junk science at its finest in the wheel space. “Deliberate low-tension wheels,” a pernicious corollary to the myth about max tension, are sometimes irresponsibly advocated by shop builders, World Cup DH mechanics (I won’t name names), and old-school roadie tinkerers. These are wheels that will last for race runs and not much more beyond that. Professional handbuilders, by contrast, build for durability as a principle concern (maybe the principle concern in hand assembly), which is why we build to a high tension figure consistently and avowedly. The place to seek different ride feels and characteristics, as I always tell folks and have said publicly many times before, is in parts selection: in spoke counts and types, in tires, rim inserts, and rim models. It’s never in the amount of tension one applies to a build. Tension should always be high, period. Full stop. Builders (like ’80s karate champions) who “aim low” are building irresponsibly and unprofessionally and should be avoided by paying clienteles.

When spokes poke through crumbling rim holes (we’ve all seen the horrifying failure pictures online), it’s extremely likely (barring crazy high spoke tensions being used) that either defective alloy in the rim or, even more likely, failure to use rim washers or failure to use butted spokes is responsible. The blame should not be laid at the doorstep of high or rim-maximum-reaching tension numbers.

Myth 3 — “Alloy nipples don’t last”

This final myth is one that generates endless forum threads in the bike world. It’s a topic riddled with personal biases and specious conclusions. Most forum threads involve someone railing against alloy and in favour of brass nipples, after finding corrosion inside their spoke nipples, whether it has led to a broken nipple/spoke/wheel or not.

In reality, alloy can corrode faster than brass — but only, inasmuch as the nipple itself is solely responsible for the failure, in very specific cases of neglect, wear, poor matching to rider, poor workmanship, or nipple-design choice. Riders and builders who tape their tubeless rims poorly, and thereby allow sealant to reach the nipples, may experience alloy corrosion faster, but only if using sealant containing ammonia. (We aim to avoid this with all builds; much of the available sealant in the industry is now ammonia-free.) Riders who build up a wheelset that places alloy nipples directly against carbon rim structures may see increased galvanic corrosion — but only if building without buffering rim washers, which also increase rim strength by spreading loads. Carbon rims should therefore always be built up with washers! Riders who never maintain their wheels may find alloy problematic (but these are customers to be avoided, in my experience, and they may be folks who even void any applicable wheel warranties…and fairly so). Alloy nipples may deform at the flats under high finishing tensions, if the builder is using — against all professional best-practice advice — a 2-or 3-sided spoke key, as opposed to the more supportive and always advisable 4-sided key (which we employ for all builds and wheel servicing at Hogtown Spokes). And riders who build with conventional 12 or 14 mm alloy nipples, versus the effective “rebar-reinforced” nipples of cycling, Double-square or Squorx-head (see FIGURE 3), may find alloy to be a wear issue, too. But doing so is, again, not typically recommended by us. The rider/builder who won’t spend a little extra for DS or Squorx-head nipples versus standard alloy, or who’s clearly driven by extreme “weight weenie” considerations in making this specification choice, has some frankly questionable mechanical judgment.

FIGURE 3 - Extended-head, internally tensioning nipples, the most structurally sound products in the industry. These designs extend the spoke further up into the inside of the rim, and they require less time for a nipple to be externally tensioned by a spoke key, helping to preserve the finish on the outside flats. From left to right: Sapim DS Alloy (16 mm), DT Squorx-head Alloy (15 mm), Sapim DS Brass (16 mm), DT Squorx-head Brass (15 mm total length). These are the vanguard products of nipple technology.

The bottom line is that when alloy fails, it’s not because of the nipple material (often aerospace-grade anodized 7075 aluminum alloy), in probably a significant majority of cases. Rather, it’s most likely due to questionable specification choices overall from the assembler or other, undiagnosed (or unadmitted) weaknesses in the wheelbuild or care attention by the rider. And yet these failings are crudely and routinely forum-dumptrucked as being the fault of alloy. This is roundly unfair and off the mark. DT Swiss and Sapim market these products as essentially their flagship nipples for good reason. This is based on extensive testing and lab time. At Hogtown Spokes, we actually prefer to build with alloy for most builds, for a number of reasons, viewing Sapim’s 16 mm Double-square Alloy nipple as the best overall nipple product in the industry.

For fun and learning purposes for our clientele, I recently conducted a “lab” experiment in my workshop. I subjected both Sapim Alloy and Brass Double-square nipples to nearly a full week of submersion in highly acidic conditions NEVER recommended for any wheels: pure vinegar and saline (sea salt). I picked a new brass nipple and a previously installed alloy one for this experiment, to really try and lean in to the zeitgeist of the online rumour factory about exposed alloy and its supposed risks.

But, iconoclastically and yet unsurprisingly for me, the results showed that alloy punches way above its rumoured weight. The brass nipple held up fine, as expected, minus some subtle de-plating of the nickel coating that lends so much to its purported and actual durability as a material. The underlying brass was showing through in micro spots. The brass nipple also developed a white, calcium-like oxide coating to protect itself against further wear. But the alloy nipple did nearly equally as well (remember that alloy is not recommended for acidic conditions, let alone extreme acidic lab conditions). The flats and most of the structure looked just fine after six days’ acid bath, largely resisting the de-anodizing that can invite and accelerate corrosion. The only “problem” spots were the areas just under the nipple head, where a blackish ring developed, and on the tool-ready internal square portion of the nipple, above the head. Tiny bits of black also developed sporadically on the larger structure. But the blacking anywhere was merely an advanced formation of chemical oxide — essentially the aluminum fortifying itself against further corrosion. Aluminum oxide is like rust for steel or iron in that respect, but essentially without adverse consequences. Rust is unstable, flaky, and weak. Aluminum oxide forms the business end of sandpaper, among other notable industrial uses!

On the whole, my informal but thorough experiment gives the lie to alloy as a weak-link nipple material. Provided one avoids extreme acidic conditions (and apparently even in extreme conditions there’s pause for optimism, though I will never officially sanction this), alloy nipples will serve most riders handedly. See below for the final visual results of my experiment.

Alloy right, brass left. Note the black ring and spots of aluminum oxide, the only real consequence for the alloy nipple of a week’s acidic submersion in this lab bath. Aluminum oxide is a hard, self-protecting chemical reaction. It’s like rust on serious, Jose Canseco-in-the-early-2000s style steroids. Note the oxidation, too, with the brass nipple (white, calcium-like film). Brass is hardly oxidation-averse.

Ultimately, both brass and alloy are fine nipple materials, for the right context and in the right rider cases — careful selection of which is part of the handbuilt process and difference. We prefer to build with alloy, though, for many builds, given its corrosion-resistance, advanced design options, anodized colour options, and ability to achieve high tensions with ease and quiet (brass squeals more). Categorically, neither material should ever be dismissed as never well suited. Unmistakably, alloy’s “never-use,” online hate squad seem to be using a sound material essentially as a scapegoat for other sources of nipple and spoke failure or consumer negligence — whether realizing so or not. A truly professional handbuilder will never rule out using 7075 anodized alloy for durable wheels and will often sing its many praises.

As always, be critical about what you read, especially in the dubious real estate of some online bike forums. Privilege science and experience.

-Jake Brennand