I have both Di2 and mechanical bikes.

Di2
My carbon Di2 is great on fast group rides and on rolling, hilly rides. It's like clicking a mouse to shift, and I can do it with my ring fingers from the hoods. The rear shifts great under load on hills, and the front shift is quite quick and drama free. I'll sometimes shift the front for just a 30 foot high climb. I'll often shift the rear for just 2 or 3 pedal revolutions, then shift again.

To keep up with fast rides, I'm always hunting for my best cadence, so I shift a lot. And when the other riders are standing and grinding up a short hill in a too-high gear, I've already dropped to my 34 chainring and a big cog. And way back in the draft, I'll go 50-11 or 50-12 and pedal very easy at a low cadence to both save my legs, and avoid coasting. (It's easier on the following riders when the rider in front doesn't pedal & coast repeatedly.)

Mechanical
I have a gravel / all-day bike with a triple crank. It's great for really low gears on steep rides. And the 39 middle chainring is really nice with a 12-25 cassette for all-day flatter rides--I have very close shifts that just change the cadence by 5 or 6 rpm. (If Di2 had a triple, I probably would have done that.)

I'm riding at a more relaxed pace, and don't need to shift quite as much, and can afford to wait longer to do a shift. So the mechanical shifting is fine for these rides. It's interesting how quickly my fingers adapt to the very different shifting when I switch bikes.

Drew,

Sorry to dis your bike. There are two points of view. One is that "It should last indefinitely if not crashed." The other was exemplified in the extreme by the recent thread "CARBON FRAMES NOT SAFE AFTER THREE YEARS!" (or somesuch). I'm an engineer, and looking at the fatigue chart for carbon fiber composite I think that the answer is probably somewhere between the extremes. If the bike was designed such that the parts are somewhat stressed (and a good design does this, else the frame is too heavy), then it will fail after a certain period of normal use. Given your Kestral frame is competitive in the weight department, it WILL fail after a period of normal use. What is that period? That's where we don't have a lot of data. IIRC the guy posting the "NOT SAFE" post had a catastrophic failure in his older (2002? 2005? It's late forgive my laziness not looking it up). So my only failure data point is that the safe period for a CF bike is 5-20 years. But there are CF bikes that are decades old and going strong. The interval will be shorter, if you weight 240# (like me) and ride a very light bike (Trek Domane Series 6). Longer if you weigh 120# and ride a 50cm frame that is really beefy. So there is a risk in riding a CF bike for decades. It may be nil, or it may be significant. The experts here are pretty positive on longevity, but the post I mention above documents a failure after 15 years or so. Point is, the useful life of CF bikes is not established in my view. I love my current CF bike, but will probably sell it after five years of use.

But you were interested and excited about trying Di2. To tell the truth, I was thinking of it. Given that I'm an engineer AND I worked in a bike shop for many years I'm pretty persnickety, and the kit for my bike is a bit more expensive (~$1400-1700 - heavens, you wouldn't expect me to use derailleurs with a different name on them, would you?*). That's put me off so far, but I definitely can see the interest and excitement about the upgrade.

I guess if you don't like the current setup, or it doesn't do something you want/need, and you can afford it you could try it. I'm thinking if you don't like it, you probably can resell it on eBay for very little loss.

Sorry to be the fart in your elevator of excitement.

*So far, I haven't gone through the bike and replaced all the nuts and bolts with titanium, as my mechanic friend (and engineer, btw) at my LBS did.

I wouldn't bother upgrading an older bike to electronic shifting (be it Di2, eTap or EPS), but I can't imagine purchasing a new bike without it. If you have the means and it turns you on I can't see a reason not to go electronic. It really is that good.

thanks for the input guys !


B@t™

Anything that is more complicated will have more complicated failures. Mechanical shifting has worked great right from the start and continues to do so.

If you love your Kestrel frame, which is awesome and unique, go for it. Get a group from one of the UK online vendors and do it yourself and save a ton on money.
Seems you can just run the wire along the bottom of the down tube, tape it nicely with clear or matching tape and it will be practically invisible.

Tap test that frame while you have the parts off. Make sure there are no areas that make thunking or dull sounds.

No, just stop it now.

CF does not typically fail unless damaged by being used in a manner for which it was NOT designed.

CF bikes that fail are because people dropped them, crashed them, banged them up on their racks or possibly in rare cases manufacturing defects.

As well, consider replacement cost. A Shimano Di2 6870 rear derailleur is about $165. A mechanical 6800 is about $55-60. A Di2 internal battery at $130 that'll probably go 5 years ?. That's a lot of shifter cables.

As much as I might desire Di2 and can see moving towards it in time, the overall cost of everything is so much higher, makes me currently stick with mechanical.

Andr0id, I disagree in theory, and in practice the situation is not as clear cut as your argument assumes.

Any material that has a fatigue failure mode will break, given high enough stress and enough cycles of flex. Any material - steel, aluminum, or CF. See here for simulated pictures of an actual event: an F-15 breaking in half in flight, due to fatigue. Recall that stress is, simply put, force per area. If we assume that the forces a bike part sees are about the same for any material (for argument's sake for right now), then the smaller the part's cross sectional area, the higher the stress. But the he smaller the area, the lower the weight. So there is a performance objective (low weight) that pushes designs to low cross sectional area in parts, which gives higher stress. So a good design will have stresses that will eventually cause fatigue failre. Your argument is that, for an undamaged bike that is not abused, carbon fiber will never break. My point is that for a properly designed bike (e.g. one with cross section low enough to be competitive in weight), there is a finite limit to the frame life due to CF composite fatigue failure. It may be 500 years, but it is finite. But that 500 year figure is based upon fresh composite with a fresh matrix resin. I'm not sure that the longevity of the resins used have been studied. The resin is almost certainly not as stable as the underlying carbon fibers. So there may be another failure mode due to resin matrix failure or delamination, which might be 5, 10, or 20 years. The bike company legal weasels are trying to limit liability with there "Hey, this here bike may fail after 3 years of use" statements. I don't think its that bad (e.g. the damned things fall apart in month 37), but CF frames will fail. Add to this that there's certainly a significant amount of variability between brand build quality and materials. So I don't think we can say a bike will not fail for 30, or even 20 years.

The practical matter is that CF frames DO fail, and catastrophically so. For some pics, see here, on the Bike Forum. We agree that most failures are probably due to local damage caused by a frame member being poked, hit, or otherwise not used as intended. But how do you tell? I have a CF frame (and I love it), but the head attached to air pump I had attached in a bracket made a little blemish on my seat tube. I noticed this before the CF itself was frayed and will rectify the issue, but will this cause my frame to fail? These small dents can be what are called stress concentrators, and can cause local stress to exceed normal stress. If the stress exceeds yield stress, in a steel bike the frame bends a bit. In a CF bike, the failure mode is brittle and you get sudden and total failure. I think that there's a lot of these small little dings and dents that folks don't even notice that may (or may not) cause CF frame failure.

But looking at what is called a fatigue chart (the stress applied, as a percent of yield strength, plotted against cycles to fail) for CF is much flatter than a lot of other materials. So undamaged CF probably does have exceptional life, as you point out. But if there's a small local bit of damage, or if the bike wasn't constructed ideally, the local stress at points may be higher than design, leading to premature and sudden failure.

i think this partly depends as well on how much you ride. I put in 7-8k miles a year, and so I'm generally putting on new cables 2-3x a season, new housing every winter, and a roll of $30 bar tape annually. Add in the cost of a $50 set of Park cable cutters to do this at home (or the labor cost to have your shop do it) and it adds up fast. In five years, I've already spent well beyond what Di2 would have cost in parts and labor to have mechanical, no doubt. I'm reminded of the pitfalls of mechanical each year when I rip off and throw away bar tape that still looks brand new.