Home » Technical part » Wheelbuilding » Bicycle wheel building basics

Bicycle wheel building basics

Updated: 22/10/2020.

In this post I’ll explain the basic terms and provide basic information important for bicycle wheel building.
I’ll be using Amazon affiliate links and earn from qualified purchases as an Amazon associate (why Bike Gremlin uses Amazon affiliate).

Contents:

  1. Introduction to bicycle wheel building
  2. Choosing good quality components
    2.1. Hubs
    2.2. Rims
    2.3. Spokes
    2.4. Nipples
  3. Spoke lacing pattern
    3.1. How cross laced wheels carry load
    3.2. The optimal number of crosses
    3.3. How radially laced wheels carry load
  4. Why optimal spoke length is important
  5. Calculating optimal spoke length
  6. The importance of achieving even spoke tension
  7. Acceptable tolerances when building a bicycle wheel
    7.1. Radial trueness
    7.2. Lateral trueness
    7.3. Rim dish trueness
    7.4. Acceptable trueness, and spoke tension tolerances
  8. Sources – highly recommended


1. Introduction to bicycle wheel building

I explained the basics of how bicycle wheels carry load and what is needed for them to be strong (the choice of good quality components) and durable in the article: Bicycle wheels – how many spokes? I also made a series of YouTube videos on how to build bicycle wheels. In the videos I explained and demonstrated most things, but I like having it all in text format as well, for easier searching and reference – it’s often faster and I like it. 🙂

In this post I’ll write about the technical part, that is – the wheel building process. I’ll also briefly recapitulate the information about the quality of components chosen.


2. Choosing good quality components

If one is replacing only a rim, or a hub, the component choice is somewhat “easier”, simpler, since the other components do limit the available choices. So for this text, I’ll start with an assumption that a wheel is being built from scratch.

I also made a YouTube video where I explain the types, quality and choice of bicycle wheel components.


2.1. Hubs

I’d start with the choice of hubs. Depending on whether a wheel is built for a road bike, or for a MTB, whether it is the front, or the rear wheel, whether one uses disc, or rim brakes – hub choice will differ. The choice of hubs depends on the frame as well – whether it accepts quick release, or requires a through axle.

Then, if it is a rear hub, it is important to know the width of frame’s rear dropouts. In a separate post I wrote about the rear hub width standards. This will further narrow down the hub choice.

Another thing to determine is whether the bicycle frame is made for rim, or disc brakes (the linked article discusses pros and cons of each). Since there are hubs for disc brakes. Disc brake hubs will also work with rim brakes, but are (needlessly, if using rim brakes) more expensive and heavier.

For rear hubs, it is also important to note how many sprockets are used and which make of shifters and derailleurs are on the bicycle (Shimano, SRAM, or Campagnolo notably). I explained this in the article about rear hub compatibility.

Within all these parameters, that are determined by the type of bicycle the wheel is being built for, it is left to choose how many spokes should the wheel have. In terms of easier sourcing of good quality, reasonably priced rims, and for more durability, I would recommend 32, or 36 spokes. The number of spoke holes in the chosen hub will affect the number of spoke holes the matching rim should have.

Here I must note that there are some “exotic” hubs that require “exotic” spokes – that, in turn, sometimes require “exotic” rims as well. Here’s an example:

Exotic hub - with straightpull spokes
Exotic hub – with straightpull spokes
Picture 1
Exotic spokes are harder to come by, and they have an exotic price tag as well
Exotic spokes are harder to come by, and they usually have an exotic price tag as well
Picture 2
Standard hub for standard ("J" bend) spokes
Standard hub for standard (“J” bend) spokes
Picture 3


2.2. Rims

For rims it is important that they have the same number of spoke holes as the chosen hubs (and vice-versa).

I think that double walled aluminium rims are among the stronger, more durable, yet not too expensive, nor too heavy.


2.3. Spokes

Inexpensive and good quality solution are the standard, with a “J” bend at the elbow, 2 mm thick, preferably thinned down in the mid-section to 1.8 mm. Such spokes are also called “swaged” (some spoke manufacturers, like DT Swiss, refer to such spokes as “double butted“, while the ones that have wider elbow part compared to the nipple thread part, with still having narrowed down mid section are referred to as “triple butted“).

Narrower spokes are a bit lighter, “more aero” (the modern buzz-word), but are less durable.

2 mm swaged spokes thinned down to 1.5 mm twist a lot when they are tightened, making the wheel build very difficult, without providing much benefits.

Thinned down mid-section of swaged spokes acts as a kind of a shock absorber, being more flexible, thus relieving some stress from the spoke ends – which are the critical parts where spokes fail from material fatigue most often. Perhaps a bit counter-intuitive, but that’s how it is. It should also be noted that this thinning down is done at the factory, by rolling, not by grinding, so that the material structure is kept in tact, without creating a weak spot prone to material fatigue.

With spokes, another important aspect is their length, but more on that later.

I’ve had great results with Sapim and DT Swiss spokes (Amazon affiliate links).

Note: modern spokes from most manufacturers are made with a head angle of 90 degrees, or even more obtuse (95 and similar).

Problem with spoke head ("J-bend") angle being too obtuse
Problem with spoke head (“J-bend”) angle being too obtuse
Picture 5

Picture 5 shows a spoke with J-bend (elbow) angle of 94 degrees. This prevents it from getting hooked to spoke flanges using the strong part of the spoke (3), putting more pressure on the head (2), the part of a spoke where there already is a lot of stress concentration from manufacturing, and less material, making spoke heads more prone to breaking off and spokes comming loose.

This design makes spoke (wheel) lacing faster and easier, but makes spokes effectively weaker. Sapim makes these angles 95 degrees, while DT Swiss is better (less bad) with 90 degrees. Still, these are the best (or least bad) spokes that can be bought nowadays.


2.4. Nipples

Nipples should match the spokes. Spokes that are 2 mm wide at the threaded end will need nipples with 2 mm wide threads as well. Note that 2 mm nipples will fit 1.8 mm wide spokes, but the fit will not be strong enough and threads could get stripped under force. 1.8 mm wide nipples, on the other hand, will not fit 2 mm wide spokes, of course.

Nipple length of 12 mm is best in practice, unless the rim is exotic and needs longer nipples. Wheels are built most quickly with these and the interface is strong and durable. There are longer nipples available – notably 14 and 16 mm.

As for the material, there are aluminium and brass nipples. Aluminium ones are lighter, but also less strong and more prone to galvanic corrosion (Wikipedia article), so I avoid them.


3. Spoke lacing pattern

In this video, at 2:48, I explained the spoke lacing patterns (link should open YouTube, starting at 2:48). But here is the explanation in textual format.

Exotics aside, there are two ways of lacing bicycle wheels: radially, and crossed. Before more detailed explanation, a couple of pictures that speak more than words:

Radially laced wheel
Radially laced wheel
Picture 6


Radially laced wheels have spokes going straight from the hub to the nearest rim spoke hole.

4 across laced wheel
4 across laced wheel
Picture 7


Cross laced wheels have spokes going at an angle, to a spoke hole in the rim, crossing over other spokes on its way. The number of crosses can be smaller, or greater. More on that later, but our wheel has 4-cross lacing pattern.

The number of spoke crosses
The number of spoke crosses
Picture 8


When counting the number of crosses, people often miss the first crossed, adjacent spoke (painted in blue in picture 8).

All else being equal, cross laced wheels can take a lot more load than radially laced ones. Radial lacing is practically only used for show (and minimal weight savings since they require shorter spokes). I’ll explain why.


3.1. How cross laced wheels carry load

There are two main types of wheel loads. One is the torque, where hub tries to rotate on its axle, relative to the rim (in case of pedalling), or, when braking using disc brakes, the rim tries to rotate relative to the hub.

Picture 9 shows an example of pedalling force – wheel is depicted from the left hand side. When braking using disc brakes, the force acts in the opposite direction, by ground “pushing” against the rim.

Affect of pedalling force and how cross-laced spokes carry that load to the rim
Affect of pedalling force and how cross-laced spokes carry that load to the rim
Picture 9


Spokes being pulled (trailing spokes) get tightened, while the ones laced in the opposite direction (leading spokes) get a bit less tight (loosened). Number of crosses can be smaller, or greater.

The part where the trailing and leading spokes cross, is the place where the further tightened trailing spokes push stronger against the leading spokes, thus reducing the amount of loosening that trailing spokes face – which is one of the benefits of the crossed lacing pattern.

The other force faced is lateral movement of the rim, relative to the hub – usually by ground pushing against the rim, over the tyre, of course.

Lateral rim load
Lateral rim load
Picture 10


For an example, look at picture 10 and imagine the force acting on the right hand side of the rim, in the direction from the viewer, towards the blue door in the background. The spokes near the point where the force is used will be mostly affected. The ones closer to the door will become loosened (since the rim is moving towards them), while the ones closer to the viewer will be tightened.

To better resist such forces, spokes are laced at an angle relative to the hub, so that the hub’s flanges can be used as a lever, to oppose this force.

Hub flange acting as a lever (L)
Hub flange acting as a lever (L)
Picture 11


The greater the spoke angle (i.e. the more spoke crosses there are), the better will a wheel cope with lateral loads. Imagine if the red spoke in picture 11 were going all the way across the head of the adjacent spoke (painted blue and marked with a red arrow). That would make an even longer lever. However, that would make replacing a broken spoke difficult, because the access would be blocked. Also, head of the spoke would create a bend, kink, in the spoke going over it, making it weaker. That is why it is important to choose the optimal number of crosses. Picture 11 shows just a bit of overlap – borderline acceptable amount.


3.2. The optimal number of crosses

The number of crosses should be such that the maximum angle of spoke relative to the rim is achieved, but without the spoke going over the head of the adjacent spoke.

The actual number of crosses that achieves this mostly depends on the total number of spokes (and on the hub flange and rim diameter, to a lesser extent). How so?

A 36 spoked wheel has each rim spoke hole at 10 degrees angle (360/36), while each hub spoke hole on each side is at twice that angle (since half spokes go from one side, the other half from the opposite hub side).

If we took a 20 spoked wheel, each rim spoke hole would be at an 18 degree angle, with hub flange holes being at 36 degrees from each other.

Spokes go in groups by 4:

Spoke pairs by 4
Spoke pairs by 4
Picture 12


First come two trailing (or leading) spokes – one from each hub side.
Then two leading (or trailing) spokes – one from each hub side.
And so on – no matter how many spokes there are in total, and regardless of the number of crosses – this 4 group lacing pattern always exists.

Crossed spokes are 1 and 3, as well as 2 and 4.

To calculate spoke angle, one must multiply the number of crosses, by the hub flange angle. For a 36 spoked wheel, that would be 20 degrees multiplied by the number of crossings.

But how to tell when it is too much and when there will be spoke head overlap? For most rims and hubs, angle below 75 degrees does not lead to any spoke head overlap.

Simple, practical way of determining the optimal number of spoke crosses:
Divide the total number of spokes by 9, then round down to the first whole number.
Exceptions are 36 spoked wheels where with 4 across (36/9) you will often get some overlap, like shown in picture 11. And 16 spoked wheels, where you can get away with 2 across (because hub flange holes are at a great angle and distance.

To make things easier, here’s a table:

Spoke countOptimal
number of crosses
161 x (2 x sometimes)
202 x
242 x
283 x
323 x
363 x (4x sometimes)
404 x
484 x
Table 1


3.3. How radially laced wheels carry load

Poorly!

With radially laced wheels, there is some (not naked eye visible) hub twisting, giving a poor copy of cross laced spoke pattern for load bearing. In the picture below, I drew the spoke position, when they are stretched (in the direction depending on the force direction). It is exaggerated in order to demonstrate (and I didn’t re-draw the whole rim as it moves with the spokes).

Torsion force on a radially laced wheel
Torsion force on a radially laced wheel
Picture 13


There’s a lot more twisting and friction, which can break spokes. That is why only front wheels for rim brake bicycles are sometimes laced this way (and non-drive side of some exotic rear wheels).

Another downside of this lacing pattern is that it puts a lot of stress on the hub, with spokes trying to rip it apart. Note in the picture below how cross laced wheel spoke tension practically tries to compress the hub flange material, while radially laced wheel’s spokes are trying to pull it outwards, rip it apart:

Radially, and cross laced wheel spoke tension force direction
Radially, and cross laced wheel spoke tension force direction
Picture 14


4. Why optimal spoke length is important

Ideally, spoke length should be such that the entire length of nipple’s threads is used for carrying the load. Besides, on rims that are not double walled, spokes that are too long will protrude from the nipples and could puncture the tube. I cut a nipple in half to demonstrate this. Marked nipple and spoke thread start and finish, since it’s not clear in every picture.

Picture (cross section) of spoke and nipple threads
Picture (cross section) of spoke and nipple threads
Picture 15
How it looks when a chosen spoke length is too short (A), too long (B), and optimal (C)
How it looks when a chosen spoke length is too short (A), too long (B), and optimal (C)
Picture 16


5. Calculating optimal spoke length

In order to calculate the optimal spoke length, the following measurements are needed:

I explained the number of crosses in this post, while for other terms I’ll be adding links as I publish relevant articles.

Finally, all the measures should be entered in a spoke length calculator (link is towards the one I mostly use).


6. The importance of achieving even spoke tension

When spokes don’t have an even tension, i.e. when some are significantly tighter, and others much looser, what happens?

  • Spokes that are tighter will carry more load than the looser spokes. Hence being more stressed, and also putting more stress on the spots where they are connected to the hub (at the flange) and the rim (where the spoke nipple is).
  • Looser spokes will be even looser as the wheel is ridden – when they are at the bottom (during every wheel revolution), and when the wheel is loaded in such way that they get lower tension (side loads from the opposite side for example). This will result in them, since they are already loose(r), having more (micro) bending and they will be more likely to break at the elbow, due to stress (fatigue).

With rear hubs, and the front ones on disc wheels, there is a caveat:

"Left" and "right" spoke tension
“Left” and “right” spoke tension
Picture 17


Right hand side spokes on the above noted hubs, have higher tension than the left hand side spokes, because their angle is sharper and the rim wouldn’t be centred otherwise. However, tension of right hand side spokes should be as even as possible, when compared to the other spokes on the same side (right hand side). Same goes for the left hand side spokes – they should have an uniform tension when compared among them.

When building a wheel and choosing between 100% true rim, and more even spoke tension, go for the even tension. Wheel built that way may not be 100% straight from the start, but it will remain as it was even after 20.000 kilometres (unless a rim is dented/bent by hitting a rock). On the other hand, if a wheel is built to be 100% true, but with uneven spoke tension, it will come out of true relatively quickly, and spokes are more likely to break on such a wheel.

With another caveat: the rim should be true within the acceptable tolerances – as explained in the next chapter. That is: 100% uniform spoke tension is not good either, if it results with the rim being too much out of true.

How to measure and compare spoke tension?
Guitar pluck can be used to pluck the spokes and get the tension surprisingly uniform by comparing sound uniformity – like plucking guitar strings.
To me it is a lot faster and easier using a proper spoke tensiometer. I’m not very musical, plus plucking the spokes does not give the information of how high the tension is (only how uniform it is) – which is often important (70 kilopond is the minimum for most wheels). Most tensiometers are not very precise. Exception is this patent (designed by late Jobst Brandt), sold by Wheel Fanatyk (their tensiometer). Or a more affordable variant, made by our Macedonian brother Filip Kralyevski – tensiometer.


7. Acceptable tolerances when building a bicycle wheel

Nothing in this world is perfect – not even the bicycle wheels (it is just as close as it gets to that 🙂 ). In mechanics, things are usually built within the acceptable tolerances. What are those when it comes to bicycle wheels, i.e. when is a wheel considered acceptably true and well built?

We will first define the terms: “radial trueness“, “lateral trueness“, “rim dish trueness“, while “spoke tension evenness” is already explained in the previous chapter.


7.1. Radial trueness

Bicycle wheel radial trueness
Bicycle wheel radial trueness
Picture 18


Rim distance from the wheel’s centre should be even along the entire rim’s circumference. Ideally:
D1 = D2 = D3 = … = Dn

Video where I demonstrate wheel radial truing.

Of course it will not be ideal, but it is important to keep within the acceptable tolerances (defined in chapter 7.4).

Until you get the feel for visually determining the gaps when checking on a truing stand, you could use feeler gauges (Amazon affiliate link).

Feeler gauges used to measure engine valves are good for this use
Feeler gauges used to measure engine valves are good for this use
Click the image for shopping on Amazon (affiliate link)
Picture 19


7.2. Lateral trueness

Lateral (un)trueness is what we all notice first – how much does the wheel wobble (sideways).

Lateral trueness - how much does the rim move left-right compared to a stationary reference point
Lateral trueness – how much does the rim move left-right compared to a stationary reference point
Picture 20

Here too you could use the above noted feeler gauges to measure the deviation, at least until you get the feeling.


7.3. Rim dish trueness

In this case, picture 21 speaks more than a 1000 words:

Rim should be centred relative to the hub (i.e. relative to the bicycle's longitudal symmetry axis)
Rim should be centred relative to the hub (i.e. relative to the bicycle’s longitudinal symmetry axis)
Picture 21

This is measured using wheel dish measuring tool (Amazon affiliate link).

Using a wheel dishing gauge
Using a wheel dishing gauge
Measuring hub locknut (see OLD explained on this link) distance from the rim on each side, then comparing if it is even
Picture 22

There are some caveats here as well. For the “exotic” stuff, see on this page the explanation what to do when using asymmetric rims, and/or offset frames.


7.4. Acceptable trueness, and spoke tension tolerances

Briefly, a list of acceptable tolerances for the above explained parameters:

  • Spoke tension uniformity (between spokes on the same side): up to +- 10% tension variation among the spokes (on the same side). Ideally +- 5%, so that the maximum tension difference between the tightest and the most loose spoke is up to 10%.
  • Radial trueness: up to +- 0.5 mm deviation.
  • Lateral trueness: up to +- 0.2 mm deviation.
  • Rim dish trueness: up to 0.5 mm left-right deviation (dish gauge will always show the deviation as doubled!)


8. Sources – highly recommended

Share this article...

1 thought on “Bicycle wheel building basics”

Leave a comment

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.