Surly Pugsley frame + Rear TC4080 motor

[thunderstorm80]

Hi,
In my never-ending search for the ultimate electric touring bike, I thought to consider the Surly Pugsley frame.
It is a fat-bike frame, fully loaded with all kinds of rack eyelets including in the fork, made of chromoly steel, and has 135mm dropouts (both front and rear) which makes it easy to mount a standard rear motor without adaptors.
BUT, the price of the narrow 135mm rear dropout is a 17.5mm dishing offset of the hub toward the drive side on the right, so there won't be tire/chain rub issues and you can use regular triple crankset gearing.
17.5mm dishing is an extreme one, and I modelled this on the spoke simulator:


It looks quite disturbing, in terms of the spoke tension ratio. Although Grin says that even 80% ratio is possible on rear wheels, I think this is better to be avoidable since we talk about a heavy loaded touring bike that might also go on light off roads. On the other hands, the tires would be Surly black floyd 26x3.8 which make the entire bike (and the wheels) to a sort of sprung mass. This is why I like the idea of fat-bike as touring bike so much.
If you think it's still ok and reliable on the long term:
1. Shall I use thicker spokes like 13g or even 12g?
2. You can also see I placed all the left spokes outside and the right ones inside to minimize that stress ratio. It means there would be no crossing support between the spokes when doing cross "1". Do you think it's ok this way, and would you choose cross "0" since the swapping might already give enough angle for torque transmition ?

It would be interesting to hear what you think.
Ideally there would be another TC4080 running the front fork (since there is 135mm at the fork as well), giving awesome 2x2 for the heavy load, and allowing to climb (or descend) at very low speeds and still have very efficient electro-mechanical conversion ratio. (Whether it's forward torque or regen)

 

[dogman dan]

Bend that sucker out to 140 or 145mm, to fix that dishing problem. Steel bike. 8)

I'm assuming the motor has an extra 5 mm of axle on each side to work with.

Don't sweat it if the chain rubs the tire when you use the inner ring up front, you won't be using it, ever. in fact, take it off the bike. but you might get it clear anyway, if you can shift the gears to the right 5 mm.

 

[neptronix]

TC40xx is an expensive, not so efficient, and difficult to drive motor.. i wouldn't bother with it.

MXUS 3kW ( really more like 2500w in a 26" ), leafmotor 1500w ( more like a 2000w rating ), or edge 1500w motor ( similar to the leaf ) are the way to go for a bike with 135mm dropouts.

 

[thunderstorm80]

TC40xx is an expensive, not so efficient, and difficult to drive motor.. i wouldn't bother with it.

MXUS 3kW ( really more like 2500w in a 26" ), leafmotor 1500w ( more like a 2000w rating ), or edge 1500w motor ( similar to the leaf ) are the way to go for a bike with 135mm dropouts.

TC40xx not efficienct and not to drive with? Why?
It has similar cogging torque like the H35xx family, but it's a much more powerful motor.
I will use it with phaserunner, PAS, and will adjust my max torque output on the fly. Why is it difficult to drive with, more than the others you wrote?

Where do you recommend to buy the MXUS ?
I am still puzzled why Grin-Tech, which is by no doubt the leading & cutting edge in E-bikes innovation and technology, doesn't have any of the ones you wrote in their stock...

 

[thunderstorm80]

Bend that sucker out to 140 or 145mm, to fix that dishing problem. Steel bike. 8)

I'm assuming the motor has an extra 5 mm of axle on each side to work with.

Don't sweat it if the chain rubs the tire when you use the inner ring up front, you won't be using it, ever. in fact, take it off the bike. but you might get it clear anyway, if you can shift the gears to the right 5 mm.

The left side of the chain-stay is bent to the right (and the right chain-stay is strait), so the dropout area is 135mm and shifted to the right. Bending this left chainstay is possible, but it will also make it a bit longer than the right side of of the dropout - therefore losing the "trueness" of the dropout axle vs the front wheel and the ground. It will be way more complicated to bend and true it, than a symmetrical bending force when compress/expand both sides of the dropout at the same time. What do you think?

 

[neptronix]

TC40xx not efficienct and not to drive with? Why?
It has similar cogging torque like the H35xx family, but it's a much more powerful motor.
I will use it with phaserunner, PAS, and will adjust my max torque output on the fly. Why is it difficult to drive with, more than the others you wrote?

Where do you recommend to buy the MXUS ?
I am still puzzled why Grin-Tech, which is by no doubt the leading & cutting edge in E-bikes innovation and technology, doesn't have any of the ones you wrote in their stock...

Look at these motors on the grin motor simulator and check out the dyno graph posted by crystalyte themselves on their 40mm motors.
Tops out at around 85% efficiency.. you can see this in the ebikes.ca simulator also.

In 2014, i posted a motor that tops out at 90.5% efficiency and produces very similar power but is dramatically lighter, smaller, and cheaper.. the difference is all in the laminations and how it is wound. Cogging is much much lower than a TC80x or H40x of course, due to the thinner lams.

Also, if you'd been following grin tech.. they were surprised by the MXUS 3kw's efficiency.. it has similarities with the leafmotor design, and lower cogging than the 0.5mm lam crystalyte motors also.

Grin has some good stuff for sale, but not everything they sell is automatically the best. The leaf motor and edge motor are both direct sales only, so there is nobody reselling them. Why grin doesn't sell the mxus 3kw motor is beyond me. You should always cross shop and look at specs anyway.

 

[thunderstorm80]

TC40xx not efficienct and not to drive with? Why?
It has similar cogging torque like the H35xx family, but it's a much more powerful motor.
I will use it with phaserunner, PAS, and will adjust my max torque output on the fly. Why is it difficult to drive with, more than the others you wrote?

Where do you recommend to buy the MXUS ?
I am still puzzled why Grin-Tech, which is by no doubt the leading & cutting edge in E-bikes innovation and technology, doesn't have any of the ones you wrote in their stock...

Look at these motors on the grin motor simulator and check out the dyno graph posted by crystalyte themselves on their 40mm motors.
Tops out at around 85% efficiency.. you can see this in the ebikes.ca simulator also.

In 2014, i posted a motor that tops out at 90.5% efficiency and produces very similar power but is dramatically lighter, smaller, and cheaper.. the difference is all in the laminations and how it is wound. Cogging is much much lower than a TC80x or H40x of course, due to the thinner lams.

Also, if you'd been following grin tech.. they were surprised by the MXUS 3kw's efficiency.. it has similarities with the leafmotor design, and lower cogging than the 0.5mm lam crystalyte motors also.

Grin has some good stuff for sale, but not everything they sell is automatically the best. The leaf motor and edge motor are both direct sales only, so there is nobody reselling them. Why grin doesn't sell the mxus 3kw motor is beyond me. You should always cross shop and look at specs anyway.

It's not only what is the top efficiency, but what is the motor's kV & it's phase resistance. Maybe the 4080 is not the best speeder on the flat (in fact the GrinHub&9C+ surpass it but it overheats faster), but it's an enormous climbing machine, and it will persist at 70%+ efficiency even at high torque outputs at low speeds, where other motors would just overheat very quickly. I assume the MXUS is better though if it's rated to 3KW, although I don't know anything about it to compare with.
I would much prefer we all compared each motor's Kv+Phase resistance along with their cogging losses rather than just saying "3KW" which really doesn't say much...

 

[neptronix]

It's not only what is the top efficiency, but what is the motor's kV & it's phase resistance. Maybe the 4080 is not the best speeder on the flat (in fact the GrinHub&9C+ surpass it but it overheats faster), but it's an enormous climbing machine, and it will persist at 70%+ efficiency even at high torque outputs at low speeds, where other motors would just overheat very quickly. I assume the MXUS is better though if it's rated to 3KW, although I don't know anything about it to compare with.
I would much prefer we all compared each motor's Kv+Phase resistance along with their cogging losses rather than just saying "3KW" which really doesn't say much...

Actually the leafmotor 4T is most likely the most high speed winding you're going to find. But you can make speed with volts or amps. So the winding choice has more to do with what speed you want at what volt, etc.

Grinhub and 9C+ overheat faster because you are simulating them at speeds above their continuous current capability. At their best, they are a bit more efficient than the TC at it's most ideal speed.

And of course the TC performs better up hill and under higher load. It's less efficient, but significantly larger.

Using phase resistance and kV to gauge motor performance is a big mistake. Here is one example.. the 35mm leafmotor and 35mm crystalyte 35xx are actually very similar. The crystalyte is realistically 1200-1500w continuous capable.. the leafmotor is realistically 1750-2000w continuous capable. The motors are the same dimensions all around,and might even have the same phase resistance and kv.. they are also roughly the same weight.. so what's the difference?

The leafmotor has windings that are pulled tighter.. and thinner laminations.

Same can be said of the 9C versus the 9C+. The old 0.5mm lamination motor is rated at 500 watts.. the newer one is rated at 750 watts. Same amount of steel and copper.. same dimensions... huge difference in efficiency.. and with less heat generated per watt in, comes more continuous AND burst power.

A modern 40mm wide stator should be producing 91-92% efficiencies. If the crown was as good as it's name, and was that good.. i'd be forking out the premium price crystalyte was asking. I just don't get why you'd want one otherwise.

PS: I owned a H35xx... i was very disappointed in it because my magic pie outperformed it.. and the magic pie isn't that great of a motor in the first place.

 

[thunderstorm80]

It's not only what is the top efficiency, but what is the motor's kV & it's phase resistance. Maybe the 4080 is not the best speeder on the flat (in fact the GrinHub&9C+ surpass it but it overheats faster), but it's an enormous climbing machine, and it will persist at 70%+ efficiency even at high torque outputs at low speeds, where other motors would just overheat very quickly. I assume the MXUS is better though if it's rated to 3KW, although I don't know anything about it to compare with.
I would much prefer we all compared each motor's Kv+Phase resistance along with their cogging losses rather than just saying "3KW" which really doesn't say much...

Actually the leafmotor 4T is most likely the most high speed winding you're going to find. But you can make speed with volts or amps. So the winding choice has more to do with what speed you want at what volt, etc.

Grinhub and 9C+ overheat faster because you are simulating them at speeds above their continuous current capability. At their best, they are a bit more efficient than the TC at it's most ideal speed.

And of course the TC performs better up hill and under higher load. It's less efficient, but significantly larger.

Using phase resistance and kV to gauge motor performance is a big mistake. Here is one example.. the 35mm leafmotor and 35mm crystalyte 35xx are actually very similar. The crystalyte is realistically 1200-1500w continuous capable.. the leafmotor is realistically 1750-2000w continuous capable. The motors are the same dimensions all around,and might even have the same phase resistance and kv.. they are also roughly the same weight.. so what's the difference?

The leafmotor has windings that are pulled tighter.. and thinner laminations.

Same can be said of the 9C versus the 9C+. The old 0.5mm lamination motor is rated at 500 watts.. the newer one is rated at 750 watts. Same amount of steel and copper.. same dimensions... huge difference in efficiency.. and with less heat generated per watt in, comes more continuous AND burst power.

A modern 40mm wide stator should be producing 91-92% efficiencies. If the crown was as good as it's name, and was that good.. i'd be forking out the premium price crystalyte was asking. I just don't get why you'd want one otherwise.

PS: I owned a H35xx... i was very disappointed in it because my magic pie outperformed it.. and the magic pie isn't that great of a motor in the first place.

The H35xx has a low power density: It weights 7.5Kg, just 1.5Kg less than the TC4080, but it's rated to 1200W while the TC4080 is rated to 2000W.
The 9C+, too, with 1000W max rated power (According to Grin), while it weights 6Kg. (With similar power density the TC4080 should have been 12Kg instead of 9Kg).
So if we judge motors by their rated power, this is a significant advantage which justifies the crown, and I also understood it's the main reason why the H35xx costs half of it. Just like spending more on a bicycle which would have same, but lighter (and more reliable) components.
I was simulating the GrinHub and the 9C+ on flats, with a load which is similar to their rated power. If you stay in that territory, they outperform the TC4080, but that's the same as installing a semi-trailer truck motor in a small private car. I still understand that the MXUS is better than the TC4080 as it has a better efficiency across the board thanks to it's smaller laminations.

 

[thunderstorm80]

Bringing back the original topic of this thread: No one knows if such extreme tension ratio can work or not? (without frame bending)

 

[neptronix]

Well, i'd say that's no problem with a steel bike. Bend 'er a bit and deal with any misalignment of disc brakes or drivetrain using angled washers or whatever.

Your pugsley has really difficult to deal with dropouts though. They are not very strong, and the wheel-slips-out-the-rear design is a disadvantage. Those weird lips are gonna get in the way too, yet are crucial for the strength of the dropout itself :? Your torque arms have to be strong and perfectly set up with less than 0.5mm of slop, period. You have one disc brake mounting point for a torque arm that is half decent, but will need to fabricate something for the side that only has a rear rack washer on a very weak spot..

Luckily the tubes on the rear look to be perfectly circular? so you could head to the hardware store and get the appropriate size clamps, and with a file, drill, and maybe an angle grinder, build a very tight bridge between the rear drop and the rear tubes at multiple points..

( here is my dropout setup for the non-disc side of my difficultly shaped Turner O2.. this actually held up to 6000W with 1000W of regen.. and this motor possibly puts out more peak torque than yours at that wattage. However, this bike has very strong dropouts to begin with. )

But.. this is not a particularly strong frame in the rear for what you are trying to do. I would run this motor at a conservative wattage and run a putzy 2.25:1 phase amp to battery amp ratio or close to, in order to flatten out the initial torque spike. Don't use a controller that isn't programmable, otherwise you will be stuck with whatever phase amp ratio the controller was set to at the factory.

 

[thunderstorm80]

Well, i'd say that's no problem with a steel bike. Bend 'er a bit and deal with any misalignment of disc brakes or drivetrain using angled washers or whatever.

Your pugsley has really difficult to deal with dropouts though. They are not very strong, and the wheel-slips-out-the-rear design is a disadvantage. Those weird lips are gonna get in the way too, yet are crucial for the strength of the dropout itself :? Your torque arms have to be strong and perfectly set up with less than 0.5mm of slop, period. You have one disc brake mounting point for a torque arm that is half decent, but will need to fabricate something for the side that only has a rear rack washer on a very weak spot..

Luckily the tubes on the rear look to be perfectly circular? so you could head to the hardware store and get the appropriate size clamps, and with a file, drill, and maybe an angle grinder, build a very tight bridge between the rear drop and the rear tubes at multiple points..

( here is my dropout setup for the non-disc side of my difficultly shaped Turner O2.. this actually held up to 6000W with 1000W of regen.. and this motor possibly puts out more peak torque than yours at that wattage. However, this bike has very strong dropouts to begin with. )

But.. this is not a particularly strong frame in the rear for what you are trying to do. I would run this motor at a conservative wattage and run a putzy 2.25:1 phase amp to battery amp ratio or close to, in order to flatten out the initial torque spike. Don't use a controller that isn't programmable, otherwise you will be stuck with whatever phase amp ratio the controller was set to at the factory.

I don't have the frame yet. I wanted to ask here because I was sure someone already did it, before I spend the money.
It seems according to what you say that the Pugsley isn't a good candidate for a rear motor.
In that case, I will take the Fat-Bike which I already have (Titan Whopper), with it's 190mm rear dropout and install the TC4080 on the steel fork. (The fork has 135mm OLD)
This is a touring bike which will have significant pedalling power too, and in that way I don't have to give up my modern 9/10 speed system for an older 7 speed freewheel.
Thanks to you, this is clear now. Thank you!

There are two more issues:

1. The road tires I will mount. I plan to go with Surly Black Floyd 3.8inch fat tires as it's the widest road tire I could find which also has treads for water channeling. Since on fatbikes the actual tire foot-print varies ALOT due to the the rims ranging from 50-100mm, I was wondering if someone have them and can measure their actual width and what rim they use. Surly has a nice table for telling what tire width you will get for each rim, but it's always better to here from someone's firsthand experience. Do you know other tables like this where users input their actual tires widths vs the rim they use? I googled this for the last 2 days and found nothing. I will give you one example: On my Fat-Bike I have Bontrager Barbagezi 4.7inch tires, and with 82mm rim they are closer to 4.2inch. A big let down...

2. I look for a 82mm (or around it) with 36h and also one 100mm with 36h. I found 80mm one on Lunacycle:
https://lunacycle.com/weinmann-dhl80-26x4-36h-fat-bike-rim/
but they don't write the ERD, and even Weinmann's website doesn't tell. How can someone buy a rim from a manufacturer without knowing it's ERD??? (This was also part of my 2 days of googling)
Do you have experience with Lunacycle? I never bought from them yet.

 

[neptronix]

Wait.. you're going to put it on the front wheel? huge mistake.
I would put this on a loosey goosey rear dropout before i put it on the front wheel.

I have had traction problems, and funky steering with a 9lb. 500w geared motor in a front wheel due to the extreme weight dis balance. I also see the front fork bending visibly under power at 900W peak.. in no way would i advocate putting a motor with 4 times the power rating and almost 3 times the weight on a front fork.

 

[thunderstorm80]

Wait.. you're going to put it on the front wheel? huge mistake.
I would put this on a loosey goosey rear dropout before i put it on the front wheel.

I have had traction problems, and funky steering with a 9lb. 500w geared motor in a front wheel due to the extreme weight dis balance. I also see the front fork bending visibly under power at 900W peak.. in no way would i advocate putting a motor with 4 times the power rating and almost 3 times the weight on a front fork.

It's the torque which bends the fork, not the power. We've had this discussion before about why I prefer front motor:
1. It's a 2AWD drive.
2. It's a rigid steel fork.
3. The motor weights 9kg on the front, and the battery 12kg on the back. I think this is quite balanced.
4. I am using Phaserunner so I am directly controlling my output torque and will be dialing my max torque limit on the fly.
5. I don't want to lose my modern drivetrain system. Why on Earth no manufacturer besides ezee are making their rear-motors with the modern Shimano/SRAM cassette hubs? This is a format already worldwide, and you can very easily design your motor according to it... The ezee motor I have was the easiest one I have ever installed, plus the joy of transferring my 10 speed XT cassette to it.

 

[Warren]

I run a TC4080 on my recumbent cargo bike. It has performed flawlessly for 12,446 miles, so far. I had to jack the steel rear stays out to accommodate the 140 mm axle, plus 5 mm spacer,

http://www.ebikes.ca/shop/electric-bicycle-parts/motor-parts/spacewash5.html

to work with a 7 speed freewheel. The chain passes the 58 x 507 tire by an eighth of an inch, in low.



But this is on bike where the chain runs 39" to the underseat idler. This is the equivalent of 39" chainstays, not 17.6" of the Pugsley. So I would guess under a sixteenth inch of clearance. Probably not a problem with a non-knobby tire.

 

[neptronix]

It's the torque which bends the fork, not the power. We've had this discussion before about why I prefer front motor:
1. It's a 2AWD drive.
2. It's a rigid steel fork.
3. The motor weights 9kg on the front, and the battery 12kg on the back. I think this is quite balanced.
4. I am using Phaserunner so I am directly controlling my output torque and will be dialing my max torque limit on the fly.
5. I don't want to lose my modern drivetrain system. Why on Earth no manufacturer besides ezee are making their rear-motors with the modern Shimano/SRAM cassette hubs? This is a format already worldwide, and you can very easily design your motor according to it... The ezee motor I have was the easiest one I have ever installed, plus the joy of transferring my 10 speed XT cassette to it.

Torque and power go together. Can't move fast without both.
I played with limiting initial torque on my 500w motor and could not find the balance between being able to accelerate properly from a stop and not slipping the front wheel.
12kg battery on the back does not change traction issues when 75% or more of the rider weight is on the rear wheel. You have the same traction problems as rear wheel drive trucks have with the engine being in the front, and the drive happening in the rear. You will learn that while trying to turn under power on water, mud, sand, etc.

By the way, i built a bike with a 'rigid steel fork' from Surly and another front geared motor with even less power. The fork bent under load and transmitted vibration and noise to the handlebars.

Multiple companies are making freewheel designs. Some of the motors i suggested have cassette options. eZee is 1 of maybe a dozen motors with cassette freewheels.

You can ignore 100% of my advice, but please heed this one.. wear protective gear when you ride this bike, every time. A fast ebike is already an inherently unsafe vehicle to travel in. Add a traction disadvantage at high speeds, and it is even less safe if you happen to have the throttle engaged and suddenly hit rough ground during a turn. Your best bet is also to design your handlebars so that you are in a forward crouched position as often as possible as well.

Good luck.

 

[thunderstorm80]

It's the torque which bends the fork, not the power. We've had this discussion before about why I prefer front motor:
1. It's a 2AWD drive.
2. It's a rigid steel fork.
3. The motor weights 9kg on the front, and the battery 12kg on the back. I think this is quite balanced.
4. I am using Phaserunner so I am directly controlling my output torque and will be dialing my max torque limit on the fly.
5. I don't want to lose my modern drivetrain system. Why on Earth no manufacturer besides ezee are making their rear-motors with the modern Shimano/SRAM cassette hubs? This is a format already worldwide, and you can very easily design your motor according to it... The ezee motor I have was the easiest one I have ever installed, plus the joy of transferring my 10 speed XT cassette to it.

Torque and power go together. Can't move fast without both.
I played with limiting initial torque on my 500w motor and could not find the balance between being able to accelerate properly from a stop and not slipping the front wheel.
12kg battery on the back does not change traction issues when 75% or more of the rider weight is on the rear wheel. You have the same traction problems as rear wheel drive trucks have with the engine being in the front, and the drive happening in the rear. You will learn that while trying to turn under power on water, mud, sand, etc.

By the way, i built a bike with a 'rigid steel fork' from Surly and another front geared motor with even less power. The fork bent under load and transmitted vibration and noise to the handlebars.

Multiple companies are making freewheel designs. Some of the motors i suggested have cassette options. eZee is 1 of maybe a dozen motors with cassette freewheels.

You can ignore 100% of my advice, but please heed this one.. wear protective gear when you ride this bike, every time. A fast ebike is already an inherently unsafe vehicle to travel in. Add a traction disadvantage at high speeds, and it is even less safe if you happen to have the throttle engaged and suddenly hit rough ground during a turn. Your best bet is also to design your handlebars so that you are in a forward crouched position as often as possible as well.

Good luck.

I agree that front traction is not as good as rear, but if I don't have any option for rear 190mm axle motors WITH modern cassette freehub, I will have to stay with the front only, for now. If you do know of such, can you show me examples?
I also totally agree with you about how easy it is to fall and lose control on such E-bike if you are not aware. I will wear protective gear when I go offroad.
I already had a front motor PAS bike before, and as long as I plan to draw few hundred watts from it at max together with very little assisting torque on uphills, I am not expecting any dangerous issues.
The motor is not going to pull the entire bike mass - I am going to pedal along and do at least half of the work. (I can output 300W continuously on a good day). Being a pedal bike, I will be already leaning forward in MTB riding position, so a greater proportion of weight is going to be on the front wheel. The front wheel will carry panniers, too. (and rear panniers on the rear)
I chose the 4080 not because it can output 2000W continuously. I chose it because it is also quite an efficient climber at low speeds without overheating in minutes, and it would be an awesome regen capable.
I would have NO PLANS at all to speed up with it to 50Kph or so... I already have A fast city commuting E-bike (my "car") which has ezee on the rear and 9C+ on the front, so I know very well how poorly the 9C+ would have handled this situation by itself, so don't worry - I will not put an even greater motor for fast city-riding...
The planned speeds for the touring bike I talk about would not surpass 30Km/h on the flats (road only), and I will regen myself on downhills to a similar speed so I don't reach 60kph or so.

 

[thunderstorm80]

Multiple companies are making freewheel designs. Some of the motors i suggested have cassette options. eZee is 1 of maybe a dozen motors with cassette freewheels.

Good luck.

Can you refer me to a company that sells rear DD motor for a Fat-bike (190mm OLD), with a modern cassette hub?
I can find results in google, but none provide any spec regarding resistance/kV/magnetic losses - only speed vs voltage but they also doesn't say if it's loaded or not. Sometimes it's so unclear that you can't tell if it's a geared motor or DD....

 

[thunderstorm80]

Your pugsley has really difficult to deal with dropouts though. They are not very strong, and the wheel-slips-out-the-rear design is a disadvantage. Those weird lips are gonna get in the way too, yet are crucial for the strength of the dropout itself :? Your torque arms have to be strong and perfectly set up with less than 0.5mm of slop, period. You have one disc brake mounting point for a torque arm that is half decent, but will need to fabricate something for the side that only has a rear rack washer on a very weak spot..

Luckily the tubes on the rear look to be perfectly circular? so you could head to the hardware store and get the appropriate size clamps, and with a file, drill, and maybe an angle grinder, build a very tight bridge between the rear drop and the rear tubes at multiple points..

It seems that all Surly framesets have those horizontal dropouts. Do you have knowledge/experience with other frames they make, and if all would have the same difficult installation issues? (due to the horizontal dropouts)
And why you say the dropouts are weak? They are made from cromoly steel which is much stronger than the common aluminium rear dropout on a regular bike.