SuDiBe Tandemseiten 


Tandem Wheels

by Bill McCready

(Über Tandem-Laufräder im Allgemeinen und die Entwicklung der 16-Speichen Tandem-Laufräder bei Santana und Shimano. Aufbereitet aus der englischsprachigen Mailingliste, 2003)

Prior to 2001, there was no test facility that scientifically measured the dynamic endurance of a bicycle wheel.

Without such a facility the best testing was either "team" or "fleet" use where dozens (if not hundreds) of wheels could be evaluated over one or more seasons of real use. In the case of tandems, since 1977 the largest identifiable fleet has been Santana tandems---which, for the first 15 years were equipped with 140mm rear wheels. From day-one these wheels came with a one-year performance warranty. And warranty claims are the surest way to track problems. Further, for most of these years Santana was the only reliable source for 40- and 48-hole replacement rims. In other words, through warranty claims and spare parts Santana was in a unique position to track the long-term performance of tens of thousands of tandem wheels.

When Santana first came to market, the best comparable fleet of tandems would have been Schwinn Paramounts. Introduced a decade earlier, these tandems had 36 heavy-gauge spokes. Within a few hundred miles a majority of Paramount owners would experience their first broken spoke. Owners of Santana's early tandems would ride many thousands of miles without changing a spoke.

The difference? While most would guess spoke count, the Paramount's 33% spoke- count deficit (36 vs. 48) was offset by thicker spokes with 32% greater cross- section. Because the total calculated spoke strength per wheel was nearly equal, the reliability advantage of the Santana wheel was NOT due to additional spokes, but was instead the result of wider symmetrical spoke bracing allowed by Santana's wider frame. This was proven by dozens of Paramount owners who installed a Phil Wood 48-hole rear wheel. Because of narrow mismatched spoke bracing angles, chronic wheel problems soon reappeared.

The reliability of Santana's original wheels was soon compromised. In 1979, to accommodate a sixth cog, Phil Wood repositioned the flanges inward. In the mid- 1980s a seventh cog required 2mm of dish. With 7-cogs, the warranty rate for our 40-spoke wheels became excessive by our standards (i.e. more than 5% of our customers experienced a second spoke-breakage incident within the first year of use).

Long before the eighth cog appeared, Santana's experience with tens of thousands of wheels allowed us to produce a formula that uses measured spoke angles and dish to calculate a wheel's "integrity index." When I published some of the scores to this forum a few years ago, some complained our index was less than 100% scientific. After using our formula for a dozen years, however, we have complete confidence in its ability to predict relative reliability.

The index showed us that adding the eighth cog within the 140mm standard would reduce the relative reliability of our wheels by 22%. Because we were already unhappy with a one-in-twenty ratio of warranty claims, we rejected this alternative.

Our calculations also indicated that evolving to 160mm would increase the reliability of our wheels by 39%.

Does Santana believe these numbers? Absolutely. We proceeded to mock-up drivetrains, work with vendors, produce new tooling, and discount hundreds of in-stock frames. By investing ten of thousands of dollars early on, Santana evolved from 7-speed 140mm directly to 8-speed 160mm. We not only avoided the mistake of cramming an eighth cog into 140mm (which caused other builders and their customers a great deal of grief), we also doubled the term of our wheel warranty. Imagine our satisfaction when the claim rate dropped from over 5% to less than one-half of one percent.

How reliable is 145mm? Instead of publishing a score that many would not care to believe, I'll ask you to consider the following. Because modern 8/9 speedcogsets are 4.2mm wider than the 7-speed cogsets they replaced, you would need to stretch the hub by 4.2mm on each side in order to maintain identical spoke bracing angles. In other words, for the current 8/9 speed standard to be equally reliable to the old 7-speed standard, Trek, Burley, Co-Motion and Ibis would have needed to agree on a width of at least 148.4mm. Because they couldn't reach an agreement on any dimension except 145mm, today's wheels from these builders and (more recently) Cannondale will have a lower "integrity index" than the wheels they delivered ten years ago.

But is a 160mm 40-spoke wheel more reliable than a 145mm 48-spoke wheel? Yes, and here's the 3-step proof. Because Santana's current 40-spoke wheels have a lower warranty claim rate than our older 48-hole 140mm wheels, we know today's 40x160 is stronger than yesterday's 48x140. Since 140mm wheels built ten years ago had a better spoke bracing geometry than today's 145mm wheels, we also know yesterday's 48x140 is stronger than today's 48x145. Logically, it follows that a 40 spoke 160mm wheel is stronger and more reliable than a 48-spoke 145mm wheel.

But is Shimano's new 16-spoke 160mm wheel stronger than both?

Synopsis thus far

Before Santana existed, I endured a decade of tandem wheel misery before working with Phil Wood, my bike shop's customers and independent frame builders to produce the 140mm tandem frame standard and the first sets of tandem wheels reliable enough to facilitate the subsequent popularity of tandeming. Twenty-five years of warranty claims and rim distribution have allowed Santana to track the reliability of many tens of thousands of tandem wheels. By 1991 Santana used this data to create a formula for an "integrity index" that predicts the relative reliability of a wheel built around any particular hub. This index caused us to realize that the emergence of 8-speed cassettes would necessitate a new standard wider than 140mm. Tandem builders who decried the need for a new standard proceeded to confirm our prediction by producing three years of relatively unreliable 8-speed wheels. Instead of joining the pre-existing 160mm standard that they had criticized (which could have been embarrassing) the brighter builders among them lobbied for 155mm. The consortium they formed, however, could not agree on a dimension beyond 145mm, a minimal step that allowed the conversion of in-stock steel frames. While the innocuous change from 140mm to 145mm has been successful to the degree that wheel reliability became acceptable, the resulting wheels are, on average, approximately 30-40% weaker than a 160mm wheel built with the same rim and spokes.

Ultra-performance tandem wheels

By 1996, a drop in warranty claims verified the incredible strength of 160mm and allowed us to consider radically lighter tandem wheels with fewer than 40 spokes. While 36 spokes was an obvious choice, our index pointed us to 32. We worked with Zipp, and were disappointed to discover that light 32-hole rims cannot tolerate the higher spoke tensions required by tandems. With the spokes properly tensioned the front rim would eventually distort into a wavy series of S-bends that prevented smooth braking.

At a subsequent Eurobike (Santana's favorite bike show) I met Rolf Dietrich, who was showing his original sets of paired-spoke wheels. I immediately realized paired spokes was the breakthrough that would allow tandems to use lighter rims with fewer spokes. A few other people noticed those wheels too. The next time I contacted Rolf he was contracted to Trek. Nevertheless, we later met and discussed the co-development of Rolf tandem wheels with lightweight rims and 28 spokes.

Before Santana could supply Rolf with the sample hubs he requested, however, I was invited to an exclusive unveiling of Shimano's prototype 16-spoke wheels.

While many enthusiasts assume that all brands of paired-spoke wheels have a comparable level of performance, those with a clear understanding of spoke bracing immediately knew Shimano's cross-over lacing had leapfrogged Rolf's efforts. The day of that industry-only preview I sat down to lunch with Shimano's product engineers, and proposed a 24-spoke 160mm dishless version of their wheelset that would be suitable for tandems. While they were excited, we imagined the small size of the project would cause Shimano to prefer to license their wheel design to Santana, who would produce wheels in the U.S.A. (A year earlier Santana had obtained a Shimano's first outside license to produce splined bottom brackets.) Because Shimano's engineers had a long backlog of projects, Santana would need to develop our own prototypes.

Development and Testing

Using only a photo of Shimano's still-unannounced 16-spoke wheels, Santana worked with Mike Hadley to create sample sets of tandem wheels with Shimano's unique lateral crossover lacing for testing and evaluation. When we brought them to Interbike, Shimano politely asked us to hide them away. The reason? Because Interbike was the first opportunity for dealers and the press to inspect Shimano's first-ever wheelset (that had already won Tour de France stages and the Polka Dot jersey for the best overall climber), they hoped we wouldn't wish to diminish the impact of their debut. We quickly agreed. Except for dealers who saw them during the first hour of show, the only attendees to inspect our wheels was a constant stream of Shimano employees who were "impressed" by the quality of the Santana/Hadley knockoffs.

Over the next few months Santana continued testing and development of what we still assumed would be a licensed-by-Shimano and produced-by-Santana wheelset. Because of mutual concern that our wheels should meet their criteria, in mid-2000 we sent a pair of these U.S.-built wheels to Japan. After some weeks of poking and prodding, Shimano came back with an interesting observation: our 24-spoke prototypes had too many spokes. By their calculations our dishless design with widely separated flanges justified a 20-spoke configuration. I politely asked them to recheck their figures.

Imagine our surprise, a few weeks later, when a package arrived at Santana with a set of Shimano 16-spoke wheels built around a 160mm disc-brake tandem hub. When I phoned my contact in Japan I learned that instead of rechecking their figures Shimano's engineers used a super-computer to model our wheels and simulate various forces. At that point they realized the combined strength of Shimano's design and Santana's 160mm spacing could produce a tandem-rated wheelset with just 16 spokes. (They did increase the gauge of the bladed spokes from 2.0mm to 2.3mm). Possibly more important, Shimano's engineers had become so captivated with our project that they now wanted to build them themselves. After negotiating a kill-fee with Hadley for 18 months of development assistance, I transferred the production of Santana's wheels to Shimano.

At the next Interbike we brought our Shimano-built samples. A few minutes after the show started we were asked by Shimano to remove the wheels from our display. It seems one Shimano's directors had been caught off-guard by the "exclusive" product and wanted to have this novel (for Shimano) arrangement reviewed at the highest levels. A few weeks later the project received the approval subject to an inspection of our factory by Shimano's V-P of Quality Control. Impressed with our operation, Shimano's first builder-specific component was re-approved.

The next surprise came when Shimano sent us a purchase order for five pairs of our stock 40-spoke wheels built around Hadley 160mm hubs. When we received this request I called to find out why they needed ten built-up tandem wheels. The answer? Because the V.P. was concerned about the "dynamic" performance and reliability of these revolutionary tandem wheels, he convinced other Shimano execs to green-light a long-envisioned project---the completion of the first-ever wheel testing facility that would duplicate the punishing lateral loads of out-of-the-saddle sprints and climbs. Once this robotic "torture chamber" for wheels was finished, the destruction of the Santana-supplied Hadley wheels would be used to set a baseline. The disappointing news was that Shimano was not moving ahead with our 16-spoke tandem wheels until their long-term reliability could be verified in this yet-unfinished facility.

Many additional months passed while Shimano built and calibrated this unique wheel lab. In the meantime we completed our second year of riding hand-built prototypes. Finally the long awaited two weeks of scientific testing took place.

Their report? The wheels failed.

The conclusion of a long explanation:

When designing competition wheels for bicycles, a starting point is the intended use (i.e. event, terrain and rider weight). Since 1996, when Santana first started development of an ultra-performance tandem wheelset, our goal had been complete reliability over uneven pavement for the strongest racers with a combined team weight as high as 150kg, or 330 pounds. When Shimano's quality control team visited our factory, this limited goal was discussed and reconfirmed.

Santana's standard wheelsets are, of course, far stronger than most teams will ever realize. Consequently, our 40-spoke wheels are guaranteed to remain trouble-free for two full years-a written warranty with no limitations on use, terrain, mileage or (especially) weight.

In spite of our limited goal for our 16-spoke wheels, at Santana we punished our prototype road-test wheels with heavy teams (up to 390 pounds) and severe terrain (including stair-step single track). When I got the report that, in spite of Shimano's computerized calculations and two-years of brutal real-world testing, the wheels had failed to meet our limited goal, I was dismayed.

"Don't worry," said my California-based contact who relayed the reported failure but could not provide many details, "they have already decided to modify the design with more spokes."

Unsatisfied, that night I called Japan. To keep a long story short, Shimano's quality-control department had unilaterally decided to ignore Santana's limited goal. In order to be approved for production the new 16-spoke tandem wheels would need to prove themselves 100% as durable as Santana's 40-spoke wheels---the wheels they had used to calibrate their wheel lab.

The most amazing part of this report was that the 16-spoke rear wheel (with wide flange separation) had exceeded this benchmark. It was the front wheel (with standard flange separation) that had fallen a bit short. I then patiently explained to my contact why Santana and most other tandem builders supply wheels with a matching number of front and rear spokes (hint: it has nothing to do with front wheel reliability). I further reminded him that for years Trek had specified mismatched sets of Shimano hubs (36-hole front and 40-hole rear) for their tandems.

Shimano agreed to lace-up some traditional 36-spoke front wheels for a retest. As was expected, the 16-spoke wheels with lateral-crossover lacing proved stronger.

While it took a long explanation to get to this point, my answer verifies the following:

Using a combination of Santana's considerable experience plus dynamic test results from the world's newest and most sophisticated wheel testing facility, Santana and Shimano can proudly stand behind the jointly-made claim that our new 16-spoke 160mm rear tandem wheel (with widely separated flanges and symmetric lateral-crossover lacing) will be more reliable than a traditional 145mm rear tandem wheel with three times as many spokes. This new wheel is also considerably lighter, laterally stiffer, a bit more comfortable, and far more aerodynamic.

This thoroughly-tested wheelset, backed by Shimano's two-year guarantee, is a significant milestone for tandem componentry.

The matching front wheel fits all tandems. It is, however, is only slightly stronger than a traditional wheel with 36 spokes.

D. Bettge; letzte Änderung: 12.10.2003