Most aspects of the Torq were covered in detail on the forum during 2006/0. This year, the emphasis for hub-motorised e-bikes has moved to other makes and to the fitting of kits. With improved battery capacity and the approach of Summer, touring on an e-bike is gaining popularity.
The Torq is maybe unusual now in having no suspension, but unsprung bikes with this type of performance still have great appeal for use on tarmac and well made up cycle tracks.
Much of the rider comfort depends on “wheel build”, and other issues can arise when a powerful hub motor is at the centre.
These notes may be useful to others who want to improve their ride and reliability themselves, or indeed use the services of a regular bike shop, to build a good wheel for their e-bike.
The rebuild focuses on three aspects that I wanted to see if I could improve:
1. Metal fatigue leading to spoke failures
2. Wrist and arm fatigue through vibration
3. Front wheel stability when braking
On completion, items 2 and 3 improved immediately, number 1 needs some more miles to be confident.
Metal fatigue in front wheel spokes
It is important to remember that spokes are very strong; much stronger than the aluminium rims that they connect to. In a straight pulling contest, the spoke nipple would pull out of the rim well before the spoke failed.
Metal fatigue takes place over a long period of time, gradually weakening the spoke. Anyone opening a tin of corned beef will know how difficult it is to just pull off the lid, but waggle that last hinge bit about 20 times and it readily snaps due to metal fatigue.
The picture shows the standard format of the Torq front wheel:
Suzhou Bafang Hub Motor, having 18 spokes on each flange in 3.2 mm diameter holes on Pitch Circle (PCD) of 132 mm. The flanges of the motor are 44 mm apart centre to centre, set approximately central in forks 125 mm apart.
The motor shaft uses an unusual Metric Fine thread, M12 x 1.25 mm pitch.
Standard spokes are 2.34 mm dia throughout, length 253 mm, into nipples 13 mm long.
Spoke lacing is “2 cross”, meaning that each spoke crosses 2 others on the same side between hub and rim.
Standard rim is 700C Weinmann XTB24 meeasuring 606 mm across spokes, with spoke to bottom of slot in nipple.
Motor rotates clockwise in this view; round green dots indicate one of the nine pairs of “pulling” spokes.
“Pulling” spokes increase in tension when the motor exerts force to pull the rim forwards.
At rest, all spokes are bowed.
Under power, the “pulling” spokes try to straighten, causing the “pushing” spokes to bow more.
Looking into the red circle, the “pulling” spoke crosses behind the “pushing” spoke. Apply the throttle, and the crossing point lifts towards us, causing a “waggle” at the bend of the “pushing” spoke near the hub.
On throttle controlled e-bikes, riding technique is critical. If the front brake (rim on the Torq) is applied before the motor torque has died away, then the tension in the pulling spoke increases greatly. Early bikes had switches in the brake levers to reduce this issue. Providing the throttle hand isn’t wracked by vibration, I am sure that we all manage to release the throttle quickly, even in an emergency braking situation. However, the software in the controller needs to kill the power instantly, rather than be set to a gentle decay.
On my Torq, 6 spokes failed at this bend point between 1200 and 1700 miles.
All 6 of them were “pushing” spokes, 5 on the near flange (red circle and arrows) and 1 on the far flange (yellow circle and arrow).
Solutions to reducing / preventing fatigue failure
1. High quality spokes, equally tensioned during the wheel build.
2. Stress relief by adjusting the bend at the red arrow during building, so that spoke is not flexing against itself.
3. Encourage spoke to flex in a straight section, away from the stress concentration of the bend at the hub – this can be assisted by varying the diameter of the spoke at different sections along its length.
Some excellent insights into the problem come from the “Tandem Club” forum. Tandems apply the torque of two riders through a single hub, often a Rohloff of similar proportions (large diameter) to the hub motor.
To select the parts for the wheel build, keep in mind the following:
1.The Suzhou Bafang motor has spoke holes of diameter 3.2 mm. This is quite large so a smaller diameter spoke might pull through the aluminium quite easily. For interest, the standard Shimano hub used in the rear wheel has holes 2.65 mm diameter for spokes of 2.34 mm.
2.The spoke flanges of the motor are quite thick in comparison to a regular hub. Spokes of 2.34 mm dia (13gauge) are usually made with a longer hook than the standard European 2.0 mm (14gauge) spokes.
3.The single eyelet of the Weinmann rim is also suited to the larger diameters of the 2.34 mm spoke nipples. The smaller 2.0 mm spokes are a little sloppy, particularly if we want to keep tensions quite high.
Wrist and arm fatigue through vibration
Without suspension, the resilience within the wheel becomes even more important. Spokes of 2.34 mm are a very rigid animal (almost a rod), with very limited capability to stretch elastically.
A little bit of elastic stretch helps each spoke pass the load to the next, avoiding the feeling of riding on a 36 sided wheel.
Spokes of smaller diameters provide this elasticity, although it must be stressed that we are designing for roads rather than mountain bikes.
Final selection
With touring in mind, I decided on the DT Swiss Alpine III spokes together with a Mavic A319 rim designed for the standard European 2.0 mm spoke entry.
The Alpine III spoke is divided into 3 sections of different diameters: 2.34 mm at the hub, 1.8 mm along the main centre section, 2.0 mm at the thread.
The A319 rim has a double eyelet, which spreads the spoke load across the inner and outer layers of the rim. Mavic recommend tensions between 70 and 90 kg for their rims. (I have used 90 kg).
Spoke length of 254 mm with 12 mm nipples in the Mavic rim brings the tip of the spoke half way up the slot. (For reference, it could have taken 255 mm). These lengths are also suitable for the standard Weinmann rim when using 12 mm nipples and good quality spokes.
I believe that this selection will provide a strong wheel for touring. Utility work might incur more sideways knocks, for which the regular DT Swiss Alpine spoke (2 diameters) might be more appropriate.
Looking down on the finished wheel, this picture shows the extent of the bow in the spokes that occurs with a large diameter hub.
Wheel stability under braking
In the construction of the new wheel, I have made one other change from the standard construction.
Looking again at the first picture in this post with the pair of “pulling” spokes marked by green dots, you will see that the nearside pulling spoke joins the hub on the outside of the flange, whereas the farside pulling spoke joins the hub on the inside of the flange.
Laced this way, when these two spokes straighten under tension, the rim will tend to shift towards us. I strongly suspect that this sets up vibration under braking.
It is much more normal for the pulling spokes to connect either to the outside on both sides or the inside on both sides, so that there is an equal and opposite lateral force cancelling out.
As shown below, I opted to mirror the lacing that was used on the farside flange with the pulling spoke on the inside face, as these angles had produced 1 fatigue failure through deflections compared with 5 failures on the nearside.
In practice, the wheel is now completely stable, but a number of changes have been done simultaneously so I am uncertain as to which is the clincher.
It would be useful to hear of situations in which the “one inner, one outer” pulling spoke has been appropriate.
Finally, I can recommend very highly a book called Wheel Building by Roger Musson which can be downloaded for a price of £9.
James
The Torq is maybe unusual now in having no suspension, but unsprung bikes with this type of performance still have great appeal for use on tarmac and well made up cycle tracks.
Much of the rider comfort depends on “wheel build”, and other issues can arise when a powerful hub motor is at the centre.
These notes may be useful to others who want to improve their ride and reliability themselves, or indeed use the services of a regular bike shop, to build a good wheel for their e-bike.
The rebuild focuses on three aspects that I wanted to see if I could improve:
1. Metal fatigue leading to spoke failures
2. Wrist and arm fatigue through vibration
3. Front wheel stability when braking
On completion, items 2 and 3 improved immediately, number 1 needs some more miles to be confident.
Metal fatigue in front wheel spokes
It is important to remember that spokes are very strong; much stronger than the aluminium rims that they connect to. In a straight pulling contest, the spoke nipple would pull out of the rim well before the spoke failed.
Metal fatigue takes place over a long period of time, gradually weakening the spoke. Anyone opening a tin of corned beef will know how difficult it is to just pull off the lid, but waggle that last hinge bit about 20 times and it readily snaps due to metal fatigue.
The picture shows the standard format of the Torq front wheel:
Suzhou Bafang Hub Motor, having 18 spokes on each flange in 3.2 mm diameter holes on Pitch Circle (PCD) of 132 mm. The flanges of the motor are 44 mm apart centre to centre, set approximately central in forks 125 mm apart.
The motor shaft uses an unusual Metric Fine thread, M12 x 1.25 mm pitch.
Standard spokes are 2.34 mm dia throughout, length 253 mm, into nipples 13 mm long.
Spoke lacing is “2 cross”, meaning that each spoke crosses 2 others on the same side between hub and rim.
Standard rim is 700C Weinmann XTB24 meeasuring 606 mm across spokes, with spoke to bottom of slot in nipple.
Motor rotates clockwise in this view; round green dots indicate one of the nine pairs of “pulling” spokes.
“Pulling” spokes increase in tension when the motor exerts force to pull the rim forwards.
At rest, all spokes are bowed.
Under power, the “pulling” spokes try to straighten, causing the “pushing” spokes to bow more.
Looking into the red circle, the “pulling” spoke crosses behind the “pushing” spoke. Apply the throttle, and the crossing point lifts towards us, causing a “waggle” at the bend of the “pushing” spoke near the hub.
On throttle controlled e-bikes, riding technique is critical. If the front brake (rim on the Torq) is applied before the motor torque has died away, then the tension in the pulling spoke increases greatly. Early bikes had switches in the brake levers to reduce this issue. Providing the throttle hand isn’t wracked by vibration, I am sure that we all manage to release the throttle quickly, even in an emergency braking situation. However, the software in the controller needs to kill the power instantly, rather than be set to a gentle decay.
On my Torq, 6 spokes failed at this bend point between 1200 and 1700 miles.
All 6 of them were “pushing” spokes, 5 on the near flange (red circle and arrows) and 1 on the far flange (yellow circle and arrow).
Solutions to reducing / preventing fatigue failure
1. High quality spokes, equally tensioned during the wheel build.
2. Stress relief by adjusting the bend at the red arrow during building, so that spoke is not flexing against itself.
3. Encourage spoke to flex in a straight section, away from the stress concentration of the bend at the hub – this can be assisted by varying the diameter of the spoke at different sections along its length.
Some excellent insights into the problem come from the “Tandem Club” forum. Tandems apply the torque of two riders through a single hub, often a Rohloff of similar proportions (large diameter) to the hub motor.
To select the parts for the wheel build, keep in mind the following:
1.The Suzhou Bafang motor has spoke holes of diameter 3.2 mm. This is quite large so a smaller diameter spoke might pull through the aluminium quite easily. For interest, the standard Shimano hub used in the rear wheel has holes 2.65 mm diameter for spokes of 2.34 mm.
2.The spoke flanges of the motor are quite thick in comparison to a regular hub. Spokes of 2.34 mm dia (13gauge) are usually made with a longer hook than the standard European 2.0 mm (14gauge) spokes.
3.The single eyelet of the Weinmann rim is also suited to the larger diameters of the 2.34 mm spoke nipples. The smaller 2.0 mm spokes are a little sloppy, particularly if we want to keep tensions quite high.
Wrist and arm fatigue through vibration
Without suspension, the resilience within the wheel becomes even more important. Spokes of 2.34 mm are a very rigid animal (almost a rod), with very limited capability to stretch elastically.
A little bit of elastic stretch helps each spoke pass the load to the next, avoiding the feeling of riding on a 36 sided wheel.
Spokes of smaller diameters provide this elasticity, although it must be stressed that we are designing for roads rather than mountain bikes.
Final selection
With touring in mind, I decided on the DT Swiss Alpine III spokes together with a Mavic A319 rim designed for the standard European 2.0 mm spoke entry.
The Alpine III spoke is divided into 3 sections of different diameters: 2.34 mm at the hub, 1.8 mm along the main centre section, 2.0 mm at the thread.
The A319 rim has a double eyelet, which spreads the spoke load across the inner and outer layers of the rim. Mavic recommend tensions between 70 and 90 kg for their rims. (I have used 90 kg).
Spoke length of 254 mm with 12 mm nipples in the Mavic rim brings the tip of the spoke half way up the slot. (For reference, it could have taken 255 mm). These lengths are also suitable for the standard Weinmann rim when using 12 mm nipples and good quality spokes.
I believe that this selection will provide a strong wheel for touring. Utility work might incur more sideways knocks, for which the regular DT Swiss Alpine spoke (2 diameters) might be more appropriate.
Looking down on the finished wheel, this picture shows the extent of the bow in the spokes that occurs with a large diameter hub.
Wheel stability under braking
In the construction of the new wheel, I have made one other change from the standard construction.
Looking again at the first picture in this post with the pair of “pulling” spokes marked by green dots, you will see that the nearside pulling spoke joins the hub on the outside of the flange, whereas the farside pulling spoke joins the hub on the inside of the flange.
Laced this way, when these two spokes straighten under tension, the rim will tend to shift towards us. I strongly suspect that this sets up vibration under braking.
It is much more normal for the pulling spokes to connect either to the outside on both sides or the inside on both sides, so that there is an equal and opposite lateral force cancelling out.
As shown below, I opted to mirror the lacing that was used on the farside flange with the pulling spoke on the inside face, as these angles had produced 1 fatigue failure through deflections compared with 5 failures on the nearside.
In practice, the wheel is now completely stable, but a number of changes have been done simultaneously so I am uncertain as to which is the clincher.
It would be useful to hear of situations in which the “one inner, one outer” pulling spoke has been appropriate.
Finally, I can recommend very highly a book called Wheel Building by Roger Musson which can be downloaded for a price of £9.
James
Last edited:
