For the Slowtwitch Mailbag this week, we’ll cover two new questions. First, do custom bikes require safety testing just like mass-produced bikes? Next, will we ever see dual-position mechanical shifting?

If you’d like to submit a triathlon-related question to our mailbag, send an email to: Mailbag (at) Slowtwitch (dot) com.

--

Jeffrey writes,

I would like to know how bike frames are tested before they are put into production. My understanding is that production bikes require testing but custom bikes do not. If that is the case how do custom bike builders test frames before they hit the road?

Thanks for a great web site and all the info you give us.

For this question, we consulted with what we’ll call “an established industry veteran who has been in the bike building business for a long time, but asked to remain anonymous.”

They reply,

“The first part of the answer should deal with what constitutes testing in the bicycle industry. Big multi-brand conglomerates like Trek or Giant have the resources to fund in-house testing of their production bikes in ways that the small builder can only dream about, using complex and expensive machinery that recreates real-world stresses on a bike frame, essentially running it through the kind of stress it would normally only experience in 20 years under a 220 pound sprinter. However, this process takes weeks or even months to test a single frame, so it's hideously expensive and well beyond the means of the small custom builder. Other manufacturers utilize destructive testing processes where a frame is loaded until it fails, such as by anchoring the bottom bracket with the frame pointed at the ceiling and loading a surrogate fork with dead weight until the top and/or down tube buckles, or a weld fails. While this is better than nothing, it really only explores one very narrow aspect of frame longevity, specifically your ability to run into solid objects without breaking your bike, though it will also identify gross design errors.

For the small builder tasked with building a one-off custom frame, two choices are available. The first (the one virtually none of them will choose) would be to build up the frame in question, then subject it to destructive testing to determine its ultimate failure points. The down side of this should be obvious. The second approach that the conscientious designer / builder can take is more pragmatic, which is to rely on a century or so of anecdotal evidence about bicycle design to guide the design and construction processes. If a builder designs and builds around materials and geometry that have been in play for decades, he knows what to expect in terms of outcome. If a nearly identical combination of tubes and geometry has worked for scores of other builders for many years, there's no reason for the builder to expect a different outcome. This assumes proper construction technique, of course - the best tubes in the world will cease flying in formation if the welds are done improperly.

Another decision the designer / builder has to make is ‘how close to the edge of the envelope he wants to fly’. With the ever-present push to make everything lighter, bike companies are forced to choose between longevity, safety margin, and sales. Shave four ounces off that frame, and you can sell more of them for a higher price, at the cost of potential failures. The cost of these failures is built into the sales model, and many riders are willing to pay for a frame that should be retired after a season or two of riding, to gain the perceived advantages. A small builder who tries to emulate these cutting-edge exercises in material brinksmanship is risking creating a frame that could be short-lived, or even dangerous, because it's possible he won't have the quality control over the construction of the frame that the big guys do. The prudent designer / builder drawing up a custom frame will include significant overkill on all aspects of the design - this won't cost much more in material or even weight, but the end result will be a frame designed for the real world, so that any deviation from the norms of design or construction won't result in the spontaneous deconstruction of the frame during the Saturday club ride.”

The short version is this: the custom builder must use their noggin. Just as a single data point, I’ll list the failures and/or problems I’ve personally had with mass-produced bikes and custom frames. I’ve owned quite a few mass produced frames/forks, 6 custom frames, and two custom forks.

Mass Produced

-Two fork safety recalls
-One warrantied carbon frame; cracked rear carbon dropout
-One recurring creaking/loosening integrated TT stem/fork, not recalled or warrantied (‘live with it’)
-Carbon frame with poorly designed seatpost clamp (wedge design). Seatpost slipped no matter what; frame warrantied.
-One set of rear dropouts out-of-alignment and poorly machined; carbon frame with alloy dropouts. Frame had to be switched from SRAM to Shimano components due to the floating upper pulley (which is more forgiving of frame misalignment than a fixed pulley).
-Many non-prepped threaded BB shells
-Improperly formed chainstays on aluminum cyclocross bike; could not run standard size CX chainrings. Had to get creative and ‘rig up’ an off-brand crank and single chainring.
-Uneven rear tire clearance on aluminum 29” mountain bike; suspected bad weld/warped seat stay or incorrectly shaped seat stay (consequence seen below)

Custom

-Bottle cage mounts placed too close together inside main triangle; titanium frame. Sent it back to be cut out, filled, and re-done
-Seat tube bottle cage mounts crooked; titanium frame (not the same frame as above). Sent back to be aligned
-Undersized seat tube (well under the 34.9mm specification); had to Coke-can-shim the front derailleur clamp
-Custom mountain bike rigid fork; disc brake tab ~3mm too far left. Used shims on both the hub and caliper. This one gets a relative ‘pass’ because it was a very unique fork/hub project that I dreamed up.

Interestingly enough, the only issues that are safety-related come from the mass produced items (the recalled forks, cracked dropout, and loosening integrated stem/fork). The bulk of my issues with both custom AND stock bikes are detail-related; how do the parts interact with that frame/fork? I find that quality in this regard does not correlate with price; my least expensive custom bikes (Habanero) are the only ones with zero problems.

--

David writes,

Do you think companies that currently manufacture mechanical gear shifters (Shimano, SRAM, etc) will ever produce a TT style brake with an integrated mechanical gear shifter?

Currently, the market only offers the standard road style shifter with hoods, or a shifter located on the end of the aero bar extension. For those of us who desire to own one 'do it all bike', these options do not offer what I feel would be the best possible set up.

I use my bike for commuting to work, group rides, and triathlons. My Shimano 105 shifters are sufficient for commuting and group rides, but when it comes to racing, the use of the road handlebars are no good. The drops on the standard rode bike are extra weight and loss of aerodynamics. However, if I were to use an aero base bar and have the shifters on the aero bar extensions, I no longer have an efficient or safe set up for commuting and group rides.

I think SRAM is in the best place to offer this type of mechanical shifting option as their double tap system already allows for a single shifter lever, where Shimano has the shifter lever and the need to push in the brake level to shift the opposite direction.

I know this desired cockpit set up is obtainable with a Shimano Di2 system, but that is not cost efficient for the mass market. Just think how awesome your Habanero Uber Training Bike [see link at the bottom of this page] would be if you had an profile design Ozero or T2 aero base bar with mechanical shifters rather than rather than the Zipp drop bar. Are you really going to use the drops on the handlebars on that bike?

That’s a big bunch of questions, David! Let’s start from the top.

As of today, I would be very surprised if any major manufacturer decided to put this idea into production. I haven’t heard any rumors in this regard from anyone in the industry. Here are my main reasons why:

1. The difficulty of the design. In order to have two shift levers actuating one shift cable, you need some sort of splitter or inline device. We see these from time to time with brakes – adding a 3rd brake lever to an aerobar, or using supplemental brake levers for cyclocross (which mount to the ‘tops’ of a drop bar; see the TRP lever below).

The reason it works with a brake is simple: You’re pulling more cable than is required to actuate a rear derailleur shift. All of these split-brake or inline designs seem to introduce a measure of ‘slop’ into the system. The supplemental lever never feels as good as a normal lever. The reason it still works ‘okay’ is because you have such a long cable pull; it gives you more room for error. If you introduced a similar precision-reducing device into a rear shift system, I think it just wouldn’t work. Indexing would probably be off and it wouldn’t work right with one of the levers. As it stands, most existing mechanical systems only give you about one quarter turn of error to the rear barrel adjuster for optimum performance.

2. It already exists with electronic shifting, as you mentioned. As time goes on, I think we’ll see this technology trickle down even lower than Ultegra-level, with a matching lower price. I’ve not used Campagnolo’s EPS system, but I have ridden with Dura Ace 7970 and Dura Ace 9000 dual-position shifting. It works so incredibly well that I just don’t see them trying to re-engineer it into a mechanical platform. I see them trying to cut cost out of the existing electronic product.

3. The system would likely be bulky. Where do you put the splitter and extra housing (see the Jtek brake splitter below)? Does it require a proprietary aerobar? Would other aerobar manufacturers support the design? Would it work with proprietary integrated front-ends (stem/bar/fork combos)?

4. Triathlon components (shifters, brake levers, handlebars), see MUCH lower sales volume than standard road product. This is a big reason why product evolution tends to be slower and options are few and far between. How long did it take Campagnolo to make an 11-speed bar-end shifter? It has to make sense for the bottom line. Think about handlebar companies that offer a single width of base bar, but four widths of road drop bars. Or seven lengths of road stems, but only two stem lengths on their integrated aerobar. How long has it taken for short cranks to really catch on with major manufacturers such as SRAM, FSA, Shimano, and many others? Part of it is cultural within the company. If you look at things like overall sales pie charts and marketing/sponsorship budgets, you’ll find that these large companies are, for the most part, focused on road and mountain bikes.

That all being said, we never know what the future holds. If there was one company that I could see maybe doing it, it’s SRAM; they appear steadfast in their support of mechanical systems. This is a total shot in the dark, but I think it would more likely be executed with a hydraulic shift system. With a closed fluid system (i.e. no fluid reservoir at the lever), you could push fluid from either the brake lever or aerobar position, provided that you plumbed in an appropriate Y-style fitting. The total volume would be fixed. Just talking round numbers, if it took 2ml of fluid to actuate a shift, the rear derailleur really shouldn’t ‘care’ which lever pushed the fluid. The obvious downside is that if air got into the system, the shift indexing would be off (because air compresses under pressure).

Finally, I left my Habanero ‘Uber Bike’ with drop bars because that bike is all about options (to the extreme), and I don’t race triathlons on it. I do use the drops on occasion. I also like having the ability to convert it to a normal cyclocross bike very easily – just remove the aerobars and flip the seatpost/saddle.