Gates Carbon Drive + Geared Hub

[georgec]

***************
NOTE: By mistake, I first emailed this directly to Cruzbike instead of posting to the forum. Doug Burton provided a very helpful response, which I have attached below, with his permission.
***************

Date: Thursday, July 12, 2012, 3:47 PM

I would like to buy a Softrider or Silvio for commuting and weekend riding. These look like bikes that will be fun to ride for a long time. But I hate chain/shifter/grease/mess/on frame and pants. I have read stories of the MBB making love to the rider's leg, which I can easily visualize it doing to my leg. I also understand that these parts run cleaner on a Cruzbike... Really! But... they are still greasy, right?

So... I would like to put a carbon belt drive / IGH (internally geared hub) on the Softrider or Silvio. When I look around the forums, I see a fair bit of angst around this idea. But I am a handy engineer with access to a good machine shop and willing to void my warranty, at least on certain parts.

As far as I can tell from the forums, the key question is how you tension the belt.

Here are a few scenarios I would like to suggest:

A) Silvio: Modify the joint between the BB and carbon stay to allow adjustment of overall length. It looks like the stay has a bead that locks into a slot on the BB. The most primitive approach I can think of is to grind (ouch!) the bead off and put a tapered shim in above it. This leaves bolt pressure alone to secure the stay to the bottom bracket.

B) Softrider: Cut the front stay forks on each side a few inches from the hub end. Thread the parts internally with right hand threads toward the bottom bracket and left hand threads toward the hub. Make a matching contra-threaded insert that allows overall stay length to be adjusted independently on each side. This makes each side of the stay into a sort of turnbuckle.

C) Silvio: Replace the carbon stay with a custom part that allows adjustment in length on both sides. Perhaps with turnbuckle action along the lines described in B) In each case, the fork needs to absorb the torque produced by the hub. Is this a big deal (provided that I attach the hub to the fork without weakening the fork)?

I am interested in any reactions/suggestions folks might have. Also, I welcome comments on which IGH might work best in each approach.

Thanks,
George

***************
UPDATE: I am most interested in doing this on the Silvio, so I have posted here. But I left the softrider related comments the original emails.
***************

From: Douglas Burton
Sent: Thursday, July 12, 2012 11:30 PM

Hello, George,

The Silvio isn't adaptable to any IGH that I am aware of. The following issues present:

1. The Silvio fork uses 130mm dropout spacing, and that dimension is on the tight side to aid front triangle rigidity. The fork won't spread to the 135mm spacing used on IG hubs happily, if at all. The 130mm dimension carries further to the carboyoke and outer carboyoke attachment parts, and these can't be spread at all.

2. The outer dropouts, that connect the carboyoke to the fork dropouts, are very wide, again for rigidity. The quick release, for example, is 170mm long instead of the usual 141-146mm. The IG hubs with which I am familiar use 182mm axles, and these are integral to the internal drivetrain and can't be switched out for longer ones. Thus there's insufficient axle length for the axle nuts to properly grab, and the outer dropouts would have to be through-drilled to pass the axle, not to mention there's no remaining clearance for the shifter unit.

The Sofrider will accept an IG hub if a tensioner is fitted.

We have no experience with belt drives, but haven't had any complaints about chainring tattoos as the chain falls quickly away to the lower side, rather than following the frame as with a conventional RWD recumbent.

We can't recommend modifying the forks and front end components. The rounded section of the Silvio carboyoke serves to positively locate the components fore-aft and rotationally, and they prevent fretting of the carbon fiber by the attaching brackets at the bottom bracket. Removing these will allow the carboyoke to move relative to the bottom bracket and likely cause eventual failure of the part.

Hope this helps.

Best,

Doug

 

[georgec]

carbon belt drive - IHG for Softrider or Silvio

On the Silvio, I believe I can address the issues that Doug has raised by making custom outer dropouts with eccentric tensioners.

The rest of the bike would remain stock, with the possible exception of "bending" or "shaving" the fork to an inside dimension of 132mm.

Assuming I can do that, what comments or suggestions do folks have?

Thanks, George

 

[Romagjack]

I've been riding Cruzbikes

I've been riding Cruzbikes for many years and currently enjoy the Quest 2 with IGH and disc brakes. Great for traffic especially when has to stop suddenly and you need to shift into a lower grear. But, I still crave the speed of a roadbike and would love to see a Silvio with IGH and disc brakes. It seems that many Cruzbike riders have evolved from DF bikes, are older or have spinal issues. Any modifications that can be done to improve comfort and ease of use while maintaining speed might appeal to a great number of Cruzbike riders. I would certainly be interested in a "Special Edition" Silvio (engineered by Cruzbike) with these features, but would not want to risk integrity of the frame. George, have you looked at the Quest? John and Doug - anything on the drawing board like a Quest 3 with 700c wheels or a Silvio with an IGH package?

 

[georgec]

Thanks for the encouragement!

Hi Connie,

When you say, "I still crave the speed of a roadbike and would love to see a Silvio with IGH and disc brakes," you have perfectly captured the reason I am fixated on the Silvio.

Thanks for the encouragement!

 

[John Tolhurst]

The belt needs to be very

The belt needs to be very tight, or it can slip. When the suspension operates, I am not sure the belt will retain sufficient tension.

 

[georgec]

John Tolhurst said, "When the

John Tolhurst said, "When the suspension operates, I am not sure the belt will retain sufficient tension."

To address this concern, we need to estimate "chain growth" and "rear triangle stiffness."


CHAIN GROWTH

Gates says, "In general, full suspension frame designs result in some form of chain growth. Chain growth being defined as a change in the resting distance between the axis of the bottom bracket axle and the axis of the rear hub axle. Because the belt does not have the ability to stretch and the tension in the belt must remain constant, even the smallest amount of chain growth during suspension travel would be detrimental to the system integrity." [1]

As I see it, the only source of chain growth is the flex in the chain stay as the bike goes over a bump. The shape of the curve in which the chain stay flexes, and total displacement, will determine chain growth. To determine the shape of the curve, we need to know the answer to a simple question:

When the front tire hits a bump, and the front shock compresses, and the chain stay flexes...

A) Do the dropouts rotate around the axle relative to the fork?

B) Or, do the dropouts remain locked at a constant angle relative to the fork?

If the answer is:

A) Then the chain stay will flex in a simple curve rooted at the BB, and the fork/dropout mating surfaces will show wear once they get some miles on them.

B) Then the chain stay will flex in a "S" curve, and there will be no wear on the bearing surfaces between the dropouts and front fork.

Do we know which it is?

Once we know how the chain stay flexes, I can calculate the amount of chain growth and ask the Gates people an intelligent question about it [2].


REAR TRIANGLE STIFFNESS

Gates also says, "Gates engineers have developed a method for directly measuring and quantifying rear triangle stiffness as it relates to belt drive performance. This frame stiffness evaluation service is offered by Gates to all frame engineers as a product development tool. It is highly recommended to submit frames for evaluation as early in the frame development process as possible." [1]

Gates measures this stiffness in terms of how much the triangle flexes in response to tension applied along the chain line.

This leads to the question:

C) Is the Silvio less stiff than a typical road bike in response to tension along the chain line?

If the answer to C) is yes, the answer to the following question should also be yes:

D) Do riders experience the Silvio crank as more squishy than normal?


Ref:
[1] 2012 Gates Carbon Drive Technical manual:
http://www.carbondrivesystems.com/downloads/tech_docs/GATES_TECH%20MANUAL_ENG_REV041.pdf

[2] Gates says, "If a full suspension frame design which utilizes the Carbon Drive system is desired, please contact Carbon Drive Systems (carbondrives@gates.com) for engineering and development assistance." [1]

 

[Andrew 1973]

My opinion

"I have read stories of the MBB making love to the rider's leg, which I can easily visualize it doing to my leg."

We're all adults here, so I won't comment on the appropriateness of the comment from the beginning of this thread, but I think you need to understand that the chain routing on the Silvio and Sofrider is such that contact with the leg is not an issue meriting any significance. As a matter of fact, I have yet to receive a "chainring tattoo" from riding a Cruzbike Sofrider. To me, your idea of a belt-driven IGH is interesting, but is overcomplicating a solution for a problem that may not even exist.

If it is simply a matter of keeping the drivetrain from potentially contacting your pants or leg, why not consider simpler options such as a chainring guard or a band or strap to keep your pant leg from getting close to the chain/chainring?

If you are interested solely in the belt-driven IGH for the sake of its perceived benefits, why not consider building a conversion and MAYBE modifying the conversion kit to accomodate the hub and tensioner mechanism? As a "handy engineer with access to a good machine shop", it is likely that you could fabricate the parts you need, if not build a MBB front end from the ground-up.

I can't speak for Mr. Tolhurst, but it seems to me that with the Silvio being intentionally designed as a road bike utilizing road bike componentry, the idea of using an IGH, which for many reasons is not favored by performance cyclists, is simply not in keeping with the Silvio's intended position in the product line.

It's just my thoughts and opinions, which you said you welcome.

 

[Romagjack]

Good Thread

John T's brilliant designs are evolving in several directions and comments on this thread show it. It's too bad that the Silvio with it's 700c wheels and dual suspension can't be easily modified to accept IGH and disc brakes - a dream bike for baby boomers like me that evolved from competitive riding to the great "ease and comfort" that I enjoy everyday on my IGH Quest 2. An IGH Silvio or a Quest 3 with 700c wheels would give me the extra speed that would get me back into weekend bike group that still use DF's. I've grown so accustomed to the Quest IGH (never thought I would having been such a purist), that I will never go back to an all derailleur drivetrain. You would think the IGH makers could engineer a product that could make a Boomer IGH Silvio just another option package rather than requiring redesign from the frame maker. Anxious to hear the thoughts of others - I know I'm not the only aging cyclist on the forum. I'll be ready for the leaning 700c IGH Cruzbike trike in 10 years!

 

[Charles.Plager]

Hi Romagjack,
This topic is

Hi Romagjack,

This topic is somewhat controversial (whether or not tire diameter affects speed), but I don't think the 26" (559) tires on the Quest slow it down much at all (although the default tires might somewhat). It's weight on the other hand, certainly when climbing, do put it at a disadvantage compared to much lighter DFs.

So maybe a carbon fiber Quest 3.0?

Cheers,
Charles

 

[John Tolhurst]

When the front tire hits a

When the front tire hits a bump, and the front shock compresses, and the chain stay flexes...

A) Do the dropouts rotate around the axle relative to the fork?

B) Or, do the dropouts remain locked at a constant angle relative to the fork?

Pretty sure it is B), if the quick release is done tightly enough. The other variable is the angle between fork and chainstay.

 

[georgec]

When the chain stay is acting

When the chain stay is acting as a compression strut, forcing it into a S-curve is actually encouraging it to go out of column, which reduces stiffness. This suggests the counterintuitive possibility that the front triangle might be stiffer if the ends of the stays were "pin" joints.

 

[georgec]

Here are the custom dropouts

Here are the custom dropouts I am thinking of making.

A tug (red) encloses a Shimano no-turn washer (blue) that keeps the hub axle from turning relative to the fork and provides positive chain tension adjustment.

I used a tug instead of the original eccentric idea because I think it provides more positive tensioning.

The axle length problem that Doug mentioned will be addressed by machining a relief (not shown) on the axle nuts that allows them to pass partially into the cover washers (teal).

 

[John Tolhurst]

the circular boss is key

Suggestion,
lose the blue washers, do it all with the red one fully surrounding the circular boss on the end of the fork.

If there were pin joins, then yes, it would retain more strength in the brief moment when it the suspension has compresssed. If there were an electric motor providing constant torque instead of a human pedalling we might run into difficulties with loading a deformed chainstay. But pin joints are costly, heavy and troublesome.

What change in length do you calculate? 1mm? It might be that a very brief reduction in the RC length will not trouble the belt's engagement with the teeth under the pulsing power of pedaling.

 

[georgec]

Good idea. The blue washers

Good idea. The blue washers are gone. The relief on the axle nut accommodates the axle length issue that Doug raised earlier.

For an S-curve induced in a 440mm chain stay by a 1.5 inch rebound, "chain growth" is 1.3mm. The Gates technical manual specifies 2mm for tensioning, so 1.3mm could be a problem. Further design mods could be:

A) If the dropout (green) is allowed to slide freely while the axle, tug (red), and fork are locked together by axle nut pressure, a spring could be added under the tug tension screw nut to help "absorb" chain growth. This would effectively substitute the "spring constant" of the Gates belt for that of the carbon stay... probably not a good thing for front triangle stiffness.

B) A bearing could be worked into the tug design to let the dropout rotate freely about the fork. This provides a pin joint at the axle end of the stay, and the stay bends in a simple curve. I estimate that this reduces chain growth by 1/2.

It's probably time to try to talk to someone at Gates.

 

[John Tolhurst]

fork dropouts

email me john@cruzbike.com and I will send you the cad of the fork dropouts. They can't be allowed to get to thin.

 

[georgec]

Fat in the fire!

Looking at John's cad files, (thank you, John) I became convinced that this is going to work, so I ordered the Silvio.

I also bought a kit of all the Gates Carbon Drive & Shumano Nexus 8 hub related parts from a new Civia Kingfield.

I am going to put a chain (horrors!) on the bike for the fall ride while I fiddle with the custom dropouts. I plan to use a Biopace 42 chainwheel and a 20 cog on the Nexus 8.

Now I need brakes, levers, & shifter.

I have read that you can't use a brifter with the Nexus 8, so I plan to use the Jtek bar end shifter from the Civia, or a Alfine RapidFire Shifter.

I need good brakes, so I am inclined to use the Shimano Ultegra BR-6700. But I read that you must use the BR-6700 brifters with these calipers. Does anyone know if this is really true?

I guess I could put the brifters on the bike and just not connect the shift wires, but this does seem a bit lame, don't you think?

Can anyone recommend...

1) Ordinary levers that will work well with the Shimano Ultegra BR-6700 calipers?

2) A stong caliper and lever combo that will work well on the Silvio?

 

[georgec]

Nexus 8 hub is in the bike and works great!

Some details:

Shaved 2mm off of the hub's dust cover nut so that the hub fits in the 130mm fork clearance. Ground down the Nexus no-turn washers to fit between the fork and dropouts. Modified the dropouts by boring to accommodate the 3/8" axle and cutting a relief for the axle nuts in each outer surfaces. Set up for now with Biopace 42 chainring, 20 cog, & chain made up neat.

Lovin the bike. Commuting to work. I am one very happy camper!

BTW: chain grease is already on my pants, so stay tuned for the belt drive dropout mods when the snow is falling.

Many thanks for all the helpful comments.

 

[ak-tux]

@georgec, Pictures speak!

@georgec, Pictures speak! Please show us how it looks now!

 

[Romagjack]

You're a genius! Photos would

You're a genius! Photos would be great.

 

[leakyduck]

Chain Grease

"BTW: chain grease is already on my pants, so stay tuned for the belt drive dropout mods when the snow is falling."

According to Sheldon Brown lubing the chain is one of the most controversial issues in cycling. That said I will put in my $.02 worth.

I clean my chain by soaking in WD-40, wipe it dry and put it back on the bike. I then put one drop of Tri-Flow on each roller. Takes a while but that is the wear point. By running the chain "dry" I don't get oil on my pants and it isn't picking up grit and dirt (which is actually where most of your wear comes from).