How to build a road bicycle wheel
Thanks to: Wiel Van Den Broek
Velofilie.nl
SPOKES AND BRAIDING PATTERNS
The spoked wheel is a structure under preload. The spokes of a standard bicycle are subjected to a tensile force of about 800N. This corresponds to a weight of 80kg hanging from the spoke. The spokes are "stretched" by approximately 2mm between the hub and the rim. In the construction, they behave like a spring; by reducing the preload, they can absorb compressive forces.
The first bicycle wheels were wagon wheels with wooden spokes and rims reinforced with iron. When the wooden spokes were replaced by steel ones, the spoke pattern remained radial, i.e., the spokes ran straight from the hub housing to the rim. Such a pattern is not suitable for absorbing the forces of propulsion and braking. The hub will first "wind up" the spokes before the force is transmitted. By mounting the spokes in a criss-cross pattern, a wheel was created that could absorb these forces.
In FIG. 1, we see the drive side of a rear wheel with a brake hub. If we examine two spoke holes in the hub flange, e.g., those of spokes D2 and B1, we see:
1. The spoke head of spoke B1 is located on the outside of the hub flange. This spoke is therefore inserted from the outside inwards. A1, B1, C1...etc. are called the inner spokes.
2. The spoke head of D2 is located on the inside and is inserted from the inside outwards. A2, B2, C2...etc. are called the outer spokes. On its way to the rim, on the drive side, spoke B1 crosses the following spokes: first D2 (the so-called low cross), then C2 and B2 (the so-called high cross). We call this lacing pattern "3X crossed" or "over 3".
3. We see that spoke D2 passes over B1 and C1, but under D1. We see that spoke B1 runs under D2 and C2, but over B2. This is wheel lacing. In robot-spoked wheels, D2 will pass over B1, C1, and D1; B1 will run under D2, C2, and B2!
4. The force of the drive (clockwise) will be transmitted through all spokes to the rim. Spokes A1, B1, etc. are subjected to tensile load: the tension spokes. Spokes A2, B2, etc. are unloaded; these are called the static spokes. In FIG. 1, the inner spokes are subjected to tensile load. When we brake, the force distribution is reversed: the outer spokes are now subjected to tensile load.
The spoke length is specified in millimeters; the thickness is given by a different unit of measurement: Standard Wire Gauge (S.W.G.). This is an old English designation for wire thickness. The S.W.G. No. 10 has a thickness of 0.128 inches; this is 3.251 mm. The next size, No. 11, becomes thinner in steps of 0.012 inches (= 0.305 mm); No. 12 and No. 13 become thinner in the same steps. From No. 14 to 18, the steps become smaller (0.008” = 0.203 mm). From No. 19 to 22, the steps are halved again.
A children's cross bike has thick No. 12 spokes; the higher the number, the thinner the spoke. In "ordinary city bikes," the front wheel has a No. 14 spoke (2 mm thick) and the rear wheel has a stronger No. 13 spoke (2.3 mm thick), because 2/3 of the weight rests on the rear wheel and, moreover, the forces from the drivetrain and any brake hub act upon it. No. 13 spokes are tougher and stiffer; they flex less when unloaded. As a result, the nipples will come loose sooner. These spokes must therefore be tightened more than No. 14s.
Due to stretching and the setting of the spoke head in the hole of the hub flange, the spoke can stretch as much as 2 become mm longer, but by not taking the thickness of the nipple above the rim bed into account, this is compensated for. For 36-spoke wheels, 3x and 4x crossing are suitable spoke patterns. Crossing 4x causes problems with a low flange hub, because the spokes will run over each other. With high flanges, it works fine. Moreover, a high flange results in a much lower spoke load than a low flange. The more spokes there are in the wheel, the more often we can cross. A 48 or 44-spoke wheel can be crossed 5 times; a 40 or 36-spoke wheel can be crossed a maximum of 4 times. A 32 or 28-spoke wheel 3 times; 24 or 20 2 times, a 16-spoke 1 time.
It is often difficult to obtain rarely used spoke sizes; although most can be ordered (per 100 or per gross!), due to delivery time and availability, we sometimes have to buy ones that deviate slightly from the desired size. There is a difference of 4mm between too short and too long. If the spokes protrude through the nipples, we should have taken 2mm shorter, and if we can still see thread on the spoke, 2mm longer. The thread on the spoke is rolled, i.e., the material from the "valleys" forms the "hills"; the thread is therefore the thickest part of the spoke! The thread on the spoke is very shallow + or - 0.25mm. The top angle of the thread is 60°, and the thread always has 56 TPI from No. 12 to 15. This is a pitch of approximately 0.45mm; so every time the nipple tensioner turns, the nipple moves 0.45mm.
Because many rims are only suitable for 14-inch nipples, there are nowadays also 14-inch nipples with a 13-inch thread; however, the wall thickness is minimal; These nipples get damaged easily. There are also 13-inch spokes with a 14-inch thread, a slightly better solution, provided the spokes are not too long; this thread is cut: the nipple jams on the spoke and there is a risk of the nipple breaking, because the thread is the weakest point!
The nipples are usually made of nickel-plated brass; expensive nipples are made of strong aluminum. Cheap spokes are made of carbon steel, protected against rust by galvanizing, nickel plating, or chrome plating. Chrome is not really suitable; it is too hard, and the elastic flexing of the spoke leads to cracks and spoke breakage. Quality spokes are made of chrome steel such as X30Cr13 (Hoshi) or chrome-nickel steel such as X5CrNi18-8 (Sapim and Alpina) and X5CrNi18-10 (DT). Tests show that double-butted spokes result in lower peak stresses and fewer vibrations that cause nipples to loosen; and, of course, they are lighter.
Wealthy weight-conscious buyers can even purchase titanium spokes from DT. For special wheels, synthetic spokes are available, including Vectran spokes for Spox wheels, Kevlar (FiberFix), Berd spokes (Dyneema), and carbon from the Taiwanese CN-spokes. Mavic uses aluminum spokes from Zicral (AA7075) for the Ksyrium SL. All these special spokes are expensive, sometimes €5 each, and they often only fit one brand of wheels or even only one type of wheel from that brand.
R1 = half rim diameter incl. bottom R2 = half hub diameter from spoke hole to spoke hole A = distance flange to wheel center K = selected braiding pattern (0, 1, 2, 3, 4, or 5 cross) N = the number of spokes in the wheel |  |
An experienced wheel builder knows which spoke length corresponds to a specific hub/rim combination and lacing pattern. For infrequently used combinations, he can use tables or calculate the spoke length. The spoke can become 1 to 2 mm longer due to stretching and the insertion of the spoke head into the hole in the hub flange, but by not taking the thickness of the nipple above the rim bed into account, this compensates for itself. Apply the cosine rule in Fig. 3a:
(L´)² =(R1)²+(R2)²-2(R1)(R2)cosɑ ; and apply Pythagoras in fig.3b: L²=(L´)² +A²;
spoke length L = √ {(R1)²+(R2)²+A²-2(R1)(R2)cos Kx720°/N)
Huns and rims
Hubs come in many shapes and sizes. Even if two hubs are of the same type (high or low flange, see Figs. 4 and 5), there can be a difference of millimeters in diameter, and therefore in spoke length. The width of the front hub is usually between 90 and 100 mm; the width of the rear hub between 115 and 140 mm. Note the centerlines in Fig. 6; here we can clearly see that the flanges are twisted relative to each other!
The height of the flange varies from 30mm for a low flange to 120mm for a drum brake. With the older types of Sturmey Archer drum brakes, both a high and a low flange were even used (see FIG. 7A). Because the height of the brake made lacing the wheel impossible, so-called loops were used to hook the spokes into. This was not a pleasant system; the load on the spoke head was high because it was only partially supported by the flange. Lacing such a wheel was difficult because the head kept slipping out of the loops; fortunately, they have now opted for two equally high flanges. An idea that Shimano dusted off for their Deore M555 disc brake hubs (see FIG. 7B).
The main source of problems for rear wheels is the fact that they are umbrella-spoked. Technically, this means that the spoke tension on the right is 50-60% higher than on the left. The distance between the flanges in a rear wheel is often smaller than in a front wheel to make room for the sprockets. As a result, the heavily loaded rear wheel is laterally less rigid.
Because the freewheel is located on the right, and the rim must run in the center of the frame anyway, the spokes in the drive shield are much tighter. The tensioning spokes on the right, in particular, operate under constant high tension. On the left side, it is mainly the static spokes that are at risk; not from excessive tension, but from excessive tension, which facilitates loosening of the nipples.
not from excessive tension, but from excessive tension, which facilitates loosening of the nipples.
Niet door overmatige spanning, maar door overmatige spanning die het ontspannen van de tepels bevordert.
not with too high, but with too low tension, which causes the nipples to come loose.
Niet met een te hoge, maar met een te lage spanning, waardoor de tepels losraken.
One method to reduce the tension difference is to install an asymmetrical rim (e.g., Ritchey rims and Rolf wheels). Here, the nipple holes are all positioned off-center to the left (FIG. 8b). Of course, this is only a rim for rear wheels! An idea from Shimano was to cross the spokes in the vertical plane (see FIG. 8c). As a result, the spoke tension on the left and right is virtually equal. The spoke heads, unfortunately still with a bend, are moved to the rim, and the nipples are placed in the hub flange. The hub flange then consists of aluminum blocks into which the nipples fit. Similar blocks can also be found on hubs for tapered spokes. In this way, the weakest point of the wheel, the spoke bend, is eliminated. However, the blocks do take up space on the hub; consequently, fewer spokes can be mounted. But that is not such a big problem; fewer spokes save weight and reduce air resistance. With front wheels, there is much greater gain to be achieved in reducing air resistance; the rear wheel sits in the wake of the seat tube.
Standard rim diameters have evolved "historically," meaning that mutual agreements were established in various countries between the bicycle and tire industries regarding the dimensions to be used. These national agreements are the cause of the enormous variety in sizes and size designations. Nowadays, ETRTO sizes are found on almost all rims (inner diameter) and tires (outer diameter). An example is: 37-590. The first number indicates the diameter of the tire in its inflated state (37mm); the second number indicates the diameter of the bead wire, i.e., the point where the tire clamps onto the rim. This size is better known as 26x1 3/8.
The standard English 28x1 1/2 size is in ETRTO: 40x635. The standard French 28x1 1/4 size is in ETRTO: 32x622. The standard English 27x1 1/4 size is in ETRTO: 32x630.
The diameter of this French 28-inch rim is smaller than the diameter of an English 27-inch rim. An English 28-inch tire does not fit on a French 28-inch rim and vice versa!
The main types of rims are shown in FIG. 9:
9A: Westwood rim; this is used on "regular" bicycles and is made of (stainless) steel.
9B: U-rim or Endrick rim; made of steel, stainless steel, or aluminum on cheaper sports bikes, semi-racing bikes, and ATBs.
9C: Tubular rim: this is often used on racing bikes. Usually made of aluminum or carbon, although wooden rims also exist. The tubulars are glued to the rim!
9D: Box rim with horned edges; these are used on supersport bikes, road bikes, and ATBs. The horned edges prevent the tire bead wire from sliding off the rim at high tire pressure (6 to 9 bar). If the pressure is too low, the horned edges can be easily damaged. Modern rims lack these edges (hookless FIG. 9E).
In the past, nipple plates were slid under the nipples to prevent the nipple from tearing through the rim. Nowadays, there is usually a small reinforcing ring around the nipple hole (see Fig. 10A). Sometimes this ring holds a bushing that pulls on both rim bases (see Fig. 10B);
Sometimes this ring holds a bushing that pulls on both rim bases (see Fig. 10B);
Soms bevat deze ring een bus die aan beide velgbasissen trekt (zie figuur 10B);
Sometimes this small ring holds a bushing that pulls on both rim bases (see Fig. 10B);
Soms bevat deze kleine ring een bus die aan beide velgbasissen trekt (zie figuur 10B);
we call this double-bushed (occurs in types 9C and D). This is the strongest construction.
Westwood rims (type 9A) usually do not have a reinforcing ring; they are supplied in front wheel (14-inch nipple holes) and rear wheel versions (13-inch nipple holes). With drum brakes, these rims are often fitted with a doweled design, i.e., the spoke hole is located in a pre-formed dimple (see Fig. 11), which already points in the direction in which the spoke will run. This is particularly convenient with high flange hubs, because the angle between spoke and rim can become very small, causing the nipple to bend the spoke. This creates an extra stress point and can lead to spoke breakage at the nipple. For capped rims, spoke lengths should be chosen a few millimeters shorter than indicated in standard tables. This may also be the case if the rim type deviates from the standard; a box rim will have a smaller rim diameter compared to a Westwood rim of the same tire size (from lug to lug, as the bead wire is naturally in the same position): in this case, choose slightly shorter spokes.
 | De spaaknippelgaten in een velg zitten uit het midden. Als we in de binnenkant van het velgbed kijken, ontstaan er t.o.v. het ventielgat twee patronen: 12a: De velg heeft het nippelgaatje boven het ventielgat RECHTS uit het midden staan: een rechtse velg; dit resulteert bij ‘t spaken in een wiel met statische buitenste spaken. 12b: Deze velg heeft het nippelgaatje boven het ventielgat LINKS uit het midden staan: een linkse velg; dit resulteert bij het spaken in een wiel met trekkende buitenste spaken. N.B. Het heeft geen zin de velg om te draaien. |
The spoke nipple holes in a rim are off-center. If we look inside the rim bed, two patterns emerge relative to the valve hole:
12a: The rim has the nipple hole positioned to the RIGHT of the center above the valve hole: a right-handed rim;
when spoking, this results in a wheel with static outer spokes.
12b: This rim has the nipple hole positioned to the LEFT of the center above the valve hole: a left-handed rim;
when spoking, this results in a wheel with pulling outer spokes. N.B. It is pointless to turn the rim over.
Pattern 12a is usually seen in touring and sports bikes with coaster brakes; pattern 12b in (semi)racing bikes. The number of spoke holes varies from 12 for a children's bike to 48 for a tandem. Because the number of spokes in the left and right flanges is the same, and there are equal numbers of inner and outer spokes, the number of spoke holes is a multiple of four. Racing rims are available with 12 to 40 holes. The 12-hole rims are used, for example, on the track or in time trials (only in the front wheel). The 40-hole rims are found in cyclocross bikes and track tandems. The standard used to be 36 holes, but we are seeing a shift towards 32-spoke wheels. Special rims are sometimes difficult to obtain; so search and/or order them!
The rim width is related to the tire diameter. Wide tires go with wide rims. However, there is a lot of leeway here, provided the rim diameter is correct. Tires that are too wide often tear along the bead wire; with tires that are too narrow, the inner tube may sometimes burst out.
Rims with the 12a pattern are mainly seen on (semi)racing bikes. The spokes in a racing wheel that are subjected to the heaviest loads are the spokes that transmit the driving forces. For the front wheel, it does not matter; for the rear wheel, it does. The sprockets rotate the wheel clockwise. The tension of the chain is transmitted via the outer spokes!
We see pattern 12b in ordinary touring and sports bikes with a coaster brake. The braking forces of a hub brake or disc brake are greater than the driving forces! The outer spokes, which are the strongest, must absorb the braking force here: this rotates counter-clockwise (left). The inner spokes must therefore absorb the driving force (clockwise). This is the reason why rims with disc brakes and coaster brakes are designed RIGHT-HANDED and spoked clockwise! The spoke hole on the rim above the valve is then off-center to the right (see FIG. 12b). We will therefore see a shift towards rim type 12b in the near future.
In FIG. 13a, the nipple hole above the valve hole is positioned LEFT of center. In FIG. 13b, it is RIGHT of center. It is pointless to turn the rim over! N.B. We are looking into the rim bed.
The hub of a wheel has two flanges through which the spokes are inserted. The first series of spokes is inserted from the outside inwards (see spoke number 1 in FIG. 13b; we call these the inner spokes). The second series is inserted from the inside of the wheel outwards (see spoke number 2 in FIG. 13b: the outer spokes). The strongest spokes in the wheel are the spokes inserted from the inside outwards. These spokes are spaced further apart, allowing them to better withstand extra load. The flanges on the hub are rotated relative to each other (see FIG. 13c). Note the centerline between the front and rear flanges. The spoke holes in the flanges are therefore staggered; This is important for braiding!
N.B. When braiding, we always start with the inner spoken.
THE FORCES ON THE SPOKES
 |
A spoke can only transmit forces in its longitudinal direction. To transfer a force from hub to rim, or vice versa, the spokes must be taut. When a spoke is subjected to a compressive load, the spoke tension decreases proportionally. This can result in the nipple coming loose due to vibrations if the preload is insufficient. Loose spokes are just as bad as broken spokes; they do not contribute to the strength of the wheel. Consequently, the other spokes will be subjected to higher loads. Assuming a 32-spoke front and rear wheel with size 14 spokes, the tensile force on the spoke at rest (static) is approximately 800N. If we place such a wheel on a 10 kg bicycle with an 80 kg rider, most of the weight will rest on the rear wheel (50 kg). The weight presses the hub downwards. Because the rim yields slightly, the effect of tension reduction at the bottom is greater than that of the extra load on the upper spokes. With a 32-spoke rear wheel under 50 kg of pressure, the "relief" is approximately 240 N and the additional load 60 N. When cycling, the spoke tension of this wheel, in each spoke, will vary between 560 and 860 N with every revolution (i.e., 500 times per kilometer). Very stiff rims result in less variation in spoke tension and are simply mandatory for wheels with few spokes. Some wheels have spokes arranged in pairs; consequently, the holes in the rim are not evenly distributed. Because two spokes are now relieved simultaneously, the variation in spoke tension is smaller. |
Spokes are very strong; these types of steel have a tensile strength of approximately 1200 N/mm². This means that a No. 14 spoke (2mm thick) will only break at a load greater than 3800 N. However, high tensile strength is not the most important characteristic of a spoke; yield strength and fatigue limit are more important. If the spoke is loaded beyond the yield limit, the preload decreases; this limit lies around 900 N/mm². For most spokes, the load must not exceed 2200 N! The fatigue limit is determined by resistance to alternating loads: for a good spoke, more than 1 million times. The spoke manufacturer Sapim even specifies 3.5 million cycles for their CX-ray. Is that a lot? Make no mistake: after 2000 km, the wheel has already turned 1 million revolutions. This causes spokes to break: not excessive tension, but metal fatigue; fortunately, steel is not very susceptible to this. In a modern 16-spoke front wheel, the spoke tension is 1000 to 1200 N (equal on both sides). In umbrella-spoked rear wheels (see Figure 8a), the spoke tension can rise considerably. On the left, the tension is sometimes 900 N, and on the right, in the spoke shield on the pinion side, it is even 1500 N! Here we are at the limit of what is permissible for spoke No. 14. Higher spoke tensions do not lead to stiffer wheels.
The variation in weight load is the same for every spoke and independent of the lacing pattern (though independent of the number of spokes). In addition to static tension and weight load, impact loads also act on the spokes. Driving through potholes and against curbs results not only in very high loads but, above all, in the relief (!) of the spokes (> 250 N!). If the static tension of the spokes is too low, the nipple will work itself loose due to these force fluctuations. Wheels are stiff relative to the tire. However, inflating the tires or tubes to 8 to 10 bar can reduce the spoke tension by 30 to 40 N.
N.B. The forces mentioned above are independent of the lacing pattern, as are the braking forces of rim brakes (so these do not pass through spokes and hubs!). The braking force acts at two points: at the top of the rim at the brake pads and under the tire. The reaction force comes from the front fork. This pushes the hub forward; the front spokes are relieved of load and the rear spokes are subjected to extra stress. These are the spokes that are subjected to the least stress while riding, which is why braking forces from rim brakes play hardly any role. What can play a role, however, is the shifting of weight. The pressure on the rear wheel decreases during braking and the pressure on the front wheel increases; in the extreme case, all the weight rests on the front wheel (be careful!).
Driving forces and braking forces from hub or disc brakes do act on the spokes; in this case, the static spokes are subjected to tensile stress. The spoke load then depends partially on the lacing pattern. Braking forces are sometimes twice as high as driving forces. It is recommended to use very strong spokes with brake hubs or disc brakes, to make the outer spokes static, and to cross them as often as possible.
THE BRAIDING PATTERN AND THE POWER DISTRIBUTION
The spoke load caused by the drive torque depends on, among other things, the rider, gear ratio, and hub flange height. If a 75 kg rider + bike combination goes up an 18% incline, the theoretical force for the drive is 750N x 0.18 = 135N. The torque delivered is then 135N x 0.335m (radius of the wheel) = 45Nm. This same torque passes through the hub flanges and the spokes. Suppose the radius of the flange is 0.025m, then the force there is: 45Nm:0.025m = 1800N! In our example with a 5 cm flange and 36 spokes, the extra load would be 50N per spoke, provided the spoke is positioned exactly perpendicular to the centerline through the hub flange.
|
When we examine a tension spoke S (see FIG. 14) in a 3x crossed wheel, we see that the driving force Fa acts perpendicular to the centerline of the hub. However, the spoke S can only absorb forces in its longitudinal direction. A force Fs arises in the spoke, which transmits the driving force, the tangential component of which, Ft, prevents the torsion of the wheel. Ft is always equal to Fa but opposite; Fs becomes as large as necessary to obtain Ft = -Fa. In a 3X intersecting wheel, angle α is approximately 60°, so Fs = 50N : sin60° = 58N. A radial component Fr is also created, which improves lateral stiffness. As angle α increases, Fr and Fs decrease in magnitude. If angle α is ninety degrees, Ft = Fs applies: the spoke load is then minimal. Unfortunately, Fr = 0 at that point, which comes at the expense of lateral stiffness. If α is greater than 90 degrees, the spoke load increases again, while Fr becomes negative. Consequently, lateral stiffness deteriorates rapidly. If the angle α becomes smaller (a less intersecting braid pattern), Fs and Fr increase, because Ft remains constant! The spoke load due to driving forces (torsion) therefore increases. In a radially spoked wheel, Fs is theoretically infinitely large. In practice, the hub will "wind up," e.g., by 2°. Then Fs = 50N: sin2° = 1433N. A wheel in which driving or braking forces act on the hub must therefore not be radially spoked! In practice, the spoke loads for the "drive shield" are higher. The hub flange on the drive side absorbs more force because the stiffness of the hub is limited. With a stiff hub, e.g. the Sturmey Archer 3-speed hub, a spoke from the right shield will absorb 60% and a left one 40%. With an aluminium hub, the values are much more unfavorable; here, the right absorbs 85% and the left 15%. |
For 36-spoke wheels, 3x and 4x crossing are suitable spoke patterns. 4x crossing causes problems with a low flange hub, because the spokes will overlap. With high flanges, it works fine. Moreover, a high flange results in a much lower spoke load than a low flange. As the wheel has more spokes, we can cross them more often. A 48 or 44-spoke wheel can be crossed 5 times without the angle α exceeding 90 degrees. A 40 or 36-spoke wheel can be crossed a maximum of 4 times. A 32 or 28-spoke wheel 3 times; 24 or 20 2 times; a 16-spoke 1 time. Note: angle α is ± 30° for a 16-spoke wheel that has been crossed once. The theoretical load due to driving forces is high: (1800N:16): sin30° = 225N per spoke! The actual forces in the right screen are higher: ± 400N.
We can spoke a front wheel without a coaster brake or disc brake in every possible way. Radially spoked wheels are the lightest, strongest, and stiffest. We can safely use 32 or fewer spokes in these wheels. With rear wheels, the matter is more complicated. We should strive to avoid umbrella spokes. This can be achieved by using wide hubs and chainstays, or by building asymmetrical frames. For a 36 or 32-spoke wheel, a 3-crossed design is desirable. For small wheels, such as 20 inches, I settle for fewer spokes and fewer crosses. Otherwise, the spoke enters the rim at too much of an angle (risk of spoke breakage in the nipple). Small wheels are stronger and stiffer anyway; moreover, they have a lower spoke load (smaller driving force and more revolutions per minute).
THE SPOKES OF BICYCLE WHEELS
FIG. 14b Minimal hand tools.
We describe the traditional spoked wheel here. There are variations on this method. If you follow a specific description, stick to that method; following multiple descriptions at once leads to confusion and errors.
There have been many new wheel designs over the past 40 years. The trend towards stiffer rims and fewer spokes is clear. The steps toward disc brakes and wider rims and tires result in somewhat heavier wheels. The lifespan of the wheel will increase due to the elimination of rim wear.
Because braking forces now run from hub to rim, radial spoking of wheels is a thing of the past.
The number of spoke holes varies from 12 for a children's bicycle to 48 for a tandem. Because the number of spokes on the left and right is the same, and there are equal numbers of inner and outer spokes, the number of spoke holes is a multiple of four. The old standard is 36 holes. N.B. Some modern variants have more spokes on one side. Always start all lacing patterns from the valve hole. Otherwise, a crossing of spokes sometimes occurs above the valve, which makes inflating the tire difficult. The lacing for all spoke patterns is virtually identical. With a specific spoke length + hub + rim, the spoke pattern is effectively fixed. An experienced wheel builder knows which spoke length corresponds to a specific hub/rim combination and lacing pattern. For infrequently used combinations, he can use tables or calculate the spoke length. At the bottom of this page, you will find a formula for calculating spoke lengths and an app in Excel.
We can spoke a front wheel without a coaster brake or disc brake in every possible way. Radially spoked wheels are the lightest, strongest, and stiffest. We can safely use fewer spokes in these wheels. With rear wheels, the matter is more complicated. We should strive to avoid umbrella spokes. This can be achieved by using wide hubs and chainstays, or by building asymmetrical frames. For a 36 or 32-spoke wheel, a 3-crossed stitch is desirable. For small wheels, such as 20 inches, I settle for fewer crosses and fewer spokes. Otherwise, the spoke enters the rim at too much of an angle (risk of spoke breakage in the nipple). Small wheels are inherently stronger and stiffer; moreover, they have a lower spoke load (lower driving force, higher revolutions per minute).
THE BRAIDING(?) : 0x en 1x
There is actually little difference between lacing a front wheel and a rear wheel. In a front wheel without a coaster brake or disc brake, we can use any lacing pattern. For a rear wheel, at least one shield, preferably multiple times, must be crossed; this depends on the number of spokes, as we have just seen.
The simplest way to spoke wheels is radial (0x crossed). It can be done in three ways:
1. All outer spokes (all spoke heads inside) 2. All inner spokes (all spoke heads outside) 3. Alternating.
I wouldn't know of a single argument for the third method. According to some, the second method looks nicer. The first method is the best; the wheel is about 10% stiffer because the spokes are spaced further apart and the spoke head is better supported by the flange, which reduces its movement. This is a weak point with radial wheels; not only do spokes break, but hub flanges do too! These are subjected to heavy loads during radial lacing. However, a radial wheel is aerodynamic, super light, and super stiff. N.B. Shimano does not provide a warranty on the standard hubs with radial spokes!
It does not matter whether the rim is left-handed or right-handed. The image in FIG.3 happens to be LEFT; for a right-handed rim, read RIGHT wherever LEFT appears. Insert the spokes from the inside out through flange A. Always start all lacing patterns from the valve hole. Insert the first spoke to the LEFT next to the valve (a1). Then proceed to the LEFT a2, a3, etc., until we are back at the valve. We now look along spoke a1 to flange B (see FIG.3), and we select the spoke hole to the LEFT of a1 on flange B. Through this, we insert the first spoke from the inside out. We mount this spoke to the LEFT of a1 in the rim between a1 and a2 in the hole b1. Now we insert the remaining spokes into the flange and mount them at b2, b3, etc., until we reach the valve again. If everything aligns correctly, we can tension and align the wheel.
The 1x cross braid pattern can also be laced using only outer spokes. Based on FIG. 3, we simply insert the first spoke at a2 and the second at a1. We go around the entire rim in this way. Naturally, on the other side, the first spoke at b2 must be inserted, and the second at b1. With a 1x crossed braid pattern, we skip a hole on the rim (for example, b1 between a1 and a2). We can now say that the low cross spans one spoke hole on the rim. As the number of crosses increases, this number increases. With 2x, there are five spoke holes on the rim, with 3x nine, and with 4x thirteen.
BRAIDING WHEELS: 2x,3x,4x EN 5x CROSSED
The lacing for all these spoke patterns is virtually identical. Given a specific spoke length + hub + rim, the spoke pattern is effectively fixed. We assume a 36-spoke hub (this eliminates 5x crossed!) and a LEFT-HAND rim. If we have a right-hand rim, read right where LEFT is written! Furthermore, spoke the wheel clockwise. If the wheel is suitable for a cassette or freewheel, the spokes on the drive side must be shorter; usually, this difference is 2mm. These spokes are subjected to 40 to 50% higher tension.
FOR A REAR HUB, WE START AT THE DRIVE SIDE. We call the flange we start with flange A.
We insert spoke A1 (see FIG. 16) from the outside inwards and secure it in the spoke hole in the rim to the LEFT of the valve hole. We now skip a hole on the hub and insert the second spoke (A2) from the outside inwards. On the rim, we now skip three holes and mount the nipple in the fourth spoke hole to the LEFT of the valve. We continue in this way until side A is full: nine spokes!
As we see in FIG. 3a, the spoke holes on the other flange (B) are slightly offset. We draw an imaginary line through the hole in the hub flange A where A1 is located. In the hole on hub flange B located to the LEFT of A1, we insert the next spoke (B1) from the outside inwards. We mount spoke B1 to the LEFT of A1. To mount B2, we again skip a hole on flange B and three holes on the rim. B2 goes to the LEFT of A2, etc., until there are nine spokes in flange B as well.
We hold side A in front of us and now turn the hub counterclockwise. Now insert the first spoke into the hub to the LEFT of A1, from the inside out. Turn this spoke to the LEFT so that it crosses 2, 3, or 4 spokes of screen A. This depends on the lacing pattern you have chosen!! We pass this spoke under the last one and mount it in the correct nipple hole in the rim (A side!). We continue in this way until flange A is full. If we have worked correctly, the spokes are now in neat groups of three. We now pass the spokes through flange B from the inside out. Here too, cross them and pass them under the last spoke. There is only one correct assembly possible; if it does not fit, a mistake has been made somewhere; this often means starting over.
TENSIONING THE SPOKES
We start by tightening all the nipples with a screwdriver until only one turn of thread is visible on the spoke. There may already be tension on the spokes at this point. If not, turn all spokes a half turn, round by round, until tension is applied. We always work from the valve! If there are a very large number of turns, the spokes are too long and will protrude through the nipple. This causes flat tires, so use shorter spokes, or cut and file down the spokes. As soon as there is tension on the spokes, we swap the screwdriver for a good nipple wrench. If all spokes are tightened to the same extent and the rim was perfectly round, the wheel will also be perfectly round; this is often not the case. If there is a lot of lateral runout, I first roughly remove it. If the rim needs to be turned to the right, tighten the right spokes a half or quarter turn over the relevant section, and loosen the left spokes a half or quarter turn. N.B. If we only tighten the right spokes, we are effectively introducing vertical runout!
If the wheel is reasonably straight, we will check for vertical runout. "Dents" are harder to remove than "bumps." Remember that the hub transmits the forces on the spokes to the other side of the wheel. While you tighten the spokes on both the left and right sides when encountering a bump, the opposite spokes must be loosened; otherwise, there is a high chance that we will go from an "egg-shaped" to an "elliptical" wheel. For a dent, for example, you loosen 2 right and 2 left spokes by half a turn, and tighten all others by a quarter turn. At the weld (often camouflaged by a sticker), a deformation may sometimes remain. We now take a wheel hub aligner and check if the rim is centered. This is a simple bracket with an adjustment screw in the middle; first hold it against the rim on one side and turn the adjustment screw all the way to the hub. Next, hold it on the other side: the adjustment screw must now touch the hub exactly again. If this is not the case: tighten all spokes on the side where the rim needs to go, and loosen the other side proportionally. As soon as the vertical runout is gone, we will remove the remaining lateral runout. If the rim has no vertical or lateral runout left, and is also centered, then we bring the spokes to final tension. How tight is that now? Tight! Just squeeze a wheel from a skilled craftsman. Spoke tension gauges are available to better monitor the process.
Once the wheel is at final tension, tighten and loosen all spokes by a quarter turn; this will remove the torsional tension. Check one last time; then put the tire on and ride. After about 500 km, we put the wheel in the truing stand again. If the spokes were set very tight (and no accidents have occurred), the wheel is still perfect. Some wheel builders then tighten their spokes by a half turn anyway. With a good wheel, the tolerances are very small: ± 0.2 mm.
Especially when repairing wheels that have been used for a long time, it is possible that a nipple becomes "rounded." There is no other option than to remove it with water pump pliers and install a new nipple. Do not make the mistake of loosening or tightening other spokes to straighten the wheel. This causes very irregular spoke tensions; one of the main causes of spoke problems. Such a wheel rarely becomes truly straight and suffers greatly from spoke breakage. A bad wheel sometimes causes problems within 5,000 kilometers. If more than 2 spokes in a wheel are broken, I re-spoke it (if the hub is worth the effort). I use new spokes and almost always a new rim; a used rim is never truly round again.
MODIFIED WHEELS
As we have seen, according to our "construction recipe", a right-hand rim results in static outer spokes, and a left-hand rim in tensioning outer spokes. To make the outer spokes in a left-hand rim static, we do the following (see Fig. 17). Insert A1 from the outside in, into the second nipple hole, to the right of the valve. Skip one hole on the flange, and 3 holes on the rim (both to the right!). Continue working clockwise until the first 9 spokes are in place. We now insert spoke B1 (note the centerline: to the left of A1!) from the outside in, and mount it next to the valve hole. Mount B2, etc., again clockwise. We now take side A again and turn the hub clockwise; we now insert a spoke between A1 and A2 from the inside out. Depending on the chosen lacing pattern, cross and pass under the last spoke. I expect that by now you know how to proceed and also how to make a right-hand rim with the outer spokes tensioned!
The spoke method described here is not the only way to build wheels. In various videos and descriptions by others, variations on my method are sometimes used.
There are more good solutions, but wrong choices are also made. With this wheel in FIG. 18, featuring a Nuvinci continuously variable transmission hub, the builder opts for an excessive number of crossings; as a result, the spoke forms a sharp angle with the rim, causing the nipple to exert uneven pressure on the spoke's thread. This bending amplifies the notch effect and leads to spoke breakage at the nipple. You can mitigate these types of problems somewhat by creating a kink in the spoke 20mm from the thread; however, that is a drastic measure and looks very ugly besides. The "snowflake" wheel (Fig. 19) also frequently suffers from this problem; incidentally, this wheel has been invented many times since.
A French patent application for a wheel with interlaced spokes (circa 1900).
Spoke thicknesses were fairly standardized over the years. Unfortunately, this practice has come to an end, and we now see spoke thicknesses of 2.16mm coming from China (with 3.6mm nipples); a horror, is nothing sacred anymore?
